pin tools like staticcheck and misspell in tools.go, update Makefile

This commit is contained in:
Shawn Smith
2021-04-24 14:30:55 +09:00
parent 5a9b17f852
commit 5e9737e8b7
676 changed files with 124863 additions and 6950 deletions

View File

@@ -3,18 +3,28 @@ all: lint build test
build:
go build ./...
install:
install:
./scripts/make-install.sh
lint:
gometalinter --exclude=vendor --exclude=repos --disable-all --enable=golint --enable=vet --enable=gofmt ./...
find . -name '*.go' | xargs gofmt -w -s
lint: fmt vet misspell
test:
fmt:
./scripts/gofmt.sh
vet:
go vet ./check ./cmd/... ./download ./handlers ./tools/...
go vet ./main.go
staticcheck:
@[ -x "$(shell which staticcheck)" ] || go install honnef.co/go/tools/cmd/staticcheck
staticcheck ./...
test:
go test -cover ./check ./handlers
start:
go run main.go
misspell:
find . -name '*.go' -not -path './vendor/*' -not -path './_repos/*' | xargs misspell -error
@[ -x "$(shell which misspell)" ] || go install ./vendor/github.com/client9/misspell/cmd/misspell
find . -name '*.go' -not -path './vendor/*' -not -path './_repos/*' -not -path './download/test_downloads/*' | xargs misspell -error

17
go.mod
View File

@@ -3,22 +3,25 @@ module github.com/gojp/goreportcard
go 1.14
require (
github.com/DataDog/zstd v1.4.5 // indirect
github.com/DataDog/zstd v1.4.8 // indirect
github.com/beorn7/perks v0.0.0-20160804104726-4c0e84591b9a // indirect
github.com/client9/misspell v0.3.4
github.com/dgraph-io/badger/v2 v2.2007.2
github.com/dgraph-io/ristretto v0.0.3 // indirect
github.com/dgryski/go-farm v0.0.0-20200201041132-a6ae2369ad13 // indirect
github.com/dustin/go-humanize v1.0.0
github.com/golang/protobuf v1.4.2 // indirect
github.com/golang/snappy v0.0.2 // indirect
github.com/golang/protobuf v1.5.2 // indirect
github.com/golang/snappy v0.0.3 // indirect
github.com/kr/pretty v0.2.1 // indirect
github.com/matttproud/golang_protobuf_extensions v1.0.1 // indirect
github.com/pkg/errors v0.9.1 // indirect
github.com/prometheus/client_golang v0.9.0-pre1.0.20171201122222-661e31bf844d
github.com/prometheus/client_model v0.0.0-20190812154241-14fe0d1b01d4 // indirect
github.com/prometheus/common v0.0.0-20171117163051-2e54d0b93cba // indirect
github.com/prometheus/procfs v0.0.0-20171221151313-8f918ac9ab4b // indirect
golang.org/x/net v0.0.0-20200927032502-5d4f70055728 // indirect
golang.org/x/sys v0.0.0-20200926100807-9d91bd62050c // indirect
golang.org/x/tools v0.0.0-20190628034336-212fb13d595e
google.golang.org/protobuf v1.25.0 // indirect
golang.org/x/lint v0.0.0-20201208152925-83fdc39ff7b5
golang.org/x/net v0.0.0-20210331212208-0fccb6fa2b5c // indirect
golang.org/x/sys v0.0.0-20210402192133-700132347e07 // indirect
golang.org/x/tools v0.1.0
honnef.co/go/tools v0.1.3
)

119
go.sum
View File

@@ -1,17 +1,16 @@
cloud.google.com/go v0.26.0/go.mod h1:aQUYkXzVsufM+DwF1aE+0xfcU+56JwCaLick0ClmMTw=
github.com/BurntSushi/toml v0.3.1 h1:WXkYYl6Yr3qBf1K79EBnL4mak0OimBfB0XUf9Vl28OQ=
github.com/BurntSushi/toml v0.3.1/go.mod h1:xHWCNGjB5oqiDr8zfno3MHue2Ht5sIBksp03qcyfWMU=
github.com/DataDog/zstd v1.4.1 h1:3oxKN3wbHibqx897utPC2LTQU4J+IHWWJO+glkAkpFM=
github.com/DataDog/zstd v1.4.1/go.mod h1:1jcaCB/ufaK+sKp1NBhlGmpz41jOoPQ35bpF36t7BBo=
github.com/DataDog/zstd v1.4.5 h1:EndNeuB0l9syBZhut0wns3gV1hL8zX8LIu6ZiVHWLIQ=
github.com/DataDog/zstd v1.4.5/go.mod h1:1jcaCB/ufaK+sKp1NBhlGmpz41jOoPQ35bpF36t7BBo=
github.com/DataDog/zstd v1.4.8 h1:Rpmta4xZ/MgZnriKNd24iZMhGpP5dvUcs/uqfBapKZY=
github.com/DataDog/zstd v1.4.8/go.mod h1:g4AWEaM3yOg3HYfnJ3YIawPnVdXJh9QME85blwSAmyw=
github.com/OneOfOne/xxhash v1.2.2 h1:KMrpdQIwFcEqXDklaen+P1axHaj9BSKzvpUUfnHldSE=
github.com/OneOfOne/xxhash v1.2.2/go.mod h1:HSdplMjZKSmBqAxg5vPj2TmRDmfkzw+cTzAElWljhcU=
github.com/armon/consul-api v0.0.0-20180202201655-eb2c6b5be1b6/go.mod h1:grANhF5doyWs3UAsr3K4I6qtAmlQcZDesFNEHPZAzj8=
github.com/beorn7/perks v0.0.0-20160804104726-4c0e84591b9a h1:BtpsbiV638WQZwhA98cEZw2BsbnQJrbd0BI7tsy0W1c=
github.com/beorn7/perks v0.0.0-20160804104726-4c0e84591b9a/go.mod h1:Dwedo/Wpr24TaqPxmxbtue+5NUziq4I4S80YR8gNf3Q=
github.com/census-instrumentation/opencensus-proto v0.2.1/go.mod h1:f6KPmirojxKA12rnyqOA5BBL4O983OfeGPqjHWSTneU=
github.com/cespare/xxhash v1.1.0 h1:a6HrQnmkObjyL+Gs60czilIUGqrzKutQD6XZog3p+ko=
github.com/cespare/xxhash v1.1.0/go.mod h1:XrSqR1VqqWfGrhpAt58auRo0WTKS1nRRg3ghfAqPWnc=
github.com/client9/misspell v0.3.4 h1:ta993UF76GwbvJcIo3Y68y/M3WxlpEHPWIGDkJYwzJI=
github.com/client9/misspell v0.3.4/go.mod h1:qj6jICC3Q7zFZvVWo7KLAzC3yx5G7kyvSDkc90ppPyw=
github.com/coreos/etcd v3.3.10+incompatible/go.mod h1:uF7uidLiAD3TWHmW31ZFd/JWoc32PjwdhPthX9715RE=
github.com/coreos/go-etcd v2.0.0+incompatible/go.mod h1:Jez6KQU2B/sWsbdaef3ED8NzMklzPG4d5KIOhIy30Tk=
@@ -22,44 +21,28 @@ github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c
github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
github.com/dgraph-io/badger/v2 v2.2007.2 h1:EjjK0KqwaFMlPin1ajhP943VPENHJdEz1KLIegjaI3k=
github.com/dgraph-io/badger/v2 v2.2007.2/go.mod h1:26P/7fbL4kUZVEVKLAKXkBXKOydDmM2p1e+NhhnBCAE=
github.com/dgraph-io/ristretto v0.0.3-0.20200630154024-f66de99634de h1:t0UHb5vdojIDUqktM6+xJAfScFBsVpXZmqC9dsgJmeA=
github.com/dgraph-io/ristretto v0.0.3-0.20200630154024-f66de99634de/go.mod h1:KPxhHT9ZxKefz+PCeOGsrHpl1qZ7i70dGTu2u+Ahh6E=
github.com/dgraph-io/ristretto v0.0.3 h1:jh22xisGBjrEVnRZ1DVTpBVQm0Xndu8sMl0CWDzSIBI=
github.com/dgraph-io/ristretto v0.0.3/go.mod h1:KPxhHT9ZxKefz+PCeOGsrHpl1qZ7i70dGTu2u+Ahh6E=
github.com/dgryski/go-farm v0.0.0-20190423205320-6a90982ecee2 h1:tdlZCpZ/P9DhczCTSixgIKmwPv6+wP5DGjqLYw5SUiA=
github.com/dgryski/go-farm v0.0.0-20190423205320-6a90982ecee2/go.mod h1:SqUrOPUnsFjfmXRMNPybcSiG0BgUW2AuFH8PAnS2iTw=
github.com/dgryski/go-farm v0.0.0-20200201041132-a6ae2369ad13 h1:fAjc9m62+UWV/WAFKLNi6ZS0675eEUC9y3AlwSbQu1Y=
github.com/dgryski/go-farm v0.0.0-20200201041132-a6ae2369ad13/go.mod h1:SqUrOPUnsFjfmXRMNPybcSiG0BgUW2AuFH8PAnS2iTw=
github.com/dustin/go-humanize v1.0.0 h1:VSnTsYCnlFHaM2/igO1h6X3HA71jcobQuxemgkq4zYo=
github.com/dustin/go-humanize v1.0.0/go.mod h1:HtrtbFcZ19U5GC7JDqmcUSB87Iq5E25KnS6fMYU6eOk=
github.com/envoyproxy/go-control-plane v0.9.1-0.20191026205805-5f8ba28d4473/go.mod h1:YTl/9mNaCwkRvm6d1a2C3ymFceY/DCBVvsKhRF0iEA4=
github.com/envoyproxy/protoc-gen-validate v0.1.0/go.mod h1:iSmxcyjqTsJpI2R4NaDN7+kN2VEUnK/pcBlmesArF7c=
github.com/fsnotify/fsnotify v1.4.7/go.mod h1:jwhsz4b93w/PPRr/qN1Yymfu8t87LnFCMoQvtojpjFo=
github.com/golang/glog v0.0.0-20160126235308-23def4e6c14b/go.mod h1:SBH7ygxi8pfUlaOkMMuAQtPIUF8ecWP5IEl/CR7VP2Q=
github.com/golang/mock v1.1.1/go.mod h1:oTYuIxOrZwtPieC+H1uAHpcLFnEyAGVDL/k47Jfbm0A=
github.com/golang/protobuf v1.2.0/go.mod h1:6lQm79b+lXiMfvg/cZm0SGofjICqVBUtrP5yJMmIC1U=
github.com/golang/protobuf v1.3.1/go.mod h1:6lQm79b+lXiMfvg/cZm0SGofjICqVBUtrP5yJMmIC1U=
github.com/golang/protobuf v1.3.2/go.mod h1:6lQm79b+lXiMfvg/cZm0SGofjICqVBUtrP5yJMmIC1U=
github.com/golang/protobuf v1.4.0-rc.1/go.mod h1:ceaxUfeHdC40wWswd/P6IGgMaK3YpKi5j83Wpe3EHw8=
github.com/golang/protobuf v1.4.0-rc.1.0.20200221234624-67d41d38c208/go.mod h1:xKAWHe0F5eneWXFV3EuXVDTCmh+JuBKY0li0aMyXATA=
github.com/golang/protobuf v1.4.0-rc.2/go.mod h1:LlEzMj4AhA7rCAGe4KMBDvJI+AwstrUpVNzEA03Pprs=
github.com/golang/protobuf v1.4.0-rc.4.0.20200313231945-b860323f09d0/go.mod h1:WU3c8KckQ9AFe+yFwt9sWVRKCVIyN9cPHBJSNnbL67w=
github.com/golang/protobuf v1.4.0/go.mod h1:jodUvKwWbYaEsadDk5Fwe5c77LiNKVO9IDvqG2KuDX0=
github.com/golang/protobuf v1.4.1/go.mod h1:U8fpvMrcmy5pZrNK1lt4xCsGvpyWQ/VVv6QDs8UjoX8=
github.com/golang/protobuf v1.4.2 h1:+Z5KGCizgyZCbGh1KZqA0fcLLkwbsjIzS4aV2v7wJX0=
github.com/golang/protobuf v1.4.2/go.mod h1:oDoupMAO8OvCJWAcko0GGGIgR6R6ocIYbsSw735rRwI=
github.com/golang/snappy v0.0.1 h1:Qgr9rKW7uDUkrbSmQeiDsGa8SjGyCOGtuasMWwvp2P4=
github.com/golang/protobuf v1.5.0/go.mod h1:FsONVRAS9T7sI+LIUmWTfcYkHO4aIWwzhcaSAoJOfIk=
github.com/golang/protobuf v1.5.2 h1:ROPKBNFfQgOUMifHyP+KYbvpjbdoFNs+aK7DXlji0Tw=
github.com/golang/protobuf v1.5.2/go.mod h1:XVQd3VNwM+JqD3oG2Ue2ip4fOMUkwXdXDdiuN0vRsmY=
github.com/golang/snappy v0.0.1/go.mod h1:/XxbfmMg8lxefKM7IXC3fBNl/7bRcc72aCRzEWrmP2Q=
github.com/golang/snappy v0.0.2 h1:aeE13tS0IiQgFjYdoL8qN3K1N2bXXtI6Vi51/y7BpMw=
github.com/golang/snappy v0.0.2/go.mod h1:/XxbfmMg8lxefKM7IXC3fBNl/7bRcc72aCRzEWrmP2Q=
github.com/google/go-cmp v0.2.0/go.mod h1:oXzfMopK8JAjlY9xF4vHSVASa0yLyX7SntLO5aqRK0M=
github.com/google/go-cmp v0.3.0/go.mod h1:8QqcDgzrUqlUb/G2PQTWiueGozuR1884gddMywk6iLU=
github.com/google/go-cmp v0.3.1/go.mod h1:8QqcDgzrUqlUb/G2PQTWiueGozuR1884gddMywk6iLU=
github.com/google/go-cmp v0.4.0/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
github.com/google/go-cmp v0.5.0 h1:/QaMHBdZ26BB3SSst0Iwl10Epc+xhTquomWX0oZEB6w=
github.com/google/go-cmp v0.5.0/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
github.com/golang/snappy v0.0.3 h1:fHPg5GQYlCeLIPB9BZqMVR5nR9A+IM5zcgeTdjMYmLA=
github.com/golang/snappy v0.0.3/go.mod h1:/XxbfmMg8lxefKM7IXC3fBNl/7bRcc72aCRzEWrmP2Q=
github.com/google/go-cmp v0.5.5 h1:Khx7svrCpmxxtHBq5j2mp/xVjsi8hQMfNLvJFAlrGgU=
github.com/google/go-cmp v0.5.5/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
github.com/hashicorp/hcl v1.0.0/go.mod h1:E5yfLk+7swimpb2L/Alb/PJmXilQ/rhwaUYs4T20WEQ=
github.com/inconshreveable/mousetrap v1.0.0/go.mod h1:PxqpIevigyE2G7u3NXJIT2ANytuPF1OarO4DADm73n8=
github.com/kisielk/gotool v1.0.0/go.mod h1:XhKaO+MFFWcvkIS/tQcRk01m1F5IRFswLeQ+oQHNcck=
github.com/kr/pretty v0.1.0/go.mod h1:dAy3ld7l9f0ibDNOQOHHMYYIIbhfbHSm3C4ZsoJORNo=
github.com/kr/pretty v0.2.1 h1:Fmg33tUaq4/8ym9TJN1x7sLJnHVwhP33CNkpYV/7rwI=
github.com/kr/pretty v0.2.1/go.mod h1:ipq/a2n7PKx3OHsz4KJII5eveXtPO4qwEXGdVfWzfnI=
@@ -101,66 +84,54 @@ github.com/stretchr/testify v1.4.0 h1:2E4SXV/wtOkTonXsotYi4li6zVWxYlZuYNCXe9XRJy
github.com/stretchr/testify v1.4.0/go.mod h1:j7eGeouHqKxXV5pUuKE4zz7dFj8WfuZ+81PSLYec5m4=
github.com/ugorji/go/codec v0.0.0-20181204163529-d75b2dcb6bc8/go.mod h1:VFNgLljTbGfSG7qAOspJ7OScBnGdDN/yBr0sguwnwf0=
github.com/xordataexchange/crypt v0.0.3-0.20170626215501-b2862e3d0a77/go.mod h1:aYKd//L2LvnjZzWKhF00oedf4jCCReLcmhLdhm1A27Q=
github.com/yuin/goldmark v1.2.1/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
golang.org/x/crypto v0.0.0-20181203042331-505ab145d0a9/go.mod h1:6SG95UA2DQfeDnfUPMdvaQW0Q7yPrPDi9nlGo2tz2b4=
golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
golang.org/x/crypto v0.0.0-20191011191535-87dc89f01550/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
golang.org/x/crypto v0.0.0-20200622213623-75b288015ac9/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
golang.org/x/exp v0.0.0-20190121172915-509febef88a4/go.mod h1:CJ0aWSM057203Lf6IL+f9T1iT9GByDxfZKAQTCR3kQA=
golang.org/x/lint v0.0.0-20181026193005-c67002cb31c3/go.mod h1:UVdnD1Gm6xHRNCYTkRU2/jEulfH38KcIWyp/GAMgvoE=
golang.org/x/lint v0.0.0-20190227174305-5b3e6a55c961/go.mod h1:wehouNa3lNwaWXcvxsM5YxQ5yQlVC4a0KAMCusXpPoU=
golang.org/x/lint v0.0.0-20190313153728-d0100b6bd8b3/go.mod h1:6SW0HCj/g11FgYtHlgUYUwCkIfeOF89ocIRzGO/8vkc=
golang.org/x/net v0.0.0-20180724234803-3673e40ba225/go.mod h1:mL1N/T3taQHkDXs73rZJwtUhF3w3ftmwwsq0BUmARs4=
golang.org/x/net v0.0.0-20180826012351-8a410e7b638d/go.mod h1:mL1N/T3taQHkDXs73rZJwtUhF3w3ftmwwsq0BUmARs4=
golang.org/x/net v0.0.0-20190213061140-3a22650c66bd/go.mod h1:mL1N/T3taQHkDXs73rZJwtUhF3w3ftmwwsq0BUmARs4=
golang.org/x/net v0.0.0-20190311183353-d8887717615a/go.mod h1:t9HGtf8HONx5eT2rtn7q6eTqICYqUVnKs3thJo3Qplg=
golang.org/x/lint v0.0.0-20201208152925-83fdc39ff7b5 h1:2M3HP5CCK1Si9FQhwnzYhXdG6DXeebvUHFpre8QvbyI=
golang.org/x/lint v0.0.0-20201208152925-83fdc39ff7b5/go.mod h1:3xt1FjdF8hUf6vQPIChWIBhFzV8gjjsPE/fR3IyQdNY=
golang.org/x/mod v0.1.1-0.20191105210325-c90efee705ee/go.mod h1:QqPTAvyqsEbceGzBzNggFXnrqF1CaUcvgkdR5Ot7KZg=
golang.org/x/mod v0.3.0 h1:RM4zey1++hCTbCVQfnWeKs9/IEsaBLA8vTkd0WVtmH4=
golang.org/x/mod v0.3.0/go.mod h1:s0Qsj1ACt9ePp/hMypM3fl4fZqREWJwdYDEqhRiZZUA=
golang.org/x/net v0.0.0-20190404232315-eb5bcb51f2a3/go.mod h1:t9HGtf8HONx5eT2rtn7q6eTqICYqUVnKs3thJo3Qplg=
golang.org/x/net v0.0.0-20190620200207-3b0461eec859/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=
golang.org/x/net v0.0.0-20200927032502-5d4f70055728 h1:5wtQIAulKU5AbLQOkjxl32UufnIOqgBX72pS0AV14H0=
golang.org/x/net v0.0.0-20200927032502-5d4f70055728/go.mod h1:/O7V0waA8r7cgGh81Ro3o1hOxt32SMVPicZroKQ2sZA=
golang.org/x/oauth2 v0.0.0-20180821212333-d2e6202438be/go.mod h1:N/0e6XlmueqKjAGxoOufVs8QHGRruUQn6yWY3a++T0U=
golang.org/x/sync v0.0.0-20180314180146-1d60e4601c6f/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
golang.org/x/net v0.0.0-20201021035429-f5854403a974/go.mod h1:sp8m0HH+o8qH0wwXwYZr8TS3Oi6o0r6Gce1SSxlDquU=
golang.org/x/net v0.0.0-20210331212208-0fccb6fa2b5c h1:KHUzaHIpjWVlVVNh65G3hhuj3KB1HnjY6Cq5cTvRQT8=
golang.org/x/net v0.0.0-20210331212208-0fccb6fa2b5c/go.mod h1:p54w0d4576C0XHj96bSt6lcn1PtDYWL6XObtHCRCNQM=
golang.org/x/sync v0.0.0-20181108010431-42b317875d0f/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
golang.org/x/sync v0.0.0-20190423024810-112230192c58/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
golang.org/x/sys v0.0.0-20180830151530-49385e6e1522/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
golang.org/x/sync v0.0.0-20201020160332-67f06af15bc9/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
golang.org/x/sys v0.0.0-20181205085412-a5c9d58dba9a/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
golang.org/x/sys v0.0.0-20190412213103-97732733099d/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
golang.org/x/sys v0.0.0-20190626221950-04f50cda93cb h1:fgwFCsaw9buMuxNd6+DQfAuSFqbNiQZpcgJQAgJsK6k=
golang.org/x/sys v0.0.0-20190626221950-04f50cda93cb/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
golang.org/x/sys v0.0.0-20200323222414-85ca7c5b95cd/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
golang.org/x/sys v0.0.0-20200926100807-9d91bd62050c h1:38q6VNPWR010vN82/SB121GujZNIfAUb4YttE2rhGuc=
golang.org/x/sys v0.0.0-20200926100807-9d91bd62050c/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
golang.org/x/sys v0.0.0-20200930185726-fdedc70b468f/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
golang.org/x/sys v0.0.0-20201119102817-f84b799fce68/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
golang.org/x/sys v0.0.0-20210119212857-b64e53b001e4/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
golang.org/x/sys v0.0.0-20210330210617-4fbd30eecc44/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
golang.org/x/sys v0.0.0-20210402192133-700132347e07 h1:4k6HsQjxj6hVMsI2Vf0yKlzt5lXxZsMW1q0zaq2k8zY=
golang.org/x/sys v0.0.0-20210402192133-700132347e07/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
golang.org/x/term v0.0.0-20201126162022-7de9c90e9dd1/go.mod h1:bj7SfCRtBDWHUb9snDiAeCFNEtKQo2Wmx5Cou7ajbmo=
golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
golang.org/x/tools v0.0.0-20190114222345-bf090417da8b/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ=
golang.org/x/tools v0.0.0-20190226205152-f727befe758c/go.mod h1:9Yl7xja0Znq3iFh3HoIrodX9oNMXvdceNzlUR8zjMvY=
golang.org/x/tools v0.0.0-20190311212946-11955173bddd/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
golang.org/x/tools v0.0.0-20190524140312-2c0ae7006135/go.mod h1:RgjU9mgBXZiqYHBnxXauZ1Gv1EHHAz9KjViQ78xBX0Q=
golang.org/x/tools v0.0.0-20190628034336-212fb13d595e h1:ZlQjfVdpDxeqxRfmO30CdqWWzTvgRCj0MxaUVfxEG1k=
golang.org/x/tools v0.0.0-20190628034336-212fb13d595e/go.mod h1:/rFqwRUd4F7ZHNgwSSTFct+R/Kf4OFW1sUzUTQQTgfc=
golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543 h1:E7g+9GITq07hpfrRu66IVDexMakfv52eLZ2CXBWiKr4=
golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ=
golang.org/x/tools v0.0.0-20191119224855-298f0cb1881e/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo=
golang.org/x/tools v0.0.0-20200130002326-2f3ba24bd6e7/go.mod h1:TB2adYChydJhpapKDTa4BR/hXlZSLoq2Wpct/0txZ28=
golang.org/x/tools v0.1.0 h1:po9/4sTYwZU9lPhi1tOrb4hCv3qrhiQ77LZfGa2OjwY=
golang.org/x/tools v0.1.0/go.mod h1:xkSsbof2nBLbhDlRMhhhyNLN/zl3eTqcnHD5viDpcZ0=
golang.org/x/xerrors v0.0.0-20190717185122-a985d3407aa7/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
golang.org/x/xerrors v0.0.0-20191011141410-1b5146add898/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
golang.org/x/xerrors v0.0.0-20191204190536-9bdfabe68543/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
google.golang.org/appengine v1.1.0/go.mod h1:EbEs0AVv82hx2wNQdGPgUI5lhzA/G0D9YwlJXL52JkM=
google.golang.org/appengine v1.4.0/go.mod h1:xpcJRLb0r/rnEns0DIKYYv+WjYCduHsrkT7/EB5XEv4=
google.golang.org/genproto v0.0.0-20180817151627-c66870c02cf8/go.mod h1:JiN7NxoALGmiZfu7CAH4rXhgtRTLTxftemlI0sWmxmc=
google.golang.org/genproto v0.0.0-20190819201941-24fa4b261c55/go.mod h1:DMBHOl98Agz4BDEuKkezgsaosCRResVns1a3J2ZsMNc=
google.golang.org/genproto v0.0.0-20200526211855-cb27e3aa2013/go.mod h1:NbSheEEYHJ7i3ixzK3sjbqSGDJWnxyFXZblF3eUsNvo=
google.golang.org/grpc v1.19.0/go.mod h1:mqu4LbDTu4XGKhr4mRzUsmM4RtVoemTSY81AxZiDr8c=
google.golang.org/grpc v1.23.0/go.mod h1:Y5yQAOtifL1yxbo5wqy6BxZv8vAUGQwXBOALyacEbxg=
google.golang.org/grpc v1.27.0/go.mod h1:qbnxyOmOxrQa7FizSgH+ReBfzJrCY1pSN7KXBS8abTk=
google.golang.org/protobuf v0.0.0-20200109180630-ec00e32a8dfd/go.mod h1:DFci5gLYBciE7Vtevhsrf46CRTquxDuWsQurQQe4oz8=
google.golang.org/protobuf v0.0.0-20200221191635-4d8936d0db64/go.mod h1:kwYJMbMJ01Woi6D6+Kah6886xMZcty6N08ah7+eCXa0=
google.golang.org/protobuf v0.0.0-20200228230310-ab0ca4ff8a60/go.mod h1:cfTl7dwQJ+fmap5saPgwCLgHXTUD7jkjRqWcaiX5VyM=
google.golang.org/protobuf v1.20.1-0.20200309200217-e05f789c0967/go.mod h1:A+miEFZTKqfCUM6K7xSMQL9OKL/b6hQv+e19PK+JZNE=
google.golang.org/protobuf v1.21.0/go.mod h1:47Nbq4nVaFHyn7ilMalzfO3qCViNmqZ2kzikPIcrTAo=
google.golang.org/protobuf v1.22.0/go.mod h1:EGpADcykh3NcUnDUJcl1+ZksZNG86OlYog2l/sGQquU=
google.golang.org/protobuf v1.23.0 h1:4MY060fB1DLGMB/7MBTLnwQUY6+F09GEiz6SsrNqyzM=
google.golang.org/protobuf v1.23.0/go.mod h1:EGpADcykh3NcUnDUJcl1+ZksZNG86OlYog2l/sGQquU=
google.golang.org/protobuf v1.23.1-0.20200526195155-81db48ad09cc/go.mod h1:EGpADcykh3NcUnDUJcl1+ZksZNG86OlYog2l/sGQquU=
google.golang.org/protobuf v1.25.0 h1:Ejskq+SyPohKW+1uil0JJMtmHCgJPJ/qWTxr8qp+R4c=
google.golang.org/protobuf v1.25.0/go.mod h1:9JNX74DMeImyA3h4bdi1ymwjUzf21/xIlbajtzgsN7c=
golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 h1:go1bK/D/BFZV2I8cIQd1NKEZ+0owSTG1fDTci4IqFcE=
golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1/go.mod h1:I/5z698sn9Ka8TeJc9MKroUUfqBBauWjQqLJ2OPfmY0=
google.golang.org/protobuf v1.26.0-rc.1/go.mod h1:jlhhOSvTdKEhbULTjvd4ARK9grFBp09yW+WbY/TyQbw=
google.golang.org/protobuf v1.26.0 h1:bxAC2xTBsZGibn2RTntX0oH50xLsqy1OxA9tTL3p/lk=
google.golang.org/protobuf v1.26.0/go.mod h1:9q0QmTI4eRPtz6boOQmLYwt+qCgq0jsYwAQnmE0givc=
gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
gopkg.in/check.v1 v1.0.0-20190902080502-41f04d3bba15 h1:YR8cESwS4TdDjEe65xsg0ogRM/Nc3DYOhEAlW+xobZo=
gopkg.in/check.v1 v1.0.0-20190902080502-41f04d3bba15/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
gopkg.in/yaml.v2 v2.2.2 h1:ZCJp+EgiOT7lHqUV2J862kp8Qj64Jo6az82+3Td9dZw=
gopkg.in/yaml.v2 v2.2.2/go.mod h1:hI93XBmqTisBFMUTm0b8Fm+jr3Dg1NNxqwp+5A1VGuI=
honnef.co/go/tools v0.0.0-20190102054323-c2f93a96b099/go.mod h1:rf3lG4BRIbNafJWhAfAdb/ePZxsR/4RtNHQocxwk9r4=
honnef.co/go/tools v0.0.0-20190523083050-ea95bdfd59fc/go.mod h1:rf3lG4BRIbNafJWhAfAdb/ePZxsR/4RtNHQocxwk9r4=
honnef.co/go/tools v0.1.3 h1:qTakTkI6ni6LFD5sBwwsdSO+AQqbSIxOauHTTQKZ/7o=
honnef.co/go/tools v0.1.3/go.mod h1:NgwopIslSNH47DimFoV78dnkksY2EFtX0ajyb3K/las=

11
scripts/gofmt.sh Executable file
View File

@@ -0,0 +1,11 @@
#!/bin/bash
need_gofmt=$(gofmt -s -l `find . -name '*.go' | grep -v vendor | grep -v _repos`)
if [[ -n ${need_gofmt} ]]; then
echo "These files fail gofmt -s:"
echo "${need_gofmt}"
exit 1
fi

9
tools/tools.go Normal file
View File

@@ -0,0 +1,9 @@
// +build tools
package tools
import (
_ "github.com/client9/misspell/cmd/misspell"
_ "golang.org/x/lint/golint"
_ "honnef.co/go/tools/cmd/staticcheck"
)

5
vendor/github.com/BurntSushi/toml/.gitignore generated vendored Normal file
View File

@@ -0,0 +1,5 @@
TAGS
tags
.*.swp
tomlcheck/tomlcheck
toml.test

15
vendor/github.com/BurntSushi/toml/.travis.yml generated vendored Normal file
View File

@@ -0,0 +1,15 @@
language: go
go:
- 1.1
- 1.2
- 1.3
- 1.4
- 1.5
- 1.6
- tip
install:
- go install ./...
- go get github.com/BurntSushi/toml-test
script:
- export PATH="$PATH:$HOME/gopath/bin"
- make test

3
vendor/github.com/BurntSushi/toml/COMPATIBLE generated vendored Normal file
View File

@@ -0,0 +1,3 @@
Compatible with TOML version
[v0.4.0](https://github.com/toml-lang/toml/blob/v0.4.0/versions/en/toml-v0.4.0.md)

21
vendor/github.com/BurntSushi/toml/COPYING generated vendored Normal file
View File

@@ -0,0 +1,21 @@
The MIT License (MIT)
Copyright (c) 2013 TOML authors
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.

19
vendor/github.com/BurntSushi/toml/Makefile generated vendored Normal file
View File

@@ -0,0 +1,19 @@
install:
go install ./...
test: install
go test -v
toml-test toml-test-decoder
toml-test -encoder toml-test-encoder
fmt:
gofmt -w *.go */*.go
colcheck *.go */*.go
tags:
find ./ -name '*.go' -print0 | xargs -0 gotags > TAGS
push:
git push origin master
git push github master

218
vendor/github.com/BurntSushi/toml/README.md generated vendored Normal file
View File

@@ -0,0 +1,218 @@
## TOML parser and encoder for Go with reflection
TOML stands for Tom's Obvious, Minimal Language. This Go package provides a
reflection interface similar to Go's standard library `json` and `xml`
packages. This package also supports the `encoding.TextUnmarshaler` and
`encoding.TextMarshaler` interfaces so that you can define custom data
representations. (There is an example of this below.)
Spec: https://github.com/toml-lang/toml
Compatible with TOML version
[v0.4.0](https://github.com/toml-lang/toml/blob/master/versions/en/toml-v0.4.0.md)
Documentation: https://godoc.org/github.com/BurntSushi/toml
Installation:
```bash
go get github.com/BurntSushi/toml
```
Try the toml validator:
```bash
go get github.com/BurntSushi/toml/cmd/tomlv
tomlv some-toml-file.toml
```
[![Build Status](https://travis-ci.org/BurntSushi/toml.svg?branch=master)](https://travis-ci.org/BurntSushi/toml) [![GoDoc](https://godoc.org/github.com/BurntSushi/toml?status.svg)](https://godoc.org/github.com/BurntSushi/toml)
### Testing
This package passes all tests in
[toml-test](https://github.com/BurntSushi/toml-test) for both the decoder
and the encoder.
### Examples
This package works similarly to how the Go standard library handles `XML`
and `JSON`. Namely, data is loaded into Go values via reflection.
For the simplest example, consider some TOML file as just a list of keys
and values:
```toml
Age = 25
Cats = [ "Cauchy", "Plato" ]
Pi = 3.14
Perfection = [ 6, 28, 496, 8128 ]
DOB = 1987-07-05T05:45:00Z
```
Which could be defined in Go as:
```go
type Config struct {
Age int
Cats []string
Pi float64
Perfection []int
DOB time.Time // requires `import time`
}
```
And then decoded with:
```go
var conf Config
if _, err := toml.Decode(tomlData, &conf); err != nil {
// handle error
}
```
You can also use struct tags if your struct field name doesn't map to a TOML
key value directly:
```toml
some_key_NAME = "wat"
```
```go
type TOML struct {
ObscureKey string `toml:"some_key_NAME"`
}
```
### Using the `encoding.TextUnmarshaler` interface
Here's an example that automatically parses duration strings into
`time.Duration` values:
```toml
[[song]]
name = "Thunder Road"
duration = "4m49s"
[[song]]
name = "Stairway to Heaven"
duration = "8m03s"
```
Which can be decoded with:
```go
type song struct {
Name string
Duration duration
}
type songs struct {
Song []song
}
var favorites songs
if _, err := toml.Decode(blob, &favorites); err != nil {
log.Fatal(err)
}
for _, s := range favorites.Song {
fmt.Printf("%s (%s)\n", s.Name, s.Duration)
}
```
And you'll also need a `duration` type that satisfies the
`encoding.TextUnmarshaler` interface:
```go
type duration struct {
time.Duration
}
func (d *duration) UnmarshalText(text []byte) error {
var err error
d.Duration, err = time.ParseDuration(string(text))
return err
}
```
### More complex usage
Here's an example of how to load the example from the official spec page:
```toml
# This is a TOML document. Boom.
title = "TOML Example"
[owner]
name = "Tom Preston-Werner"
organization = "GitHub"
bio = "GitHub Cofounder & CEO\nLikes tater tots and beer."
dob = 1979-05-27T07:32:00Z # First class dates? Why not?
[database]
server = "192.168.1.1"
ports = [ 8001, 8001, 8002 ]
connection_max = 5000
enabled = true
[servers]
# You can indent as you please. Tabs or spaces. TOML don't care.
[servers.alpha]
ip = "10.0.0.1"
dc = "eqdc10"
[servers.beta]
ip = "10.0.0.2"
dc = "eqdc10"
[clients]
data = [ ["gamma", "delta"], [1, 2] ] # just an update to make sure parsers support it
# Line breaks are OK when inside arrays
hosts = [
"alpha",
"omega"
]
```
And the corresponding Go types are:
```go
type tomlConfig struct {
Title string
Owner ownerInfo
DB database `toml:"database"`
Servers map[string]server
Clients clients
}
type ownerInfo struct {
Name string
Org string `toml:"organization"`
Bio string
DOB time.Time
}
type database struct {
Server string
Ports []int
ConnMax int `toml:"connection_max"`
Enabled bool
}
type server struct {
IP string
DC string
}
type clients struct {
Data [][]interface{}
Hosts []string
}
```
Note that a case insensitive match will be tried if an exact match can't be
found.
A working example of the above can be found in `_examples/example.{go,toml}`.

509
vendor/github.com/BurntSushi/toml/decode.go generated vendored Normal file
View File

@@ -0,0 +1,509 @@
package toml
import (
"fmt"
"io"
"io/ioutil"
"math"
"reflect"
"strings"
"time"
)
func e(format string, args ...interface{}) error {
return fmt.Errorf("toml: "+format, args...)
}
// Unmarshaler is the interface implemented by objects that can unmarshal a
// TOML description of themselves.
type Unmarshaler interface {
UnmarshalTOML(interface{}) error
}
// Unmarshal decodes the contents of `p` in TOML format into a pointer `v`.
func Unmarshal(p []byte, v interface{}) error {
_, err := Decode(string(p), v)
return err
}
// Primitive is a TOML value that hasn't been decoded into a Go value.
// When using the various `Decode*` functions, the type `Primitive` may
// be given to any value, and its decoding will be delayed.
//
// A `Primitive` value can be decoded using the `PrimitiveDecode` function.
//
// The underlying representation of a `Primitive` value is subject to change.
// Do not rely on it.
//
// N.B. Primitive values are still parsed, so using them will only avoid
// the overhead of reflection. They can be useful when you don't know the
// exact type of TOML data until run time.
type Primitive struct {
undecoded interface{}
context Key
}
// DEPRECATED!
//
// Use MetaData.PrimitiveDecode instead.
func PrimitiveDecode(primValue Primitive, v interface{}) error {
md := MetaData{decoded: make(map[string]bool)}
return md.unify(primValue.undecoded, rvalue(v))
}
// PrimitiveDecode is just like the other `Decode*` functions, except it
// decodes a TOML value that has already been parsed. Valid primitive values
// can *only* be obtained from values filled by the decoder functions,
// including this method. (i.e., `v` may contain more `Primitive`
// values.)
//
// Meta data for primitive values is included in the meta data returned by
// the `Decode*` functions with one exception: keys returned by the Undecoded
// method will only reflect keys that were decoded. Namely, any keys hidden
// behind a Primitive will be considered undecoded. Executing this method will
// update the undecoded keys in the meta data. (See the example.)
func (md *MetaData) PrimitiveDecode(primValue Primitive, v interface{}) error {
md.context = primValue.context
defer func() { md.context = nil }()
return md.unify(primValue.undecoded, rvalue(v))
}
// Decode will decode the contents of `data` in TOML format into a pointer
// `v`.
//
// TOML hashes correspond to Go structs or maps. (Dealer's choice. They can be
// used interchangeably.)
//
// TOML arrays of tables correspond to either a slice of structs or a slice
// of maps.
//
// TOML datetimes correspond to Go `time.Time` values.
//
// All other TOML types (float, string, int, bool and array) correspond
// to the obvious Go types.
//
// An exception to the above rules is if a type implements the
// encoding.TextUnmarshaler interface. In this case, any primitive TOML value
// (floats, strings, integers, booleans and datetimes) will be converted to
// a byte string and given to the value's UnmarshalText method. See the
// Unmarshaler example for a demonstration with time duration strings.
//
// Key mapping
//
// TOML keys can map to either keys in a Go map or field names in a Go
// struct. The special `toml` struct tag may be used to map TOML keys to
// struct fields that don't match the key name exactly. (See the example.)
// A case insensitive match to struct names will be tried if an exact match
// can't be found.
//
// The mapping between TOML values and Go values is loose. That is, there
// may exist TOML values that cannot be placed into your representation, and
// there may be parts of your representation that do not correspond to
// TOML values. This loose mapping can be made stricter by using the IsDefined
// and/or Undecoded methods on the MetaData returned.
//
// This decoder will not handle cyclic types. If a cyclic type is passed,
// `Decode` will not terminate.
func Decode(data string, v interface{}) (MetaData, error) {
rv := reflect.ValueOf(v)
if rv.Kind() != reflect.Ptr {
return MetaData{}, e("Decode of non-pointer %s", reflect.TypeOf(v))
}
if rv.IsNil() {
return MetaData{}, e("Decode of nil %s", reflect.TypeOf(v))
}
p, err := parse(data)
if err != nil {
return MetaData{}, err
}
md := MetaData{
p.mapping, p.types, p.ordered,
make(map[string]bool, len(p.ordered)), nil,
}
return md, md.unify(p.mapping, indirect(rv))
}
// DecodeFile is just like Decode, except it will automatically read the
// contents of the file at `fpath` and decode it for you.
func DecodeFile(fpath string, v interface{}) (MetaData, error) {
bs, err := ioutil.ReadFile(fpath)
if err != nil {
return MetaData{}, err
}
return Decode(string(bs), v)
}
// DecodeReader is just like Decode, except it will consume all bytes
// from the reader and decode it for you.
func DecodeReader(r io.Reader, v interface{}) (MetaData, error) {
bs, err := ioutil.ReadAll(r)
if err != nil {
return MetaData{}, err
}
return Decode(string(bs), v)
}
// unify performs a sort of type unification based on the structure of `rv`,
// which is the client representation.
//
// Any type mismatch produces an error. Finding a type that we don't know
// how to handle produces an unsupported type error.
func (md *MetaData) unify(data interface{}, rv reflect.Value) error {
// Special case. Look for a `Primitive` value.
if rv.Type() == reflect.TypeOf((*Primitive)(nil)).Elem() {
// Save the undecoded data and the key context into the primitive
// value.
context := make(Key, len(md.context))
copy(context, md.context)
rv.Set(reflect.ValueOf(Primitive{
undecoded: data,
context: context,
}))
return nil
}
// Special case. Unmarshaler Interface support.
if rv.CanAddr() {
if v, ok := rv.Addr().Interface().(Unmarshaler); ok {
return v.UnmarshalTOML(data)
}
}
// Special case. Handle time.Time values specifically.
// TODO: Remove this code when we decide to drop support for Go 1.1.
// This isn't necessary in Go 1.2 because time.Time satisfies the encoding
// interfaces.
if rv.Type().AssignableTo(rvalue(time.Time{}).Type()) {
return md.unifyDatetime(data, rv)
}
// Special case. Look for a value satisfying the TextUnmarshaler interface.
if v, ok := rv.Interface().(TextUnmarshaler); ok {
return md.unifyText(data, v)
}
// BUG(burntsushi)
// The behavior here is incorrect whenever a Go type satisfies the
// encoding.TextUnmarshaler interface but also corresponds to a TOML
// hash or array. In particular, the unmarshaler should only be applied
// to primitive TOML values. But at this point, it will be applied to
// all kinds of values and produce an incorrect error whenever those values
// are hashes or arrays (including arrays of tables).
k := rv.Kind()
// laziness
if k >= reflect.Int && k <= reflect.Uint64 {
return md.unifyInt(data, rv)
}
switch k {
case reflect.Ptr:
elem := reflect.New(rv.Type().Elem())
err := md.unify(data, reflect.Indirect(elem))
if err != nil {
return err
}
rv.Set(elem)
return nil
case reflect.Struct:
return md.unifyStruct(data, rv)
case reflect.Map:
return md.unifyMap(data, rv)
case reflect.Array:
return md.unifyArray(data, rv)
case reflect.Slice:
return md.unifySlice(data, rv)
case reflect.String:
return md.unifyString(data, rv)
case reflect.Bool:
return md.unifyBool(data, rv)
case reflect.Interface:
// we only support empty interfaces.
if rv.NumMethod() > 0 {
return e("unsupported type %s", rv.Type())
}
return md.unifyAnything(data, rv)
case reflect.Float32:
fallthrough
case reflect.Float64:
return md.unifyFloat64(data, rv)
}
return e("unsupported type %s", rv.Kind())
}
func (md *MetaData) unifyStruct(mapping interface{}, rv reflect.Value) error {
tmap, ok := mapping.(map[string]interface{})
if !ok {
if mapping == nil {
return nil
}
return e("type mismatch for %s: expected table but found %T",
rv.Type().String(), mapping)
}
for key, datum := range tmap {
var f *field
fields := cachedTypeFields(rv.Type())
for i := range fields {
ff := &fields[i]
if ff.name == key {
f = ff
break
}
if f == nil && strings.EqualFold(ff.name, key) {
f = ff
}
}
if f != nil {
subv := rv
for _, i := range f.index {
subv = indirect(subv.Field(i))
}
if isUnifiable(subv) {
md.decoded[md.context.add(key).String()] = true
md.context = append(md.context, key)
if err := md.unify(datum, subv); err != nil {
return err
}
md.context = md.context[0 : len(md.context)-1]
} else if f.name != "" {
// Bad user! No soup for you!
return e("cannot write unexported field %s.%s",
rv.Type().String(), f.name)
}
}
}
return nil
}
func (md *MetaData) unifyMap(mapping interface{}, rv reflect.Value) error {
tmap, ok := mapping.(map[string]interface{})
if !ok {
if tmap == nil {
return nil
}
return badtype("map", mapping)
}
if rv.IsNil() {
rv.Set(reflect.MakeMap(rv.Type()))
}
for k, v := range tmap {
md.decoded[md.context.add(k).String()] = true
md.context = append(md.context, k)
rvkey := indirect(reflect.New(rv.Type().Key()))
rvval := reflect.Indirect(reflect.New(rv.Type().Elem()))
if err := md.unify(v, rvval); err != nil {
return err
}
md.context = md.context[0 : len(md.context)-1]
rvkey.SetString(k)
rv.SetMapIndex(rvkey, rvval)
}
return nil
}
func (md *MetaData) unifyArray(data interface{}, rv reflect.Value) error {
datav := reflect.ValueOf(data)
if datav.Kind() != reflect.Slice {
if !datav.IsValid() {
return nil
}
return badtype("slice", data)
}
sliceLen := datav.Len()
if sliceLen != rv.Len() {
return e("expected array length %d; got TOML array of length %d",
rv.Len(), sliceLen)
}
return md.unifySliceArray(datav, rv)
}
func (md *MetaData) unifySlice(data interface{}, rv reflect.Value) error {
datav := reflect.ValueOf(data)
if datav.Kind() != reflect.Slice {
if !datav.IsValid() {
return nil
}
return badtype("slice", data)
}
n := datav.Len()
if rv.IsNil() || rv.Cap() < n {
rv.Set(reflect.MakeSlice(rv.Type(), n, n))
}
rv.SetLen(n)
return md.unifySliceArray(datav, rv)
}
func (md *MetaData) unifySliceArray(data, rv reflect.Value) error {
sliceLen := data.Len()
for i := 0; i < sliceLen; i++ {
v := data.Index(i).Interface()
sliceval := indirect(rv.Index(i))
if err := md.unify(v, sliceval); err != nil {
return err
}
}
return nil
}
func (md *MetaData) unifyDatetime(data interface{}, rv reflect.Value) error {
if _, ok := data.(time.Time); ok {
rv.Set(reflect.ValueOf(data))
return nil
}
return badtype("time.Time", data)
}
func (md *MetaData) unifyString(data interface{}, rv reflect.Value) error {
if s, ok := data.(string); ok {
rv.SetString(s)
return nil
}
return badtype("string", data)
}
func (md *MetaData) unifyFloat64(data interface{}, rv reflect.Value) error {
if num, ok := data.(float64); ok {
switch rv.Kind() {
case reflect.Float32:
fallthrough
case reflect.Float64:
rv.SetFloat(num)
default:
panic("bug")
}
return nil
}
return badtype("float", data)
}
func (md *MetaData) unifyInt(data interface{}, rv reflect.Value) error {
if num, ok := data.(int64); ok {
if rv.Kind() >= reflect.Int && rv.Kind() <= reflect.Int64 {
switch rv.Kind() {
case reflect.Int, reflect.Int64:
// No bounds checking necessary.
case reflect.Int8:
if num < math.MinInt8 || num > math.MaxInt8 {
return e("value %d is out of range for int8", num)
}
case reflect.Int16:
if num < math.MinInt16 || num > math.MaxInt16 {
return e("value %d is out of range for int16", num)
}
case reflect.Int32:
if num < math.MinInt32 || num > math.MaxInt32 {
return e("value %d is out of range for int32", num)
}
}
rv.SetInt(num)
} else if rv.Kind() >= reflect.Uint && rv.Kind() <= reflect.Uint64 {
unum := uint64(num)
switch rv.Kind() {
case reflect.Uint, reflect.Uint64:
// No bounds checking necessary.
case reflect.Uint8:
if num < 0 || unum > math.MaxUint8 {
return e("value %d is out of range for uint8", num)
}
case reflect.Uint16:
if num < 0 || unum > math.MaxUint16 {
return e("value %d is out of range for uint16", num)
}
case reflect.Uint32:
if num < 0 || unum > math.MaxUint32 {
return e("value %d is out of range for uint32", num)
}
}
rv.SetUint(unum)
} else {
panic("unreachable")
}
return nil
}
return badtype("integer", data)
}
func (md *MetaData) unifyBool(data interface{}, rv reflect.Value) error {
if b, ok := data.(bool); ok {
rv.SetBool(b)
return nil
}
return badtype("boolean", data)
}
func (md *MetaData) unifyAnything(data interface{}, rv reflect.Value) error {
rv.Set(reflect.ValueOf(data))
return nil
}
func (md *MetaData) unifyText(data interface{}, v TextUnmarshaler) error {
var s string
switch sdata := data.(type) {
case TextMarshaler:
text, err := sdata.MarshalText()
if err != nil {
return err
}
s = string(text)
case fmt.Stringer:
s = sdata.String()
case string:
s = sdata
case bool:
s = fmt.Sprintf("%v", sdata)
case int64:
s = fmt.Sprintf("%d", sdata)
case float64:
s = fmt.Sprintf("%f", sdata)
default:
return badtype("primitive (string-like)", data)
}
if err := v.UnmarshalText([]byte(s)); err != nil {
return err
}
return nil
}
// rvalue returns a reflect.Value of `v`. All pointers are resolved.
func rvalue(v interface{}) reflect.Value {
return indirect(reflect.ValueOf(v))
}
// indirect returns the value pointed to by a pointer.
// Pointers are followed until the value is not a pointer.
// New values are allocated for each nil pointer.
//
// An exception to this rule is if the value satisfies an interface of
// interest to us (like encoding.TextUnmarshaler).
func indirect(v reflect.Value) reflect.Value {
if v.Kind() != reflect.Ptr {
if v.CanSet() {
pv := v.Addr()
if _, ok := pv.Interface().(TextUnmarshaler); ok {
return pv
}
}
return v
}
if v.IsNil() {
v.Set(reflect.New(v.Type().Elem()))
}
return indirect(reflect.Indirect(v))
}
func isUnifiable(rv reflect.Value) bool {
if rv.CanSet() {
return true
}
if _, ok := rv.Interface().(TextUnmarshaler); ok {
return true
}
return false
}
func badtype(expected string, data interface{}) error {
return e("cannot load TOML value of type %T into a Go %s", data, expected)
}

121
vendor/github.com/BurntSushi/toml/decode_meta.go generated vendored Normal file
View File

@@ -0,0 +1,121 @@
package toml
import "strings"
// MetaData allows access to meta information about TOML data that may not
// be inferrable via reflection. In particular, whether a key has been defined
// and the TOML type of a key.
type MetaData struct {
mapping map[string]interface{}
types map[string]tomlType
keys []Key
decoded map[string]bool
context Key // Used only during decoding.
}
// IsDefined returns true if the key given exists in the TOML data. The key
// should be specified hierarchially. e.g.,
//
// // access the TOML key 'a.b.c'
// IsDefined("a", "b", "c")
//
// IsDefined will return false if an empty key given. Keys are case sensitive.
func (md *MetaData) IsDefined(key ...string) bool {
if len(key) == 0 {
return false
}
var hash map[string]interface{}
var ok bool
var hashOrVal interface{} = md.mapping
for _, k := range key {
if hash, ok = hashOrVal.(map[string]interface{}); !ok {
return false
}
if hashOrVal, ok = hash[k]; !ok {
return false
}
}
return true
}
// Type returns a string representation of the type of the key specified.
//
// Type will return the empty string if given an empty key or a key that
// does not exist. Keys are case sensitive.
func (md *MetaData) Type(key ...string) string {
fullkey := strings.Join(key, ".")
if typ, ok := md.types[fullkey]; ok {
return typ.typeString()
}
return ""
}
// Key is the type of any TOML key, including key groups. Use (MetaData).Keys
// to get values of this type.
type Key []string
func (k Key) String() string {
return strings.Join(k, ".")
}
func (k Key) maybeQuotedAll() string {
var ss []string
for i := range k {
ss = append(ss, k.maybeQuoted(i))
}
return strings.Join(ss, ".")
}
func (k Key) maybeQuoted(i int) string {
quote := false
for _, c := range k[i] {
if !isBareKeyChar(c) {
quote = true
break
}
}
if quote {
return "\"" + strings.Replace(k[i], "\"", "\\\"", -1) + "\""
}
return k[i]
}
func (k Key) add(piece string) Key {
newKey := make(Key, len(k)+1)
copy(newKey, k)
newKey[len(k)] = piece
return newKey
}
// Keys returns a slice of every key in the TOML data, including key groups.
// Each key is itself a slice, where the first element is the top of the
// hierarchy and the last is the most specific.
//
// The list will have the same order as the keys appeared in the TOML data.
//
// All keys returned are non-empty.
func (md *MetaData) Keys() []Key {
return md.keys
}
// Undecoded returns all keys that have not been decoded in the order in which
// they appear in the original TOML document.
//
// This includes keys that haven't been decoded because of a Primitive value.
// Once the Primitive value is decoded, the keys will be considered decoded.
//
// Also note that decoding into an empty interface will result in no decoding,
// and so no keys will be considered decoded.
//
// In this sense, the Undecoded keys correspond to keys in the TOML document
// that do not have a concrete type in your representation.
func (md *MetaData) Undecoded() []Key {
undecoded := make([]Key, 0, len(md.keys))
for _, key := range md.keys {
if !md.decoded[key.String()] {
undecoded = append(undecoded, key)
}
}
return undecoded
}

27
vendor/github.com/BurntSushi/toml/doc.go generated vendored Normal file
View File

@@ -0,0 +1,27 @@
/*
Package toml provides facilities for decoding and encoding TOML configuration
files via reflection. There is also support for delaying decoding with
the Primitive type, and querying the set of keys in a TOML document with the
MetaData type.
The specification implemented: https://github.com/toml-lang/toml
The sub-command github.com/BurntSushi/toml/cmd/tomlv can be used to verify
whether a file is a valid TOML document. It can also be used to print the
type of each key in a TOML document.
Testing
There are two important types of tests used for this package. The first is
contained inside '*_test.go' files and uses the standard Go unit testing
framework. These tests are primarily devoted to holistically testing the
decoder and encoder.
The second type of testing is used to verify the implementation's adherence
to the TOML specification. These tests have been factored into their own
project: https://github.com/BurntSushi/toml-test
The reason the tests are in a separate project is so that they can be used by
any implementation of TOML. Namely, it is language agnostic.
*/
package toml

568
vendor/github.com/BurntSushi/toml/encode.go generated vendored Normal file
View File

@@ -0,0 +1,568 @@
package toml
import (
"bufio"
"errors"
"fmt"
"io"
"reflect"
"sort"
"strconv"
"strings"
"time"
)
type tomlEncodeError struct{ error }
var (
errArrayMixedElementTypes = errors.New(
"toml: cannot encode array with mixed element types")
errArrayNilElement = errors.New(
"toml: cannot encode array with nil element")
errNonString = errors.New(
"toml: cannot encode a map with non-string key type")
errAnonNonStruct = errors.New(
"toml: cannot encode an anonymous field that is not a struct")
errArrayNoTable = errors.New(
"toml: TOML array element cannot contain a table")
errNoKey = errors.New(
"toml: top-level values must be Go maps or structs")
errAnything = errors.New("") // used in testing
)
var quotedReplacer = strings.NewReplacer(
"\t", "\\t",
"\n", "\\n",
"\r", "\\r",
"\"", "\\\"",
"\\", "\\\\",
)
// Encoder controls the encoding of Go values to a TOML document to some
// io.Writer.
//
// The indentation level can be controlled with the Indent field.
type Encoder struct {
// A single indentation level. By default it is two spaces.
Indent string
// hasWritten is whether we have written any output to w yet.
hasWritten bool
w *bufio.Writer
}
// NewEncoder returns a TOML encoder that encodes Go values to the io.Writer
// given. By default, a single indentation level is 2 spaces.
func NewEncoder(w io.Writer) *Encoder {
return &Encoder{
w: bufio.NewWriter(w),
Indent: " ",
}
}
// Encode writes a TOML representation of the Go value to the underlying
// io.Writer. If the value given cannot be encoded to a valid TOML document,
// then an error is returned.
//
// The mapping between Go values and TOML values should be precisely the same
// as for the Decode* functions. Similarly, the TextMarshaler interface is
// supported by encoding the resulting bytes as strings. (If you want to write
// arbitrary binary data then you will need to use something like base64 since
// TOML does not have any binary types.)
//
// When encoding TOML hashes (i.e., Go maps or structs), keys without any
// sub-hashes are encoded first.
//
// If a Go map is encoded, then its keys are sorted alphabetically for
// deterministic output. More control over this behavior may be provided if
// there is demand for it.
//
// Encoding Go values without a corresponding TOML representation---like map
// types with non-string keys---will cause an error to be returned. Similarly
// for mixed arrays/slices, arrays/slices with nil elements, embedded
// non-struct types and nested slices containing maps or structs.
// (e.g., [][]map[string]string is not allowed but []map[string]string is OK
// and so is []map[string][]string.)
func (enc *Encoder) Encode(v interface{}) error {
rv := eindirect(reflect.ValueOf(v))
if err := enc.safeEncode(Key([]string{}), rv); err != nil {
return err
}
return enc.w.Flush()
}
func (enc *Encoder) safeEncode(key Key, rv reflect.Value) (err error) {
defer func() {
if r := recover(); r != nil {
if terr, ok := r.(tomlEncodeError); ok {
err = terr.error
return
}
panic(r)
}
}()
enc.encode(key, rv)
return nil
}
func (enc *Encoder) encode(key Key, rv reflect.Value) {
// Special case. Time needs to be in ISO8601 format.
// Special case. If we can marshal the type to text, then we used that.
// Basically, this prevents the encoder for handling these types as
// generic structs (or whatever the underlying type of a TextMarshaler is).
switch rv.Interface().(type) {
case time.Time, TextMarshaler:
enc.keyEqElement(key, rv)
return
}
k := rv.Kind()
switch k {
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32,
reflect.Int64,
reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32,
reflect.Uint64,
reflect.Float32, reflect.Float64, reflect.String, reflect.Bool:
enc.keyEqElement(key, rv)
case reflect.Array, reflect.Slice:
if typeEqual(tomlArrayHash, tomlTypeOfGo(rv)) {
enc.eArrayOfTables(key, rv)
} else {
enc.keyEqElement(key, rv)
}
case reflect.Interface:
if rv.IsNil() {
return
}
enc.encode(key, rv.Elem())
case reflect.Map:
if rv.IsNil() {
return
}
enc.eTable(key, rv)
case reflect.Ptr:
if rv.IsNil() {
return
}
enc.encode(key, rv.Elem())
case reflect.Struct:
enc.eTable(key, rv)
default:
panic(e("unsupported type for key '%s': %s", key, k))
}
}
// eElement encodes any value that can be an array element (primitives and
// arrays).
func (enc *Encoder) eElement(rv reflect.Value) {
switch v := rv.Interface().(type) {
case time.Time:
// Special case time.Time as a primitive. Has to come before
// TextMarshaler below because time.Time implements
// encoding.TextMarshaler, but we need to always use UTC.
enc.wf(v.UTC().Format("2006-01-02T15:04:05Z"))
return
case TextMarshaler:
// Special case. Use text marshaler if it's available for this value.
if s, err := v.MarshalText(); err != nil {
encPanic(err)
} else {
enc.writeQuoted(string(s))
}
return
}
switch rv.Kind() {
case reflect.Bool:
enc.wf(strconv.FormatBool(rv.Bool()))
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32,
reflect.Int64:
enc.wf(strconv.FormatInt(rv.Int(), 10))
case reflect.Uint, reflect.Uint8, reflect.Uint16,
reflect.Uint32, reflect.Uint64:
enc.wf(strconv.FormatUint(rv.Uint(), 10))
case reflect.Float32:
enc.wf(floatAddDecimal(strconv.FormatFloat(rv.Float(), 'f', -1, 32)))
case reflect.Float64:
enc.wf(floatAddDecimal(strconv.FormatFloat(rv.Float(), 'f', -1, 64)))
case reflect.Array, reflect.Slice:
enc.eArrayOrSliceElement(rv)
case reflect.Interface:
enc.eElement(rv.Elem())
case reflect.String:
enc.writeQuoted(rv.String())
default:
panic(e("unexpected primitive type: %s", rv.Kind()))
}
}
// By the TOML spec, all floats must have a decimal with at least one
// number on either side.
func floatAddDecimal(fstr string) string {
if !strings.Contains(fstr, ".") {
return fstr + ".0"
}
return fstr
}
func (enc *Encoder) writeQuoted(s string) {
enc.wf("\"%s\"", quotedReplacer.Replace(s))
}
func (enc *Encoder) eArrayOrSliceElement(rv reflect.Value) {
length := rv.Len()
enc.wf("[")
for i := 0; i < length; i++ {
elem := rv.Index(i)
enc.eElement(elem)
if i != length-1 {
enc.wf(", ")
}
}
enc.wf("]")
}
func (enc *Encoder) eArrayOfTables(key Key, rv reflect.Value) {
if len(key) == 0 {
encPanic(errNoKey)
}
for i := 0; i < rv.Len(); i++ {
trv := rv.Index(i)
if isNil(trv) {
continue
}
panicIfInvalidKey(key)
enc.newline()
enc.wf("%s[[%s]]", enc.indentStr(key), key.maybeQuotedAll())
enc.newline()
enc.eMapOrStruct(key, trv)
}
}
func (enc *Encoder) eTable(key Key, rv reflect.Value) {
panicIfInvalidKey(key)
if len(key) == 1 {
// Output an extra newline between top-level tables.
// (The newline isn't written if nothing else has been written though.)
enc.newline()
}
if len(key) > 0 {
enc.wf("%s[%s]", enc.indentStr(key), key.maybeQuotedAll())
enc.newline()
}
enc.eMapOrStruct(key, rv)
}
func (enc *Encoder) eMapOrStruct(key Key, rv reflect.Value) {
switch rv := eindirect(rv); rv.Kind() {
case reflect.Map:
enc.eMap(key, rv)
case reflect.Struct:
enc.eStruct(key, rv)
default:
panic("eTable: unhandled reflect.Value Kind: " + rv.Kind().String())
}
}
func (enc *Encoder) eMap(key Key, rv reflect.Value) {
rt := rv.Type()
if rt.Key().Kind() != reflect.String {
encPanic(errNonString)
}
// Sort keys so that we have deterministic output. And write keys directly
// underneath this key first, before writing sub-structs or sub-maps.
var mapKeysDirect, mapKeysSub []string
for _, mapKey := range rv.MapKeys() {
k := mapKey.String()
if typeIsHash(tomlTypeOfGo(rv.MapIndex(mapKey))) {
mapKeysSub = append(mapKeysSub, k)
} else {
mapKeysDirect = append(mapKeysDirect, k)
}
}
var writeMapKeys = func(mapKeys []string) {
sort.Strings(mapKeys)
for _, mapKey := range mapKeys {
mrv := rv.MapIndex(reflect.ValueOf(mapKey))
if isNil(mrv) {
// Don't write anything for nil fields.
continue
}
enc.encode(key.add(mapKey), mrv)
}
}
writeMapKeys(mapKeysDirect)
writeMapKeys(mapKeysSub)
}
func (enc *Encoder) eStruct(key Key, rv reflect.Value) {
// Write keys for fields directly under this key first, because if we write
// a field that creates a new table, then all keys under it will be in that
// table (not the one we're writing here).
rt := rv.Type()
var fieldsDirect, fieldsSub [][]int
var addFields func(rt reflect.Type, rv reflect.Value, start []int)
addFields = func(rt reflect.Type, rv reflect.Value, start []int) {
for i := 0; i < rt.NumField(); i++ {
f := rt.Field(i)
// skip unexported fields
if f.PkgPath != "" && !f.Anonymous {
continue
}
frv := rv.Field(i)
if f.Anonymous {
t := f.Type
switch t.Kind() {
case reflect.Struct:
// Treat anonymous struct fields with
// tag names as though they are not
// anonymous, like encoding/json does.
if getOptions(f.Tag).name == "" {
addFields(t, frv, f.Index)
continue
}
case reflect.Ptr:
if t.Elem().Kind() == reflect.Struct &&
getOptions(f.Tag).name == "" {
if !frv.IsNil() {
addFields(t.Elem(), frv.Elem(), f.Index)
}
continue
}
// Fall through to the normal field encoding logic below
// for non-struct anonymous fields.
}
}
if typeIsHash(tomlTypeOfGo(frv)) {
fieldsSub = append(fieldsSub, append(start, f.Index...))
} else {
fieldsDirect = append(fieldsDirect, append(start, f.Index...))
}
}
}
addFields(rt, rv, nil)
var writeFields = func(fields [][]int) {
for _, fieldIndex := range fields {
sft := rt.FieldByIndex(fieldIndex)
sf := rv.FieldByIndex(fieldIndex)
if isNil(sf) {
// Don't write anything for nil fields.
continue
}
opts := getOptions(sft.Tag)
if opts.skip {
continue
}
keyName := sft.Name
if opts.name != "" {
keyName = opts.name
}
if opts.omitempty && isEmpty(sf) {
continue
}
if opts.omitzero && isZero(sf) {
continue
}
enc.encode(key.add(keyName), sf)
}
}
writeFields(fieldsDirect)
writeFields(fieldsSub)
}
// tomlTypeName returns the TOML type name of the Go value's type. It is
// used to determine whether the types of array elements are mixed (which is
// forbidden). If the Go value is nil, then it is illegal for it to be an array
// element, and valueIsNil is returned as true.
// Returns the TOML type of a Go value. The type may be `nil`, which means
// no concrete TOML type could be found.
func tomlTypeOfGo(rv reflect.Value) tomlType {
if isNil(rv) || !rv.IsValid() {
return nil
}
switch rv.Kind() {
case reflect.Bool:
return tomlBool
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32,
reflect.Int64,
reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32,
reflect.Uint64:
return tomlInteger
case reflect.Float32, reflect.Float64:
return tomlFloat
case reflect.Array, reflect.Slice:
if typeEqual(tomlHash, tomlArrayType(rv)) {
return tomlArrayHash
}
return tomlArray
case reflect.Ptr, reflect.Interface:
return tomlTypeOfGo(rv.Elem())
case reflect.String:
return tomlString
case reflect.Map:
return tomlHash
case reflect.Struct:
switch rv.Interface().(type) {
case time.Time:
return tomlDatetime
case TextMarshaler:
return tomlString
default:
return tomlHash
}
default:
panic("unexpected reflect.Kind: " + rv.Kind().String())
}
}
// tomlArrayType returns the element type of a TOML array. The type returned
// may be nil if it cannot be determined (e.g., a nil slice or a zero length
// slize). This function may also panic if it finds a type that cannot be
// expressed in TOML (such as nil elements, heterogeneous arrays or directly
// nested arrays of tables).
func tomlArrayType(rv reflect.Value) tomlType {
if isNil(rv) || !rv.IsValid() || rv.Len() == 0 {
return nil
}
firstType := tomlTypeOfGo(rv.Index(0))
if firstType == nil {
encPanic(errArrayNilElement)
}
rvlen := rv.Len()
for i := 1; i < rvlen; i++ {
elem := rv.Index(i)
switch elemType := tomlTypeOfGo(elem); {
case elemType == nil:
encPanic(errArrayNilElement)
case !typeEqual(firstType, elemType):
encPanic(errArrayMixedElementTypes)
}
}
// If we have a nested array, then we must make sure that the nested
// array contains ONLY primitives.
// This checks arbitrarily nested arrays.
if typeEqual(firstType, tomlArray) || typeEqual(firstType, tomlArrayHash) {
nest := tomlArrayType(eindirect(rv.Index(0)))
if typeEqual(nest, tomlHash) || typeEqual(nest, tomlArrayHash) {
encPanic(errArrayNoTable)
}
}
return firstType
}
type tagOptions struct {
skip bool // "-"
name string
omitempty bool
omitzero bool
}
func getOptions(tag reflect.StructTag) tagOptions {
t := tag.Get("toml")
if t == "-" {
return tagOptions{skip: true}
}
var opts tagOptions
parts := strings.Split(t, ",")
opts.name = parts[0]
for _, s := range parts[1:] {
switch s {
case "omitempty":
opts.omitempty = true
case "omitzero":
opts.omitzero = true
}
}
return opts
}
func isZero(rv reflect.Value) bool {
switch rv.Kind() {
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
return rv.Int() == 0
case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
return rv.Uint() == 0
case reflect.Float32, reflect.Float64:
return rv.Float() == 0.0
}
return false
}
func isEmpty(rv reflect.Value) bool {
switch rv.Kind() {
case reflect.Array, reflect.Slice, reflect.Map, reflect.String:
return rv.Len() == 0
case reflect.Bool:
return !rv.Bool()
}
return false
}
func (enc *Encoder) newline() {
if enc.hasWritten {
enc.wf("\n")
}
}
func (enc *Encoder) keyEqElement(key Key, val reflect.Value) {
if len(key) == 0 {
encPanic(errNoKey)
}
panicIfInvalidKey(key)
enc.wf("%s%s = ", enc.indentStr(key), key.maybeQuoted(len(key)-1))
enc.eElement(val)
enc.newline()
}
func (enc *Encoder) wf(format string, v ...interface{}) {
if _, err := fmt.Fprintf(enc.w, format, v...); err != nil {
encPanic(err)
}
enc.hasWritten = true
}
func (enc *Encoder) indentStr(key Key) string {
return strings.Repeat(enc.Indent, len(key)-1)
}
func encPanic(err error) {
panic(tomlEncodeError{err})
}
func eindirect(v reflect.Value) reflect.Value {
switch v.Kind() {
case reflect.Ptr, reflect.Interface:
return eindirect(v.Elem())
default:
return v
}
}
func isNil(rv reflect.Value) bool {
switch rv.Kind() {
case reflect.Interface, reflect.Map, reflect.Ptr, reflect.Slice:
return rv.IsNil()
default:
return false
}
}
func panicIfInvalidKey(key Key) {
for _, k := range key {
if len(k) == 0 {
encPanic(e("Key '%s' is not a valid table name. Key names "+
"cannot be empty.", key.maybeQuotedAll()))
}
}
}
func isValidKeyName(s string) bool {
return len(s) != 0
}

19
vendor/github.com/BurntSushi/toml/encoding_types.go generated vendored Normal file
View File

@@ -0,0 +1,19 @@
// +build go1.2
package toml
// In order to support Go 1.1, we define our own TextMarshaler and
// TextUnmarshaler types. For Go 1.2+, we just alias them with the
// standard library interfaces.
import (
"encoding"
)
// TextMarshaler is a synonym for encoding.TextMarshaler. It is defined here
// so that Go 1.1 can be supported.
type TextMarshaler encoding.TextMarshaler
// TextUnmarshaler is a synonym for encoding.TextUnmarshaler. It is defined
// here so that Go 1.1 can be supported.
type TextUnmarshaler encoding.TextUnmarshaler

View File

@@ -0,0 +1,18 @@
// +build !go1.2
package toml
// These interfaces were introduced in Go 1.2, so we add them manually when
// compiling for Go 1.1.
// TextMarshaler is a synonym for encoding.TextMarshaler. It is defined here
// so that Go 1.1 can be supported.
type TextMarshaler interface {
MarshalText() (text []byte, err error)
}
// TextUnmarshaler is a synonym for encoding.TextUnmarshaler. It is defined
// here so that Go 1.1 can be supported.
type TextUnmarshaler interface {
UnmarshalText(text []byte) error
}

953
vendor/github.com/BurntSushi/toml/lex.go generated vendored Normal file
View File

@@ -0,0 +1,953 @@
package toml
import (
"fmt"
"strings"
"unicode"
"unicode/utf8"
)
type itemType int
const (
itemError itemType = iota
itemNIL // used in the parser to indicate no type
itemEOF
itemText
itemString
itemRawString
itemMultilineString
itemRawMultilineString
itemBool
itemInteger
itemFloat
itemDatetime
itemArray // the start of an array
itemArrayEnd
itemTableStart
itemTableEnd
itemArrayTableStart
itemArrayTableEnd
itemKeyStart
itemCommentStart
itemInlineTableStart
itemInlineTableEnd
)
const (
eof = 0
comma = ','
tableStart = '['
tableEnd = ']'
arrayTableStart = '['
arrayTableEnd = ']'
tableSep = '.'
keySep = '='
arrayStart = '['
arrayEnd = ']'
commentStart = '#'
stringStart = '"'
stringEnd = '"'
rawStringStart = '\''
rawStringEnd = '\''
inlineTableStart = '{'
inlineTableEnd = '}'
)
type stateFn func(lx *lexer) stateFn
type lexer struct {
input string
start int
pos int
line int
state stateFn
items chan item
// Allow for backing up up to three runes.
// This is necessary because TOML contains 3-rune tokens (""" and ''').
prevWidths [3]int
nprev int // how many of prevWidths are in use
// If we emit an eof, we can still back up, but it is not OK to call
// next again.
atEOF bool
// A stack of state functions used to maintain context.
// The idea is to reuse parts of the state machine in various places.
// For example, values can appear at the top level or within arbitrarily
// nested arrays. The last state on the stack is used after a value has
// been lexed. Similarly for comments.
stack []stateFn
}
type item struct {
typ itemType
val string
line int
}
func (lx *lexer) nextItem() item {
for {
select {
case item := <-lx.items:
return item
default:
lx.state = lx.state(lx)
}
}
}
func lex(input string) *lexer {
lx := &lexer{
input: input,
state: lexTop,
line: 1,
items: make(chan item, 10),
stack: make([]stateFn, 0, 10),
}
return lx
}
func (lx *lexer) push(state stateFn) {
lx.stack = append(lx.stack, state)
}
func (lx *lexer) pop() stateFn {
if len(lx.stack) == 0 {
return lx.errorf("BUG in lexer: no states to pop")
}
last := lx.stack[len(lx.stack)-1]
lx.stack = lx.stack[0 : len(lx.stack)-1]
return last
}
func (lx *lexer) current() string {
return lx.input[lx.start:lx.pos]
}
func (lx *lexer) emit(typ itemType) {
lx.items <- item{typ, lx.current(), lx.line}
lx.start = lx.pos
}
func (lx *lexer) emitTrim(typ itemType) {
lx.items <- item{typ, strings.TrimSpace(lx.current()), lx.line}
lx.start = lx.pos
}
func (lx *lexer) next() (r rune) {
if lx.atEOF {
panic("next called after EOF")
}
if lx.pos >= len(lx.input) {
lx.atEOF = true
return eof
}
if lx.input[lx.pos] == '\n' {
lx.line++
}
lx.prevWidths[2] = lx.prevWidths[1]
lx.prevWidths[1] = lx.prevWidths[0]
if lx.nprev < 3 {
lx.nprev++
}
r, w := utf8.DecodeRuneInString(lx.input[lx.pos:])
lx.prevWidths[0] = w
lx.pos += w
return r
}
// ignore skips over the pending input before this point.
func (lx *lexer) ignore() {
lx.start = lx.pos
}
// backup steps back one rune. Can be called only twice between calls to next.
func (lx *lexer) backup() {
if lx.atEOF {
lx.atEOF = false
return
}
if lx.nprev < 1 {
panic("backed up too far")
}
w := lx.prevWidths[0]
lx.prevWidths[0] = lx.prevWidths[1]
lx.prevWidths[1] = lx.prevWidths[2]
lx.nprev--
lx.pos -= w
if lx.pos < len(lx.input) && lx.input[lx.pos] == '\n' {
lx.line--
}
}
// accept consumes the next rune if it's equal to `valid`.
func (lx *lexer) accept(valid rune) bool {
if lx.next() == valid {
return true
}
lx.backup()
return false
}
// peek returns but does not consume the next rune in the input.
func (lx *lexer) peek() rune {
r := lx.next()
lx.backup()
return r
}
// skip ignores all input that matches the given predicate.
func (lx *lexer) skip(pred func(rune) bool) {
for {
r := lx.next()
if pred(r) {
continue
}
lx.backup()
lx.ignore()
return
}
}
// errorf stops all lexing by emitting an error and returning `nil`.
// Note that any value that is a character is escaped if it's a special
// character (newlines, tabs, etc.).
func (lx *lexer) errorf(format string, values ...interface{}) stateFn {
lx.items <- item{
itemError,
fmt.Sprintf(format, values...),
lx.line,
}
return nil
}
// lexTop consumes elements at the top level of TOML data.
func lexTop(lx *lexer) stateFn {
r := lx.next()
if isWhitespace(r) || isNL(r) {
return lexSkip(lx, lexTop)
}
switch r {
case commentStart:
lx.push(lexTop)
return lexCommentStart
case tableStart:
return lexTableStart
case eof:
if lx.pos > lx.start {
return lx.errorf("unexpected EOF")
}
lx.emit(itemEOF)
return nil
}
// At this point, the only valid item can be a key, so we back up
// and let the key lexer do the rest.
lx.backup()
lx.push(lexTopEnd)
return lexKeyStart
}
// lexTopEnd is entered whenever a top-level item has been consumed. (A value
// or a table.) It must see only whitespace, and will turn back to lexTop
// upon a newline. If it sees EOF, it will quit the lexer successfully.
func lexTopEnd(lx *lexer) stateFn {
r := lx.next()
switch {
case r == commentStart:
// a comment will read to a newline for us.
lx.push(lexTop)
return lexCommentStart
case isWhitespace(r):
return lexTopEnd
case isNL(r):
lx.ignore()
return lexTop
case r == eof:
lx.emit(itemEOF)
return nil
}
return lx.errorf("expected a top-level item to end with a newline, "+
"comment, or EOF, but got %q instead", r)
}
// lexTable lexes the beginning of a table. Namely, it makes sure that
// it starts with a character other than '.' and ']'.
// It assumes that '[' has already been consumed.
// It also handles the case that this is an item in an array of tables.
// e.g., '[[name]]'.
func lexTableStart(lx *lexer) stateFn {
if lx.peek() == arrayTableStart {
lx.next()
lx.emit(itemArrayTableStart)
lx.push(lexArrayTableEnd)
} else {
lx.emit(itemTableStart)
lx.push(lexTableEnd)
}
return lexTableNameStart
}
func lexTableEnd(lx *lexer) stateFn {
lx.emit(itemTableEnd)
return lexTopEnd
}
func lexArrayTableEnd(lx *lexer) stateFn {
if r := lx.next(); r != arrayTableEnd {
return lx.errorf("expected end of table array name delimiter %q, "+
"but got %q instead", arrayTableEnd, r)
}
lx.emit(itemArrayTableEnd)
return lexTopEnd
}
func lexTableNameStart(lx *lexer) stateFn {
lx.skip(isWhitespace)
switch r := lx.peek(); {
case r == tableEnd || r == eof:
return lx.errorf("unexpected end of table name " +
"(table names cannot be empty)")
case r == tableSep:
return lx.errorf("unexpected table separator " +
"(table names cannot be empty)")
case r == stringStart || r == rawStringStart:
lx.ignore()
lx.push(lexTableNameEnd)
return lexValue // reuse string lexing
default:
return lexBareTableName
}
}
// lexBareTableName lexes the name of a table. It assumes that at least one
// valid character for the table has already been read.
func lexBareTableName(lx *lexer) stateFn {
r := lx.next()
if isBareKeyChar(r) {
return lexBareTableName
}
lx.backup()
lx.emit(itemText)
return lexTableNameEnd
}
// lexTableNameEnd reads the end of a piece of a table name, optionally
// consuming whitespace.
func lexTableNameEnd(lx *lexer) stateFn {
lx.skip(isWhitespace)
switch r := lx.next(); {
case isWhitespace(r):
return lexTableNameEnd
case r == tableSep:
lx.ignore()
return lexTableNameStart
case r == tableEnd:
return lx.pop()
default:
return lx.errorf("expected '.' or ']' to end table name, "+
"but got %q instead", r)
}
}
// lexKeyStart consumes a key name up until the first non-whitespace character.
// lexKeyStart will ignore whitespace.
func lexKeyStart(lx *lexer) stateFn {
r := lx.peek()
switch {
case r == keySep:
return lx.errorf("unexpected key separator %q", keySep)
case isWhitespace(r) || isNL(r):
lx.next()
return lexSkip(lx, lexKeyStart)
case r == stringStart || r == rawStringStart:
lx.ignore()
lx.emit(itemKeyStart)
lx.push(lexKeyEnd)
return lexValue // reuse string lexing
default:
lx.ignore()
lx.emit(itemKeyStart)
return lexBareKey
}
}
// lexBareKey consumes the text of a bare key. Assumes that the first character
// (which is not whitespace) has not yet been consumed.
func lexBareKey(lx *lexer) stateFn {
switch r := lx.next(); {
case isBareKeyChar(r):
return lexBareKey
case isWhitespace(r):
lx.backup()
lx.emit(itemText)
return lexKeyEnd
case r == keySep:
lx.backup()
lx.emit(itemText)
return lexKeyEnd
default:
return lx.errorf("bare keys cannot contain %q", r)
}
}
// lexKeyEnd consumes the end of a key and trims whitespace (up to the key
// separator).
func lexKeyEnd(lx *lexer) stateFn {
switch r := lx.next(); {
case r == keySep:
return lexSkip(lx, lexValue)
case isWhitespace(r):
return lexSkip(lx, lexKeyEnd)
default:
return lx.errorf("expected key separator %q, but got %q instead",
keySep, r)
}
}
// lexValue starts the consumption of a value anywhere a value is expected.
// lexValue will ignore whitespace.
// After a value is lexed, the last state on the next is popped and returned.
func lexValue(lx *lexer) stateFn {
// We allow whitespace to precede a value, but NOT newlines.
// In array syntax, the array states are responsible for ignoring newlines.
r := lx.next()
switch {
case isWhitespace(r):
return lexSkip(lx, lexValue)
case isDigit(r):
lx.backup() // avoid an extra state and use the same as above
return lexNumberOrDateStart
}
switch r {
case arrayStart:
lx.ignore()
lx.emit(itemArray)
return lexArrayValue
case inlineTableStart:
lx.ignore()
lx.emit(itemInlineTableStart)
return lexInlineTableValue
case stringStart:
if lx.accept(stringStart) {
if lx.accept(stringStart) {
lx.ignore() // Ignore """
return lexMultilineString
}
lx.backup()
}
lx.ignore() // ignore the '"'
return lexString
case rawStringStart:
if lx.accept(rawStringStart) {
if lx.accept(rawStringStart) {
lx.ignore() // Ignore """
return lexMultilineRawString
}
lx.backup()
}
lx.ignore() // ignore the "'"
return lexRawString
case '+', '-':
return lexNumberStart
case '.': // special error case, be kind to users
return lx.errorf("floats must start with a digit, not '.'")
}
if unicode.IsLetter(r) {
// Be permissive here; lexBool will give a nice error if the
// user wrote something like
// x = foo
// (i.e. not 'true' or 'false' but is something else word-like.)
lx.backup()
return lexBool
}
return lx.errorf("expected value but found %q instead", r)
}
// lexArrayValue consumes one value in an array. It assumes that '[' or ','
// have already been consumed. All whitespace and newlines are ignored.
func lexArrayValue(lx *lexer) stateFn {
r := lx.next()
switch {
case isWhitespace(r) || isNL(r):
return lexSkip(lx, lexArrayValue)
case r == commentStart:
lx.push(lexArrayValue)
return lexCommentStart
case r == comma:
return lx.errorf("unexpected comma")
case r == arrayEnd:
// NOTE(caleb): The spec isn't clear about whether you can have
// a trailing comma or not, so we'll allow it.
return lexArrayEnd
}
lx.backup()
lx.push(lexArrayValueEnd)
return lexValue
}
// lexArrayValueEnd consumes everything between the end of an array value and
// the next value (or the end of the array): it ignores whitespace and newlines
// and expects either a ',' or a ']'.
func lexArrayValueEnd(lx *lexer) stateFn {
r := lx.next()
switch {
case isWhitespace(r) || isNL(r):
return lexSkip(lx, lexArrayValueEnd)
case r == commentStart:
lx.push(lexArrayValueEnd)
return lexCommentStart
case r == comma:
lx.ignore()
return lexArrayValue // move on to the next value
case r == arrayEnd:
return lexArrayEnd
}
return lx.errorf(
"expected a comma or array terminator %q, but got %q instead",
arrayEnd, r,
)
}
// lexArrayEnd finishes the lexing of an array.
// It assumes that a ']' has just been consumed.
func lexArrayEnd(lx *lexer) stateFn {
lx.ignore()
lx.emit(itemArrayEnd)
return lx.pop()
}
// lexInlineTableValue consumes one key/value pair in an inline table.
// It assumes that '{' or ',' have already been consumed. Whitespace is ignored.
func lexInlineTableValue(lx *lexer) stateFn {
r := lx.next()
switch {
case isWhitespace(r):
return lexSkip(lx, lexInlineTableValue)
case isNL(r):
return lx.errorf("newlines not allowed within inline tables")
case r == commentStart:
lx.push(lexInlineTableValue)
return lexCommentStart
case r == comma:
return lx.errorf("unexpected comma")
case r == inlineTableEnd:
return lexInlineTableEnd
}
lx.backup()
lx.push(lexInlineTableValueEnd)
return lexKeyStart
}
// lexInlineTableValueEnd consumes everything between the end of an inline table
// key/value pair and the next pair (or the end of the table):
// it ignores whitespace and expects either a ',' or a '}'.
func lexInlineTableValueEnd(lx *lexer) stateFn {
r := lx.next()
switch {
case isWhitespace(r):
return lexSkip(lx, lexInlineTableValueEnd)
case isNL(r):
return lx.errorf("newlines not allowed within inline tables")
case r == commentStart:
lx.push(lexInlineTableValueEnd)
return lexCommentStart
case r == comma:
lx.ignore()
return lexInlineTableValue
case r == inlineTableEnd:
return lexInlineTableEnd
}
return lx.errorf("expected a comma or an inline table terminator %q, "+
"but got %q instead", inlineTableEnd, r)
}
// lexInlineTableEnd finishes the lexing of an inline table.
// It assumes that a '}' has just been consumed.
func lexInlineTableEnd(lx *lexer) stateFn {
lx.ignore()
lx.emit(itemInlineTableEnd)
return lx.pop()
}
// lexString consumes the inner contents of a string. It assumes that the
// beginning '"' has already been consumed and ignored.
func lexString(lx *lexer) stateFn {
r := lx.next()
switch {
case r == eof:
return lx.errorf("unexpected EOF")
case isNL(r):
return lx.errorf("strings cannot contain newlines")
case r == '\\':
lx.push(lexString)
return lexStringEscape
case r == stringEnd:
lx.backup()
lx.emit(itemString)
lx.next()
lx.ignore()
return lx.pop()
}
return lexString
}
// lexMultilineString consumes the inner contents of a string. It assumes that
// the beginning '"""' has already been consumed and ignored.
func lexMultilineString(lx *lexer) stateFn {
switch lx.next() {
case eof:
return lx.errorf("unexpected EOF")
case '\\':
return lexMultilineStringEscape
case stringEnd:
if lx.accept(stringEnd) {
if lx.accept(stringEnd) {
lx.backup()
lx.backup()
lx.backup()
lx.emit(itemMultilineString)
lx.next()
lx.next()
lx.next()
lx.ignore()
return lx.pop()
}
lx.backup()
}
}
return lexMultilineString
}
// lexRawString consumes a raw string. Nothing can be escaped in such a string.
// It assumes that the beginning "'" has already been consumed and ignored.
func lexRawString(lx *lexer) stateFn {
r := lx.next()
switch {
case r == eof:
return lx.errorf("unexpected EOF")
case isNL(r):
return lx.errorf("strings cannot contain newlines")
case r == rawStringEnd:
lx.backup()
lx.emit(itemRawString)
lx.next()
lx.ignore()
return lx.pop()
}
return lexRawString
}
// lexMultilineRawString consumes a raw string. Nothing can be escaped in such
// a string. It assumes that the beginning "'''" has already been consumed and
// ignored.
func lexMultilineRawString(lx *lexer) stateFn {
switch lx.next() {
case eof:
return lx.errorf("unexpected EOF")
case rawStringEnd:
if lx.accept(rawStringEnd) {
if lx.accept(rawStringEnd) {
lx.backup()
lx.backup()
lx.backup()
lx.emit(itemRawMultilineString)
lx.next()
lx.next()
lx.next()
lx.ignore()
return lx.pop()
}
lx.backup()
}
}
return lexMultilineRawString
}
// lexMultilineStringEscape consumes an escaped character. It assumes that the
// preceding '\\' has already been consumed.
func lexMultilineStringEscape(lx *lexer) stateFn {
// Handle the special case first:
if isNL(lx.next()) {
return lexMultilineString
}
lx.backup()
lx.push(lexMultilineString)
return lexStringEscape(lx)
}
func lexStringEscape(lx *lexer) stateFn {
r := lx.next()
switch r {
case 'b':
fallthrough
case 't':
fallthrough
case 'n':
fallthrough
case 'f':
fallthrough
case 'r':
fallthrough
case '"':
fallthrough
case '\\':
return lx.pop()
case 'u':
return lexShortUnicodeEscape
case 'U':
return lexLongUnicodeEscape
}
return lx.errorf("invalid escape character %q; only the following "+
"escape characters are allowed: "+
`\b, \t, \n, \f, \r, \", \\, \uXXXX, and \UXXXXXXXX`, r)
}
func lexShortUnicodeEscape(lx *lexer) stateFn {
var r rune
for i := 0; i < 4; i++ {
r = lx.next()
if !isHexadecimal(r) {
return lx.errorf(`expected four hexadecimal digits after '\u', `+
"but got %q instead", lx.current())
}
}
return lx.pop()
}
func lexLongUnicodeEscape(lx *lexer) stateFn {
var r rune
for i := 0; i < 8; i++ {
r = lx.next()
if !isHexadecimal(r) {
return lx.errorf(`expected eight hexadecimal digits after '\U', `+
"but got %q instead", lx.current())
}
}
return lx.pop()
}
// lexNumberOrDateStart consumes either an integer, a float, or datetime.
func lexNumberOrDateStart(lx *lexer) stateFn {
r := lx.next()
if isDigit(r) {
return lexNumberOrDate
}
switch r {
case '_':
return lexNumber
case 'e', 'E':
return lexFloat
case '.':
return lx.errorf("floats must start with a digit, not '.'")
}
return lx.errorf("expected a digit but got %q", r)
}
// lexNumberOrDate consumes either an integer, float or datetime.
func lexNumberOrDate(lx *lexer) stateFn {
r := lx.next()
if isDigit(r) {
return lexNumberOrDate
}
switch r {
case '-':
return lexDatetime
case '_':
return lexNumber
case '.', 'e', 'E':
return lexFloat
}
lx.backup()
lx.emit(itemInteger)
return lx.pop()
}
// lexDatetime consumes a Datetime, to a first approximation.
// The parser validates that it matches one of the accepted formats.
func lexDatetime(lx *lexer) stateFn {
r := lx.next()
if isDigit(r) {
return lexDatetime
}
switch r {
case '-', 'T', ':', '.', 'Z', '+':
return lexDatetime
}
lx.backup()
lx.emit(itemDatetime)
return lx.pop()
}
// lexNumberStart consumes either an integer or a float. It assumes that a sign
// has already been read, but that *no* digits have been consumed.
// lexNumberStart will move to the appropriate integer or float states.
func lexNumberStart(lx *lexer) stateFn {
// We MUST see a digit. Even floats have to start with a digit.
r := lx.next()
if !isDigit(r) {
if r == '.' {
return lx.errorf("floats must start with a digit, not '.'")
}
return lx.errorf("expected a digit but got %q", r)
}
return lexNumber
}
// lexNumber consumes an integer or a float after seeing the first digit.
func lexNumber(lx *lexer) stateFn {
r := lx.next()
if isDigit(r) {
return lexNumber
}
switch r {
case '_':
return lexNumber
case '.', 'e', 'E':
return lexFloat
}
lx.backup()
lx.emit(itemInteger)
return lx.pop()
}
// lexFloat consumes the elements of a float. It allows any sequence of
// float-like characters, so floats emitted by the lexer are only a first
// approximation and must be validated by the parser.
func lexFloat(lx *lexer) stateFn {
r := lx.next()
if isDigit(r) {
return lexFloat
}
switch r {
case '_', '.', '-', '+', 'e', 'E':
return lexFloat
}
lx.backup()
lx.emit(itemFloat)
return lx.pop()
}
// lexBool consumes a bool string: 'true' or 'false.
func lexBool(lx *lexer) stateFn {
var rs []rune
for {
r := lx.next()
if !unicode.IsLetter(r) {
lx.backup()
break
}
rs = append(rs, r)
}
s := string(rs)
switch s {
case "true", "false":
lx.emit(itemBool)
return lx.pop()
}
return lx.errorf("expected value but found %q instead", s)
}
// lexCommentStart begins the lexing of a comment. It will emit
// itemCommentStart and consume no characters, passing control to lexComment.
func lexCommentStart(lx *lexer) stateFn {
lx.ignore()
lx.emit(itemCommentStart)
return lexComment
}
// lexComment lexes an entire comment. It assumes that '#' has been consumed.
// It will consume *up to* the first newline character, and pass control
// back to the last state on the stack.
func lexComment(lx *lexer) stateFn {
r := lx.peek()
if isNL(r) || r == eof {
lx.emit(itemText)
return lx.pop()
}
lx.next()
return lexComment
}
// lexSkip ignores all slurped input and moves on to the next state.
func lexSkip(lx *lexer, nextState stateFn) stateFn {
return func(lx *lexer) stateFn {
lx.ignore()
return nextState
}
}
// isWhitespace returns true if `r` is a whitespace character according
// to the spec.
func isWhitespace(r rune) bool {
return r == '\t' || r == ' '
}
func isNL(r rune) bool {
return r == '\n' || r == '\r'
}
func isDigit(r rune) bool {
return r >= '0' && r <= '9'
}
func isHexadecimal(r rune) bool {
return (r >= '0' && r <= '9') ||
(r >= 'a' && r <= 'f') ||
(r >= 'A' && r <= 'F')
}
func isBareKeyChar(r rune) bool {
return (r >= 'A' && r <= 'Z') ||
(r >= 'a' && r <= 'z') ||
(r >= '0' && r <= '9') ||
r == '_' ||
r == '-'
}
func (itype itemType) String() string {
switch itype {
case itemError:
return "Error"
case itemNIL:
return "NIL"
case itemEOF:
return "EOF"
case itemText:
return "Text"
case itemString, itemRawString, itemMultilineString, itemRawMultilineString:
return "String"
case itemBool:
return "Bool"
case itemInteger:
return "Integer"
case itemFloat:
return "Float"
case itemDatetime:
return "DateTime"
case itemTableStart:
return "TableStart"
case itemTableEnd:
return "TableEnd"
case itemKeyStart:
return "KeyStart"
case itemArray:
return "Array"
case itemArrayEnd:
return "ArrayEnd"
case itemCommentStart:
return "CommentStart"
}
panic(fmt.Sprintf("BUG: Unknown type '%d'.", int(itype)))
}
func (item item) String() string {
return fmt.Sprintf("(%s, %s)", item.typ.String(), item.val)
}

592
vendor/github.com/BurntSushi/toml/parse.go generated vendored Normal file
View File

@@ -0,0 +1,592 @@
package toml
import (
"fmt"
"strconv"
"strings"
"time"
"unicode"
"unicode/utf8"
)
type parser struct {
mapping map[string]interface{}
types map[string]tomlType
lx *lexer
// A list of keys in the order that they appear in the TOML data.
ordered []Key
// the full key for the current hash in scope
context Key
// the base key name for everything except hashes
currentKey string
// rough approximation of line number
approxLine int
// A map of 'key.group.names' to whether they were created implicitly.
implicits map[string]bool
}
type parseError string
func (pe parseError) Error() string {
return string(pe)
}
func parse(data string) (p *parser, err error) {
defer func() {
if r := recover(); r != nil {
var ok bool
if err, ok = r.(parseError); ok {
return
}
panic(r)
}
}()
p = &parser{
mapping: make(map[string]interface{}),
types: make(map[string]tomlType),
lx: lex(data),
ordered: make([]Key, 0),
implicits: make(map[string]bool),
}
for {
item := p.next()
if item.typ == itemEOF {
break
}
p.topLevel(item)
}
return p, nil
}
func (p *parser) panicf(format string, v ...interface{}) {
msg := fmt.Sprintf("Near line %d (last key parsed '%s'): %s",
p.approxLine, p.current(), fmt.Sprintf(format, v...))
panic(parseError(msg))
}
func (p *parser) next() item {
it := p.lx.nextItem()
if it.typ == itemError {
p.panicf("%s", it.val)
}
return it
}
func (p *parser) bug(format string, v ...interface{}) {
panic(fmt.Sprintf("BUG: "+format+"\n\n", v...))
}
func (p *parser) expect(typ itemType) item {
it := p.next()
p.assertEqual(typ, it.typ)
return it
}
func (p *parser) assertEqual(expected, got itemType) {
if expected != got {
p.bug("Expected '%s' but got '%s'.", expected, got)
}
}
func (p *parser) topLevel(item item) {
switch item.typ {
case itemCommentStart:
p.approxLine = item.line
p.expect(itemText)
case itemTableStart:
kg := p.next()
p.approxLine = kg.line
var key Key
for ; kg.typ != itemTableEnd && kg.typ != itemEOF; kg = p.next() {
key = append(key, p.keyString(kg))
}
p.assertEqual(itemTableEnd, kg.typ)
p.establishContext(key, false)
p.setType("", tomlHash)
p.ordered = append(p.ordered, key)
case itemArrayTableStart:
kg := p.next()
p.approxLine = kg.line
var key Key
for ; kg.typ != itemArrayTableEnd && kg.typ != itemEOF; kg = p.next() {
key = append(key, p.keyString(kg))
}
p.assertEqual(itemArrayTableEnd, kg.typ)
p.establishContext(key, true)
p.setType("", tomlArrayHash)
p.ordered = append(p.ordered, key)
case itemKeyStart:
kname := p.next()
p.approxLine = kname.line
p.currentKey = p.keyString(kname)
val, typ := p.value(p.next())
p.setValue(p.currentKey, val)
p.setType(p.currentKey, typ)
p.ordered = append(p.ordered, p.context.add(p.currentKey))
p.currentKey = ""
default:
p.bug("Unexpected type at top level: %s", item.typ)
}
}
// Gets a string for a key (or part of a key in a table name).
func (p *parser) keyString(it item) string {
switch it.typ {
case itemText:
return it.val
case itemString, itemMultilineString,
itemRawString, itemRawMultilineString:
s, _ := p.value(it)
return s.(string)
default:
p.bug("Unexpected key type: %s", it.typ)
panic("unreachable")
}
}
// value translates an expected value from the lexer into a Go value wrapped
// as an empty interface.
func (p *parser) value(it item) (interface{}, tomlType) {
switch it.typ {
case itemString:
return p.replaceEscapes(it.val), p.typeOfPrimitive(it)
case itemMultilineString:
trimmed := stripFirstNewline(stripEscapedWhitespace(it.val))
return p.replaceEscapes(trimmed), p.typeOfPrimitive(it)
case itemRawString:
return it.val, p.typeOfPrimitive(it)
case itemRawMultilineString:
return stripFirstNewline(it.val), p.typeOfPrimitive(it)
case itemBool:
switch it.val {
case "true":
return true, p.typeOfPrimitive(it)
case "false":
return false, p.typeOfPrimitive(it)
}
p.bug("Expected boolean value, but got '%s'.", it.val)
case itemInteger:
if !numUnderscoresOK(it.val) {
p.panicf("Invalid integer %q: underscores must be surrounded by digits",
it.val)
}
val := strings.Replace(it.val, "_", "", -1)
num, err := strconv.ParseInt(val, 10, 64)
if err != nil {
// Distinguish integer values. Normally, it'd be a bug if the lexer
// provides an invalid integer, but it's possible that the number is
// out of range of valid values (which the lexer cannot determine).
// So mark the former as a bug but the latter as a legitimate user
// error.
if e, ok := err.(*strconv.NumError); ok &&
e.Err == strconv.ErrRange {
p.panicf("Integer '%s' is out of the range of 64-bit "+
"signed integers.", it.val)
} else {
p.bug("Expected integer value, but got '%s'.", it.val)
}
}
return num, p.typeOfPrimitive(it)
case itemFloat:
parts := strings.FieldsFunc(it.val, func(r rune) bool {
switch r {
case '.', 'e', 'E':
return true
}
return false
})
for _, part := range parts {
if !numUnderscoresOK(part) {
p.panicf("Invalid float %q: underscores must be "+
"surrounded by digits", it.val)
}
}
if !numPeriodsOK(it.val) {
// As a special case, numbers like '123.' or '1.e2',
// which are valid as far as Go/strconv are concerned,
// must be rejected because TOML says that a fractional
// part consists of '.' followed by 1+ digits.
p.panicf("Invalid float %q: '.' must be followed "+
"by one or more digits", it.val)
}
val := strings.Replace(it.val, "_", "", -1)
num, err := strconv.ParseFloat(val, 64)
if err != nil {
if e, ok := err.(*strconv.NumError); ok &&
e.Err == strconv.ErrRange {
p.panicf("Float '%s' is out of the range of 64-bit "+
"IEEE-754 floating-point numbers.", it.val)
} else {
p.panicf("Invalid float value: %q", it.val)
}
}
return num, p.typeOfPrimitive(it)
case itemDatetime:
var t time.Time
var ok bool
var err error
for _, format := range []string{
"2006-01-02T15:04:05Z07:00",
"2006-01-02T15:04:05",
"2006-01-02",
} {
t, err = time.ParseInLocation(format, it.val, time.Local)
if err == nil {
ok = true
break
}
}
if !ok {
p.panicf("Invalid TOML Datetime: %q.", it.val)
}
return t, p.typeOfPrimitive(it)
case itemArray:
array := make([]interface{}, 0)
types := make([]tomlType, 0)
for it = p.next(); it.typ != itemArrayEnd; it = p.next() {
if it.typ == itemCommentStart {
p.expect(itemText)
continue
}
val, typ := p.value(it)
array = append(array, val)
types = append(types, typ)
}
return array, p.typeOfArray(types)
case itemInlineTableStart:
var (
hash = make(map[string]interface{})
outerContext = p.context
outerKey = p.currentKey
)
p.context = append(p.context, p.currentKey)
p.currentKey = ""
for it := p.next(); it.typ != itemInlineTableEnd; it = p.next() {
if it.typ != itemKeyStart {
p.bug("Expected key start but instead found %q, around line %d",
it.val, p.approxLine)
}
if it.typ == itemCommentStart {
p.expect(itemText)
continue
}
// retrieve key
k := p.next()
p.approxLine = k.line
kname := p.keyString(k)
// retrieve value
p.currentKey = kname
val, typ := p.value(p.next())
// make sure we keep metadata up to date
p.setType(kname, typ)
p.ordered = append(p.ordered, p.context.add(p.currentKey))
hash[kname] = val
}
p.context = outerContext
p.currentKey = outerKey
return hash, tomlHash
}
p.bug("Unexpected value type: %s", it.typ)
panic("unreachable")
}
// numUnderscoresOK checks whether each underscore in s is surrounded by
// characters that are not underscores.
func numUnderscoresOK(s string) bool {
accept := false
for _, r := range s {
if r == '_' {
if !accept {
return false
}
accept = false
continue
}
accept = true
}
return accept
}
// numPeriodsOK checks whether every period in s is followed by a digit.
func numPeriodsOK(s string) bool {
period := false
for _, r := range s {
if period && !isDigit(r) {
return false
}
period = r == '.'
}
return !period
}
// establishContext sets the current context of the parser,
// where the context is either a hash or an array of hashes. Which one is
// set depends on the value of the `array` parameter.
//
// Establishing the context also makes sure that the key isn't a duplicate, and
// will create implicit hashes automatically.
func (p *parser) establishContext(key Key, array bool) {
var ok bool
// Always start at the top level and drill down for our context.
hashContext := p.mapping
keyContext := make(Key, 0)
// We only need implicit hashes for key[0:-1]
for _, k := range key[0 : len(key)-1] {
_, ok = hashContext[k]
keyContext = append(keyContext, k)
// No key? Make an implicit hash and move on.
if !ok {
p.addImplicit(keyContext)
hashContext[k] = make(map[string]interface{})
}
// If the hash context is actually an array of tables, then set
// the hash context to the last element in that array.
//
// Otherwise, it better be a table, since this MUST be a key group (by
// virtue of it not being the last element in a key).
switch t := hashContext[k].(type) {
case []map[string]interface{}:
hashContext = t[len(t)-1]
case map[string]interface{}:
hashContext = t
default:
p.panicf("Key '%s' was already created as a hash.", keyContext)
}
}
p.context = keyContext
if array {
// If this is the first element for this array, then allocate a new
// list of tables for it.
k := key[len(key)-1]
if _, ok := hashContext[k]; !ok {
hashContext[k] = make([]map[string]interface{}, 0, 5)
}
// Add a new table. But make sure the key hasn't already been used
// for something else.
if hash, ok := hashContext[k].([]map[string]interface{}); ok {
hashContext[k] = append(hash, make(map[string]interface{}))
} else {
p.panicf("Key '%s' was already created and cannot be used as "+
"an array.", keyContext)
}
} else {
p.setValue(key[len(key)-1], make(map[string]interface{}))
}
p.context = append(p.context, key[len(key)-1])
}
// setValue sets the given key to the given value in the current context.
// It will make sure that the key hasn't already been defined, account for
// implicit key groups.
func (p *parser) setValue(key string, value interface{}) {
var tmpHash interface{}
var ok bool
hash := p.mapping
keyContext := make(Key, 0)
for _, k := range p.context {
keyContext = append(keyContext, k)
if tmpHash, ok = hash[k]; !ok {
p.bug("Context for key '%s' has not been established.", keyContext)
}
switch t := tmpHash.(type) {
case []map[string]interface{}:
// The context is a table of hashes. Pick the most recent table
// defined as the current hash.
hash = t[len(t)-1]
case map[string]interface{}:
hash = t
default:
p.bug("Expected hash to have type 'map[string]interface{}', but "+
"it has '%T' instead.", tmpHash)
}
}
keyContext = append(keyContext, key)
if _, ok := hash[key]; ok {
// Typically, if the given key has already been set, then we have
// to raise an error since duplicate keys are disallowed. However,
// it's possible that a key was previously defined implicitly. In this
// case, it is allowed to be redefined concretely. (See the
// `tests/valid/implicit-and-explicit-after.toml` test in `toml-test`.)
//
// But we have to make sure to stop marking it as an implicit. (So that
// another redefinition provokes an error.)
//
// Note that since it has already been defined (as a hash), we don't
// want to overwrite it. So our business is done.
if p.isImplicit(keyContext) {
p.removeImplicit(keyContext)
return
}
// Otherwise, we have a concrete key trying to override a previous
// key, which is *always* wrong.
p.panicf("Key '%s' has already been defined.", keyContext)
}
hash[key] = value
}
// setType sets the type of a particular value at a given key.
// It should be called immediately AFTER setValue.
//
// Note that if `key` is empty, then the type given will be applied to the
// current context (which is either a table or an array of tables).
func (p *parser) setType(key string, typ tomlType) {
keyContext := make(Key, 0, len(p.context)+1)
for _, k := range p.context {
keyContext = append(keyContext, k)
}
if len(key) > 0 { // allow type setting for hashes
keyContext = append(keyContext, key)
}
p.types[keyContext.String()] = typ
}
// addImplicit sets the given Key as having been created implicitly.
func (p *parser) addImplicit(key Key) {
p.implicits[key.String()] = true
}
// removeImplicit stops tagging the given key as having been implicitly
// created.
func (p *parser) removeImplicit(key Key) {
p.implicits[key.String()] = false
}
// isImplicit returns true if the key group pointed to by the key was created
// implicitly.
func (p *parser) isImplicit(key Key) bool {
return p.implicits[key.String()]
}
// current returns the full key name of the current context.
func (p *parser) current() string {
if len(p.currentKey) == 0 {
return p.context.String()
}
if len(p.context) == 0 {
return p.currentKey
}
return fmt.Sprintf("%s.%s", p.context, p.currentKey)
}
func stripFirstNewline(s string) string {
if len(s) == 0 || s[0] != '\n' {
return s
}
return s[1:]
}
func stripEscapedWhitespace(s string) string {
esc := strings.Split(s, "\\\n")
if len(esc) > 1 {
for i := 1; i < len(esc); i++ {
esc[i] = strings.TrimLeftFunc(esc[i], unicode.IsSpace)
}
}
return strings.Join(esc, "")
}
func (p *parser) replaceEscapes(str string) string {
var replaced []rune
s := []byte(str)
r := 0
for r < len(s) {
if s[r] != '\\' {
c, size := utf8.DecodeRune(s[r:])
r += size
replaced = append(replaced, c)
continue
}
r += 1
if r >= len(s) {
p.bug("Escape sequence at end of string.")
return ""
}
switch s[r] {
default:
p.bug("Expected valid escape code after \\, but got %q.", s[r])
return ""
case 'b':
replaced = append(replaced, rune(0x0008))
r += 1
case 't':
replaced = append(replaced, rune(0x0009))
r += 1
case 'n':
replaced = append(replaced, rune(0x000A))
r += 1
case 'f':
replaced = append(replaced, rune(0x000C))
r += 1
case 'r':
replaced = append(replaced, rune(0x000D))
r += 1
case '"':
replaced = append(replaced, rune(0x0022))
r += 1
case '\\':
replaced = append(replaced, rune(0x005C))
r += 1
case 'u':
// At this point, we know we have a Unicode escape of the form
// `uXXXX` at [r, r+5). (Because the lexer guarantees this
// for us.)
escaped := p.asciiEscapeToUnicode(s[r+1 : r+5])
replaced = append(replaced, escaped)
r += 5
case 'U':
// At this point, we know we have a Unicode escape of the form
// `uXXXX` at [r, r+9). (Because the lexer guarantees this
// for us.)
escaped := p.asciiEscapeToUnicode(s[r+1 : r+9])
replaced = append(replaced, escaped)
r += 9
}
}
return string(replaced)
}
func (p *parser) asciiEscapeToUnicode(bs []byte) rune {
s := string(bs)
hex, err := strconv.ParseUint(strings.ToLower(s), 16, 32)
if err != nil {
p.bug("Could not parse '%s' as a hexadecimal number, but the "+
"lexer claims it's OK: %s", s, err)
}
if !utf8.ValidRune(rune(hex)) {
p.panicf("Escaped character '\\u%s' is not valid UTF-8.", s)
}
return rune(hex)
}
func isStringType(ty itemType) bool {
return ty == itemString || ty == itemMultilineString ||
ty == itemRawString || ty == itemRawMultilineString
}

1
vendor/github.com/BurntSushi/toml/session.vim generated vendored Normal file
View File

@@ -0,0 +1 @@
au BufWritePost *.go silent!make tags > /dev/null 2>&1

91
vendor/github.com/BurntSushi/toml/type_check.go generated vendored Normal file
View File

@@ -0,0 +1,91 @@
package toml
// tomlType represents any Go type that corresponds to a TOML type.
// While the first draft of the TOML spec has a simplistic type system that
// probably doesn't need this level of sophistication, we seem to be militating
// toward adding real composite types.
type tomlType interface {
typeString() string
}
// typeEqual accepts any two types and returns true if they are equal.
func typeEqual(t1, t2 tomlType) bool {
if t1 == nil || t2 == nil {
return false
}
return t1.typeString() == t2.typeString()
}
func typeIsHash(t tomlType) bool {
return typeEqual(t, tomlHash) || typeEqual(t, tomlArrayHash)
}
type tomlBaseType string
func (btype tomlBaseType) typeString() string {
return string(btype)
}
func (btype tomlBaseType) String() string {
return btype.typeString()
}
var (
tomlInteger tomlBaseType = "Integer"
tomlFloat tomlBaseType = "Float"
tomlDatetime tomlBaseType = "Datetime"
tomlString tomlBaseType = "String"
tomlBool tomlBaseType = "Bool"
tomlArray tomlBaseType = "Array"
tomlHash tomlBaseType = "Hash"
tomlArrayHash tomlBaseType = "ArrayHash"
)
// typeOfPrimitive returns a tomlType of any primitive value in TOML.
// Primitive values are: Integer, Float, Datetime, String and Bool.
//
// Passing a lexer item other than the following will cause a BUG message
// to occur: itemString, itemBool, itemInteger, itemFloat, itemDatetime.
func (p *parser) typeOfPrimitive(lexItem item) tomlType {
switch lexItem.typ {
case itemInteger:
return tomlInteger
case itemFloat:
return tomlFloat
case itemDatetime:
return tomlDatetime
case itemString:
return tomlString
case itemMultilineString:
return tomlString
case itemRawString:
return tomlString
case itemRawMultilineString:
return tomlString
case itemBool:
return tomlBool
}
p.bug("Cannot infer primitive type of lex item '%s'.", lexItem)
panic("unreachable")
}
// typeOfArray returns a tomlType for an array given a list of types of its
// values.
//
// In the current spec, if an array is homogeneous, then its type is always
// "Array". If the array is not homogeneous, an error is generated.
func (p *parser) typeOfArray(types []tomlType) tomlType {
// Empty arrays are cool.
if len(types) == 0 {
return tomlArray
}
theType := types[0]
for _, t := range types[1:] {
if !typeEqual(theType, t) {
p.panicf("Array contains values of type '%s' and '%s', but "+
"arrays must be homogeneous.", theType, t)
}
}
return tomlArray
}

242
vendor/github.com/BurntSushi/toml/type_fields.go generated vendored Normal file
View File

@@ -0,0 +1,242 @@
package toml
// Struct field handling is adapted from code in encoding/json:
//
// Copyright 2010 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the Go distribution.
import (
"reflect"
"sort"
"sync"
)
// A field represents a single field found in a struct.
type field struct {
name string // the name of the field (`toml` tag included)
tag bool // whether field has a `toml` tag
index []int // represents the depth of an anonymous field
typ reflect.Type // the type of the field
}
// byName sorts field by name, breaking ties with depth,
// then breaking ties with "name came from toml tag", then
// breaking ties with index sequence.
type byName []field
func (x byName) Len() int { return len(x) }
func (x byName) Swap(i, j int) { x[i], x[j] = x[j], x[i] }
func (x byName) Less(i, j int) bool {
if x[i].name != x[j].name {
return x[i].name < x[j].name
}
if len(x[i].index) != len(x[j].index) {
return len(x[i].index) < len(x[j].index)
}
if x[i].tag != x[j].tag {
return x[i].tag
}
return byIndex(x).Less(i, j)
}
// byIndex sorts field by index sequence.
type byIndex []field
func (x byIndex) Len() int { return len(x) }
func (x byIndex) Swap(i, j int) { x[i], x[j] = x[j], x[i] }
func (x byIndex) Less(i, j int) bool {
for k, xik := range x[i].index {
if k >= len(x[j].index) {
return false
}
if xik != x[j].index[k] {
return xik < x[j].index[k]
}
}
return len(x[i].index) < len(x[j].index)
}
// typeFields returns a list of fields that TOML should recognize for the given
// type. The algorithm is breadth-first search over the set of structs to
// include - the top struct and then any reachable anonymous structs.
func typeFields(t reflect.Type) []field {
// Anonymous fields to explore at the current level and the next.
current := []field{}
next := []field{{typ: t}}
// Count of queued names for current level and the next.
count := map[reflect.Type]int{}
nextCount := map[reflect.Type]int{}
// Types already visited at an earlier level.
visited := map[reflect.Type]bool{}
// Fields found.
var fields []field
for len(next) > 0 {
current, next = next, current[:0]
count, nextCount = nextCount, map[reflect.Type]int{}
for _, f := range current {
if visited[f.typ] {
continue
}
visited[f.typ] = true
// Scan f.typ for fields to include.
for i := 0; i < f.typ.NumField(); i++ {
sf := f.typ.Field(i)
if sf.PkgPath != "" && !sf.Anonymous { // unexported
continue
}
opts := getOptions(sf.Tag)
if opts.skip {
continue
}
index := make([]int, len(f.index)+1)
copy(index, f.index)
index[len(f.index)] = i
ft := sf.Type
if ft.Name() == "" && ft.Kind() == reflect.Ptr {
// Follow pointer.
ft = ft.Elem()
}
// Record found field and index sequence.
if opts.name != "" || !sf.Anonymous || ft.Kind() != reflect.Struct {
tagged := opts.name != ""
name := opts.name
if name == "" {
name = sf.Name
}
fields = append(fields, field{name, tagged, index, ft})
if count[f.typ] > 1 {
// If there were multiple instances, add a second,
// so that the annihilation code will see a duplicate.
// It only cares about the distinction between 1 or 2,
// so don't bother generating any more copies.
fields = append(fields, fields[len(fields)-1])
}
continue
}
// Record new anonymous struct to explore in next round.
nextCount[ft]++
if nextCount[ft] == 1 {
f := field{name: ft.Name(), index: index, typ: ft}
next = append(next, f)
}
}
}
}
sort.Sort(byName(fields))
// Delete all fields that are hidden by the Go rules for embedded fields,
// except that fields with TOML tags are promoted.
// The fields are sorted in primary order of name, secondary order
// of field index length. Loop over names; for each name, delete
// hidden fields by choosing the one dominant field that survives.
out := fields[:0]
for advance, i := 0, 0; i < len(fields); i += advance {
// One iteration per name.
// Find the sequence of fields with the name of this first field.
fi := fields[i]
name := fi.name
for advance = 1; i+advance < len(fields); advance++ {
fj := fields[i+advance]
if fj.name != name {
break
}
}
if advance == 1 { // Only one field with this name
out = append(out, fi)
continue
}
dominant, ok := dominantField(fields[i : i+advance])
if ok {
out = append(out, dominant)
}
}
fields = out
sort.Sort(byIndex(fields))
return fields
}
// dominantField looks through the fields, all of which are known to
// have the same name, to find the single field that dominates the
// others using Go's embedding rules, modified by the presence of
// TOML tags. If there are multiple top-level fields, the boolean
// will be false: This condition is an error in Go and we skip all
// the fields.
func dominantField(fields []field) (field, bool) {
// The fields are sorted in increasing index-length order. The winner
// must therefore be one with the shortest index length. Drop all
// longer entries, which is easy: just truncate the slice.
length := len(fields[0].index)
tagged := -1 // Index of first tagged field.
for i, f := range fields {
if len(f.index) > length {
fields = fields[:i]
break
}
if f.tag {
if tagged >= 0 {
// Multiple tagged fields at the same level: conflict.
// Return no field.
return field{}, false
}
tagged = i
}
}
if tagged >= 0 {
return fields[tagged], true
}
// All remaining fields have the same length. If there's more than one,
// we have a conflict (two fields named "X" at the same level) and we
// return no field.
if len(fields) > 1 {
return field{}, false
}
return fields[0], true
}
var fieldCache struct {
sync.RWMutex
m map[reflect.Type][]field
}
// cachedTypeFields is like typeFields but uses a cache to avoid repeated work.
func cachedTypeFields(t reflect.Type) []field {
fieldCache.RLock()
f := fieldCache.m[t]
fieldCache.RUnlock()
if f != nil {
return f
}
// Compute fields without lock.
// Might duplicate effort but won't hold other computations back.
f = typeFields(t)
if f == nil {
f = []field{}
}
fieldCache.Lock()
if fieldCache.m == nil {
fieldCache.m = map[reflect.Type][]field{}
}
fieldCache.m[t] = f
fieldCache.Unlock()
return f
}

View File

@@ -60,6 +60,9 @@ NewWriterLevelDict(w io.Writer, level int, dict []byte) *Writer
// Write compresses the input data and write it to the underlying writer
(w *Writer) Write(p []byte) (int, error)
// Flush writes any unwritten data to the underlying writer
(w *Writer) Flush() error
// Close flushes the buffer and frees C zstd objects
(w *Writer) Close() error
```

View File

@@ -1,35 +1,15 @@
/* ******************************************************************
bitstream
Part of FSE library
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- Source repository : https://github.com/Cyan4973/FiniteStateEntropy
* bitstream
* Part of FSE library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
#ifndef BITSTREAM_H_MODULE
#define BITSTREAM_H_MODULE
@@ -37,7 +17,6 @@
#if defined (__cplusplus)
extern "C" {
#endif
/*
* This API consists of small unitary functions, which must be inlined for best performance.
* Since link-time-optimization is not available for all compilers,
@@ -48,6 +27,7 @@ extern "C" {
* Dependencies
******************************************/
#include "mem.h" /* unaligned access routines */
#include "compiler.h" /* UNLIKELY() */
#include "debug.h" /* assert(), DEBUGLOG(), RAWLOG() */
#include "error_private.h" /* error codes and messages */
@@ -55,10 +35,12 @@ extern "C" {
/*=========================================
* Target specific
=========================================*/
#if defined(__BMI__) && defined(__GNUC__)
# include <immintrin.h> /* support for bextr (experimental) */
#elif defined(__ICCARM__)
# include <intrinsics.h>
#ifndef ZSTD_NO_INTRINSICS
# if defined(__BMI__) && defined(__GNUC__)
# include <immintrin.h> /* support for bextr (experimental) */
# elif defined(__ICCARM__)
# include <intrinsics.h>
# endif
#endif
#define STREAM_ACCUMULATOR_MIN_32 25
@@ -160,9 +142,12 @@ MEM_STATIC unsigned BIT_highbit32 (U32 val)
assert(val != 0);
{
# if defined(_MSC_VER) /* Visual */
unsigned long r=0;
_BitScanReverse ( &r, val );
return (unsigned) r;
# if STATIC_BMI2 == 1
return _lzcnt_u32(val) ^ 31;
# else
unsigned long r = 0;
return _BitScanReverse(&r, val) ? (unsigned)r : 0;
# endif
# elif defined(__GNUC__) && (__GNUC__ >= 3) /* Use GCC Intrinsic */
return __builtin_clz (val) ^ 31;
# elif defined(__ICCARM__) /* IAR Intrinsic */
@@ -218,7 +203,7 @@ MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC,
MEM_STATIC void BIT_addBits(BIT_CStream_t* bitC,
size_t value, unsigned nbBits)
{
MEM_STATIC_ASSERT(BIT_MASK_SIZE == 32);
DEBUG_STATIC_ASSERT(BIT_MASK_SIZE == 32);
assert(nbBits < BIT_MASK_SIZE);
assert(nbBits + bitC->bitPos < sizeof(bitC->bitContainer) * 8);
bitC->bitContainer |= (value & BIT_mask[nbBits]) << bitC->bitPos;
@@ -291,7 +276,7 @@ MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC)
*/
MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, size_t srcSize)
{
if (srcSize < 1) { memset(bitD, 0, sizeof(*bitD)); return ERROR(srcSize_wrong); }
if (srcSize < 1) { ZSTD_memset(bitD, 0, sizeof(*bitD)); return ERROR(srcSize_wrong); }
bitD->start = (const char*)srcBuffer;
bitD->limitPtr = bitD->start + sizeof(bitD->bitContainer);
@@ -337,12 +322,12 @@ MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, si
return srcSize;
}
MEM_STATIC size_t BIT_getUpperBits(size_t bitContainer, U32 const start)
MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getUpperBits(size_t bitContainer, U32 const start)
{
return bitContainer >> start;
}
MEM_STATIC size_t BIT_getMiddleBits(size_t bitContainer, U32 const start, U32 const nbBits)
MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getMiddleBits(size_t bitContainer, U32 const start, U32 const nbBits)
{
U32 const regMask = sizeof(bitContainer)*8 - 1;
/* if start > regMask, bitstream is corrupted, and result is undefined */
@@ -350,10 +335,14 @@ MEM_STATIC size_t BIT_getMiddleBits(size_t bitContainer, U32 const start, U32 co
return (bitContainer >> (start & regMask)) & BIT_mask[nbBits];
}
MEM_STATIC size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits)
MEM_STATIC FORCE_INLINE_ATTR size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits)
{
#if defined(STATIC_BMI2) && STATIC_BMI2 == 1
return _bzhi_u64(bitContainer, nbBits);
#else
assert(nbBits < BIT_MASK_SIZE);
return bitContainer & BIT_mask[nbBits];
#endif
}
/*! BIT_lookBits() :
@@ -362,7 +351,7 @@ MEM_STATIC size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits)
* On 32-bits, maxNbBits==24.
* On 64-bits, maxNbBits==56.
* @return : value extracted */
MEM_STATIC size_t BIT_lookBits(const BIT_DStream_t* bitD, U32 nbBits)
MEM_STATIC FORCE_INLINE_ATTR size_t BIT_lookBits(const BIT_DStream_t* bitD, U32 nbBits)
{
/* arbitrate between double-shift and shift+mask */
#if 1
@@ -385,7 +374,7 @@ MEM_STATIC size_t BIT_lookBitsFast(const BIT_DStream_t* bitD, U32 nbBits)
return (bitD->bitContainer << (bitD->bitsConsumed & regMask)) >> (((regMask+1)-nbBits) & regMask);
}
MEM_STATIC void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits)
MEM_STATIC FORCE_INLINE_ATTR void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits)
{
bitD->bitsConsumed += nbBits;
}
@@ -394,7 +383,7 @@ MEM_STATIC void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits)
* Read (consume) next n bits from local register and update.
* Pay attention to not read more than nbBits contained into local register.
* @return : extracted value. */
MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits)
MEM_STATIC FORCE_INLINE_ATTR size_t BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits)
{
size_t const value = BIT_lookBits(bitD, nbBits);
BIT_skipBits(bitD, nbBits);
@@ -411,6 +400,23 @@ MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, unsigned nbBits)
return value;
}
/*! BIT_reloadDStreamFast() :
* Similar to BIT_reloadDStream(), but with two differences:
* 1. bitsConsumed <= sizeof(bitD->bitContainer)*8 must hold!
* 2. Returns BIT_DStream_overflow when bitD->ptr < bitD->limitPtr, at this
* point you must use BIT_reloadDStream() to reload.
*/
MEM_STATIC BIT_DStream_status BIT_reloadDStreamFast(BIT_DStream_t* bitD)
{
if (UNLIKELY(bitD->ptr < bitD->limitPtr))
return BIT_DStream_overflow;
assert(bitD->bitsConsumed <= sizeof(bitD->bitContainer)*8);
bitD->ptr -= bitD->bitsConsumed >> 3;
bitD->bitsConsumed &= 7;
bitD->bitContainer = MEM_readLEST(bitD->ptr);
return BIT_DStream_unfinished;
}
/*! BIT_reloadDStream() :
* Refill `bitD` from buffer previously set in BIT_initDStream() .
* This function is safe, it guarantees it will not read beyond src buffer.
@@ -422,10 +428,7 @@ MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD)
return BIT_DStream_overflow;
if (bitD->ptr >= bitD->limitPtr) {
bitD->ptr -= bitD->bitsConsumed >> 3;
bitD->bitsConsumed &= 7;
bitD->bitContainer = MEM_readLEST(bitD->ptr);
return BIT_DStream_unfinished;
return BIT_reloadDStreamFast(bitD);
}
if (bitD->ptr == bitD->start) {
if (bitD->bitsConsumed < sizeof(bitD->bitContainer)*8) return BIT_DStream_endOfBuffer;

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -17,7 +17,7 @@
/* force inlining */
#if !defined(ZSTD_NO_INLINE)
#if defined (__GNUC__) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */
#if (defined(__GNUC__) && !defined(__STRICT_ANSI__)) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */
# define INLINE_KEYWORD inline
#else
# define INLINE_KEYWORD
@@ -38,6 +38,17 @@
#endif
/**
On MSVC qsort requires that functions passed into it use the __cdecl calling conversion(CC).
This explictly marks such functions as __cdecl so that the code will still compile
if a CC other than __cdecl has been made the default.
*/
#if defined(_MSC_VER)
# define WIN_CDECL __cdecl
#else
# define WIN_CDECL
#endif
/**
* FORCE_INLINE_TEMPLATE is used to define C "templates", which take constant
* parameters. They must be inlined for the compiler to eliminate the constant
@@ -117,6 +128,9 @@
# elif defined(__GNUC__) && ( (__GNUC__ >= 4) || ( (__GNUC__ == 3) && (__GNUC_MINOR__ >= 1) ) )
# define PREFETCH_L1(ptr) __builtin_prefetch((ptr), 0 /* rw==read */, 3 /* locality */)
# define PREFETCH_L2(ptr) __builtin_prefetch((ptr), 0 /* rw==read */, 2 /* locality */)
# elif defined(__aarch64__)
# define PREFETCH_L1(ptr) __asm__ __volatile__("prfm pldl1keep, %0" ::"Q"(*(ptr)))
# define PREFETCH_L2(ptr) __asm__ __volatile__("prfm pldl2keep, %0" ::"Q"(*(ptr)))
# else
# define PREFETCH_L1(ptr) (void)(ptr) /* disabled */
# define PREFETCH_L2(ptr) (void)(ptr) /* disabled */
@@ -136,7 +150,7 @@
/* vectorization
* older GCC (pre gcc-4.3 picked as the cutoff) uses a different syntax */
#if !defined(__clang__) && defined(__GNUC__)
#if !defined(__INTEL_COMPILER) && !defined(__clang__) && defined(__GNUC__)
# if (__GNUC__ == 4 && __GNUC_MINOR__ > 3) || (__GNUC__ >= 5)
# define DONT_VECTORIZE __attribute__((optimize("no-tree-vectorize")))
# else
@@ -146,6 +160,19 @@
# define DONT_VECTORIZE
#endif
/* Tell the compiler that a branch is likely or unlikely.
* Only use these macros if it causes the compiler to generate better code.
* If you can remove a LIKELY/UNLIKELY annotation without speed changes in gcc
* and clang, please do.
*/
#if defined(__GNUC__)
#define LIKELY(x) (__builtin_expect((x), 1))
#define UNLIKELY(x) (__builtin_expect((x), 0))
#else
#define LIKELY(x) (x)
#define UNLIKELY(x) (x)
#endif
/* disable warnings */
#ifdef _MSC_VER /* Visual Studio */
# include <intrin.h> /* For Visual 2005 */
@@ -156,4 +183,106 @@
# pragma warning(disable : 4324) /* disable: C4324: padded structure */
#endif
/*Like DYNAMIC_BMI2 but for compile time determination of BMI2 support*/
#ifndef STATIC_BMI2
# if defined(_MSC_VER) && (defined(_M_X64) || defined(_M_I86))
# ifdef __AVX2__ //MSVC does not have a BMI2 specific flag, but every CPU that supports AVX2 also supports BMI2
# define STATIC_BMI2 1
# endif
# endif
#endif
#ifndef STATIC_BMI2
#define STATIC_BMI2 0
#endif
/* compat. with non-clang compilers */
#ifndef __has_builtin
# define __has_builtin(x) 0
#endif
/* compat. with non-clang compilers */
#ifndef __has_feature
# define __has_feature(x) 0
#endif
/* detects whether we are being compiled under msan */
#ifndef ZSTD_MEMORY_SANITIZER
# if __has_feature(memory_sanitizer)
# define ZSTD_MEMORY_SANITIZER 1
# else
# define ZSTD_MEMORY_SANITIZER 0
# endif
#endif
#if ZSTD_MEMORY_SANITIZER
/* Not all platforms that support msan provide sanitizers/msan_interface.h.
* We therefore declare the functions we need ourselves, rather than trying to
* include the header file... */
#include <stddef.h> /* size_t */
#define ZSTD_DEPS_NEED_STDINT
#include "zstd_deps.h" /* intptr_t */
/* Make memory region fully initialized (without changing its contents). */
void __msan_unpoison(const volatile void *a, size_t size);
/* Make memory region fully uninitialized (without changing its contents).
This is a legacy interface that does not update origin information. Use
__msan_allocated_memory() instead. */
void __msan_poison(const volatile void *a, size_t size);
/* Returns the offset of the first (at least partially) poisoned byte in the
memory range, or -1 if the whole range is good. */
intptr_t __msan_test_shadow(const volatile void *x, size_t size);
#endif
/* detects whether we are being compiled under asan */
#ifndef ZSTD_ADDRESS_SANITIZER
# if __has_feature(address_sanitizer)
# define ZSTD_ADDRESS_SANITIZER 1
# elif defined(__SANITIZE_ADDRESS__)
# define ZSTD_ADDRESS_SANITIZER 1
# else
# define ZSTD_ADDRESS_SANITIZER 0
# endif
#endif
#if ZSTD_ADDRESS_SANITIZER
/* Not all platforms that support asan provide sanitizers/asan_interface.h.
* We therefore declare the functions we need ourselves, rather than trying to
* include the header file... */
#include <stddef.h> /* size_t */
/**
* Marks a memory region (<c>[addr, addr+size)</c>) as unaddressable.
*
* This memory must be previously allocated by your program. Instrumented
* code is forbidden from accessing addresses in this region until it is
* unpoisoned. This function is not guaranteed to poison the entire region -
* it could poison only a subregion of <c>[addr, addr+size)</c> due to ASan
* alignment restrictions.
*
* \note This function is not thread-safe because no two threads can poison or
* unpoison memory in the same memory region simultaneously.
*
* \param addr Start of memory region.
* \param size Size of memory region. */
void __asan_poison_memory_region(void const volatile *addr, size_t size);
/**
* Marks a memory region (<c>[addr, addr+size)</c>) as addressable.
*
* This memory must be previously allocated by your program. Accessing
* addresses in this region is allowed until this region is poisoned again.
* This function could unpoison a super-region of <c>[addr, addr+size)</c> due
* to ASan alignment restrictions.
*
* \note This function is not thread-safe because no two threads can
* poison or unpoison memory in the same memory region simultaneously.
*
* \param addr Start of memory region.
* \param size Size of memory region. */
void __asan_unpoison_memory_region(void const volatile *addr, size_t size);
#endif
#endif /* ZSTD_COMPILER_H */

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -40,33 +40,42 @@
* Constants
***************************************/
#define COVER_MAX_SAMPLES_SIZE (sizeof(size_t) == 8 ? ((unsigned)-1) : ((unsigned)1 GB))
#define DEFAULT_SPLITPOINT 1.0
#define COVER_DEFAULT_SPLITPOINT 1.0
/*-*************************************
* Console display
***************************************/
#ifndef LOCALDISPLAYLEVEL
static int g_displayLevel = 2;
#endif
#undef DISPLAY
#define DISPLAY(...) \
{ \
fprintf(stderr, __VA_ARGS__); \
fflush(stderr); \
}
#undef LOCALDISPLAYLEVEL
#define LOCALDISPLAYLEVEL(displayLevel, l, ...) \
if (displayLevel >= l) { \
DISPLAY(__VA_ARGS__); \
} /* 0 : no display; 1: errors; 2: default; 3: details; 4: debug */
#undef DISPLAYLEVEL
#define DISPLAYLEVEL(l, ...) LOCALDISPLAYLEVEL(g_displayLevel, l, __VA_ARGS__)
#ifndef LOCALDISPLAYUPDATE
static const clock_t g_refreshRate = CLOCKS_PER_SEC * 15 / 100;
static clock_t g_time = 0;
#endif
#undef LOCALDISPLAYUPDATE
#define LOCALDISPLAYUPDATE(displayLevel, l, ...) \
if (displayLevel >= l) { \
if ((clock() - g_time > refreshRate) || (displayLevel >= 4)) { \
if ((clock() - g_time > g_refreshRate) || (displayLevel >= 4)) { \
g_time = clock(); \
DISPLAY(__VA_ARGS__); \
} \
}
#undef DISPLAYUPDATE
#define DISPLAYUPDATE(l, ...) LOCALDISPLAYUPDATE(g_displayLevel, l, __VA_ARGS__)
static const clock_t refreshRate = CLOCKS_PER_SEC * 15 / 100;
static clock_t g_time = 0;
/*-*************************************
* Hash table
@@ -120,9 +129,9 @@ static int COVER_map_init(COVER_map_t *map, U32 size) {
/**
* Internal hash function
*/
static const U32 prime4bytes = 2654435761U;
static const U32 COVER_prime4bytes = 2654435761U;
static U32 COVER_map_hash(COVER_map_t *map, U32 key) {
return (key * prime4bytes) >> (32 - map->sizeLog);
return (key * COVER_prime4bytes) >> (32 - map->sizeLog);
}
/**
@@ -215,7 +224,7 @@ typedef struct {
} COVER_ctx_t;
/* We need a global context for qsort... */
static COVER_ctx_t *g_ctx = NULL;
static COVER_ctx_t *g_coverCtx = NULL;
/*-*************************************
* Helper functions
@@ -258,11 +267,11 @@ static int COVER_cmp8(COVER_ctx_t *ctx, const void *lp, const void *rp) {
/**
* Same as COVER_cmp() except ties are broken by pointer value
* NOTE: g_ctx must be set to call this function. A global is required because
* NOTE: g_coverCtx must be set to call this function. A global is required because
* qsort doesn't take an opaque pointer.
*/
static int COVER_strict_cmp(const void *lp, const void *rp) {
int result = COVER_cmp(g_ctx, lp, rp);
static int WIN_CDECL COVER_strict_cmp(const void *lp, const void *rp) {
int result = COVER_cmp(g_coverCtx, lp, rp);
if (result == 0) {
result = lp < rp ? -1 : 1;
}
@@ -271,8 +280,8 @@ static int COVER_strict_cmp(const void *lp, const void *rp) {
/**
* Faster version for d <= 8.
*/
static int COVER_strict_cmp8(const void *lp, const void *rp) {
int result = COVER_cmp8(g_ctx, lp, rp);
static int WIN_CDECL COVER_strict_cmp8(const void *lp, const void *rp) {
int result = COVER_cmp8(g_coverCtx, lp, rp);
if (result == 0) {
result = lp < rp ? -1 : 1;
}
@@ -603,7 +612,7 @@ static size_t COVER_ctx_init(COVER_ctx_t *ctx, const void *samplesBuffer,
/* qsort doesn't take an opaque pointer, so pass as a global.
* On OpenBSD qsort() is not guaranteed to be stable, their mergesort() is.
*/
g_ctx = ctx;
g_coverCtx = ctx;
#if defined(__OpenBSD__)
mergesort(ctx->suffix, ctx->suffixSize, sizeof(U32),
(ctx->d <= 8 ? &COVER_strict_cmp8 : &COVER_strict_cmp));
@@ -946,7 +955,7 @@ void COVER_dictSelectionFree(COVER_dictSelection_t selection){
free(selection.dictContent);
}
COVER_dictSelection_t COVER_selectDict(BYTE* customDictContent,
COVER_dictSelection_t COVER_selectDict(BYTE* customDictContent, size_t dictBufferCapacity,
size_t dictContentSize, const BYTE* samplesBuffer, const size_t* samplesSizes, unsigned nbFinalizeSamples,
size_t nbCheckSamples, size_t nbSamples, ZDICT_cover_params_t params, size_t* offsets, size_t totalCompressedSize) {
@@ -954,8 +963,8 @@ COVER_dictSelection_t COVER_selectDict(BYTE* customDictContent,
size_t largestCompressed = 0;
BYTE* customDictContentEnd = customDictContent + dictContentSize;
BYTE * largestDictbuffer = (BYTE *)malloc(dictContentSize);
BYTE * candidateDictBuffer = (BYTE *)malloc(dictContentSize);
BYTE * largestDictbuffer = (BYTE *)malloc(dictBufferCapacity);
BYTE * candidateDictBuffer = (BYTE *)malloc(dictBufferCapacity);
double regressionTolerance = ((double)params.shrinkDictMaxRegression / 100.0) + 1.00;
if (!largestDictbuffer || !candidateDictBuffer) {
@@ -967,7 +976,7 @@ COVER_dictSelection_t COVER_selectDict(BYTE* customDictContent,
/* Initial dictionary size and compressed size */
memcpy(largestDictbuffer, customDictContent, dictContentSize);
dictContentSize = ZDICT_finalizeDictionary(
largestDictbuffer, dictContentSize, customDictContent, dictContentSize,
largestDictbuffer, dictBufferCapacity, customDictContent, dictContentSize,
samplesBuffer, samplesSizes, nbFinalizeSamples, params.zParams);
if (ZDICT_isError(dictContentSize)) {
@@ -1001,7 +1010,7 @@ COVER_dictSelection_t COVER_selectDict(BYTE* customDictContent,
while (dictContentSize < largestDict) {
memcpy(candidateDictBuffer, largestDictbuffer, largestDict);
dictContentSize = ZDICT_finalizeDictionary(
candidateDictBuffer, dictContentSize, customDictContentEnd - dictContentSize, dictContentSize,
candidateDictBuffer, dictBufferCapacity, customDictContentEnd - dictContentSize, dictContentSize,
samplesBuffer, samplesSizes, nbFinalizeSamples, params.zParams);
if (ZDICT_isError(dictContentSize)) {
@@ -1079,7 +1088,7 @@ static void COVER_tryParameters(void *opaque) {
{
const size_t tail = COVER_buildDictionary(ctx, freqs, &activeDmers, dict,
dictBufferCapacity, parameters);
selection = COVER_selectDict(dict + tail, dictBufferCapacity - tail,
selection = COVER_selectDict(dict + tail, dictBufferCapacity, dictBufferCapacity - tail,
ctx->samples, ctx->samplesSizes, (unsigned)ctx->nbTrainSamples, ctx->nbTrainSamples, ctx->nbSamples, parameters, ctx->offsets,
totalCompressedSize);
@@ -1106,7 +1115,7 @@ ZDICTLIB_API size_t ZDICT_optimizeTrainFromBuffer_cover(
/* constants */
const unsigned nbThreads = parameters->nbThreads;
const double splitPoint =
parameters->splitPoint <= 0.0 ? DEFAULT_SPLITPOINT : parameters->splitPoint;
parameters->splitPoint <= 0.0 ? COVER_DEFAULT_SPLITPOINT : parameters->splitPoint;
const unsigned kMinD = parameters->d == 0 ? 6 : parameters->d;
const unsigned kMaxD = parameters->d == 0 ? 8 : parameters->d;
const unsigned kMinK = parameters->k == 0 ? 50 : parameters->k;

View File

@@ -1,3 +1,13 @@
/*
* Copyright (c) 2017-2020, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#include <stdio.h> /* fprintf */
#include <stdlib.h> /* malloc, free, qsort */
#include <string.h> /* memset */
@@ -142,6 +152,6 @@ void COVER_dictSelectionFree(COVER_dictSelection_t selection);
* smallest dictionary within a specified regression of the compressed size
* from the largest dictionary.
*/
COVER_dictSelection_t COVER_selectDict(BYTE* customDictContent,
COVER_dictSelection_t COVER_selectDict(BYTE* customDictContent, size_t dictBufferCapacity,
size_t dictContentSize, const BYTE* samplesBuffer, const size_t* samplesSizes, unsigned nbFinalizeSamples,
size_t nbCheckSamples, size_t nbSamples, ZDICT_cover_params_t params, size_t* offsets, size_t totalCompressedSize);

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2018-present, Facebook, Inc.
* Copyright (c) 2018-2020, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -16,8 +16,6 @@
* https://github.com/facebook/folly/blob/master/folly/CpuId.h
*/
#include <string.h>
#include "mem.h"
#ifdef _MSC_VER

View File

@@ -1,35 +1,15 @@
/* ******************************************************************
debug
Part of FSE library
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- Source repository : https://github.com/Cyan4973/FiniteStateEntropy
* debug
* Part of FSE library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */

View File

@@ -1,35 +1,15 @@
/* ******************************************************************
debug
Part of FSE library
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- Source repository : https://github.com/Cyan4973/FiniteStateEntropy
* debug
* Part of FSE library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
@@ -71,15 +51,6 @@ extern "C" {
#endif
/* DEBUGFILE can be defined externally,
* typically through compiler command line.
* note : currently useless.
* Value must be stderr or stdout */
#ifndef DEBUGFILE
# define DEBUGFILE stderr
#endif
/* recommended values for DEBUGLEVEL :
* 0 : release mode, no debug, all run-time checks disabled
* 1 : enables assert() only, no display
@@ -96,7 +67,8 @@ extern "C" {
*/
#if (DEBUGLEVEL>=1)
# include <assert.h>
# define ZSTD_DEPS_NEED_ASSERT
# include "zstd_deps.h"
#else
# ifndef assert /* assert may be already defined, due to prior #include <assert.h> */
# define assert(condition) ((void)0) /* disable assert (default) */
@@ -104,7 +76,8 @@ extern "C" {
#endif
#if (DEBUGLEVEL>=2)
# include <stdio.h>
# define ZSTD_DEPS_NEED_IO
# include "zstd_deps.h"
extern int g_debuglevel; /* the variable is only declared,
it actually lives in debug.c,
and is shared by the whole process.
@@ -112,14 +85,14 @@ extern int g_debuglevel; /* the variable is only declared,
It's useful when enabling very verbose levels
on selective conditions (such as position in src) */
# define RAWLOG(l, ...) { \
if (l<=g_debuglevel) { \
fprintf(stderr, __VA_ARGS__); \
# define RAWLOG(l, ...) { \
if (l<=g_debuglevel) { \
ZSTD_DEBUG_PRINT(__VA_ARGS__); \
} }
# define DEBUGLOG(l, ...) { \
if (l<=g_debuglevel) { \
fprintf(stderr, __FILE__ ": " __VA_ARGS__); \
fprintf(stderr, " \n"); \
# define DEBUGLOG(l, ...) { \
if (l<=g_debuglevel) { \
ZSTD_DEBUG_PRINT(__FILE__ ": " __VA_ARGS__); \
ZSTD_DEBUG_PRINT(" \n"); \
} }
#else
# define RAWLOG(l, ...) {} /* disabled */

View File

@@ -1,36 +1,16 @@
/*
Common functions of New Generation Entropy library
Copyright (C) 2016, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
*************************************************************************** */
/* ******************************************************************
* Common functions of New Generation Entropy library
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* *************************************
* Dependencies
@@ -58,8 +38,31 @@ const char* HUF_getErrorName(size_t code) { return ERR_getErrorName(code); }
/*-**************************************************************
* FSE NCount encoding-decoding
****************************************************************/
size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
const void* headerBuffer, size_t hbSize)
static U32 FSE_ctz(U32 val)
{
assert(val != 0);
{
# if defined(_MSC_VER) /* Visual */
unsigned long r=0;
return _BitScanForward(&r, val) ? (unsigned)r : 0;
# elif defined(__GNUC__) && (__GNUC__ >= 3) /* GCC Intrinsic */
return __builtin_ctz(val);
# elif defined(__ICCARM__) /* IAR Intrinsic */
return __CTZ(val);
# else /* Software version */
U32 count = 0;
while ((val & 1) == 0) {
val >>= 1;
++count;
}
return count;
# endif
}
}
FORCE_INLINE_TEMPLATE
size_t FSE_readNCount_body(short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
const void* headerBuffer, size_t hbSize)
{
const BYTE* const istart = (const BYTE*) headerBuffer;
const BYTE* const iend = istart + hbSize;
@@ -70,23 +73,23 @@ size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSVPtr, unsigned* t
U32 bitStream;
int bitCount;
unsigned charnum = 0;
unsigned const maxSV1 = *maxSVPtr + 1;
int previous0 = 0;
if (hbSize < 4) {
/* This function only works when hbSize >= 4 */
char buffer[4];
memset(buffer, 0, sizeof(buffer));
memcpy(buffer, headerBuffer, hbSize);
if (hbSize < 8) {
/* This function only works when hbSize >= 8 */
char buffer[8] = {0};
ZSTD_memcpy(buffer, headerBuffer, hbSize);
{ size_t const countSize = FSE_readNCount(normalizedCounter, maxSVPtr, tableLogPtr,
buffer, sizeof(buffer));
if (FSE_isError(countSize)) return countSize;
if (countSize > hbSize) return ERROR(corruption_detected);
return countSize;
} }
assert(hbSize >= 4);
assert(hbSize >= 8);
/* init */
memset(normalizedCounter, 0, (*maxSVPtr+1) * sizeof(normalizedCounter[0])); /* all symbols not present in NCount have a frequency of 0 */
ZSTD_memset(normalizedCounter, 0, (*maxSVPtr+1) * sizeof(normalizedCounter[0])); /* all symbols not present in NCount have a frequency of 0 */
bitStream = MEM_readLE32(ip);
nbBits = (bitStream & 0xF) + FSE_MIN_TABLELOG; /* extract tableLog */
if (nbBits > FSE_TABLELOG_ABSOLUTE_MAX) return ERROR(tableLog_tooLarge);
@@ -97,36 +100,58 @@ size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSVPtr, unsigned* t
threshold = 1<<nbBits;
nbBits++;
while ((remaining>1) & (charnum<=*maxSVPtr)) {
for (;;) {
if (previous0) {
unsigned n0 = charnum;
while ((bitStream & 0xFFFF) == 0xFFFF) {
n0 += 24;
if (ip < iend-5) {
ip += 2;
bitStream = MEM_readLE32(ip) >> bitCount;
/* Count the number of repeats. Each time the
* 2-bit repeat code is 0b11 there is another
* repeat.
* Avoid UB by setting the high bit to 1.
*/
int repeats = FSE_ctz(~bitStream | 0x80000000) >> 1;
while (repeats >= 12) {
charnum += 3 * 12;
if (LIKELY(ip <= iend-7)) {
ip += 3;
} else {
bitStream >>= 16;
bitCount += 16;
} }
while ((bitStream & 3) == 3) {
n0 += 3;
bitStream >>= 2;
bitCount += 2;
bitCount -= (int)(8 * (iend - 7 - ip));
bitCount &= 31;
ip = iend - 4;
}
bitStream = MEM_readLE32(ip) >> bitCount;
repeats = FSE_ctz(~bitStream | 0x80000000) >> 1;
}
n0 += bitStream & 3;
charnum += 3 * repeats;
bitStream >>= 2 * repeats;
bitCount += 2 * repeats;
/* Add the final repeat which isn't 0b11. */
assert((bitStream & 3) < 3);
charnum += bitStream & 3;
bitCount += 2;
if (n0 > *maxSVPtr) return ERROR(maxSymbolValue_tooSmall);
while (charnum < n0) normalizedCounter[charnum++] = 0;
if ((ip <= iend-7) || (ip + (bitCount>>3) <= iend-4)) {
/* This is an error, but break and return an error
* at the end, because returning out of a loop makes
* it harder for the compiler to optimize.
*/
if (charnum >= maxSV1) break;
/* We don't need to set the normalized count to 0
* because we already memset the whole buffer to 0.
*/
if (LIKELY(ip <= iend-7) || (ip + (bitCount>>3) <= iend-4)) {
assert((bitCount >> 3) <= 3); /* For first condition to work */
ip += bitCount>>3;
bitCount &= 7;
bitStream = MEM_readLE32(ip) >> bitCount;
} else {
bitStream >>= 2;
} }
{ int const max = (2*threshold-1) - remaining;
bitCount -= (int)(8 * (iend - 4 - ip));
bitCount &= 31;
ip = iend - 4;
}
bitStream = MEM_readLE32(ip) >> bitCount;
}
{
int const max = (2*threshold-1) - remaining;
int count;
if ((bitStream & (threshold-1)) < (U32)max) {
@@ -139,24 +164,43 @@ size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSVPtr, unsigned* t
}
count--; /* extra accuracy */
remaining -= count < 0 ? -count : count; /* -1 means +1 */
/* When it matters (small blocks), this is a
* predictable branch, because we don't use -1.
*/
if (count >= 0) {
remaining -= count;
} else {
assert(count == -1);
remaining += count;
}
normalizedCounter[charnum++] = (short)count;
previous0 = !count;
while (remaining < threshold) {
nbBits--;
threshold >>= 1;
}
if ((ip <= iend-7) || (ip + (bitCount>>3) <= iend-4)) {
assert(threshold > 1);
if (remaining < threshold) {
/* This branch can be folded into the
* threshold update condition because we
* know that threshold > 1.
*/
if (remaining <= 1) break;
nbBits = BIT_highbit32(remaining) + 1;
threshold = 1 << (nbBits - 1);
}
if (charnum >= maxSV1) break;
if (LIKELY(ip <= iend-7) || (ip + (bitCount>>3) <= iend-4)) {
ip += bitCount>>3;
bitCount &= 7;
} else {
bitCount -= (int)(8 * (iend - 4 - ip));
bitCount &= 31;
ip = iend - 4;
}
bitStream = MEM_readLE32(ip) >> (bitCount & 31);
} } /* while ((remaining>1) & (charnum<=*maxSVPtr)) */
bitStream = MEM_readLE32(ip) >> bitCount;
} }
if (remaining != 1) return ERROR(corruption_detected);
/* Only possible when there are too many zeros. */
if (charnum > maxSV1) return ERROR(maxSymbolValue_tooSmall);
if (bitCount > 32) return ERROR(corruption_detected);
*maxSVPtr = charnum-1;
@@ -164,6 +208,43 @@ size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSVPtr, unsigned* t
return ip-istart;
}
/* Avoids the FORCE_INLINE of the _body() function. */
static size_t FSE_readNCount_body_default(
short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
const void* headerBuffer, size_t hbSize)
{
return FSE_readNCount_body(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize);
}
#if DYNAMIC_BMI2
TARGET_ATTRIBUTE("bmi2") static size_t FSE_readNCount_body_bmi2(
short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
const void* headerBuffer, size_t hbSize)
{
return FSE_readNCount_body(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize);
}
#endif
size_t FSE_readNCount_bmi2(
short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
const void* headerBuffer, size_t hbSize, int bmi2)
{
#if DYNAMIC_BMI2
if (bmi2) {
return FSE_readNCount_body_bmi2(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize);
}
#endif
(void)bmi2;
return FSE_readNCount_body_default(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize);
}
size_t FSE_readNCount(
short* normalizedCounter, unsigned* maxSVPtr, unsigned* tableLogPtr,
const void* headerBuffer, size_t hbSize)
{
return FSE_readNCount_bmi2(normalizedCounter, maxSVPtr, tableLogPtr, headerBuffer, hbSize, /* bmi2 */ 0);
}
/*! HUF_readStats() :
Read compact Huffman tree, saved by HUF_writeCTable().
@@ -175,6 +256,17 @@ size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSVPtr, unsigned* t
size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
U32* nbSymbolsPtr, U32* tableLogPtr,
const void* src, size_t srcSize)
{
U32 wksp[HUF_READ_STATS_WORKSPACE_SIZE_U32];
return HUF_readStats_wksp(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, wksp, sizeof(wksp), /* bmi2 */ 0);
}
FORCE_INLINE_TEMPLATE size_t
HUF_readStats_body(BYTE* huffWeight, size_t hwSize, U32* rankStats,
U32* nbSymbolsPtr, U32* tableLogPtr,
const void* src, size_t srcSize,
void* workSpace, size_t wkspSize,
int bmi2)
{
U32 weightTotal;
const BYTE* ip = (const BYTE*) src;
@@ -183,7 +275,7 @@ size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
if (!srcSize) return ERROR(srcSize_wrong);
iSize = ip[0];
/* memset(huffWeight, 0, hwSize); *//* is not necessary, even though some analyzer complain ... */
/* ZSTD_memset(huffWeight, 0, hwSize); *//* is not necessary, even though some analyzer complain ... */
if (iSize >= 128) { /* special header */
oSize = iSize - 127;
@@ -197,14 +289,14 @@ size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
huffWeight[n+1] = ip[n/2] & 15;
} } }
else { /* header compressed with FSE (normal case) */
FSE_DTable fseWorkspace[FSE_DTABLE_SIZE_U32(6)]; /* 6 is max possible tableLog for HUF header (maybe even 5, to be tested) */
if (iSize+1 > srcSize) return ERROR(srcSize_wrong);
oSize = FSE_decompress_wksp(huffWeight, hwSize-1, ip+1, iSize, fseWorkspace, 6); /* max (hwSize-1) values decoded, as last one is implied */
/* max (hwSize-1) values decoded, as last one is implied */
oSize = FSE_decompress_wksp_bmi2(huffWeight, hwSize-1, ip+1, iSize, 6, workSpace, wkspSize, bmi2);
if (FSE_isError(oSize)) return oSize;
}
/* collect weight stats */
memset(rankStats, 0, (HUF_TABLELOG_MAX + 1) * sizeof(U32));
ZSTD_memset(rankStats, 0, (HUF_TABLELOG_MAX + 1) * sizeof(U32));
weightTotal = 0;
{ U32 n; for (n=0; n<oSize; n++) {
if (huffWeight[n] >= HUF_TABLELOG_MAX) return ERROR(corruption_detected);
@@ -234,3 +326,37 @@ size_t HUF_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
*nbSymbolsPtr = (U32)(oSize+1);
return iSize+1;
}
/* Avoids the FORCE_INLINE of the _body() function. */
static size_t HUF_readStats_body_default(BYTE* huffWeight, size_t hwSize, U32* rankStats,
U32* nbSymbolsPtr, U32* tableLogPtr,
const void* src, size_t srcSize,
void* workSpace, size_t wkspSize)
{
return HUF_readStats_body(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize, 0);
}
#if DYNAMIC_BMI2
static TARGET_ATTRIBUTE("bmi2") size_t HUF_readStats_body_bmi2(BYTE* huffWeight, size_t hwSize, U32* rankStats,
U32* nbSymbolsPtr, U32* tableLogPtr,
const void* src, size_t srcSize,
void* workSpace, size_t wkspSize)
{
return HUF_readStats_body(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize, 1);
}
#endif
size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize, U32* rankStats,
U32* nbSymbolsPtr, U32* tableLogPtr,
const void* src, size_t srcSize,
void* workSpace, size_t wkspSize,
int bmi2)
{
#if DYNAMIC_BMI2
if (bmi2) {
return HUF_readStats_body_bmi2(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize);
}
#endif
(void)bmi2;
return HUF_readStats_body_default(huffWeight, hwSize, rankStats, nbSymbolsPtr, tableLogPtr, src, srcSize, workSpace, wkspSize);
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -47,6 +47,8 @@ const char* ERR_getErrorString(ERR_enum code)
/* following error codes are not stable and may be removed or changed in a future version */
case PREFIX(frameIndex_tooLarge): return "Frame index is too large";
case PREFIX(seekableIO): return "An I/O error occurred when reading/seeking";
case PREFIX(dstBuffer_wrong): return "Destination buffer is wrong";
case PREFIX(srcBuffer_wrong): return "Source buffer is wrong";
case PREFIX(maxCode):
default: return notErrorCode;
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -21,7 +21,7 @@ extern "C" {
/* ****************************************
* Dependencies
******************************************/
#include <stddef.h> /* size_t */
#include "zstd_deps.h" /* size_t */
#include "zstd_errors.h" /* enum list */
@@ -49,7 +49,7 @@ typedef ZSTD_ErrorCode ERR_enum;
/*-****************************************
* Error codes handling
******************************************/
#undef ERROR /* reported already defined on VS 2015 (Rich Geldreich) */
#undef ERROR /* already defined on Visual Studio */
#define ERROR(name) ZSTD_ERROR(name)
#define ZSTD_ERROR(name) ((size_t)-PREFIX(name))
@@ -57,6 +57,10 @@ ERR_STATIC unsigned ERR_isError(size_t code) { return (code > ERROR(maxCode)); }
ERR_STATIC ERR_enum ERR_getErrorCode(size_t code) { if (!ERR_isError(code)) return (ERR_enum)0; return (ERR_enum) (0-code); }
/* check and forward error code */
#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return e
#define CHECK_F(f) { CHECK_V_F(_var_err__, f); }
/*-****************************************
* Error Strings

View File

@@ -1,3 +1,13 @@
/*
* Copyright (c) 2018-2020, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
/*-*************************************
* Dependencies
***************************************/
@@ -11,6 +21,7 @@
#include "threading.h"
#include "cover.h"
#include "zstd_internal.h" /* includes zstd.h */
#include "zstd_compress_internal.h" /* ZSTD_hash*() */
#ifndef ZDICT_STATIC_LINKING_ONLY
#define ZDICT_STATIC_LINKING_ONLY
#endif
@@ -23,7 +34,7 @@
#define FASTCOVER_MAX_SAMPLES_SIZE (sizeof(size_t) == 8 ? ((unsigned)-1) : ((unsigned)1 GB))
#define FASTCOVER_MAX_F 31
#define FASTCOVER_MAX_ACCEL 10
#define DEFAULT_SPLITPOINT 0.75
#define FASTCOVER_DEFAULT_SPLITPOINT 0.75
#define DEFAULT_F 20
#define DEFAULT_ACCEL 1
@@ -31,50 +42,50 @@
/*-*************************************
* Console display
***************************************/
#ifndef LOCALDISPLAYLEVEL
static int g_displayLevel = 2;
#endif
#undef DISPLAY
#define DISPLAY(...) \
{ \
fprintf(stderr, __VA_ARGS__); \
fflush(stderr); \
}
#undef LOCALDISPLAYLEVEL
#define LOCALDISPLAYLEVEL(displayLevel, l, ...) \
if (displayLevel >= l) { \
DISPLAY(__VA_ARGS__); \
} /* 0 : no display; 1: errors; 2: default; 3: details; 4: debug */
#undef DISPLAYLEVEL
#define DISPLAYLEVEL(l, ...) LOCALDISPLAYLEVEL(g_displayLevel, l, __VA_ARGS__)
#ifndef LOCALDISPLAYUPDATE
static const clock_t g_refreshRate = CLOCKS_PER_SEC * 15 / 100;
static clock_t g_time = 0;
#endif
#undef LOCALDISPLAYUPDATE
#define LOCALDISPLAYUPDATE(displayLevel, l, ...) \
if (displayLevel >= l) { \
if ((clock() - g_time > refreshRate) || (displayLevel >= 4)) { \
if ((clock() - g_time > g_refreshRate) || (displayLevel >= 4)) { \
g_time = clock(); \
DISPLAY(__VA_ARGS__); \
} \
}
#undef DISPLAYUPDATE
#define DISPLAYUPDATE(l, ...) LOCALDISPLAYUPDATE(g_displayLevel, l, __VA_ARGS__)
static const clock_t refreshRate = CLOCKS_PER_SEC * 15 / 100;
static clock_t g_time = 0;
/*-*************************************
* Hash Functions
***************************************/
static const U64 prime6bytes = 227718039650203ULL;
static size_t ZSTD_hash6(U64 u, U32 h) { return (size_t)(((u << (64-48)) * prime6bytes) >> (64-h)) ; }
static size_t ZSTD_hash6Ptr(const void* p, U32 h) { return ZSTD_hash6(MEM_readLE64(p), h); }
static const U64 prime8bytes = 0xCF1BBCDCB7A56463ULL;
static size_t ZSTD_hash8(U64 u, U32 h) { return (size_t)(((u) * prime8bytes) >> (64-h)) ; }
static size_t ZSTD_hash8Ptr(const void* p, U32 h) { return ZSTD_hash8(MEM_readLE64(p), h); }
/**
* Hash the d-byte value pointed to by p and mod 2^f
* Hash the d-byte value pointed to by p and mod 2^f into the frequency vector
*/
static size_t FASTCOVER_hashPtrToIndex(const void* p, U32 h, unsigned d) {
static size_t FASTCOVER_hashPtrToIndex(const void* p, U32 f, unsigned d) {
if (d == 6) {
return ZSTD_hash6Ptr(p, h) & ((1 << h) - 1);
return ZSTD_hash6Ptr(p, f);
}
return ZSTD_hash8Ptr(p, h) & ((1 << h) - 1);
return ZSTD_hash8Ptr(p, f);
}
@@ -476,7 +487,7 @@ static void FASTCOVER_tryParameters(void *opaque)
parameters, segmentFreqs);
const unsigned nbFinalizeSamples = (unsigned)(ctx->nbTrainSamples * ctx->accelParams.finalize / 100);
selection = COVER_selectDict(dict + tail, dictBufferCapacity - tail,
selection = COVER_selectDict(dict + tail, dictBufferCapacity, dictBufferCapacity - tail,
ctx->samples, ctx->samplesSizes, nbFinalizeSamples, ctx->nbTrainSamples, ctx->nbSamples, parameters, ctx->offsets,
totalCompressedSize);
@@ -607,7 +618,7 @@ ZDICT_optimizeTrainFromBuffer_fastCover(
/* constants */
const unsigned nbThreads = parameters->nbThreads;
const double splitPoint =
parameters->splitPoint <= 0.0 ? DEFAULT_SPLITPOINT : parameters->splitPoint;
parameters->splitPoint <= 0.0 ? FASTCOVER_DEFAULT_SPLITPOINT : parameters->splitPoint;
const unsigned kMinD = parameters->d == 0 ? 6 : parameters->d;
const unsigned kMaxD = parameters->d == 0 ? 8 : parameters->d;
const unsigned kMinK = parameters->k == 0 ? 50 : parameters->k;

92
vendor/github.com/DataDog/zstd/fse.h generated vendored
View File

@@ -1,35 +1,15 @@
/* ******************************************************************
FSE : Finite State Entropy codec
Public Prototypes declaration
Copyright (C) 2013-2016, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- Source repository : https://github.com/Cyan4973/FiniteStateEntropy
* FSE : Finite State Entropy codec
* Public Prototypes declaration
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
#if defined (__cplusplus)
@@ -43,7 +23,7 @@ extern "C" {
/*-*****************************************
* Dependencies
******************************************/
#include <stddef.h> /* size_t, ptrdiff_t */
#include "zstd_deps.h" /* size_t, ptrdiff_t */
/*-*****************************************
@@ -157,10 +137,16 @@ FSE_PUBLIC_API unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize
/*! FSE_normalizeCount():
normalize counts so that sum(count[]) == Power_of_2 (2^tableLog)
'normalizedCounter' is a table of short, of minimum size (maxSymbolValue+1).
useLowProbCount is a boolean parameter which trades off compressed size for
faster header decoding. When it is set to 1, the compressed data will be slightly
smaller. And when it is set to 0, FSE_readNCount() and FSE_buildDTable() will be
faster. If you are compressing a small amount of data (< 2 KB) then useLowProbCount=0
is a good default, since header deserialization makes a big speed difference.
Otherwise, useLowProbCount=1 is a good default, since the speed difference is small.
@return : tableLog,
or an errorCode, which can be tested using FSE_isError() */
FSE_PUBLIC_API size_t FSE_normalizeCount(short* normalizedCounter, unsigned tableLog,
const unsigned* count, size_t srcSize, unsigned maxSymbolValue);
const unsigned* count, size_t srcSize, unsigned maxSymbolValue, unsigned useLowProbCount);
/*! FSE_NCountWriteBound():
Provides the maximum possible size of an FSE normalized table, given 'maxSymbolValue' and 'tableLog'.
@@ -248,6 +234,13 @@ FSE_PUBLIC_API size_t FSE_readNCount (short* normalizedCounter,
unsigned* maxSymbolValuePtr, unsigned* tableLogPtr,
const void* rBuffer, size_t rBuffSize);
/*! FSE_readNCount_bmi2():
* Same as FSE_readNCount() but pass bmi2=1 when your CPU supports BMI2 and 0 otherwise.
*/
FSE_PUBLIC_API size_t FSE_readNCount_bmi2(short* normalizedCounter,
unsigned* maxSymbolValuePtr, unsigned* tableLogPtr,
const void* rBuffer, size_t rBuffSize, int bmi2);
/*! Constructor and Destructor of FSE_DTable.
Note that its size depends on 'tableLog' */
typedef unsigned FSE_DTable; /* don't allocate that. It's just a way to be more restrictive than void* */
@@ -308,12 +301,12 @@ If there is an error, the function will return an error code, which can be teste
*******************************************/
/* FSE buffer bounds */
#define FSE_NCOUNTBOUND 512
#define FSE_BLOCKBOUND(size) (size + (size>>7) + 4 /* fse states */ + sizeof(size_t) /* bitContainer */)
#define FSE_BLOCKBOUND(size) ((size) + ((size)>>7) + 4 /* fse states */ + sizeof(size_t) /* bitContainer */)
#define FSE_COMPRESSBOUND(size) (FSE_NCOUNTBOUND + FSE_BLOCKBOUND(size)) /* Macro version, useful for static allocation */
/* It is possible to statically allocate FSE CTable/DTable as a table of FSE_CTable/FSE_DTable using below macros */
#define FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) (1 + (1<<(maxTableLog-1)) + ((maxSymbolValue+1)*2))
#define FSE_DTABLE_SIZE_U32(maxTableLog) (1 + (1<<maxTableLog))
#define FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) (1 + (1<<((maxTableLog)-1)) + (((maxSymbolValue)+1)*2))
#define FSE_DTABLE_SIZE_U32(maxTableLog) (1 + (1<<(maxTableLog)))
/* or use the size to malloc() space directly. Pay attention to alignment restrictions though */
#define FSE_CTABLE_SIZE(maxTableLog, maxSymbolValue) (FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) * sizeof(FSE_CTable))
@@ -329,9 +322,9 @@ unsigned FSE_optimalTableLog_internal(unsigned maxTableLog, size_t srcSize, unsi
/* FSE_compress_wksp() :
* Same as FSE_compress2(), but using an externally allocated scratch buffer (`workSpace`).
* FSE_WKSP_SIZE_U32() provides the minimum size required for `workSpace` as a table of FSE_CTable.
* FSE_COMPRESS_WKSP_SIZE_U32() provides the minimum size required for `workSpace` as a table of FSE_CTable.
*/
#define FSE_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) ( FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) + ((maxTableLog > 12) ? (1 << (maxTableLog - 2)) : 1024) )
#define FSE_COMPRESS_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) ( FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) + ((maxTableLog > 12) ? (1 << (maxTableLog - 2)) : 1024) )
size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize);
size_t FSE_buildCTable_raw (FSE_CTable* ct, unsigned nbBits);
@@ -342,18 +335,30 @@ size_t FSE_buildCTable_rle (FSE_CTable* ct, unsigned char symbolValue);
/* FSE_buildCTable_wksp() :
* Same as FSE_buildCTable(), but using an externally allocated scratch buffer (`workSpace`).
* `wkspSize` must be >= `(1<<tableLog)`.
* `wkspSize` must be >= `FSE_BUILD_CTABLE_WORKSPACE_SIZE_U32(maxSymbolValue, tableLog)` of `unsigned`.
*/
#define FSE_BUILD_CTABLE_WORKSPACE_SIZE_U32(maxSymbolValue, tableLog) (maxSymbolValue + 2 + (1ull << (tableLog - 2)))
#define FSE_BUILD_CTABLE_WORKSPACE_SIZE(maxSymbolValue, tableLog) (sizeof(unsigned) * FSE_BUILD_CTABLE_WORKSPACE_SIZE_U32(maxSymbolValue, tableLog))
size_t FSE_buildCTable_wksp(FSE_CTable* ct, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize);
#define FSE_BUILD_DTABLE_WKSP_SIZE(maxTableLog, maxSymbolValue) (sizeof(short) * (maxSymbolValue + 1) + (1ULL << maxTableLog) + 8)
#define FSE_BUILD_DTABLE_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) ((FSE_BUILD_DTABLE_WKSP_SIZE(maxTableLog, maxSymbolValue) + sizeof(unsigned) - 1) / sizeof(unsigned))
FSE_PUBLIC_API size_t FSE_buildDTable_wksp(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize);
/**< Same as FSE_buildDTable(), using an externally allocated `workspace` produced with `FSE_BUILD_DTABLE_WKSP_SIZE_U32(maxSymbolValue)` */
size_t FSE_buildDTable_raw (FSE_DTable* dt, unsigned nbBits);
/**< build a fake FSE_DTable, designed to read a flat distribution where each symbol uses nbBits */
size_t FSE_buildDTable_rle (FSE_DTable* dt, unsigned char symbolValue);
/**< build a fake FSE_DTable, designed to always generate the same symbolValue */
size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, FSE_DTable* workSpace, unsigned maxLog);
/**< same as FSE_decompress(), using an externally allocated `workSpace` produced with `FSE_DTABLE_SIZE_U32(maxLog)` */
#define FSE_DECOMPRESS_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) (FSE_DTABLE_SIZE_U32(maxTableLog) + FSE_BUILD_DTABLE_WKSP_SIZE_U32(maxTableLog, maxSymbolValue))
#define FSE_DECOMPRESS_WKSP_SIZE(maxTableLog, maxSymbolValue) (FSE_DECOMPRESS_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) * sizeof(unsigned))
size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize);
/**< same as FSE_decompress(), using an externally allocated `workSpace` produced with `FSE_DECOMPRESS_WKSP_SIZE_U32(maxLog, maxSymbolValue)` */
size_t FSE_decompress_wksp_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize, int bmi2);
/**< Same as FSE_decompress_wksp() but with dynamic BMI2 support. Pass 1 if your CPU supports BMI2 or 0 if it doesn't. */
typedef enum {
FSE_repeat_none, /**< Cannot use the previous table */
@@ -664,6 +669,9 @@ MEM_STATIC unsigned FSE_endOfDState(const FSE_DState_t* DStatePtr)
#ifndef FSE_DEFAULT_MEMORY_USAGE
# define FSE_DEFAULT_MEMORY_USAGE 13
#endif
#if (FSE_DEFAULT_MEMORY_USAGE > FSE_MAX_MEMORY_USAGE)
# error "FSE_DEFAULT_MEMORY_USAGE must be <= FSE_MAX_MEMORY_USAGE"
#endif
/*!FSE_MAX_SYMBOL_VALUE :
* Maximum symbol value authorized.
@@ -697,7 +705,7 @@ MEM_STATIC unsigned FSE_endOfDState(const FSE_DState_t* DStatePtr)
# error "FSE_MAX_TABLELOG > FSE_TABLELOG_ABSOLUTE_MAX is not supported"
#endif
#define FSE_TABLESTEP(tableSize) ((tableSize>>1) + (tableSize>>3) + 3)
#define FSE_TABLESTEP(tableSize) (((tableSize)>>1) + ((tableSize)>>3) + 3)
#endif /* FSE_STATIC_LINKING_ONLY */

View File

@@ -1,42 +1,20 @@
/* ******************************************************************
FSE : Finite State Entropy encoder
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* FSE : Finite State Entropy encoder
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* **************************************************************
* Includes
****************************************************************/
#include <stdlib.h> /* malloc, free, qsort */
#include <string.h> /* memcpy, memset */
#include "compiler.h"
#include "mem.h" /* U32, U16, etc. */
#include "debug.h" /* assert, DEBUGLOG */
@@ -45,6 +23,9 @@
#define FSE_STATIC_LINKING_ONLY
#include "fse.h"
#include "error_private.h"
#define ZSTD_DEPS_NEED_MALLOC
#define ZSTD_DEPS_NEED_MATH64
#include "zstd_deps.h" /* ZSTD_malloc, ZSTD_free, ZSTD_memcpy, ZSTD_memset */
/* **************************************************************
@@ -94,13 +75,15 @@ size_t FSE_buildCTable_wksp(FSE_CTable* ct,
void* const FSCT = ((U32*)ptr) + 1 /* header */ + (tableLog ? tableSize>>1 : 1) ;
FSE_symbolCompressionTransform* const symbolTT = (FSE_symbolCompressionTransform*) (FSCT);
U32 const step = FSE_TABLESTEP(tableSize);
U32 cumul[FSE_MAX_SYMBOL_VALUE+2];
FSE_FUNCTION_TYPE* const tableSymbol = (FSE_FUNCTION_TYPE*)workSpace;
U32* cumul = (U32*)workSpace;
FSE_FUNCTION_TYPE* tableSymbol = (FSE_FUNCTION_TYPE*)(cumul + (maxSymbolValue + 2));
U32 highThreshold = tableSize-1;
if ((size_t)workSpace & 3) return ERROR(GENERIC); /* Must be 4 byte aligned */
if (FSE_BUILD_CTABLE_WORKSPACE_SIZE(maxSymbolValue, tableLog) > wkspSize) return ERROR(tableLog_tooLarge);
/* CTable header */
if (((size_t)1 << tableLog) * sizeof(FSE_FUNCTION_TYPE) > wkspSize) return ERROR(tableLog_tooLarge);
tableU16[-2] = (U16) tableLog;
tableU16[-1] = (U16) maxSymbolValue;
assert(tableLog < 16); /* required for threshold strategy to work */
@@ -109,7 +92,7 @@ size_t FSE_buildCTable_wksp(FSE_CTable* ct,
* http://fastcompression.blogspot.fr/2014/02/fse-distributing-symbol-values.html */
#ifdef __clang_analyzer__
memset(tableSymbol, 0, sizeof(*tableSymbol) * tableSize); /* useless initialization, just to keep scan-build happy */
ZSTD_memset(tableSymbol, 0, sizeof(*tableSymbol) * tableSize); /* useless initialization, just to keep scan-build happy */
#endif
/* symbol start positions */
@@ -188,12 +171,13 @@ size_t FSE_buildCTable_wksp(FSE_CTable* ct,
return 0;
}
#ifndef ZSTD_NO_UNUSED_FUNCTIONS
size_t FSE_buildCTable(FSE_CTable* ct, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog)
{
FSE_FUNCTION_TYPE tableSymbol[FSE_MAX_TABLESIZE]; /* memset() is not necessary, even if static analyzer complain about it */
return FSE_buildCTable_wksp(ct, normalizedCounter, maxSymbolValue, tableLog, tableSymbol, sizeof(tableSymbol));
}
#endif
@@ -327,10 +311,10 @@ FSE_CTable* FSE_createCTable (unsigned maxSymbolValue, unsigned tableLog)
size_t size;
if (tableLog > FSE_TABLELOG_ABSOLUTE_MAX) tableLog = FSE_TABLELOG_ABSOLUTE_MAX;
size = FSE_CTABLE_SIZE_U32 (tableLog, maxSymbolValue) * sizeof(U32);
return (FSE_CTable*)malloc(size);
return (FSE_CTable*)ZSTD_malloc(size);
}
void FSE_freeCTable (FSE_CTable* ct) { free(ct); }
void FSE_freeCTable (FSE_CTable* ct) { ZSTD_free(ct); }
/* provides the minimum logSize to safely represent a distribution */
static unsigned FSE_minTableLog(size_t srcSize, unsigned maxSymbolValue)
@@ -361,11 +345,10 @@ unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxS
return FSE_optimalTableLog_internal(maxTableLog, srcSize, maxSymbolValue, 2);
}
/* Secondary normalization method.
To be used when primary method fails. */
static size_t FSE_normalizeM2(short* norm, U32 tableLog, const unsigned* count, size_t total, U32 maxSymbolValue)
static size_t FSE_normalizeM2(short* norm, U32 tableLog, const unsigned* count, size_t total, U32 maxSymbolValue, short lowProbCount)
{
short const NOT_YET_ASSIGNED = -2;
U32 s;
@@ -382,7 +365,7 @@ static size_t FSE_normalizeM2(short* norm, U32 tableLog, const unsigned* count,
continue;
}
if (count[s] <= lowThreshold) {
norm[s] = -1;
norm[s] = lowProbCount;
distributed++;
total -= count[s];
continue;
@@ -434,7 +417,7 @@ static size_t FSE_normalizeM2(short* norm, U32 tableLog, const unsigned* count,
{ U64 const vStepLog = 62 - tableLog;
U64 const mid = (1ULL << (vStepLog-1)) - 1;
U64 const rStep = ((((U64)1<<vStepLog) * ToDistribute) + mid) / total; /* scale on remaining */
U64 const rStep = ZSTD_div64((((U64)1<<vStepLog) * ToDistribute) + mid, (U32)total); /* scale on remaining */
U64 tmpTotal = mid;
for (s=0; s<=maxSymbolValue; s++) {
if (norm[s]==NOT_YET_ASSIGNED) {
@@ -451,10 +434,9 @@ static size_t FSE_normalizeM2(short* norm, U32 tableLog, const unsigned* count,
return 0;
}
size_t FSE_normalizeCount (short* normalizedCounter, unsigned tableLog,
const unsigned* count, size_t total,
unsigned maxSymbolValue)
unsigned maxSymbolValue, unsigned useLowProbCount)
{
/* Sanity checks */
if (tableLog==0) tableLog = FSE_DEFAULT_TABLELOG;
@@ -463,8 +445,9 @@ size_t FSE_normalizeCount (short* normalizedCounter, unsigned tableLog,
if (tableLog < FSE_minTableLog(total, maxSymbolValue)) return ERROR(GENERIC); /* Too small tableLog, compression potentially impossible */
{ static U32 const rtbTable[] = { 0, 473195, 504333, 520860, 550000, 700000, 750000, 830000 };
short const lowProbCount = useLowProbCount ? -1 : 1;
U64 const scale = 62 - tableLog;
U64 const step = ((U64)1<<62) / total; /* <== here, one division ! */
U64 const step = ZSTD_div64((U64)1<<62, (U32)total); /* <== here, one division ! */
U64 const vStep = 1ULL<<(scale-20);
int stillToDistribute = 1<<tableLog;
unsigned s;
@@ -476,7 +459,7 @@ size_t FSE_normalizeCount (short* normalizedCounter, unsigned tableLog,
if (count[s] == total) return 0; /* rle special case */
if (count[s] == 0) { normalizedCounter[s]=0; continue; }
if (count[s] <= lowThreshold) {
normalizedCounter[s] = -1;
normalizedCounter[s] = lowProbCount;
stillToDistribute--;
} else {
short proba = (short)((count[s]*step) >> scale);
@@ -490,7 +473,7 @@ size_t FSE_normalizeCount (short* normalizedCounter, unsigned tableLog,
} }
if (-stillToDistribute >= (normalizedCounter[largest] >> 1)) {
/* corner case, need another normalization method */
size_t const errorCode = FSE_normalizeM2(normalizedCounter, tableLog, count, total, maxSymbolValue);
size_t const errorCode = FSE_normalizeM2(normalizedCounter, tableLog, count, total, maxSymbolValue, lowProbCount);
if (FSE_isError(errorCode)) return errorCode;
}
else normalizedCounter[largest] += (short)stillToDistribute;
@@ -645,9 +628,7 @@ size_t FSE_compress_usingCTable (void* dst, size_t dstSize,
size_t FSE_compressBound(size_t size) { return FSE_COMPRESSBOUND(size); }
#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return e
#define CHECK_F(f) { CHECK_V_F(_var_err__, f); }
#ifndef ZSTD_NO_UNUSED_FUNCTIONS
/* FSE_compress_wksp() :
* Same as FSE_compress2(), but using an externally allocated scratch buffer (`workSpace`).
* `wkspSize` size must be `(1<<tableLog)`.
@@ -666,7 +647,7 @@ size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t src
size_t const scratchBufferSize = wkspSize - (CTableSize * sizeof(FSE_CTable));
/* init conditions */
if (wkspSize < FSE_WKSP_SIZE_U32(tableLog, maxSymbolValue)) return ERROR(tableLog_tooLarge);
if (wkspSize < FSE_COMPRESS_WKSP_SIZE_U32(tableLog, maxSymbolValue)) return ERROR(tableLog_tooLarge);
if (srcSize <= 1) return 0; /* Not compressible */
if (!maxSymbolValue) maxSymbolValue = FSE_MAX_SYMBOL_VALUE;
if (!tableLog) tableLog = FSE_DEFAULT_TABLELOG;
@@ -679,7 +660,7 @@ size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t src
}
tableLog = FSE_optimalTableLog(tableLog, srcSize, maxSymbolValue);
CHECK_F( FSE_normalizeCount(norm, tableLog, count, srcSize, maxSymbolValue) );
CHECK_F( FSE_normalizeCount(norm, tableLog, count, srcSize, maxSymbolValue, /* useLowProbCount */ srcSize >= 2048) );
/* Write table description header */
{ CHECK_V_F(nc_err, FSE_writeNCount(op, oend-op, norm, maxSymbolValue, tableLog) );
@@ -701,13 +682,16 @@ size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t src
typedef struct {
FSE_CTable CTable_max[FSE_CTABLE_SIZE_U32(FSE_MAX_TABLELOG, FSE_MAX_SYMBOL_VALUE)];
BYTE scratchBuffer[1 << FSE_MAX_TABLELOG];
union {
U32 hist_wksp[HIST_WKSP_SIZE_U32];
BYTE scratchBuffer[1 << FSE_MAX_TABLELOG];
} workspace;
} fseWkspMax_t;
size_t FSE_compress2 (void* dst, size_t dstCapacity, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog)
{
fseWkspMax_t scratchBuffer;
DEBUG_STATIC_ASSERT(sizeof(scratchBuffer) >= FSE_WKSP_SIZE_U32(FSE_MAX_TABLELOG, FSE_MAX_SYMBOL_VALUE)); /* compilation failures here means scratchBuffer is not large enough */
DEBUG_STATIC_ASSERT(sizeof(scratchBuffer) >= FSE_COMPRESS_WKSP_SIZE_U32(FSE_MAX_TABLELOG, FSE_MAX_SYMBOL_VALUE)); /* compilation failures here means scratchBuffer is not large enough */
if (tableLog > FSE_MAX_TABLELOG) return ERROR(tableLog_tooLarge);
return FSE_compress_wksp(dst, dstCapacity, src, srcSize, maxSymbolValue, tableLog, &scratchBuffer, sizeof(scratchBuffer));
}
@@ -716,6 +700,6 @@ size_t FSE_compress (void* dst, size_t dstCapacity, const void* src, size_t srcS
{
return FSE_compress2(dst, dstCapacity, src, srcSize, FSE_MAX_SYMBOL_VALUE, FSE_DEFAULT_TABLELOG);
}
#endif
#endif /* FSE_COMMONDEFS_ONLY */

View File

@@ -1,48 +1,29 @@
/* ******************************************************************
FSE : Finite State Entropy decoder
Copyright (C) 2013-2015, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* FSE : Finite State Entropy decoder
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* **************************************************************
* Includes
****************************************************************/
#include <stdlib.h> /* malloc, free, qsort */
#include <string.h> /* memcpy, memset */
#include "debug.h" /* assert */
#include "bitstream.h"
#include "compiler.h"
#define FSE_STATIC_LINKING_ONLY
#include "fse.h"
#include "error_private.h"
#define ZSTD_DEPS_NEED_MALLOC
#include "zstd_deps.h"
/* **************************************************************
@@ -51,11 +32,6 @@
#define FSE_isError ERR_isError
#define FSE_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c) /* use only *after* variable declarations */
/* check and forward error code */
#ifndef CHECK_F
#define CHECK_F(f) { size_t const e = f; if (FSE_isError(e)) return e; }
#endif
/* **************************************************************
* Templates
@@ -84,25 +60,27 @@
FSE_DTable* FSE_createDTable (unsigned tableLog)
{
if (tableLog > FSE_TABLELOG_ABSOLUTE_MAX) tableLog = FSE_TABLELOG_ABSOLUTE_MAX;
return (FSE_DTable*)malloc( FSE_DTABLE_SIZE_U32(tableLog) * sizeof (U32) );
return (FSE_DTable*)ZSTD_malloc( FSE_DTABLE_SIZE_U32(tableLog) * sizeof (U32) );
}
void FSE_freeDTable (FSE_DTable* dt)
{
free(dt);
ZSTD_free(dt);
}
size_t FSE_buildDTable(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog)
static size_t FSE_buildDTable_internal(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize)
{
void* const tdPtr = dt+1; /* because *dt is unsigned, 32-bits aligned on 32-bits */
FSE_DECODE_TYPE* const tableDecode = (FSE_DECODE_TYPE*) (tdPtr);
U16 symbolNext[FSE_MAX_SYMBOL_VALUE+1];
U16* symbolNext = (U16*)workSpace;
BYTE* spread = (BYTE*)(symbolNext + maxSymbolValue + 1);
U32 const maxSV1 = maxSymbolValue + 1;
U32 const tableSize = 1 << tableLog;
U32 highThreshold = tableSize-1;
/* Sanity Checks */
if (FSE_BUILD_DTABLE_WKSP_SIZE(tableLog, maxSymbolValue) > wkspSize) return ERROR(maxSymbolValue_tooLarge);
if (maxSymbolValue > FSE_MAX_SYMBOL_VALUE) return ERROR(maxSymbolValue_tooLarge);
if (tableLog > FSE_MAX_TABLELOG) return ERROR(tableLog_tooLarge);
@@ -120,11 +98,57 @@ size_t FSE_buildDTable(FSE_DTable* dt, const short* normalizedCounter, unsigned
if (normalizedCounter[s] >= largeLimit) DTableH.fastMode=0;
symbolNext[s] = normalizedCounter[s];
} } }
memcpy(dt, &DTableH, sizeof(DTableH));
ZSTD_memcpy(dt, &DTableH, sizeof(DTableH));
}
/* Spread symbols */
{ U32 const tableMask = tableSize-1;
if (highThreshold == tableSize - 1) {
size_t const tableMask = tableSize-1;
size_t const step = FSE_TABLESTEP(tableSize);
/* First lay down the symbols in order.
* We use a uint64_t to lay down 8 bytes at a time. This reduces branch
* misses since small blocks generally have small table logs, so nearly
* all symbols have counts <= 8. We ensure we have 8 bytes at the end of
* our buffer to handle the over-write.
*/
{
U64 const add = 0x0101010101010101ull;
size_t pos = 0;
U64 sv = 0;
U32 s;
for (s=0; s<maxSV1; ++s, sv += add) {
int i;
int const n = normalizedCounter[s];
MEM_write64(spread + pos, sv);
for (i = 8; i < n; i += 8) {
MEM_write64(spread + pos + i, sv);
}
pos += n;
}
}
/* Now we spread those positions across the table.
* The benefit of doing it in two stages is that we avoid the the
* variable size inner loop, which caused lots of branch misses.
* Now we can run through all the positions without any branch misses.
* We unroll the loop twice, since that is what emperically worked best.
*/
{
size_t position = 0;
size_t s;
size_t const unroll = 2;
assert(tableSize % unroll == 0); /* FSE_MIN_TABLELOG is 5 */
for (s = 0; s < (size_t)tableSize; s += unroll) {
size_t u;
for (u = 0; u < unroll; ++u) {
size_t const uPosition = (position + (u * step)) & tableMask;
tableDecode[uPosition].symbol = spread[s + u];
}
position = (position + (unroll * step)) & tableMask;
}
assert(position == 0);
}
} else {
U32 const tableMask = tableSize-1;
U32 const step = FSE_TABLESTEP(tableSize);
U32 s, position = 0;
for (s=0; s<maxSV1; s++) {
@@ -149,6 +173,11 @@ size_t FSE_buildDTable(FSE_DTable* dt, const short* normalizedCounter, unsigned
return 0;
}
size_t FSE_buildDTable_wksp(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize)
{
return FSE_buildDTable_internal(dt, normalizedCounter, maxSymbolValue, tableLog, workSpace, wkspSize);
}
#ifndef FSE_COMMONDEFS_ONLY
@@ -276,36 +305,89 @@ size_t FSE_decompress_usingDTable(void* dst, size_t originalSize,
}
size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, FSE_DTable* workSpace, unsigned maxLog)
size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize)
{
return FSE_decompress_wksp_bmi2(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, /* bmi2 */ 0);
}
FORCE_INLINE_TEMPLATE size_t FSE_decompress_wksp_body(
void* dst, size_t dstCapacity,
const void* cSrc, size_t cSrcSize,
unsigned maxLog, void* workSpace, size_t wkspSize,
int bmi2)
{
const BYTE* const istart = (const BYTE*)cSrc;
const BYTE* ip = istart;
short counting[FSE_MAX_SYMBOL_VALUE+1];
unsigned tableLog;
unsigned maxSymbolValue = FSE_MAX_SYMBOL_VALUE;
FSE_DTable* const dtable = (FSE_DTable*)workSpace;
/* normal FSE decoding mode */
size_t const NCountLength = FSE_readNCount (counting, &maxSymbolValue, &tableLog, istart, cSrcSize);
size_t const NCountLength = FSE_readNCount_bmi2(counting, &maxSymbolValue, &tableLog, istart, cSrcSize, bmi2);
if (FSE_isError(NCountLength)) return NCountLength;
//if (NCountLength >= cSrcSize) return ERROR(srcSize_wrong); /* too small input size; supposed to be already checked in NCountLength, only remaining case : NCountLength==cSrcSize */
if (tableLog > maxLog) return ERROR(tableLog_tooLarge);
assert(NCountLength <= cSrcSize);
ip += NCountLength;
cSrcSize -= NCountLength;
CHECK_F( FSE_buildDTable (workSpace, counting, maxSymbolValue, tableLog) );
if (FSE_DECOMPRESS_WKSP_SIZE(tableLog, maxSymbolValue) > wkspSize) return ERROR(tableLog_tooLarge);
workSpace = dtable + FSE_DTABLE_SIZE_U32(tableLog);
wkspSize -= FSE_DTABLE_SIZE(tableLog);
return FSE_decompress_usingDTable (dst, dstCapacity, ip, cSrcSize, workSpace); /* always return, even if it is an error code */
CHECK_F( FSE_buildDTable_internal(dtable, counting, maxSymbolValue, tableLog, workSpace, wkspSize) );
{
const void* ptr = dtable;
const FSE_DTableHeader* DTableH = (const FSE_DTableHeader*)ptr;
const U32 fastMode = DTableH->fastMode;
/* select fast mode (static) */
if (fastMode) return FSE_decompress_usingDTable_generic(dst, dstCapacity, ip, cSrcSize, dtable, 1);
return FSE_decompress_usingDTable_generic(dst, dstCapacity, ip, cSrcSize, dtable, 0);
}
}
/* Avoids the FORCE_INLINE of the _body() function. */
static size_t FSE_decompress_wksp_body_default(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize)
{
return FSE_decompress_wksp_body(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, 0);
}
#if DYNAMIC_BMI2
TARGET_ATTRIBUTE("bmi2") static size_t FSE_decompress_wksp_body_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize)
{
return FSE_decompress_wksp_body(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize, 1);
}
#endif
size_t FSE_decompress_wksp_bmi2(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, unsigned maxLog, void* workSpace, size_t wkspSize, int bmi2)
{
#if DYNAMIC_BMI2
if (bmi2) {
return FSE_decompress_wksp_body_bmi2(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize);
}
#endif
(void)bmi2;
return FSE_decompress_wksp_body_default(dst, dstCapacity, cSrc, cSrcSize, maxLog, workSpace, wkspSize);
}
typedef FSE_DTable DTable_max_t[FSE_DTABLE_SIZE_U32(FSE_MAX_TABLELOG)];
size_t FSE_decompress(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize)
{
DTable_max_t dt; /* Static analyzer seems unable to understand this table will be properly initialized later */
return FSE_decompress_wksp(dst, dstCapacity, cSrc, cSrcSize, dt, FSE_MAX_TABLELOG);
#ifndef ZSTD_NO_UNUSED_FUNCTIONS
size_t FSE_buildDTable(FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog) {
U32 wksp[FSE_BUILD_DTABLE_WKSP_SIZE_U32(FSE_TABLELOG_ABSOLUTE_MAX, FSE_MAX_SYMBOL_VALUE)];
return FSE_buildDTable_wksp(dt, normalizedCounter, maxSymbolValue, tableLog, wksp, sizeof(wksp));
}
size_t FSE_decompress(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize)
{
/* Static analyzer seems unable to understand this table will be properly initialized later */
U32 wksp[FSE_DECOMPRESS_WKSP_SIZE_U32(FSE_MAX_TABLELOG, FSE_MAX_SYMBOL_VALUE)];
return FSE_decompress_wksp(dst, dstCapacity, cSrc, cSrcSize, FSE_MAX_TABLELOG, wksp, sizeof(wksp));
}
#endif
#endif /* FSE_COMMONDEFS_ONLY */

3
vendor/github.com/DataDog/zstd/go.mod generated vendored Normal file
View File

@@ -0,0 +1,3 @@
module github.com/DataDog/zstd
go 1.14

View File

@@ -1,36 +1,16 @@
/* ******************************************************************
hist : Histogram functions
part of Finite State Entropy project
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* hist : Histogram functions
* part of Finite State Entropy project
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* --- dependencies --- */
@@ -54,7 +34,7 @@ unsigned HIST_count_simple(unsigned* count, unsigned* maxSymbolValuePtr,
unsigned maxSymbolValue = *maxSymbolValuePtr;
unsigned largestCount=0;
memset(count, 0, (maxSymbolValue+1) * sizeof(*count));
ZSTD_memset(count, 0, (maxSymbolValue+1) * sizeof(*count));
if (srcSize==0) { *maxSymbolValuePtr = 0; return 0; }
while (ip<end) {
@@ -80,9 +60,9 @@ typedef enum { trustInput, checkMaxSymbolValue } HIST_checkInput_e;
* this design makes better use of OoO cpus,
* and is noticeably faster when some values are heavily repeated.
* But it needs some additional workspace for intermediate tables.
* `workSpace` size must be a table of size >= HIST_WKSP_SIZE_U32.
* `workSpace` must be a U32 table of size >= HIST_WKSP_SIZE_U32.
* @return : largest histogram frequency,
* or an error code (notably when histogram would be larger than *maxSymbolValuePtr). */
* or an error code (notably when histogram's alphabet is larger than *maxSymbolValuePtr) */
static size_t HIST_count_parallel_wksp(
unsigned* count, unsigned* maxSymbolValuePtr,
const void* source, size_t sourceSize,
@@ -91,22 +71,21 @@ static size_t HIST_count_parallel_wksp(
{
const BYTE* ip = (const BYTE*)source;
const BYTE* const iend = ip+sourceSize;
unsigned maxSymbolValue = *maxSymbolValuePtr;
size_t const countSize = (*maxSymbolValuePtr + 1) * sizeof(*count);
unsigned max=0;
U32* const Counting1 = workSpace;
U32* const Counting2 = Counting1 + 256;
U32* const Counting3 = Counting2 + 256;
U32* const Counting4 = Counting3 + 256;
memset(workSpace, 0, 4*256*sizeof(unsigned));
/* safety checks */
assert(*maxSymbolValuePtr <= 255);
if (!sourceSize) {
memset(count, 0, maxSymbolValue + 1);
ZSTD_memset(count, 0, countSize);
*maxSymbolValuePtr = 0;
return 0;
}
if (!maxSymbolValue) maxSymbolValue = 255; /* 0 == default */
ZSTD_memset(workSpace, 0, 4*256*sizeof(unsigned));
/* by stripes of 16 bytes */
{ U32 cached = MEM_read32(ip); ip += 4;
@@ -138,21 +117,18 @@ static size_t HIST_count_parallel_wksp(
/* finish last symbols */
while (ip<iend) Counting1[*ip++]++;
if (check) { /* verify stats will fit into destination table */
U32 s; for (s=255; s>maxSymbolValue; s--) {
Counting1[s] += Counting2[s] + Counting3[s] + Counting4[s];
if (Counting1[s]) return ERROR(maxSymbolValue_tooSmall);
} }
{ U32 s;
if (maxSymbolValue > 255) maxSymbolValue = 255;
for (s=0; s<=maxSymbolValue; s++) {
count[s] = Counting1[s] + Counting2[s] + Counting3[s] + Counting4[s];
if (count[s] > max) max = count[s];
for (s=0; s<256; s++) {
Counting1[s] += Counting2[s] + Counting3[s] + Counting4[s];
if (Counting1[s] > max) max = Counting1[s];
} }
while (!count[maxSymbolValue]) maxSymbolValue--;
*maxSymbolValuePtr = maxSymbolValue;
{ unsigned maxSymbolValue = 255;
while (!Counting1[maxSymbolValue]) maxSymbolValue--;
if (check && maxSymbolValue > *maxSymbolValuePtr) return ERROR(maxSymbolValue_tooSmall);
*maxSymbolValuePtr = maxSymbolValue;
ZSTD_memmove(count, Counting1, countSize); /* in case count & Counting1 are overlapping */
}
return (size_t)max;
}
@@ -172,14 +148,6 @@ size_t HIST_countFast_wksp(unsigned* count, unsigned* maxSymbolValuePtr,
return HIST_count_parallel_wksp(count, maxSymbolValuePtr, source, sourceSize, trustInput, (U32*)workSpace);
}
/* fast variant (unsafe : won't check if src contains values beyond count[] limit) */
size_t HIST_countFast(unsigned* count, unsigned* maxSymbolValuePtr,
const void* source, size_t sourceSize)
{
unsigned tmpCounters[HIST_WKSP_SIZE_U32];
return HIST_countFast_wksp(count, maxSymbolValuePtr, source, sourceSize, tmpCounters, sizeof(tmpCounters));
}
/* HIST_count_wksp() :
* Same as HIST_count(), but using an externally provided scratch buffer.
* `workSpace` size must be table of >= HIST_WKSP_SIZE_U32 unsigned */
@@ -195,9 +163,19 @@ size_t HIST_count_wksp(unsigned* count, unsigned* maxSymbolValuePtr,
return HIST_countFast_wksp(count, maxSymbolValuePtr, source, sourceSize, workSpace, workSpaceSize);
}
#ifndef ZSTD_NO_UNUSED_FUNCTIONS
/* fast variant (unsafe : won't check if src contains values beyond count[] limit) */
size_t HIST_countFast(unsigned* count, unsigned* maxSymbolValuePtr,
const void* source, size_t sourceSize)
{
unsigned tmpCounters[HIST_WKSP_SIZE_U32];
return HIST_countFast_wksp(count, maxSymbolValuePtr, source, sourceSize, tmpCounters, sizeof(tmpCounters));
}
size_t HIST_count(unsigned* count, unsigned* maxSymbolValuePtr,
const void* src, size_t srcSize)
{
unsigned tmpCounters[HIST_WKSP_SIZE_U32];
return HIST_count_wksp(count, maxSymbolValuePtr, src, srcSize, tmpCounters, sizeof(tmpCounters));
}
#endif

View File

@@ -1,40 +1,20 @@
/* ******************************************************************
hist : Histogram functions
part of Finite State Entropy project
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* hist : Histogram functions
* part of Finite State Entropy project
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* --- dependencies --- */
#include <stddef.h> /* size_t */
#include "zstd_deps.h" /* size_t */
/* --- simple histogram functions --- */

79
vendor/github.com/DataDog/zstd/huf.h generated vendored
View File

@@ -1,35 +1,15 @@
/* ******************************************************************
huff0 huffman codec,
part of Finite State Entropy library
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- Source repository : https://github.com/Cyan4973/FiniteStateEntropy
* huff0 huffman codec,
* part of Finite State Entropy library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
#if defined (__cplusplus)
@@ -40,7 +20,7 @@ extern "C" {
#define HUF_H_298734234
/* *** Dependencies *** */
#include <stddef.h> /* size_t */
#include "zstd_deps.h" /* size_t */
/* *** library symbols visibility *** */
@@ -110,7 +90,7 @@ HUF_PUBLIC_API size_t HUF_compress2 (void* dst, size_t dstCapacity,
/** HUF_compress4X_wksp() :
* Same as HUF_compress2(), but uses externally allocated `workSpace`.
* `workspace` must have minimum alignment of 4, and be at least as large as HUF_WORKSPACE_SIZE */
#define HUF_WORKSPACE_SIZE (6 << 10)
#define HUF_WORKSPACE_SIZE ((6 << 10) + 256)
#define HUF_WORKSPACE_SIZE_U32 (HUF_WORKSPACE_SIZE / sizeof(U32))
HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
@@ -131,6 +111,8 @@ HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity,
/* *** Dependencies *** */
#include "mem.h" /* U32 */
#define FSE_STATIC_LINKING_ONLY
#include "fse.h"
/* *** Constants *** */
@@ -153,12 +135,16 @@ HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity,
#define HUF_COMPRESSBOUND(size) (HUF_CTABLEBOUND + HUF_BLOCKBOUND(size)) /* Macro version, useful for static allocation */
/* static allocation of HUF's Compression Table */
/* this is a private definition, just exposed for allocation and strict aliasing purpose. never EVER access its members directly */
struct HUF_CElt_s {
U16 val;
BYTE nbBits;
}; /* typedef'd to HUF_CElt */
typedef struct HUF_CElt_s HUF_CElt; /* consider it an incomplete type */
#define HUF_CTABLE_SIZE_U32(maxSymbolValue) ((maxSymbolValue)+1) /* Use tables of U32, for proper alignment */
#define HUF_CTABLE_SIZE(maxSymbolValue) (HUF_CTABLE_SIZE_U32(maxSymbolValue) * sizeof(U32))
#define HUF_CREATE_STATIC_CTABLE(name, maxSymbolValue) \
U32 name##hb[HUF_CTABLE_SIZE_U32(maxSymbolValue)]; \
void* name##hv = &(name##hb); \
HUF_CElt* name = (HUF_CElt*)(name##hv) /* no final ; */
HUF_CElt name[HUF_CTABLE_SIZE_U32(maxSymbolValue)] /* no final ; */
/* static allocation of HUF's DTable */
typedef U32 HUF_DTable;
@@ -204,10 +190,11 @@ size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize,
* or to save and regenerate 'CTable' using external methods.
*/
unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue);
typedef struct HUF_CElt_s HUF_CElt; /* incomplete type */
size_t HUF_buildCTable (HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits); /* @return : maxNbBits; CTable and count can overlap. In which case, CTable will overwrite count content */
size_t HUF_writeCTable (void* dst, size_t maxDstSize, const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog);
size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable);
size_t HUF_estimateCompressedSize(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue);
int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue);
typedef enum {
HUF_repeat_none, /**< Cannot use the previous table */
@@ -244,9 +231,22 @@ size_t HUF_readStats(BYTE* huffWeight, size_t hwSize,
U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr,
const void* src, size_t srcSize);
/*! HUF_readStats_wksp() :
* Same as HUF_readStats() but takes an external workspace which must be
* 4-byte aligned and its size must be >= HUF_READ_STATS_WORKSPACE_SIZE.
* If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0.
*/
#define HUF_READ_STATS_WORKSPACE_SIZE_U32 FSE_DECOMPRESS_WKSP_SIZE_U32(6, HUF_TABLELOG_MAX-1)
#define HUF_READ_STATS_WORKSPACE_SIZE (HUF_READ_STATS_WORKSPACE_SIZE_U32 * sizeof(unsigned))
size_t HUF_readStats_wksp(BYTE* huffWeight, size_t hwSize,
U32* rankStats, U32* nbSymbolsPtr, U32* tableLogPtr,
const void* src, size_t srcSize,
void* workspace, size_t wkspSize,
int bmi2);
/** HUF_readCTable() :
* Loading a CTable saved with HUF_writeCTable() */
size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize);
size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize, unsigned *hasZeroWeights);
/** HUF_getNbBits() :
* Read nbBits from CTable symbolTable, for symbol `symbolValue` presumed <= HUF_SYMBOLVALUE_MAX
@@ -350,6 +350,9 @@ size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstS
#endif
size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2);
size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2);
#ifndef HUF_FORCE_DECOMPRESS_X2
size_t HUF_readDTableX1_wksp_bmi2(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int bmi2);
#endif
#endif /* HUF_STATIC_LINKING_ONLY */

View File

@@ -1,35 +1,15 @@
/* ******************************************************************
Huffman encoder, part of New Generation Entropy library
Copyright (C) 2013-2016, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* Huffman encoder, part of New Generation Entropy library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* **************************************************************
@@ -43,8 +23,7 @@
/* **************************************************************
* Includes
****************************************************************/
#include <string.h> /* memcpy, memset */
#include <stdio.h> /* printf (debug) */
#include "zstd_deps.h" /* ZSTD_memcpy, ZSTD_memset */
#include "compiler.h"
#include "bitstream.h"
#include "hist.h"
@@ -60,8 +39,6 @@
****************************************************************/
#define HUF_isError ERR_isError
#define HUF_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c) /* use only *after* variable declarations */
#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return e
#define CHECK_F(f) { CHECK_V_F(_var_err__, f); }
/* **************************************************************
@@ -92,7 +69,7 @@ static size_t HUF_compressWeights (void* dst, size_t dstSize, const void* weight
U32 tableLog = MAX_FSE_TABLELOG_FOR_HUFF_HEADER;
FSE_CTable CTable[FSE_CTABLE_SIZE_U32(MAX_FSE_TABLELOG_FOR_HUFF_HEADER, HUF_TABLELOG_MAX)];
BYTE scratchBuffer[1<<MAX_FSE_TABLELOG_FOR_HUFF_HEADER];
U32 scratchBuffer[FSE_BUILD_CTABLE_WORKSPACE_SIZE_U32(HUF_TABLELOG_MAX, MAX_FSE_TABLELOG_FOR_HUFF_HEADER)];
unsigned count[HUF_TABLELOG_MAX+1];
S16 norm[HUF_TABLELOG_MAX+1];
@@ -107,29 +84,24 @@ static size_t HUF_compressWeights (void* dst, size_t dstSize, const void* weight
}
tableLog = FSE_optimalTableLog(tableLog, wtSize, maxSymbolValue);
CHECK_F( FSE_normalizeCount(norm, tableLog, count, wtSize, maxSymbolValue) );
CHECK_F( FSE_normalizeCount(norm, tableLog, count, wtSize, maxSymbolValue, /* useLowProbCount */ 0) );
/* Write table description header */
{ CHECK_V_F(hSize, FSE_writeNCount(op, oend-op, norm, maxSymbolValue, tableLog) );
{ CHECK_V_F(hSize, FSE_writeNCount(op, (size_t)(oend-op), norm, maxSymbolValue, tableLog) );
op += hSize;
}
/* Compress */
CHECK_F( FSE_buildCTable_wksp(CTable, norm, maxSymbolValue, tableLog, scratchBuffer, sizeof(scratchBuffer)) );
{ CHECK_V_F(cSize, FSE_compress_usingCTable(op, oend - op, weightTable, wtSize, CTable) );
{ CHECK_V_F(cSize, FSE_compress_usingCTable(op, (size_t)(oend - op), weightTable, wtSize, CTable) );
if (cSize == 0) return 0; /* not enough space for compressed data */
op += cSize;
}
return op-ostart;
return (size_t)(op-ostart);
}
struct HUF_CElt_s {
U16 val;
BYTE nbBits;
}; /* typedef'd to HUF_CElt within "huf.h" */
/*! HUF_writeCTable() :
`CTable` : Huffman tree to save, using huf representation.
@return : size of saved CTable */
@@ -169,7 +141,7 @@ size_t HUF_writeCTable (void* dst, size_t maxDstSize,
}
size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize)
size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize, unsigned* hasZeroWeights)
{
BYTE huffWeight[HUF_SYMBOLVALUE_MAX + 1]; /* init not required, even though some static analyzer may complain */
U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1]; /* large enough for values from 0 to 16 */
@@ -178,6 +150,7 @@ size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void
/* get symbol weights */
CHECK_V_F(readSize, HUF_readStats(huffWeight, HUF_SYMBOLVALUE_MAX+1, rankVal, &nbSymbols, &tableLog, src, srcSize));
*hasZeroWeights = (rankVal[0] > 0);
/* check result */
if (tableLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge);
@@ -186,15 +159,15 @@ size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void
/* Prepare base value per rank */
{ U32 n, nextRankStart = 0;
for (n=1; n<=tableLog; n++) {
U32 current = nextRankStart;
U32 curr = nextRankStart;
nextRankStart += (rankVal[n] << (n-1));
rankVal[n] = current;
rankVal[n] = curr;
} }
/* fill nbBits */
{ U32 n; for (n=0; n<nbSymbols; n++) {
const U32 w = huffWeight[n];
CTable[n].nbBits = (BYTE)(tableLog + 1 - w);
CTable[n].nbBits = (BYTE)(tableLog + 1 - w) & -(w != 0);
} }
/* fill val */
@@ -232,106 +205,202 @@ typedef struct nodeElt_s {
BYTE nbBits;
} nodeElt;
/**
* HUF_setMaxHeight():
* Enforces maxNbBits on the Huffman tree described in huffNode.
*
* It sets all nodes with nbBits > maxNbBits to be maxNbBits. Then it adjusts
* the tree to so that it is a valid canonical Huffman tree.
*
* @pre The sum of the ranks of each symbol == 2^largestBits,
* where largestBits == huffNode[lastNonNull].nbBits.
* @post The sum of the ranks of each symbol == 2^largestBits,
* where largestBits is the return value <= maxNbBits.
*
* @param huffNode The Huffman tree modified in place to enforce maxNbBits.
* @param lastNonNull The symbol with the lowest count in the Huffman tree.
* @param maxNbBits The maximum allowed number of bits, which the Huffman tree
* may not respect. After this function the Huffman tree will
* respect maxNbBits.
* @return The maximum number of bits of the Huffman tree after adjustment,
* necessarily no more than maxNbBits.
*/
static U32 HUF_setMaxHeight(nodeElt* huffNode, U32 lastNonNull, U32 maxNbBits)
{
const U32 largestBits = huffNode[lastNonNull].nbBits;
if (largestBits <= maxNbBits) return largestBits; /* early exit : no elt > maxNbBits */
/* early exit : no elt > maxNbBits, so the tree is already valid. */
if (largestBits <= maxNbBits) return largestBits;
/* there are several too large elements (at least >= 2) */
{ int totalCost = 0;
const U32 baseCost = 1 << (largestBits - maxNbBits);
U32 n = lastNonNull;
int n = (int)lastNonNull;
/* Adjust any ranks > maxNbBits to maxNbBits.
* Compute totalCost, which is how far the sum of the ranks is
* we are over 2^largestBits after adjust the offending ranks.
*/
while (huffNode[n].nbBits > maxNbBits) {
totalCost += baseCost - (1 << (largestBits - huffNode[n].nbBits));
huffNode[n].nbBits = (BYTE)maxNbBits;
n --;
} /* n stops at huffNode[n].nbBits <= maxNbBits */
while (huffNode[n].nbBits == maxNbBits) n--; /* n end at index of smallest symbol using < maxNbBits */
n--;
}
/* n stops at huffNode[n].nbBits <= maxNbBits */
assert(huffNode[n].nbBits <= maxNbBits);
/* n end at index of smallest symbol using < maxNbBits */
while (huffNode[n].nbBits == maxNbBits) --n;
/* renorm totalCost */
totalCost >>= (largestBits - maxNbBits); /* note : totalCost is necessarily a multiple of baseCost */
/* renorm totalCost from 2^largestBits to 2^maxNbBits
* note : totalCost is necessarily a multiple of baseCost */
assert((totalCost & (baseCost - 1)) == 0);
totalCost >>= (largestBits - maxNbBits);
assert(totalCost > 0);
/* repay normalized cost */
{ U32 const noSymbol = 0xF0F0F0F0;
U32 rankLast[HUF_TABLELOG_MAX+2];
int pos;
/* Get pos of last (smallest) symbol per rank */
memset(rankLast, 0xF0, sizeof(rankLast));
/* Get pos of last (smallest = lowest cum. count) symbol per rank */
ZSTD_memset(rankLast, 0xF0, sizeof(rankLast));
{ U32 currentNbBits = maxNbBits;
int pos;
for (pos=n ; pos >= 0; pos--) {
if (huffNode[pos].nbBits >= currentNbBits) continue;
currentNbBits = huffNode[pos].nbBits; /* < maxNbBits */
rankLast[maxNbBits-currentNbBits] = pos;
rankLast[maxNbBits-currentNbBits] = (U32)pos;
} }
while (totalCost > 0) {
U32 nBitsToDecrease = BIT_highbit32(totalCost) + 1;
/* Try to reduce the next power of 2 above totalCost because we
* gain back half the rank.
*/
U32 nBitsToDecrease = BIT_highbit32((U32)totalCost) + 1;
for ( ; nBitsToDecrease > 1; nBitsToDecrease--) {
U32 highPos = rankLast[nBitsToDecrease];
U32 lowPos = rankLast[nBitsToDecrease-1];
U32 const highPos = rankLast[nBitsToDecrease];
U32 const lowPos = rankLast[nBitsToDecrease-1];
if (highPos == noSymbol) continue;
/* Decrease highPos if no symbols of lowPos or if it is
* not cheaper to remove 2 lowPos than highPos.
*/
if (lowPos == noSymbol) break;
{ U32 const highTotal = huffNode[highPos].count;
U32 const lowTotal = 2 * huffNode[lowPos].count;
if (highTotal <= lowTotal) break;
} }
/* only triggered when no more rank 1 symbol left => find closest one (note : there is necessarily at least one !) */
assert(rankLast[nBitsToDecrease] != noSymbol || nBitsToDecrease == 1);
/* HUF_MAX_TABLELOG test just to please gcc 5+; but it should not be necessary */
while ((nBitsToDecrease<=HUF_TABLELOG_MAX) && (rankLast[nBitsToDecrease] == noSymbol))
nBitsToDecrease ++;
nBitsToDecrease++;
assert(rankLast[nBitsToDecrease] != noSymbol);
/* Increase the number of bits to gain back half the rank cost. */
totalCost -= 1 << (nBitsToDecrease-1);
huffNode[rankLast[nBitsToDecrease]].nbBits++;
/* Fix up the new rank.
* If the new rank was empty, this symbol is now its smallest.
* Otherwise, this symbol will be the largest in the new rank so no adjustment.
*/
if (rankLast[nBitsToDecrease-1] == noSymbol)
rankLast[nBitsToDecrease-1] = rankLast[nBitsToDecrease]; /* this rank is no longer empty */
huffNode[rankLast[nBitsToDecrease]].nbBits ++;
rankLast[nBitsToDecrease-1] = rankLast[nBitsToDecrease];
/* Fix up the old rank.
* If the symbol was at position 0, meaning it was the highest weight symbol in the tree,
* it must be the only symbol in its rank, so the old rank now has no symbols.
* Otherwise, since the Huffman nodes are sorted by count, the previous position is now
* the smallest node in the rank. If the previous position belongs to a different rank,
* then the rank is now empty.
*/
if (rankLast[nBitsToDecrease] == 0) /* special case, reached largest symbol */
rankLast[nBitsToDecrease] = noSymbol;
else {
rankLast[nBitsToDecrease]--;
if (huffNode[rankLast[nBitsToDecrease]].nbBits != maxNbBits-nBitsToDecrease)
rankLast[nBitsToDecrease] = noSymbol; /* this rank is now empty */
} } /* while (totalCost > 0) */
}
} /* while (totalCost > 0) */
/* If we've removed too much weight, then we have to add it back.
* To avoid overshooting again, we only adjust the smallest rank.
* We take the largest nodes from the lowest rank 0 and move them
* to rank 1. There's guaranteed to be enough rank 0 symbols because
* TODO.
*/
while (totalCost < 0) { /* Sometimes, cost correction overshoot */
if (rankLast[1] == noSymbol) { /* special case : no rank 1 symbol (using maxNbBits-1); let's create one from largest rank 0 (using maxNbBits) */
/* special case : no rank 1 symbol (using maxNbBits-1);
* let's create one from largest rank 0 (using maxNbBits).
*/
if (rankLast[1] == noSymbol) {
while (huffNode[n].nbBits == maxNbBits) n--;
huffNode[n+1].nbBits--;
rankLast[1] = n+1;
assert(n >= 0);
rankLast[1] = (U32)(n+1);
totalCost++;
continue;
}
huffNode[ rankLast[1] + 1 ].nbBits--;
rankLast[1]++;
totalCost ++;
} } } /* there are several too large elements (at least >= 2) */
}
} /* repay normalized cost */
} /* there are several too large elements (at least >= 2) */
return maxNbBits;
}
typedef struct {
U32 base;
U32 current;
U32 curr;
} rankPos;
static void HUF_sort(nodeElt* huffNode, const unsigned* count, U32 maxSymbolValue)
{
rankPos rank[32];
U32 n;
typedef nodeElt huffNodeTable[HUF_CTABLE_WORKSPACE_SIZE_U32];
memset(rank, 0, sizeof(rank));
for (n=0; n<=maxSymbolValue; n++) {
U32 r = BIT_highbit32(count[n] + 1);
rank[r].base ++;
#define RANK_POSITION_TABLE_SIZE 32
typedef struct {
huffNodeTable huffNodeTbl;
rankPos rankPosition[RANK_POSITION_TABLE_SIZE];
} HUF_buildCTable_wksp_tables;
/**
* HUF_sort():
* Sorts the symbols [0, maxSymbolValue] by count[symbol] in decreasing order.
*
* @param[out] huffNode Sorted symbols by decreasing count. Only members `.count` and `.byte` are filled.
* Must have (maxSymbolValue + 1) entries.
* @param[in] count Histogram of the symbols.
* @param[in] maxSymbolValue Maximum symbol value.
* @param rankPosition This is a scratch workspace. Must have RANK_POSITION_TABLE_SIZE entries.
*/
static void HUF_sort(nodeElt* huffNode, const unsigned* count, U32 maxSymbolValue, rankPos* rankPosition)
{
int n;
int const maxSymbolValue1 = (int)maxSymbolValue + 1;
/* Compute base and set curr to base.
* For symbol s let lowerRank = BIT_highbit32(count[n]+1) and rank = lowerRank + 1.
* Then 2^lowerRank <= count[n]+1 <= 2^rank.
* We attribute each symbol to lowerRank's base value, because we want to know where
* each rank begins in the output, so for rank R we want to count ranks R+1 and above.
*/
ZSTD_memset(rankPosition, 0, sizeof(*rankPosition) * RANK_POSITION_TABLE_SIZE);
for (n = 0; n < maxSymbolValue1; ++n) {
U32 lowerRank = BIT_highbit32(count[n] + 1);
rankPosition[lowerRank].base++;
}
for (n=30; n>0; n--) rank[n-1].base += rank[n].base;
for (n=0; n<32; n++) rank[n].current = rank[n].base;
for (n=0; n<=maxSymbolValue; n++) {
assert(rankPosition[RANK_POSITION_TABLE_SIZE - 1].base == 0);
for (n = RANK_POSITION_TABLE_SIZE - 1; n > 0; --n) {
rankPosition[n-1].base += rankPosition[n].base;
rankPosition[n-1].curr = rankPosition[n-1].base;
}
/* Sort */
for (n = 0; n < maxSymbolValue1; ++n) {
U32 const c = count[n];
U32 const r = BIT_highbit32(c+1) + 1;
U32 pos = rank[r].current++;
while ((pos > rank[r].base) && (c > huffNode[pos-1].count)) {
U32 pos = rankPosition[r].curr++;
/* Insert into the correct position in the rank.
* We have at most 256 symbols, so this insertion should be fine.
*/
while ((pos > rankPosition[r].base) && (c > huffNode[pos-1].count)) {
huffNode[pos] = huffNode[pos-1];
pos--;
}
@@ -343,45 +412,40 @@ static void HUF_sort(nodeElt* huffNode, const unsigned* count, U32 maxSymbolValu
/** HUF_buildCTable_wksp() :
* Same as HUF_buildCTable(), but using externally allocated scratch buffer.
* `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as a table of HUF_CTABLE_WORKSPACE_SIZE_U32 unsigned.
* `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as sizeof(HUF_buildCTable_wksp_tables).
*/
#define STARTNODE (HUF_SYMBOLVALUE_MAX+1)
typedef nodeElt huffNodeTable[HUF_CTABLE_WORKSPACE_SIZE_U32];
size_t HUF_buildCTable_wksp (HUF_CElt* tree, const unsigned* count, U32 maxSymbolValue, U32 maxNbBits, void* workSpace, size_t wkspSize)
/* HUF_buildTree():
* Takes the huffNode array sorted by HUF_sort() and builds an unlimited-depth Huffman tree.
*
* @param huffNode The array sorted by HUF_sort(). Builds the Huffman tree in this array.
* @param maxSymbolValue The maximum symbol value.
* @return The smallest node in the Huffman tree (by count).
*/
static int HUF_buildTree(nodeElt* huffNode, U32 maxSymbolValue)
{
nodeElt* const huffNode0 = (nodeElt*)workSpace;
nodeElt* const huffNode = huffNode0+1;
U32 n, nonNullRank;
nodeElt* const huffNode0 = huffNode - 1;
int nonNullRank;
int lowS, lowN;
U16 nodeNb = STARTNODE;
U32 nodeRoot;
/* safety checks */
if (((size_t)workSpace & 3) != 0) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */
if (wkspSize < sizeof(huffNodeTable)) return ERROR(workSpace_tooSmall);
if (maxNbBits == 0) maxNbBits = HUF_TABLELOG_DEFAULT;
if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(maxSymbolValue_tooLarge);
memset(huffNode0, 0, sizeof(huffNodeTable));
/* sort, decreasing order */
HUF_sort(huffNode, count, maxSymbolValue);
int nodeNb = STARTNODE;
int n, nodeRoot;
/* init for parents */
nonNullRank = maxSymbolValue;
nonNullRank = (int)maxSymbolValue;
while(huffNode[nonNullRank].count == 0) nonNullRank--;
lowS = nonNullRank; nodeRoot = nodeNb + lowS - 1; lowN = nodeNb;
huffNode[nodeNb].count = huffNode[lowS].count + huffNode[lowS-1].count;
huffNode[lowS].parent = huffNode[lowS-1].parent = nodeNb;
huffNode[lowS].parent = huffNode[lowS-1].parent = (U16)nodeNb;
nodeNb++; lowS-=2;
for (n=nodeNb; n<=nodeRoot; n++) huffNode[n].count = (U32)(1U<<30);
huffNode0[0].count = (U32)(1U<<31); /* fake entry, strong barrier */
/* create parents */
while (nodeNb <= nodeRoot) {
U32 n1 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
U32 n2 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
int const n1 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
int const n2 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
huffNode[nodeNb].count = huffNode[n1].count + huffNode[n2].count;
huffNode[n1].parent = huffNode[n2].parent = nodeNb;
huffNode[n1].parent = huffNode[n2].parent = (U16)nodeNb;
nodeNb++;
}
@@ -392,42 +456,73 @@ size_t HUF_buildCTable_wksp (HUF_CElt* tree, const unsigned* count, U32 maxSymbo
for (n=0; n<=nonNullRank; n++)
huffNode[n].nbBits = huffNode[ huffNode[n].parent ].nbBits + 1;
/* enforce maxTableLog */
maxNbBits = HUF_setMaxHeight(huffNode, nonNullRank, maxNbBits);
return nonNullRank;
}
/* fill result into tree (val, nbBits) */
{ U16 nbPerRank[HUF_TABLELOG_MAX+1] = {0};
U16 valPerRank[HUF_TABLELOG_MAX+1] = {0};
if (maxNbBits > HUF_TABLELOG_MAX) return ERROR(GENERIC); /* check fit into table */
for (n=0; n<=nonNullRank; n++)
nbPerRank[huffNode[n].nbBits]++;
/* determine stating value per rank */
{ U16 min = 0;
for (n=maxNbBits; n>0; n--) {
valPerRank[n] = min; /* get starting value within each rank */
min += nbPerRank[n];
min >>= 1;
} }
for (n=0; n<=maxSymbolValue; n++)
tree[huffNode[n].byte].nbBits = huffNode[n].nbBits; /* push nbBits per symbol, symbol order */
for (n=0; n<=maxSymbolValue; n++)
tree[n].val = valPerRank[tree[n].nbBits]++; /* assign value within rank, symbol order */
}
/**
* HUF_buildCTableFromTree():
* Build the CTable given the Huffman tree in huffNode.
*
* @param[out] CTable The output Huffman CTable.
* @param huffNode The Huffman tree.
* @param nonNullRank The last and smallest node in the Huffman tree.
* @param maxSymbolValue The maximum symbol value.
* @param maxNbBits The exact maximum number of bits used in the Huffman tree.
*/
static void HUF_buildCTableFromTree(HUF_CElt* CTable, nodeElt const* huffNode, int nonNullRank, U32 maxSymbolValue, U32 maxNbBits)
{
/* fill result into ctable (val, nbBits) */
int n;
U16 nbPerRank[HUF_TABLELOG_MAX+1] = {0};
U16 valPerRank[HUF_TABLELOG_MAX+1] = {0};
int const alphabetSize = (int)(maxSymbolValue + 1);
for (n=0; n<=nonNullRank; n++)
nbPerRank[huffNode[n].nbBits]++;
/* determine starting value per rank */
{ U16 min = 0;
for (n=(int)maxNbBits; n>0; n--) {
valPerRank[n] = min; /* get starting value within each rank */
min += nbPerRank[n];
min >>= 1;
} }
for (n=0; n<alphabetSize; n++)
CTable[huffNode[n].byte].nbBits = huffNode[n].nbBits; /* push nbBits per symbol, symbol order */
for (n=0; n<alphabetSize; n++)
CTable[n].val = valPerRank[CTable[n].nbBits]++; /* assign value within rank, symbol order */
}
size_t HUF_buildCTable_wksp (HUF_CElt* tree, const unsigned* count, U32 maxSymbolValue, U32 maxNbBits, void* workSpace, size_t wkspSize)
{
HUF_buildCTable_wksp_tables* const wksp_tables = (HUF_buildCTable_wksp_tables*)workSpace;
nodeElt* const huffNode0 = wksp_tables->huffNodeTbl;
nodeElt* const huffNode = huffNode0+1;
int nonNullRank;
/* safety checks */
if (((size_t)workSpace & 3) != 0) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */
if (wkspSize < sizeof(HUF_buildCTable_wksp_tables))
return ERROR(workSpace_tooSmall);
if (maxNbBits == 0) maxNbBits = HUF_TABLELOG_DEFAULT;
if (maxSymbolValue > HUF_SYMBOLVALUE_MAX)
return ERROR(maxSymbolValue_tooLarge);
ZSTD_memset(huffNode0, 0, sizeof(huffNodeTable));
/* sort, decreasing order */
HUF_sort(huffNode, count, maxSymbolValue, wksp_tables->rankPosition);
/* build tree */
nonNullRank = HUF_buildTree(huffNode, maxSymbolValue);
/* enforce maxTableLog */
maxNbBits = HUF_setMaxHeight(huffNode, (U32)nonNullRank, maxNbBits);
if (maxNbBits > HUF_TABLELOG_MAX) return ERROR(GENERIC); /* check fit into table */
HUF_buildCTableFromTree(tree, huffNode, nonNullRank, maxSymbolValue, maxNbBits);
return maxNbBits;
}
/** HUF_buildCTable() :
* @return : maxNbBits
* Note : count is used before tree is written, so they can safely overlap
*/
size_t HUF_buildCTable (HUF_CElt* tree, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits)
{
huffNodeTable nodeTable;
return HUF_buildCTable_wksp(tree, count, maxSymbolValue, maxNbBits, nodeTable, sizeof(nodeTable));
}
static size_t HUF_estimateCompressedSize(HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue)
size_t HUF_estimateCompressedSize(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue)
{
size_t nbBits = 0;
int s;
@@ -437,7 +532,7 @@ static size_t HUF_estimateCompressedSize(HUF_CElt* CTable, const unsigned* count
return nbBits >> 3;
}
static int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue) {
int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue) {
int bad = 0;
int s;
for (s = 0; s <= (int)maxSymbolValue; ++s) {
@@ -476,7 +571,7 @@ HUF_compress1X_usingCTable_internal_body(void* dst, size_t dstSize,
/* init */
if (dstSize < 8) return 0; /* not enough space to compress */
{ size_t const initErr = BIT_initCStream(&bitC, op, oend-op);
{ size_t const initErr = BIT_initCStream(&bitC, op, (size_t)(oend-op));
if (HUF_isError(initErr)) return 0; }
n = srcSize & ~3; /* join to mod 4 */
@@ -573,7 +668,8 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
if (srcSize < 12) return 0; /* no saving possible : too small input */
op += 6; /* jumpTable */
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, oend-op, ip, segmentSize, CTable, bmi2) );
assert(op <= oend);
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) );
if (cSize==0) return 0;
assert(cSize <= 65535);
MEM_writeLE16(ostart, (U16)cSize);
@@ -581,7 +677,8 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
}
ip += segmentSize;
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, oend-op, ip, segmentSize, CTable, bmi2) );
assert(op <= oend);
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) );
if (cSize==0) return 0;
assert(cSize <= 65535);
MEM_writeLE16(ostart+2, (U16)cSize);
@@ -589,7 +686,8 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
}
ip += segmentSize;
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, oend-op, ip, segmentSize, CTable, bmi2) );
assert(op <= oend);
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) );
if (cSize==0) return 0;
assert(cSize <= 65535);
MEM_writeLE16(ostart+4, (U16)cSize);
@@ -597,12 +695,14 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
}
ip += segmentSize;
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, oend-op, ip, iend-ip, CTable, bmi2) );
assert(op <= oend);
assert(ip <= iend);
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, (size_t)(iend-ip), CTable, bmi2) );
if (cSize==0) return 0;
op += cSize;
}
return op-ostart;
return (size_t)(op-ostart);
}
size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable)
@@ -618,40 +718,44 @@ static size_t HUF_compressCTable_internal(
HUF_nbStreams_e nbStreams, const HUF_CElt* CTable, const int bmi2)
{
size_t const cSize = (nbStreams==HUF_singleStream) ?
HUF_compress1X_usingCTable_internal(op, oend - op, src, srcSize, CTable, bmi2) :
HUF_compress4X_usingCTable_internal(op, oend - op, src, srcSize, CTable, bmi2);
HUF_compress1X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, bmi2) :
HUF_compress4X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, bmi2);
if (HUF_isError(cSize)) { return cSize; }
if (cSize==0) { return 0; } /* uncompressible */
op += cSize;
/* check compressibility */
assert(op >= ostart);
if ((size_t)(op-ostart) >= srcSize-1) { return 0; }
return op-ostart;
return (size_t)(op-ostart);
}
typedef struct {
unsigned count[HUF_SYMBOLVALUE_MAX + 1];
HUF_CElt CTable[HUF_SYMBOLVALUE_MAX + 1];
huffNodeTable nodeTable;
HUF_buildCTable_wksp_tables buildCTable_wksp;
} HUF_compress_tables_t;
/* HUF_compress_internal() :
* `workSpace` must a table of at least HUF_WORKSPACE_SIZE_U32 unsigned */
* `workSpace_align4` must be aligned on 4-bytes boundaries,
* and occupies the same space as a table of HUF_WORKSPACE_SIZE_U32 unsigned */
static size_t
HUF_compress_internal (void* dst, size_t dstSize,
const void* src, size_t srcSize,
unsigned maxSymbolValue, unsigned huffLog,
HUF_nbStreams_e nbStreams,
void* workSpace, size_t wkspSize,
void* workSpace_align4, size_t wkspSize,
HUF_CElt* oldHufTable, HUF_repeat* repeat, int preferRepeat,
const int bmi2)
{
HUF_compress_tables_t* const table = (HUF_compress_tables_t*)workSpace;
HUF_compress_tables_t* const table = (HUF_compress_tables_t*)workSpace_align4;
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstSize;
BYTE* op = ostart;
HUF_STATIC_ASSERT(sizeof(*table) <= HUF_WORKSPACE_SIZE);
assert(((size_t)workSpace_align4 & 3) == 0); /* must be aligned on 4-bytes boundaries */
/* checks & inits */
if (((size_t)workSpace & 3) != 0) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */
if (wkspSize < HUF_WORKSPACE_SIZE) return ERROR(workSpace_tooSmall);
if (!srcSize) return 0; /* Uncompressed */
if (!dstSize) return 0; /* cannot fit anything within dst budget */
@@ -669,7 +773,7 @@ HUF_compress_internal (void* dst, size_t dstSize,
}
/* Scan input and build symbol stats */
{ CHECK_V_F(largest, HIST_count_wksp (table->count, &maxSymbolValue, (const BYTE*)src, srcSize, workSpace, wkspSize) );
{ CHECK_V_F(largest, HIST_count_wksp (table->count, &maxSymbolValue, (const BYTE*)src, srcSize, workSpace_align4, wkspSize) );
if (largest == srcSize) { *ostart = ((const BYTE*)src)[0]; return 1; } /* single symbol, rle */
if (largest <= (srcSize >> 7)+4) return 0; /* heuristic : probably not compressible enough */
}
@@ -691,11 +795,11 @@ HUF_compress_internal (void* dst, size_t dstSize,
huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue);
{ size_t const maxBits = HUF_buildCTable_wksp(table->CTable, table->count,
maxSymbolValue, huffLog,
table->nodeTable, sizeof(table->nodeTable));
&table->buildCTable_wksp, sizeof(table->buildCTable_wksp));
CHECK_F(maxBits);
huffLog = (U32)maxBits;
/* Zero unused symbols in CTable, so we can check it for validity */
memset(table->CTable + (maxSymbolValue + 1), 0,
ZSTD_memset(table->CTable + (maxSymbolValue + 1), 0,
sizeof(table->CTable) - ((maxSymbolValue + 1) * sizeof(HUF_CElt)));
}
@@ -716,7 +820,7 @@ HUF_compress_internal (void* dst, size_t dstSize,
op += hSize;
if (repeat) { *repeat = HUF_repeat_none; }
if (oldHufTable)
memcpy(oldHufTable, table->CTable, sizeof(table->CTable)); /* Save new table */
ZSTD_memcpy(oldHufTable, table->CTable, sizeof(table->CTable)); /* Save new table */
}
return HUF_compressCTable_internal(ostart, op, oend,
src, srcSize,
@@ -747,14 +851,6 @@ size_t HUF_compress1X_repeat (void* dst, size_t dstSize,
repeat, preferRepeat, bmi2);
}
size_t HUF_compress1X (void* dst, size_t dstSize,
const void* src, size_t srcSize,
unsigned maxSymbolValue, unsigned huffLog)
{
unsigned workSpace[HUF_WORKSPACE_SIZE_U32];
return HUF_compress1X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(workSpace));
}
/* HUF_compress4X_repeat():
* compress input using 4 streams.
* provide workspace to generate compression tables */
@@ -784,6 +880,25 @@ size_t HUF_compress4X_repeat (void* dst, size_t dstSize,
hufTable, repeat, preferRepeat, bmi2);
}
#ifndef ZSTD_NO_UNUSED_FUNCTIONS
/** HUF_buildCTable() :
* @return : maxNbBits
* Note : count is used before tree is written, so they can safely overlap
*/
size_t HUF_buildCTable (HUF_CElt* tree, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits)
{
HUF_buildCTable_wksp_tables workspace;
return HUF_buildCTable_wksp(tree, count, maxSymbolValue, maxNbBits, &workspace, sizeof(workspace));
}
size_t HUF_compress1X (void* dst, size_t dstSize,
const void* src, size_t srcSize,
unsigned maxSymbolValue, unsigned huffLog)
{
unsigned workSpace[HUF_WORKSPACE_SIZE_U32];
return HUF_compress1X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(workSpace));
}
size_t HUF_compress2 (void* dst, size_t dstSize,
const void* src, size_t srcSize,
unsigned maxSymbolValue, unsigned huffLog)
@@ -796,3 +911,4 @@ size_t HUF_compress (void* dst, size_t maxDstSize, const void* src, size_t srcSi
{
return HUF_compress2(dst, maxDstSize, src, srcSize, 255, HUF_TABLELOG_DEFAULT);
}
#endif

View File

@@ -1,41 +1,21 @@
/* ******************************************************************
huff0 huffman decoder,
part of Finite State Entropy library
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
* huff0 huffman decoder,
* part of Finite State Entropy library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* **************************************************************
* Dependencies
****************************************************************/
#include <string.h> /* memcpy, memset */
#include "zstd_deps.h" /* ZSTD_memcpy, ZSTD_memset */
#include "compiler.h"
#include "bitstream.h" /* BIT_* */
#include "fse.h" /* to compress headers */
@@ -61,9 +41,6 @@
* Error Management
****************************************************************/
#define HUF_isError ERR_isError
#ifndef CHECK_F
#define CHECK_F(f) { size_t const err_ = (f); if (HUF_isError(err_)) return err_; }
#endif
/* **************************************************************
@@ -126,7 +103,7 @@ typedef struct { BYTE maxTableLog; BYTE tableType; BYTE tableLog; BYTE reserved;
static DTableDesc HUF_getDTableDesc(const HUF_DTable* table)
{
DTableDesc dtd;
memcpy(&dtd, table, sizeof(dtd));
ZSTD_memcpy(&dtd, table, sizeof(dtd));
return dtd;
}
@@ -138,29 +115,51 @@ static DTableDesc HUF_getDTableDesc(const HUF_DTable* table)
/*-***************************/
typedef struct { BYTE byte; BYTE nbBits; } HUF_DEltX1; /* single-symbol decoding */
/**
* Packs 4 HUF_DEltX1 structs into a U64. This is used to lay down 4 entries at
* a time.
*/
static U64 HUF_DEltX1_set4(BYTE symbol, BYTE nbBits) {
U64 D4;
if (MEM_isLittleEndian()) {
D4 = symbol + (nbBits << 8);
} else {
D4 = (symbol << 8) + nbBits;
}
D4 *= 0x0001000100010001ULL;
return D4;
}
typedef struct {
U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1];
U32 rankStart[HUF_TABLELOG_ABSOLUTEMAX + 1];
U32 statsWksp[HUF_READ_STATS_WORKSPACE_SIZE_U32];
BYTE symbols[HUF_SYMBOLVALUE_MAX + 1];
BYTE huffWeight[HUF_SYMBOLVALUE_MAX + 1];
} HUF_ReadDTableX1_Workspace;
size_t HUF_readDTableX1_wksp(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize)
{
return HUF_readDTableX1_wksp_bmi2(DTable, src, srcSize, workSpace, wkspSize, /* bmi2 */ 0);
}
size_t HUF_readDTableX1_wksp_bmi2(HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize, int bmi2)
{
U32 tableLog = 0;
U32 nbSymbols = 0;
size_t iSize;
void* const dtPtr = DTable + 1;
HUF_DEltX1* const dt = (HUF_DEltX1*)dtPtr;
HUF_ReadDTableX1_Workspace* wksp = (HUF_ReadDTableX1_Workspace*)workSpace;
U32* rankVal;
BYTE* huffWeight;
size_t spaceUsed32 = 0;
rankVal = (U32 *)workSpace + spaceUsed32;
spaceUsed32 += HUF_TABLELOG_ABSOLUTEMAX + 1;
huffWeight = (BYTE *)((U32 *)workSpace + spaceUsed32);
spaceUsed32 += HUF_ALIGN(HUF_SYMBOLVALUE_MAX + 1, sizeof(U32)) >> 2;
if ((spaceUsed32 << 2) > wkspSize) return ERROR(tableLog_tooLarge);
DEBUG_STATIC_ASSERT(HUF_DECOMPRESS_WORKSPACE_SIZE >= sizeof(*wksp));
if (sizeof(*wksp) > wkspSize) return ERROR(tableLog_tooLarge);
DEBUG_STATIC_ASSERT(sizeof(DTableDesc) == sizeof(HUF_DTable));
/* memset(huffWeight, 0, sizeof(huffWeight)); */ /* is not necessary, even though some analyzer complain ... */
/* ZSTD_memset(huffWeight, 0, sizeof(huffWeight)); */ /* is not necessary, even though some analyzer complain ... */
iSize = HUF_readStats(huffWeight, HUF_SYMBOLVALUE_MAX + 1, rankVal, &nbSymbols, &tableLog, src, srcSize);
iSize = HUF_readStats_wksp(wksp->huffWeight, HUF_SYMBOLVALUE_MAX + 1, wksp->rankVal, &nbSymbols, &tableLog, src, srcSize, wksp->statsWksp, sizeof(wksp->statsWksp), bmi2);
if (HUF_isError(iSize)) return iSize;
/* Table header */
@@ -168,40 +167,117 @@ size_t HUF_readDTableX1_wksp(HUF_DTable* DTable, const void* src, size_t srcSize
if (tableLog > (U32)(dtd.maxTableLog+1)) return ERROR(tableLog_tooLarge); /* DTable too small, Huffman tree cannot fit in */
dtd.tableType = 0;
dtd.tableLog = (BYTE)tableLog;
memcpy(DTable, &dtd, sizeof(dtd));
ZSTD_memcpy(DTable, &dtd, sizeof(dtd));
}
/* Calculate starting value for each rank */
{ U32 n, nextRankStart = 0;
for (n=1; n<tableLog+1; n++) {
U32 const current = nextRankStart;
nextRankStart += (rankVal[n] << (n-1));
rankVal[n] = current;
} }
/* fill DTable */
{ U32 n;
for (n=0; n<nbSymbols; n++) {
U32 const w = huffWeight[n];
U32 const length = (1 << w) >> 1;
U32 u;
HUF_DEltX1 D;
D.byte = (BYTE)n; D.nbBits = (BYTE)(tableLog + 1 - w);
for (u = rankVal[w]; u < rankVal[w] + length; u++)
dt[u] = D;
rankVal[w] += length;
} }
/* Compute symbols and rankStart given rankVal:
*
* rankVal already contains the number of values of each weight.
*
* symbols contains the symbols ordered by weight. First are the rankVal[0]
* weight 0 symbols, followed by the rankVal[1] weight 1 symbols, and so on.
* symbols[0] is filled (but unused) to avoid a branch.
*
* rankStart contains the offset where each rank belongs in the DTable.
* rankStart[0] is not filled because there are no entries in the table for
* weight 0.
*/
{
int n;
int nextRankStart = 0;
int const unroll = 4;
int const nLimit = (int)nbSymbols - unroll + 1;
for (n=0; n<(int)tableLog+1; n++) {
U32 const curr = nextRankStart;
nextRankStart += wksp->rankVal[n];
wksp->rankStart[n] = curr;
}
for (n=0; n < nLimit; n += unroll) {
int u;
for (u=0; u < unroll; ++u) {
size_t const w = wksp->huffWeight[n+u];
wksp->symbols[wksp->rankStart[w]++] = (BYTE)(n+u);
}
}
for (; n < (int)nbSymbols; ++n) {
size_t const w = wksp->huffWeight[n];
wksp->symbols[wksp->rankStart[w]++] = (BYTE)n;
}
}
/* fill DTable
* We fill all entries of each weight in order.
* That way length is a constant for each iteration of the outter loop.
* We can switch based on the length to a different inner loop which is
* optimized for that particular case.
*/
{
U32 w;
int symbol=wksp->rankVal[0];
int rankStart=0;
for (w=1; w<tableLog+1; ++w) {
int const symbolCount = wksp->rankVal[w];
int const length = (1 << w) >> 1;
int uStart = rankStart;
BYTE const nbBits = (BYTE)(tableLog + 1 - w);
int s;
int u;
switch (length) {
case 1:
for (s=0; s<symbolCount; ++s) {
HUF_DEltX1 D;
D.byte = wksp->symbols[symbol + s];
D.nbBits = nbBits;
dt[uStart] = D;
uStart += 1;
}
break;
case 2:
for (s=0; s<symbolCount; ++s) {
HUF_DEltX1 D;
D.byte = wksp->symbols[symbol + s];
D.nbBits = nbBits;
dt[uStart+0] = D;
dt[uStart+1] = D;
uStart += 2;
}
break;
case 4:
for (s=0; s<symbolCount; ++s) {
U64 const D4 = HUF_DEltX1_set4(wksp->symbols[symbol + s], nbBits);
MEM_write64(dt + uStart, D4);
uStart += 4;
}
break;
case 8:
for (s=0; s<symbolCount; ++s) {
U64 const D4 = HUF_DEltX1_set4(wksp->symbols[symbol + s], nbBits);
MEM_write64(dt + uStart, D4);
MEM_write64(dt + uStart + 4, D4);
uStart += 8;
}
break;
default:
for (s=0; s<symbolCount; ++s) {
U64 const D4 = HUF_DEltX1_set4(wksp->symbols[symbol + s], nbBits);
for (u=0; u < length; u += 16) {
MEM_write64(dt + uStart + u + 0, D4);
MEM_write64(dt + uStart + u + 4, D4);
MEM_write64(dt + uStart + u + 8, D4);
MEM_write64(dt + uStart + u + 12, D4);
}
assert(u == length);
uStart += length;
}
break;
}
symbol += symbolCount;
rankStart += symbolCount * length;
}
}
return iSize;
}
size_t HUF_readDTableX1(HUF_DTable* DTable, const void* src, size_t srcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_readDTableX1_wksp(DTable, src, srcSize,
workSpace, sizeof(workSpace));
}
FORCE_INLINE_TEMPLATE BYTE
HUF_decodeSymbolX1(BIT_DStream_t* Dstream, const HUF_DEltX1* dt, const U32 dtLog)
{
@@ -282,6 +358,7 @@ HUF_decompress4X1_usingDTable_internal_body(
{ const BYTE* const istart = (const BYTE*) cSrc;
BYTE* const ostart = (BYTE*) dst;
BYTE* const oend = ostart + dstSize;
BYTE* const olimit = oend - 3;
const void* const dtPtr = DTable + 1;
const HUF_DEltX1* const dt = (const HUF_DEltX1*)dtPtr;
@@ -306,9 +383,9 @@ HUF_decompress4X1_usingDTable_internal_body(
BYTE* op2 = opStart2;
BYTE* op3 = opStart3;
BYTE* op4 = opStart4;
U32 endSignal = BIT_DStream_unfinished;
DTableDesc const dtd = HUF_getDTableDesc(DTable);
U32 const dtLog = dtd.tableLog;
U32 endSignal = 1;
if (length4 > cSrcSize) return ERROR(corruption_detected); /* overflow */
CHECK_F( BIT_initDStream(&bitD1, istart1, length1) );
@@ -317,8 +394,7 @@ HUF_decompress4X1_usingDTable_internal_body(
CHECK_F( BIT_initDStream(&bitD4, istart4, length4) );
/* up to 16 symbols per loop (4 symbols per stream) in 64-bit mode */
endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4);
while ( (endSignal==BIT_DStream_unfinished) && (op4<(oend-3)) ) {
for ( ; (endSignal) & (op4 < olimit) ; ) {
HUF_DECODE_SYMBOLX1_2(op1, &bitD1);
HUF_DECODE_SYMBOLX1_2(op2, &bitD2);
HUF_DECODE_SYMBOLX1_2(op3, &bitD3);
@@ -335,10 +411,10 @@ HUF_decompress4X1_usingDTable_internal_body(
HUF_DECODE_SYMBOLX1_0(op2, &bitD2);
HUF_DECODE_SYMBOLX1_0(op3, &bitD3);
HUF_DECODE_SYMBOLX1_0(op4, &bitD4);
BIT_reloadDStream(&bitD1);
BIT_reloadDStream(&bitD2);
BIT_reloadDStream(&bitD3);
BIT_reloadDStream(&bitD4);
endSignal &= BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD4) == BIT_DStream_unfinished;
}
/* check corruption */
@@ -400,20 +476,6 @@ size_t HUF_decompress1X1_DCtx_wksp(HUF_DTable* DCtx, void* dst, size_t dstSize,
}
size_t HUF_decompress1X1_DCtx(HUF_DTable* DCtx, void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress1X1_DCtx_wksp(DCtx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress1X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX1(DTable, HUF_TABLELOG_MAX);
return HUF_decompress1X1_DCtx (DTable, dst, dstSize, cSrc, cSrcSize);
}
size_t HUF_decompress4X1_usingDTable(
void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize,
@@ -430,8 +492,7 @@ static size_t HUF_decompress4X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size
{
const BYTE* ip = (const BYTE*) cSrc;
size_t const hSize = HUF_readDTableX1_wksp (dctx, cSrc, cSrcSize,
workSpace, wkspSize);
size_t const hSize = HUF_readDTableX1_wksp_bmi2(dctx, cSrc, cSrcSize, workSpace, wkspSize, bmi2);
if (HUF_isError(hSize)) return hSize;
if (hSize >= cSrcSize) return ERROR(srcSize_wrong);
ip += hSize; cSrcSize -= hSize;
@@ -447,18 +508,6 @@ size_t HUF_decompress4X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize,
}
size_t HUF_decompress4X1_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress4X1_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress4X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX1(DTable, HUF_TABLELOG_MAX);
return HUF_decompress4X1_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
}
#endif /* HUF_FORCE_DECOMPRESS_X2 */
@@ -485,7 +534,7 @@ static void HUF_fillDTableX2Level2(HUF_DEltX2* DTable, U32 sizeLog, const U32 co
U32 rankVal[HUF_TABLELOG_MAX + 1];
/* get pre-calculated rankVal */
memcpy(rankVal, rankValOrigin, sizeof(rankVal));
ZSTD_memcpy(rankVal, rankValOrigin, sizeof(rankVal));
/* fill skipped values */
if (minWeight>1) {
@@ -527,7 +576,7 @@ static void HUF_fillDTableX2(HUF_DEltX2* DTable, const U32 targetLog,
const U32 minBits = nbBitsBaseline - maxWeight;
U32 s;
memcpy(rankVal, rankValOrigin, sizeof(rankVal));
ZSTD_memcpy(rankVal, rankValOrigin, sizeof(rankVal));
/* fill DTable */
for (s=0; s<sortedListSize; s++) {
@@ -592,11 +641,11 @@ size_t HUF_readDTableX2_wksp(HUF_DTable* DTable,
if ((spaceUsed32 << 2) > wkspSize) return ERROR(tableLog_tooLarge);
rankStart = rankStart0 + 1;
memset(rankStats, 0, sizeof(U32) * (2 * HUF_TABLELOG_MAX + 2 + 1));
ZSTD_memset(rankStats, 0, sizeof(U32) * (2 * HUF_TABLELOG_MAX + 2 + 1));
DEBUG_STATIC_ASSERT(sizeof(HUF_DEltX2) == sizeof(HUF_DTable)); /* if compiler fails here, assertion is wrong */
if (maxTableLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge);
/* memset(weightList, 0, sizeof(weightList)); */ /* is not necessary, even though some analyzer complain ... */
/* ZSTD_memset(weightList, 0, sizeof(weightList)); */ /* is not necessary, even though some analyzer complain ... */
iSize = HUF_readStats(weightList, HUF_SYMBOLVALUE_MAX + 1, rankStats, &nbSymbols, &tableLog, src, srcSize);
if (HUF_isError(iSize)) return iSize;
@@ -610,9 +659,9 @@ size_t HUF_readDTableX2_wksp(HUF_DTable* DTable,
/* Get start index of each weight */
{ U32 w, nextRankStart = 0;
for (w=1; w<maxW+1; w++) {
U32 current = nextRankStart;
U32 curr = nextRankStart;
nextRankStart += rankStats[w];
rankStart[w] = current;
rankStart[w] = curr;
}
rankStart[0] = nextRankStart; /* put all 0w symbols at the end of sorted list*/
sizeOfSort = nextRankStart;
@@ -635,9 +684,9 @@ size_t HUF_readDTableX2_wksp(HUF_DTable* DTable,
U32 nextRankVal = 0;
U32 w;
for (w=1; w<maxW+1; w++) {
U32 current = nextRankVal;
U32 curr = nextRankVal;
nextRankVal += rankStats[w] << (w+rescale);
rankVal0[w] = current;
rankVal0[w] = curr;
} }
{ U32 const minBits = tableLog+1 - maxW;
U32 consumed;
@@ -655,23 +704,16 @@ size_t HUF_readDTableX2_wksp(HUF_DTable* DTable,
dtd.tableLog = (BYTE)maxTableLog;
dtd.tableType = 1;
memcpy(DTable, &dtd, sizeof(dtd));
ZSTD_memcpy(DTable, &dtd, sizeof(dtd));
return iSize;
}
size_t HUF_readDTableX2(HUF_DTable* DTable, const void* src, size_t srcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_readDTableX2_wksp(DTable, src, srcSize,
workSpace, sizeof(workSpace));
}
FORCE_INLINE_TEMPLATE U32
HUF_decodeSymbolX2(void* op, BIT_DStream_t* DStream, const HUF_DEltX2* dt, const U32 dtLog)
{
size_t const val = BIT_lookBitsFast(DStream, dtLog); /* note : dtLog >= 1 */
memcpy(op, dt+val, 2);
ZSTD_memcpy(op, dt+val, 2);
BIT_skipBits(DStream, dt[val].nbBits);
return dt[val].length;
}
@@ -680,7 +722,7 @@ FORCE_INLINE_TEMPLATE U32
HUF_decodeLastSymbolX2(void* op, BIT_DStream_t* DStream, const HUF_DEltX2* dt, const U32 dtLog)
{
size_t const val = BIT_lookBitsFast(DStream, dtLog); /* note : dtLog >= 1 */
memcpy(op, dt+val, 1);
ZSTD_memcpy(op, dt+val, 1);
if (dt[val].length==1) BIT_skipBits(DStream, dt[val].nbBits);
else {
if (DStream->bitsConsumed < (sizeof(DStream->bitContainer)*8)) {
@@ -757,7 +799,6 @@ HUF_decompress1X2_usingDTable_internal_body(
return dstSize;
}
FORCE_INLINE_TEMPLATE size_t
HUF_decompress4X2_usingDTable_internal_body(
void* dst, size_t dstSize,
@@ -769,6 +810,7 @@ HUF_decompress4X2_usingDTable_internal_body(
{ const BYTE* const istart = (const BYTE*) cSrc;
BYTE* const ostart = (BYTE*) dst;
BYTE* const oend = ostart + dstSize;
BYTE* const olimit = oend - (sizeof(size_t)-1);
const void* const dtPtr = DTable+1;
const HUF_DEltX2* const dt = (const HUF_DEltX2*)dtPtr;
@@ -793,7 +835,7 @@ HUF_decompress4X2_usingDTable_internal_body(
BYTE* op2 = opStart2;
BYTE* op3 = opStart3;
BYTE* op4 = opStart4;
U32 endSignal;
U32 endSignal = 1;
DTableDesc const dtd = HUF_getDTableDesc(DTable);
U32 const dtLog = dtd.tableLog;
@@ -804,8 +846,29 @@ HUF_decompress4X2_usingDTable_internal_body(
CHECK_F( BIT_initDStream(&bitD4, istart4, length4) );
/* 16-32 symbols per loop (4-8 symbols per stream) */
endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4);
for ( ; (endSignal==BIT_DStream_unfinished) & (op4<(oend-(sizeof(bitD4.bitContainer)-1))) ; ) {
for ( ; (endSignal) & (op4 < olimit); ) {
#if defined(__clang__) && (defined(__x86_64__) || defined(__i386__))
HUF_DECODE_SYMBOLX2_2(op1, &bitD1);
HUF_DECODE_SYMBOLX2_1(op1, &bitD1);
HUF_DECODE_SYMBOLX2_2(op1, &bitD1);
HUF_DECODE_SYMBOLX2_0(op1, &bitD1);
HUF_DECODE_SYMBOLX2_2(op2, &bitD2);
HUF_DECODE_SYMBOLX2_1(op2, &bitD2);
HUF_DECODE_SYMBOLX2_2(op2, &bitD2);
HUF_DECODE_SYMBOLX2_0(op2, &bitD2);
endSignal &= BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished;
HUF_DECODE_SYMBOLX2_2(op3, &bitD3);
HUF_DECODE_SYMBOLX2_1(op3, &bitD3);
HUF_DECODE_SYMBOLX2_2(op3, &bitD3);
HUF_DECODE_SYMBOLX2_0(op3, &bitD3);
HUF_DECODE_SYMBOLX2_2(op4, &bitD4);
HUF_DECODE_SYMBOLX2_1(op4, &bitD4);
HUF_DECODE_SYMBOLX2_2(op4, &bitD4);
HUF_DECODE_SYMBOLX2_0(op4, &bitD4);
endSignal &= BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD4) == BIT_DStream_unfinished;
#else
HUF_DECODE_SYMBOLX2_2(op1, &bitD1);
HUF_DECODE_SYMBOLX2_2(op2, &bitD2);
HUF_DECODE_SYMBOLX2_2(op3, &bitD3);
@@ -822,8 +885,12 @@ HUF_decompress4X2_usingDTable_internal_body(
HUF_DECODE_SYMBOLX2_0(op2, &bitD2);
HUF_DECODE_SYMBOLX2_0(op3, &bitD3);
HUF_DECODE_SYMBOLX2_0(op4, &bitD4);
endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4);
endSignal = (U32)LIKELY(
(BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished)
& (BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished)
& (BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished)
& (BIT_reloadDStreamFast(&bitD4) == BIT_DStream_unfinished));
#endif
}
/* check corruption */
@@ -876,20 +943,6 @@ size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable* DCtx, void* dst, size_t dstSize,
}
size_t HUF_decompress1X2_DCtx(HUF_DTable* DCtx, void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress1X2_DCtx_wksp(DCtx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress1X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_TABLELOG_MAX);
return HUF_decompress1X2_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
}
size_t HUF_decompress4X2_usingDTable(
void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize,
@@ -923,20 +976,6 @@ size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize,
}
size_t HUF_decompress4X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_TABLELOG_MAX);
return HUF_decompress4X2_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
}
#endif /* HUF_FORCE_DECOMPRESS_X1 */
@@ -1037,67 +1076,6 @@ U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize)
}
typedef size_t (*decompressionAlgo)(void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);
size_t HUF_decompress (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
#if !defined(HUF_FORCE_DECOMPRESS_X1) && !defined(HUF_FORCE_DECOMPRESS_X2)
static const decompressionAlgo decompress[2] = { HUF_decompress4X1, HUF_decompress4X2 };
#endif
/* validation checks */
if (dstSize == 0) return ERROR(dstSize_tooSmall);
if (cSrcSize > dstSize) return ERROR(corruption_detected); /* invalid */
if (cSrcSize == dstSize) { memcpy(dst, cSrc, dstSize); return dstSize; } /* not compressed */
if (cSrcSize == 1) { memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)algoNb;
assert(algoNb == 0);
return HUF_decompress4X1(dst, dstSize, cSrc, cSrcSize);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)algoNb;
assert(algoNb == 1);
return HUF_decompress4X2(dst, dstSize, cSrc, cSrcSize);
#else
return decompress[algoNb](dst, dstSize, cSrc, cSrcSize);
#endif
}
}
size_t HUF_decompress4X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
/* validation checks */
if (dstSize == 0) return ERROR(dstSize_tooSmall);
if (cSrcSize > dstSize) return ERROR(corruption_detected); /* invalid */
if (cSrcSize == dstSize) { memcpy(dst, cSrc, dstSize); return dstSize; } /* not compressed */
if (cSrcSize == 1) { memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)algoNb;
assert(algoNb == 0);
return HUF_decompress4X1_DCtx(dctx, dst, dstSize, cSrc, cSrcSize);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)algoNb;
assert(algoNb == 1);
return HUF_decompress4X2_DCtx(dctx, dst, dstSize, cSrc, cSrcSize);
#else
return algoNb ? HUF_decompress4X2_DCtx(dctx, dst, dstSize, cSrc, cSrcSize) :
HUF_decompress4X1_DCtx(dctx, dst, dstSize, cSrc, cSrcSize) ;
#endif
}
}
size_t HUF_decompress4X_hufOnly(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress4X_hufOnly_wksp(dctx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst,
size_t dstSize, const void* cSrc,
size_t cSrcSize, void* workSpace,
@@ -1131,8 +1109,8 @@ size_t HUF_decompress1X_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize,
/* validation checks */
if (dstSize == 0) return ERROR(dstSize_tooSmall);
if (cSrcSize > dstSize) return ERROR(corruption_detected); /* invalid */
if (cSrcSize == dstSize) { memcpy(dst, cSrc, dstSize); return dstSize; } /* not compressed */
if (cSrcSize == 1) { memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */
if (cSrcSize == dstSize) { ZSTD_memcpy(dst, cSrc, dstSize); return dstSize; } /* not compressed */
if (cSrcSize == 1) { ZSTD_memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
#if defined(HUF_FORCE_DECOMPRESS_X1)
@@ -1154,14 +1132,6 @@ size_t HUF_decompress1X_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize,
}
}
size_t HUF_decompress1X_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress1X_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress1X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2)
{
@@ -1185,7 +1155,7 @@ size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstS
{
const BYTE* ip = (const BYTE*) cSrc;
size_t const hSize = HUF_readDTableX1_wksp(dctx, cSrc, cSrcSize, workSpace, wkspSize);
size_t const hSize = HUF_readDTableX1_wksp_bmi2(dctx, cSrc, cSrcSize, workSpace, wkspSize, bmi2);
if (HUF_isError(hSize)) return hSize;
if (hSize >= cSrcSize) return ERROR(srcSize_wrong);
ip += hSize; cSrcSize -= hSize;
@@ -1232,3 +1202,149 @@ size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t ds
#endif
}
}
#ifndef ZSTD_NO_UNUSED_FUNCTIONS
#ifndef HUF_FORCE_DECOMPRESS_X2
size_t HUF_readDTableX1(HUF_DTable* DTable, const void* src, size_t srcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_readDTableX1_wksp(DTable, src, srcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress1X1_DCtx(HUF_DTable* DCtx, void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress1X1_DCtx_wksp(DCtx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress1X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX1(DTable, HUF_TABLELOG_MAX);
return HUF_decompress1X1_DCtx (DTable, dst, dstSize, cSrc, cSrcSize);
}
#endif
#ifndef HUF_FORCE_DECOMPRESS_X1
size_t HUF_readDTableX2(HUF_DTable* DTable, const void* src, size_t srcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_readDTableX2_wksp(DTable, src, srcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress1X2_DCtx(HUF_DTable* DCtx, void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress1X2_DCtx_wksp(DCtx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress1X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_TABLELOG_MAX);
return HUF_decompress1X2_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
}
#endif
#ifndef HUF_FORCE_DECOMPRESS_X2
size_t HUF_decompress4X1_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress4X1_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress4X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX1(DTable, HUF_TABLELOG_MAX);
return HUF_decompress4X1_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
}
#endif
#ifndef HUF_FORCE_DECOMPRESS_X1
size_t HUF_decompress4X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_TABLELOG_MAX);
return HUF_decompress4X2_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
}
#endif
typedef size_t (*decompressionAlgo)(void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);
size_t HUF_decompress (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
#if !defined(HUF_FORCE_DECOMPRESS_X1) && !defined(HUF_FORCE_DECOMPRESS_X2)
static const decompressionAlgo decompress[2] = { HUF_decompress4X1, HUF_decompress4X2 };
#endif
/* validation checks */
if (dstSize == 0) return ERROR(dstSize_tooSmall);
if (cSrcSize > dstSize) return ERROR(corruption_detected); /* invalid */
if (cSrcSize == dstSize) { ZSTD_memcpy(dst, cSrc, dstSize); return dstSize; } /* not compressed */
if (cSrcSize == 1) { ZSTD_memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)algoNb;
assert(algoNb == 0);
return HUF_decompress4X1(dst, dstSize, cSrc, cSrcSize);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)algoNb;
assert(algoNb == 1);
return HUF_decompress4X2(dst, dstSize, cSrc, cSrcSize);
#else
return decompress[algoNb](dst, dstSize, cSrc, cSrcSize);
#endif
}
}
size_t HUF_decompress4X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
/* validation checks */
if (dstSize == 0) return ERROR(dstSize_tooSmall);
if (cSrcSize > dstSize) return ERROR(corruption_detected); /* invalid */
if (cSrcSize == dstSize) { ZSTD_memcpy(dst, cSrc, dstSize); return dstSize; } /* not compressed */
if (cSrcSize == 1) { ZSTD_memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)algoNb;
assert(algoNb == 0);
return HUF_decompress4X1_DCtx(dctx, dst, dstSize, cSrc, cSrcSize);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)algoNb;
assert(algoNb == 1);
return HUF_decompress4X2_DCtx(dctx, dst, dstSize, cSrc, cSrcSize);
#else
return algoNb ? HUF_decompress4X2_DCtx(dctx, dst, dstSize, cSrc, cSrcSize) :
HUF_decompress4X1_DCtx(dctx, dst, dstSize, cSrc, cSrcSize) ;
#endif
}
}
size_t HUF_decompress4X_hufOnly(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress4X_hufOnly_wksp(dctx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
size_t HUF_decompress1X_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize,
const void* cSrc, size_t cSrcSize)
{
U32 workSpace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32];
return HUF_decompress1X_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize,
workSpace, sizeof(workSpace));
}
#endif

161
vendor/github.com/DataDog/zstd/mem.h generated vendored
View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -18,8 +18,10 @@ extern "C" {
/*-****************************************
* Dependencies
******************************************/
#include <stddef.h> /* size_t, ptrdiff_t */
#include <string.h> /* memcpy */
#include <stddef.h> /* size_t, ptrdiff_t */
#include "compiler.h" /* __has_builtin */
#include "debug.h" /* DEBUG_STATIC_ASSERT */
#include "zstd_deps.h" /* ZSTD_memcpy */
/*-****************************************
@@ -39,93 +41,15 @@ extern "C" {
# define MEM_STATIC static /* this version may generate warnings for unused static functions; disable the relevant warning */
#endif
#ifndef __has_builtin
# define __has_builtin(x) 0 /* compat. with non-clang compilers */
#endif
/* code only tested on 32 and 64 bits systems */
#define MEM_STATIC_ASSERT(c) { enum { MEM_static_assert = 1/(int)(!!(c)) }; }
MEM_STATIC void MEM_check(void) { MEM_STATIC_ASSERT((sizeof(size_t)==4) || (sizeof(size_t)==8)); }
/* detects whether we are being compiled under msan */
#if defined (__has_feature)
# if __has_feature(memory_sanitizer)
# define MEMORY_SANITIZER 1
# endif
#endif
#if defined (MEMORY_SANITIZER)
/* Not all platforms that support msan provide sanitizers/msan_interface.h.
* We therefore declare the functions we need ourselves, rather than trying to
* include the header file... */
#include <stdint.h> /* intptr_t */
/* Make memory region fully initialized (without changing its contents). */
void __msan_unpoison(const volatile void *a, size_t size);
/* Make memory region fully uninitialized (without changing its contents).
This is a legacy interface that does not update origin information. Use
__msan_allocated_memory() instead. */
void __msan_poison(const volatile void *a, size_t size);
/* Returns the offset of the first (at least partially) poisoned byte in the
memory range, or -1 if the whole range is good. */
intptr_t __msan_test_shadow(const volatile void *x, size_t size);
#endif
/* detects whether we are being compiled under asan */
#if defined (__has_feature)
# if __has_feature(address_sanitizer)
# define ADDRESS_SANITIZER 1
# endif
#elif defined(__SANITIZE_ADDRESS__)
# define ADDRESS_SANITIZER 1
#endif
#if defined (ADDRESS_SANITIZER)
/* Not all platforms that support asan provide sanitizers/asan_interface.h.
* We therefore declare the functions we need ourselves, rather than trying to
* include the header file... */
/**
* Marks a memory region (<c>[addr, addr+size)</c>) as unaddressable.
*
* This memory must be previously allocated by your program. Instrumented
* code is forbidden from accessing addresses in this region until it is
* unpoisoned. This function is not guaranteed to poison the entire region -
* it could poison only a subregion of <c>[addr, addr+size)</c> due to ASan
* alignment restrictions.
*
* \note This function is not thread-safe because no two threads can poison or
* unpoison memory in the same memory region simultaneously.
*
* \param addr Start of memory region.
* \param size Size of memory region. */
void __asan_poison_memory_region(void const volatile *addr, size_t size);
/**
* Marks a memory region (<c>[addr, addr+size)</c>) as addressable.
*
* This memory must be previously allocated by your program. Accessing
* addresses in this region is allowed until this region is poisoned again.
* This function could unpoison a super-region of <c>[addr, addr+size)</c> due
* to ASan alignment restrictions.
*
* \note This function is not thread-safe because no two threads can
* poison or unpoison memory in the same memory region simultaneously.
*
* \param addr Start of memory region.
* \param size Size of memory region. */
void __asan_unpoison_memory_region(void const volatile *addr, size_t size);
#endif
/*-**************************************************************
* Basic Types
*****************************************************************/
#if !defined (__VMS) && (defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) )
# include <stdint.h>
# if defined(_AIX)
# include <inttypes.h>
# else
# include <stdint.h> /* intptr_t */
# endif
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef int16_t S16;
@@ -157,7 +81,53 @@ void __asan_unpoison_memory_region(void const volatile *addr, size_t size);
/*-**************************************************************
* Memory I/O
* Memory I/O API
*****************************************************************/
/*=== Static platform detection ===*/
MEM_STATIC unsigned MEM_32bits(void);
MEM_STATIC unsigned MEM_64bits(void);
MEM_STATIC unsigned MEM_isLittleEndian(void);
/*=== Native unaligned read/write ===*/
MEM_STATIC U16 MEM_read16(const void* memPtr);
MEM_STATIC U32 MEM_read32(const void* memPtr);
MEM_STATIC U64 MEM_read64(const void* memPtr);
MEM_STATIC size_t MEM_readST(const void* memPtr);
MEM_STATIC void MEM_write16(void* memPtr, U16 value);
MEM_STATIC void MEM_write32(void* memPtr, U32 value);
MEM_STATIC void MEM_write64(void* memPtr, U64 value);
/*=== Little endian unaligned read/write ===*/
MEM_STATIC U16 MEM_readLE16(const void* memPtr);
MEM_STATIC U32 MEM_readLE24(const void* memPtr);
MEM_STATIC U32 MEM_readLE32(const void* memPtr);
MEM_STATIC U64 MEM_readLE64(const void* memPtr);
MEM_STATIC size_t MEM_readLEST(const void* memPtr);
MEM_STATIC void MEM_writeLE16(void* memPtr, U16 val);
MEM_STATIC void MEM_writeLE24(void* memPtr, U32 val);
MEM_STATIC void MEM_writeLE32(void* memPtr, U32 val32);
MEM_STATIC void MEM_writeLE64(void* memPtr, U64 val64);
MEM_STATIC void MEM_writeLEST(void* memPtr, size_t val);
/*=== Big endian unaligned read/write ===*/
MEM_STATIC U32 MEM_readBE32(const void* memPtr);
MEM_STATIC U64 MEM_readBE64(const void* memPtr);
MEM_STATIC size_t MEM_readBEST(const void* memPtr);
MEM_STATIC void MEM_writeBE32(void* memPtr, U32 val32);
MEM_STATIC void MEM_writeBE64(void* memPtr, U64 val64);
MEM_STATIC void MEM_writeBEST(void* memPtr, size_t val);
/*=== Byteswap ===*/
MEM_STATIC U32 MEM_swap32(U32 in);
MEM_STATIC U64 MEM_swap64(U64 in);
MEM_STATIC size_t MEM_swapST(size_t in);
/*-**************************************************************
* Memory I/O Implementation
*****************************************************************/
/* MEM_FORCE_MEMORY_ACCESS :
* By default, access to unaligned memory is controlled by `memcpy()`, which is safe and portable.
@@ -236,37 +206,37 @@ MEM_STATIC void MEM_write64(void* memPtr, U64 value) { ((unalign64*)memPtr)->v =
MEM_STATIC U16 MEM_read16(const void* memPtr)
{
U16 val; memcpy(&val, memPtr, sizeof(val)); return val;
U16 val; ZSTD_memcpy(&val, memPtr, sizeof(val)); return val;
}
MEM_STATIC U32 MEM_read32(const void* memPtr)
{
U32 val; memcpy(&val, memPtr, sizeof(val)); return val;
U32 val; ZSTD_memcpy(&val, memPtr, sizeof(val)); return val;
}
MEM_STATIC U64 MEM_read64(const void* memPtr)
{
U64 val; memcpy(&val, memPtr, sizeof(val)); return val;
U64 val; ZSTD_memcpy(&val, memPtr, sizeof(val)); return val;
}
MEM_STATIC size_t MEM_readST(const void* memPtr)
{
size_t val; memcpy(&val, memPtr, sizeof(val)); return val;
size_t val; ZSTD_memcpy(&val, memPtr, sizeof(val)); return val;
}
MEM_STATIC void MEM_write16(void* memPtr, U16 value)
{
memcpy(memPtr, &value, sizeof(value));
ZSTD_memcpy(memPtr, &value, sizeof(value));
}
MEM_STATIC void MEM_write32(void* memPtr, U32 value)
{
memcpy(memPtr, &value, sizeof(value));
ZSTD_memcpy(memPtr, &value, sizeof(value));
}
MEM_STATIC void MEM_write64(void* memPtr, U64 value)
{
memcpy(memPtr, &value, sizeof(value));
ZSTD_memcpy(memPtr, &value, sizeof(value));
}
#endif /* MEM_FORCE_MEMORY_ACCESS */
@@ -445,6 +415,9 @@ MEM_STATIC void MEM_writeBEST(void* memPtr, size_t val)
MEM_writeBE64(memPtr, (U64)val);
}
/* code only tested on 32 and 64 bits systems */
MEM_STATIC void MEM_check(void) { DEBUG_STATIC_ASSERT((sizeof(size_t)==4) || (sizeof(size_t)==8)); }
#if defined (__cplusplus)
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -10,9 +10,9 @@
/* ====== Dependencies ======= */
#include <stddef.h> /* size_t */
#include "zstd_deps.h" /* size_t */
#include "debug.h" /* assert */
#include "zstd_internal.h" /* ZSTD_malloc, ZSTD_free */
#include "zstd_internal.h" /* ZSTD_customMalloc, ZSTD_customFree */
#include "pool.h"
/* ====== Compiler specifics ====== */
@@ -105,6 +105,10 @@ static void* POOL_thread(void* opaque) {
assert(0); /* Unreachable */
}
POOL_ctx* ZSTD_createThreadPool(size_t numThreads) {
return POOL_create (numThreads, 0);
}
POOL_ctx* POOL_create(size_t numThreads, size_t queueSize) {
return POOL_create_advanced(numThreads, queueSize, ZSTD_defaultCMem);
}
@@ -115,14 +119,14 @@ POOL_ctx* POOL_create_advanced(size_t numThreads, size_t queueSize,
/* Check parameters */
if (!numThreads) { return NULL; }
/* Allocate the context and zero initialize */
ctx = (POOL_ctx*)ZSTD_calloc(sizeof(POOL_ctx), customMem);
ctx = (POOL_ctx*)ZSTD_customCalloc(sizeof(POOL_ctx), customMem);
if (!ctx) { return NULL; }
/* Initialize the job queue.
* It needs one extra space since one space is wasted to differentiate
* empty and full queues.
*/
ctx->queueSize = queueSize + 1;
ctx->queue = (POOL_job*)ZSTD_malloc(ctx->queueSize * sizeof(POOL_job), customMem);
ctx->queue = (POOL_job*)ZSTD_customMalloc(ctx->queueSize * sizeof(POOL_job), customMem);
ctx->queueHead = 0;
ctx->queueTail = 0;
ctx->numThreadsBusy = 0;
@@ -136,7 +140,7 @@ POOL_ctx* POOL_create_advanced(size_t numThreads, size_t queueSize,
}
ctx->shutdown = 0;
/* Allocate space for the thread handles */
ctx->threads = (ZSTD_pthread_t*)ZSTD_malloc(numThreads * sizeof(ZSTD_pthread_t), customMem);
ctx->threads = (ZSTD_pthread_t*)ZSTD_customMalloc(numThreads * sizeof(ZSTD_pthread_t), customMem);
ctx->threadCapacity = 0;
ctx->customMem = customMem;
/* Check for errors */
@@ -179,12 +183,14 @@ void POOL_free(POOL_ctx *ctx) {
ZSTD_pthread_mutex_destroy(&ctx->queueMutex);
ZSTD_pthread_cond_destroy(&ctx->queuePushCond);
ZSTD_pthread_cond_destroy(&ctx->queuePopCond);
ZSTD_free(ctx->queue, ctx->customMem);
ZSTD_free(ctx->threads, ctx->customMem);
ZSTD_free(ctx, ctx->customMem);
ZSTD_customFree(ctx->queue, ctx->customMem);
ZSTD_customFree(ctx->threads, ctx->customMem);
ZSTD_customFree(ctx, ctx->customMem);
}
void ZSTD_freeThreadPool (ZSTD_threadPool* pool) {
POOL_free (pool);
}
size_t POOL_sizeof(POOL_ctx *ctx) {
if (ctx==NULL) return 0; /* supports sizeof NULL */
@@ -203,11 +209,11 @@ static int POOL_resize_internal(POOL_ctx* ctx, size_t numThreads)
return 0;
}
/* numThreads > threadCapacity */
{ ZSTD_pthread_t* const threadPool = (ZSTD_pthread_t*)ZSTD_malloc(numThreads * sizeof(ZSTD_pthread_t), ctx->customMem);
{ ZSTD_pthread_t* const threadPool = (ZSTD_pthread_t*)ZSTD_customMalloc(numThreads * sizeof(ZSTD_pthread_t), ctx->customMem);
if (!threadPool) return 1;
/* replace existing thread pool */
memcpy(threadPool, ctx->threads, ctx->threadCapacity * sizeof(*threadPool));
ZSTD_free(ctx->threads, ctx->customMem);
ZSTD_memcpy(threadPool, ctx->threads, ctx->threadCapacity * sizeof(*threadPool));
ZSTD_customFree(ctx->threads, ctx->customMem);
ctx->threads = threadPool;
/* Initialize additional threads */
{ size_t threadId;
@@ -301,7 +307,7 @@ int POOL_tryAdd(POOL_ctx* ctx, POOL_function function, void* opaque)
struct POOL_ctx_s {
int dummy;
};
static POOL_ctx g_ctx;
static POOL_ctx g_poolCtx;
POOL_ctx* POOL_create(size_t numThreads, size_t queueSize) {
return POOL_create_advanced(numThreads, queueSize, ZSTD_defaultCMem);
@@ -311,11 +317,11 @@ POOL_ctx* POOL_create_advanced(size_t numThreads, size_t queueSize, ZSTD_customM
(void)numThreads;
(void)queueSize;
(void)customMem;
return &g_ctx;
return &g_poolCtx;
}
void POOL_free(POOL_ctx* ctx) {
assert(!ctx || ctx == &g_ctx);
assert(!ctx || ctx == &g_poolCtx);
(void)ctx;
}
@@ -337,7 +343,7 @@ int POOL_tryAdd(POOL_ctx* ctx, POOL_function function, void* opaque) {
size_t POOL_sizeof(POOL_ctx* ctx) {
if (ctx==NULL) return 0; /* supports sizeof NULL */
assert(ctx == &g_ctx);
assert(ctx == &g_poolCtx);
return sizeof(*ctx);
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -16,7 +16,7 @@ extern "C" {
#endif
#include <stddef.h> /* size_t */
#include "zstd_deps.h"
#define ZSTD_STATIC_LINKING_ONLY /* ZSTD_customMem */
#include "zstd.h"

View File

@@ -2,12 +2,13 @@
* Copyright (c) 2016 Tino Reichardt
* All rights reserved.
*
* You can contact the author at:
* - zstdmt source repository: https://github.com/mcmilk/zstdmt
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
*
* You can contact the author at:
* - zstdmt source repository: https://github.com/mcmilk/zstdmt
* You may select, at your option, one of the above-listed licenses.
*/
/**
@@ -77,11 +78,12 @@ int ZSTD_pthread_join(ZSTD_pthread_t thread, void **value_ptr)
#if defined(ZSTD_MULTITHREAD) && DEBUGLEVEL >= 1 && !defined(_WIN32)
#include <stdlib.h>
#define ZSTD_DEPS_NEED_MALLOC
#include "zstd_deps.h"
int ZSTD_pthread_mutex_init(ZSTD_pthread_mutex_t* mutex, pthread_mutexattr_t const* attr)
{
*mutex = (pthread_mutex_t*)malloc(sizeof(pthread_mutex_t));
*mutex = (pthread_mutex_t*)ZSTD_malloc(sizeof(pthread_mutex_t));
if (!*mutex)
return 1;
return pthread_mutex_init(*mutex, attr);
@@ -93,14 +95,14 @@ int ZSTD_pthread_mutex_destroy(ZSTD_pthread_mutex_t* mutex)
return 0;
{
int const ret = pthread_mutex_destroy(*mutex);
free(*mutex);
ZSTD_free(*mutex);
return ret;
}
}
int ZSTD_pthread_cond_init(ZSTD_pthread_cond_t* cond, pthread_condattr_t const* attr)
{
*cond = (pthread_cond_t*)malloc(sizeof(pthread_cond_t));
*cond = (pthread_cond_t*)ZSTD_malloc(sizeof(pthread_cond_t));
if (!*cond)
return 1;
return pthread_cond_init(*cond, attr);
@@ -112,7 +114,7 @@ int ZSTD_pthread_cond_destroy(ZSTD_pthread_cond_t* cond)
return 0;
{
int const ret = pthread_cond_destroy(*cond);
free(*cond);
ZSTD_free(*cond);
return ret;
}
}

View File

@@ -2,12 +2,13 @@
* Copyright (c) 2016 Tino Reichardt
* All rights reserved.
*
* You can contact the author at:
* - zstdmt source repository: https://github.com/mcmilk/zstdmt
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
*
* You can contact the author at:
* - zstdmt source repository: https://github.com/mcmilk/zstdmt
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef THREADING_H_938743

View File

@@ -1,6 +1,8 @@
#!/bin/bash
# Get utilities
yum -y -q -e 0 install wget tar unzip gcc
#yum -y -q -e 0 install wget tar unzip gcc
apt-get update
apt-get -y install wget tar unzip gcc
# Get Go
wget -q https://dl.google.com/go/go1.13.linux-386.tar.gz
@@ -13,5 +15,5 @@ unzip mr.zip
# Build and run tests
go build
PAYLOAD=$(pwd)/mr go test -v
PAYLOAD=$(pwd)/mr go test -bench .
DISABLE_BIG_TESTS=1 PAYLOAD=$(pwd)/mr go test -v
DISABLE_BIG_TESTS=1 PAYLOAD=$(pwd)/mr go test -bench .

View File

@@ -1,35 +1,15 @@
/*
* xxHash - Fast Hash algorithm
* Copyright (C) 2012-2016, Yann Collet
*
* BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are
* met:
*
* * Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following disclaimer
* in the documentation and/or other materials provided with the
* distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* You can contact the author at :
* - xxHash homepage: http://www.xxhash.com
* - xxHash source repository : https://github.com/Cyan4973/xxHash
* xxHash - Fast Hash algorithm
* Copyright (c) 2012-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - xxHash homepage: http://www.xxhash.com
* - xxHash source repository : https://github.com/Cyan4973/xxHash
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
@@ -97,14 +77,12 @@
* Includes & Memory related functions
***************************************/
/* Modify the local functions below should you wish to use some other memory routines */
/* for malloc(), free() */
#include <stdlib.h>
#include <stddef.h> /* size_t */
static void* XXH_malloc(size_t s) { return malloc(s); }
static void XXH_free (void* p) { free(p); }
/* for memcpy() */
#include <string.h>
static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcpy(dest,src,size); }
/* for ZSTD_malloc(), ZSTD_free() */
#define ZSTD_DEPS_NEED_MALLOC
#include "zstd_deps.h" /* size_t, ZSTD_malloc, ZSTD_free, ZSTD_memcpy */
static void* XXH_malloc(size_t s) { return ZSTD_malloc(s); }
static void XXH_free (void* p) { ZSTD_free(p); }
static void* XXH_memcpy(void* dest, const void* src, size_t size) { return ZSTD_memcpy(dest,src,size); }
#ifndef XXH_STATIC_LINKING_ONLY
# define XXH_STATIC_LINKING_ONLY
@@ -115,49 +93,13 @@ static void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcp
/* *************************************
* Compiler Specific Options
***************************************/
#if defined (__GNUC__) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */
# define INLINE_KEYWORD inline
#else
# define INLINE_KEYWORD
#endif
#if defined(__GNUC__) || defined(__ICCARM__)
# define FORCE_INLINE_ATTR __attribute__((always_inline))
#elif defined(_MSC_VER)
# define FORCE_INLINE_ATTR __forceinline
#else
# define FORCE_INLINE_ATTR
#endif
#define FORCE_INLINE_TEMPLATE static INLINE_KEYWORD FORCE_INLINE_ATTR
#ifdef _MSC_VER
# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */
#endif
#include "compiler.h"
/* *************************************
* Basic Types
***************************************/
#ifndef MEM_MODULE
# define MEM_MODULE
# if !defined (__VMS) && (defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) )
# include <stdint.h>
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef uint32_t U32;
typedef int32_t S32;
typedef uint64_t U64;
# else
typedef unsigned char BYTE;
typedef unsigned short U16;
typedef unsigned int U32;
typedef signed int S32;
typedef unsigned long long U64; /* if your compiler doesn't support unsigned long long, replace by another 64-bit type here. Note that xxhash.h will also need to be updated. */
# endif
#endif
#include "mem.h" /* BYTE, U32, U64, size_t */
#if (defined(XXH_FORCE_MEMORY_ACCESS) && (XXH_FORCE_MEMORY_ACCESS==2))
@@ -183,14 +125,14 @@ static U64 XXH_read64(const void* ptr) { return ((const unalign*)ptr)->u64; }
static U32 XXH_read32(const void* memPtr)
{
U32 val;
memcpy(&val, memPtr, sizeof(val));
ZSTD_memcpy(&val, memPtr, sizeof(val));
return val;
}
static U64 XXH_read64(const void* memPtr)
{
U64 val;
memcpy(&val, memPtr, sizeof(val));
ZSTD_memcpy(&val, memPtr, sizeof(val));
return val;
}
@@ -327,12 +269,12 @@ XXH_PUBLIC_API unsigned XXH_versionNumber (void) { return XXH_VERSION_NUMBER; }
****************************/
XXH_PUBLIC_API void XXH32_copyState(XXH32_state_t* restrict dstState, const XXH32_state_t* restrict srcState)
{
memcpy(dstState, srcState, sizeof(*dstState));
ZSTD_memcpy(dstState, srcState, sizeof(*dstState));
}
XXH_PUBLIC_API void XXH64_copyState(XXH64_state_t* restrict dstState, const XXH64_state_t* restrict srcState)
{
memcpy(dstState, srcState, sizeof(*dstState));
ZSTD_memcpy(dstState, srcState, sizeof(*dstState));
}
@@ -574,12 +516,12 @@ XXH_PUBLIC_API XXH_errorcode XXH64_freeState(XXH64_state_t* statePtr)
XXH_PUBLIC_API XXH_errorcode XXH32_reset(XXH32_state_t* statePtr, unsigned int seed)
{
XXH32_state_t state; /* using a local state to memcpy() in order to avoid strict-aliasing warnings */
memset(&state, 0, sizeof(state)-4); /* do not write into reserved, for future removal */
ZSTD_memset(&state, 0, sizeof(state)-4); /* do not write into reserved, for future removal */
state.v1 = seed + PRIME32_1 + PRIME32_2;
state.v2 = seed + PRIME32_2;
state.v3 = seed + 0;
state.v4 = seed - PRIME32_1;
memcpy(statePtr, &state, sizeof(state));
ZSTD_memcpy(statePtr, &state, sizeof(state));
return XXH_OK;
}
@@ -587,12 +529,12 @@ XXH_PUBLIC_API XXH_errorcode XXH32_reset(XXH32_state_t* statePtr, unsigned int s
XXH_PUBLIC_API XXH_errorcode XXH64_reset(XXH64_state_t* statePtr, unsigned long long seed)
{
XXH64_state_t state; /* using a local state to memcpy() in order to avoid strict-aliasing warnings */
memset(&state, 0, sizeof(state)-8); /* do not write into reserved, for future removal */
ZSTD_memset(&state, 0, sizeof(state)-8); /* do not write into reserved, for future removal */
state.v1 = seed + PRIME64_1 + PRIME64_2;
state.v2 = seed + PRIME64_2;
state.v3 = seed + 0;
state.v4 = seed - PRIME64_1;
memcpy(statePtr, &state, sizeof(state));
ZSTD_memcpy(statePtr, &state, sizeof(state));
return XXH_OK;
}
@@ -729,7 +671,9 @@ FORCE_INLINE_TEMPLATE XXH_errorcode XXH64_update_endian (XXH64_state_t* state, c
state->total_len += len;
if (state->memsize + len < 32) { /* fill in tmp buffer */
XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, len);
if (input != NULL) {
XXH_memcpy(((BYTE*)state->mem64) + state->memsize, input, len);
}
state->memsize += (U32)len;
return XXH_OK;
}
@@ -861,14 +805,14 @@ XXH_PUBLIC_API void XXH32_canonicalFromHash(XXH32_canonical_t* dst, XXH32_hash_t
{
XXH_STATIC_ASSERT(sizeof(XXH32_canonical_t) == sizeof(XXH32_hash_t));
if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap32(hash);
memcpy(dst, &hash, sizeof(*dst));
ZSTD_memcpy(dst, &hash, sizeof(*dst));
}
XXH_PUBLIC_API void XXH64_canonicalFromHash(XXH64_canonical_t* dst, XXH64_hash_t hash)
{
XXH_STATIC_ASSERT(sizeof(XXH64_canonical_t) == sizeof(XXH64_hash_t));
if (XXH_CPU_LITTLE_ENDIAN) hash = XXH_swap64(hash);
memcpy(dst, &hash, sizeof(*dst));
ZSTD_memcpy(dst, &hash, sizeof(*dst));
}
XXH_PUBLIC_API XXH32_hash_t XXH32_hashFromCanonical(const XXH32_canonical_t* src)

View File

@@ -1,35 +1,15 @@
/*
xxHash - Extremely Fast Hash algorithm
Header File
Copyright (C) 2012-2016, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- xxHash source repository : https://github.com/Cyan4973/xxHash
* xxHash - Extremely Fast Hash algorithm
* Header File
* Copyright (c) 2012-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - xxHash source repository : https://github.com/Cyan4973/xxHash
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
/* Notice extracted from xxHash homepage :
@@ -75,7 +55,7 @@ extern "C" {
/* ****************************
* Definitions
******************************/
#include <stddef.h> /* size_t */
#include "zstd_deps.h"
typedef enum { XXH_OK=0, XXH_ERROR } XXH_errorcode;

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -48,6 +48,7 @@
# define ZDICT_STATIC_LINKING_ONLY
#endif
#include "zdict.h"
#include "zstd_compress_internal.h" /* ZSTD_loadCEntropy() */
/*-*************************************
@@ -61,14 +62,15 @@
#define NOISELENGTH 32
static const int g_compressionLevel_default = 3;
static const U32 g_selectivity_default = 9;
/*-*************************************
* Console display
***************************************/
#undef DISPLAY
#define DISPLAY(...) { fprintf(stderr, __VA_ARGS__); fflush( stderr ); }
#undef DISPLAYLEVEL
#define DISPLAYLEVEL(l, ...) if (notificationLevel>=l) { DISPLAY(__VA_ARGS__); } /* 0 : no display; 1: errors; 2: default; 3: details; 4: debug */
static clock_t ZDICT_clockSpan(clock_t nPrevious) { return clock() - nPrevious; }
@@ -99,6 +101,26 @@ unsigned ZDICT_getDictID(const void* dictBuffer, size_t dictSize)
return MEM_readLE32((const char*)dictBuffer + 4);
}
size_t ZDICT_getDictHeaderSize(const void* dictBuffer, size_t dictSize)
{
size_t headerSize;
if (dictSize <= 8 || MEM_readLE32(dictBuffer) != ZSTD_MAGIC_DICTIONARY) return ERROR(dictionary_corrupted);
{ ZSTD_compressedBlockState_t* bs = (ZSTD_compressedBlockState_t*)malloc(sizeof(ZSTD_compressedBlockState_t));
U32* wksp = (U32*)malloc(HUF_WORKSPACE_SIZE);
if (!bs || !wksp) {
headerSize = ERROR(memory_allocation);
} else {
ZSTD_reset_compressedBlockState(bs);
headerSize = ZSTD_loadCEntropy(bs, wksp, dictBuffer, dictSize);
}
free(bs);
free(wksp);
}
return headerSize;
}
/*-********************************************************
* Dictionary training functions
@@ -508,6 +530,7 @@ static size_t ZDICT_trainBuffer_legacy(dictItem* dictList, U32 dictListSize,
clock_t displayClock = 0;
clock_t const refreshRate = CLOCKS_PER_SEC * 3 / 10;
# undef DISPLAYUPDATE
# define DISPLAYUPDATE(l, ...) if (notificationLevel>=l) { \
if (ZDICT_clockSpan(displayClock) > refreshRate) \
{ displayClock = clock(); DISPLAY(__VA_ARGS__); \
@@ -588,12 +611,12 @@ typedef struct
#define MAXREPOFFSET 1024
static void ZDICT_countEStats(EStats_ress_t esr, ZSTD_parameters params,
static void ZDICT_countEStats(EStats_ress_t esr, const ZSTD_parameters* params,
unsigned* countLit, unsigned* offsetcodeCount, unsigned* matchlengthCount, unsigned* litlengthCount, U32* repOffsets,
const void* src, size_t srcSize,
U32 notificationLevel)
{
size_t const blockSizeMax = MIN (ZSTD_BLOCKSIZE_MAX, 1 << params.cParams.windowLog);
size_t const blockSizeMax = MIN (ZSTD_BLOCKSIZE_MAX, 1 << params->cParams.windowLog);
size_t cSize;
if (srcSize > blockSizeMax) srcSize = blockSizeMax; /* protection vs large samples */
@@ -682,7 +705,7 @@ static void ZDICT_flatLit(unsigned* countLit)
#define OFFCODE_MAX 30 /* only applicable to first block */
static size_t ZDICT_analyzeEntropy(void* dstBuffer, size_t maxDstSize,
unsigned compressionLevel,
int compressionLevel,
const void* srcBuffer, const size_t* fileSizes, unsigned nbFiles,
const void* dictBuffer, size_t dictBufferSize,
unsigned notificationLevel)
@@ -717,7 +740,7 @@ static size_t ZDICT_analyzeEntropy(void* dstBuffer, size_t maxDstSize,
memset(repOffset, 0, sizeof(repOffset));
repOffset[1] = repOffset[4] = repOffset[8] = 1;
memset(bestRepOffset, 0, sizeof(bestRepOffset));
if (compressionLevel==0) compressionLevel = g_compressionLevel_default;
if (compressionLevel==0) compressionLevel = ZSTD_CLEVEL_DEFAULT;
params = ZSTD_getParams(compressionLevel, averageSampleSize, dictBufferSize);
esr.dict = ZSTD_createCDict_advanced(dictBuffer, dictBufferSize, ZSTD_dlm_byRef, ZSTD_dct_rawContent, params.cParams, ZSTD_defaultCMem);
@@ -731,7 +754,7 @@ static size_t ZDICT_analyzeEntropy(void* dstBuffer, size_t maxDstSize,
/* collect stats on all samples */
for (u=0; u<nbFiles; u++) {
ZDICT_countEStats(esr, params,
ZDICT_countEStats(esr, &params,
countLit, offcodeCount, matchLengthCount, litLengthCount, repOffset,
(const char*)srcBuffer + pos, fileSizes[u],
notificationLevel);
@@ -762,7 +785,7 @@ static size_t ZDICT_analyzeEntropy(void* dstBuffer, size_t maxDstSize,
/* note : the result of this phase should be used to better appreciate the impact on statistics */
total=0; for (u=0; u<=offcodeMax; u++) total+=offcodeCount[u];
errorCode = FSE_normalizeCount(offcodeNCount, Offlog, offcodeCount, total, offcodeMax);
errorCode = FSE_normalizeCount(offcodeNCount, Offlog, offcodeCount, total, offcodeMax, /* useLowProbCount */ 1);
if (FSE_isError(errorCode)) {
eSize = errorCode;
DISPLAYLEVEL(1, "FSE_normalizeCount error with offcodeCount \n");
@@ -771,7 +794,7 @@ static size_t ZDICT_analyzeEntropy(void* dstBuffer, size_t maxDstSize,
Offlog = (U32)errorCode;
total=0; for (u=0; u<=MaxML; u++) total+=matchLengthCount[u];
errorCode = FSE_normalizeCount(matchLengthNCount, mlLog, matchLengthCount, total, MaxML);
errorCode = FSE_normalizeCount(matchLengthNCount, mlLog, matchLengthCount, total, MaxML, /* useLowProbCount */ 1);
if (FSE_isError(errorCode)) {
eSize = errorCode;
DISPLAYLEVEL(1, "FSE_normalizeCount error with matchLengthCount \n");
@@ -780,7 +803,7 @@ static size_t ZDICT_analyzeEntropy(void* dstBuffer, size_t maxDstSize,
mlLog = (U32)errorCode;
total=0; for (u=0; u<=MaxLL; u++) total+=litLengthCount[u];
errorCode = FSE_normalizeCount(litLengthNCount, llLog, litLengthCount, total, MaxLL);
errorCode = FSE_normalizeCount(litLengthNCount, llLog, litLengthCount, total, MaxLL, /* useLowProbCount */ 1);
if (FSE_isError(errorCode)) {
eSize = errorCode;
DISPLAYLEVEL(1, "FSE_normalizeCount error with litLengthCount \n");
@@ -869,7 +892,7 @@ size_t ZDICT_finalizeDictionary(void* dictBuffer, size_t dictBufferCapacity,
size_t hSize;
#define HBUFFSIZE 256 /* should prove large enough for all entropy headers */
BYTE header[HBUFFSIZE];
int const compressionLevel = (params.compressionLevel == 0) ? g_compressionLevel_default : params.compressionLevel;
int const compressionLevel = (params.compressionLevel == 0) ? ZSTD_CLEVEL_DEFAULT : params.compressionLevel;
U32 const notificationLevel = params.notificationLevel;
/* check conditions */
@@ -915,7 +938,7 @@ static size_t ZDICT_addEntropyTablesFromBuffer_advanced(
const void* samplesBuffer, const size_t* samplesSizes, unsigned nbSamples,
ZDICT_params_t params)
{
int const compressionLevel = (params.compressionLevel == 0) ? g_compressionLevel_default : params.compressionLevel;
int const compressionLevel = (params.compressionLevel == 0) ? ZSTD_CLEVEL_DEFAULT : params.compressionLevel;
U32 const notificationLevel = params.notificationLevel;
size_t hSize = 8;
@@ -1090,8 +1113,8 @@ size_t ZDICT_trainFromBuffer(void* dictBuffer, size_t dictBufferCapacity,
memset(&params, 0, sizeof(params));
params.d = 8;
params.steps = 4;
/* Default to level 6 since no compression level information is available */
params.zParams.compressionLevel = 3;
/* Use default level since no compression level information is available */
params.zParams.compressionLevel = ZSTD_CLEVEL_DEFAULT;
#if defined(DEBUGLEVEL) && (DEBUGLEVEL>=1)
params.zParams.notificationLevel = DEBUGLEVEL;
#endif

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -61,9 +61,57 @@ ZDICTLIB_API size_t ZDICT_trainFromBuffer(void* dictBuffer, size_t dictBufferCap
const void* samplesBuffer,
const size_t* samplesSizes, unsigned nbSamples);
typedef struct {
int compressionLevel; /*< optimize for a specific zstd compression level; 0 means default */
unsigned notificationLevel; /*< Write log to stderr; 0 = none (default); 1 = errors; 2 = progression; 3 = details; 4 = debug; */
unsigned dictID; /*< force dictID value; 0 means auto mode (32-bits random value) */
} ZDICT_params_t;
/*! ZDICT_finalizeDictionary():
* Given a custom content as a basis for dictionary, and a set of samples,
* finalize dictionary by adding headers and statistics according to the zstd
* dictionary format.
*
* Samples must be stored concatenated in a flat buffer `samplesBuffer`,
* supplied with an array of sizes `samplesSizes`, providing the size of each
* sample in order. The samples are used to construct the statistics, so they
* should be representative of what you will compress with this dictionary.
*
* The compression level can be set in `parameters`. You should pass the
* compression level you expect to use in production. The statistics for each
* compression level differ, so tuning the dictionary for the compression level
* can help quite a bit.
*
* You can set an explicit dictionary ID in `parameters`, or allow us to pick
* a random dictionary ID for you, but we can't guarantee no collisions.
*
* The dstDictBuffer and the dictContent may overlap, and the content will be
* appended to the end of the header. If the header + the content doesn't fit in
* maxDictSize the beginning of the content is truncated to make room, since it
* is presumed that the most profitable content is at the end of the dictionary,
* since that is the cheapest to reference.
*
* `dictContentSize` must be >= ZDICT_CONTENTSIZE_MIN bytes.
* `maxDictSize` must be >= max(dictContentSize, ZSTD_DICTSIZE_MIN).
*
* @return: size of dictionary stored into `dstDictBuffer` (<= `maxDictSize`),
* or an error code, which can be tested by ZDICT_isError().
* Note: ZDICT_finalizeDictionary() will push notifications into stderr if
* instructed to, using notificationLevel>0.
* NOTE: This function currently may fail in several edge cases including:
* * Not enough samples
* * Samples are uncompressible
* * Samples are all exactly the same
*/
ZDICTLIB_API size_t ZDICT_finalizeDictionary(void* dstDictBuffer, size_t maxDictSize,
const void* dictContent, size_t dictContentSize,
const void* samplesBuffer, const size_t* samplesSizes, unsigned nbSamples,
ZDICT_params_t parameters);
/*====== Helper functions ======*/
ZDICTLIB_API unsigned ZDICT_getDictID(const void* dictBuffer, size_t dictSize); /**< extracts dictID; @return zero if error (not a valid dictionary) */
ZDICTLIB_API size_t ZDICT_getDictHeaderSize(const void* dictBuffer, size_t dictSize); /* returns dict header size; returns a ZSTD error code on failure */
ZDICTLIB_API unsigned ZDICT_isError(size_t errorCode);
ZDICTLIB_API const char* ZDICT_getErrorName(size_t errorCode);
@@ -78,11 +126,8 @@ ZDICTLIB_API const char* ZDICT_getErrorName(size_t errorCode);
* Use them only in association with static linking.
* ==================================================================================== */
typedef struct {
int compressionLevel; /* optimize for a specific zstd compression level; 0 means default */
unsigned notificationLevel; /* Write log to stderr; 0 = none (default); 1 = errors; 2 = progression; 3 = details; 4 = debug; */
unsigned dictID; /* force dictID value; 0 means auto mode (32-bits random value) */
} ZDICT_params_t;
#define ZDICT_CONTENTSIZE_MIN 128
#define ZDICT_DICTSIZE_MIN 256
/*! ZDICT_cover_params_t:
* k and d are the only required parameters.
@@ -198,28 +243,6 @@ ZDICTLIB_API size_t ZDICT_optimizeTrainFromBuffer_fastCover(void* dictBuffer,
const size_t* samplesSizes, unsigned nbSamples,
ZDICT_fastCover_params_t* parameters);
/*! ZDICT_finalizeDictionary():
* Given a custom content as a basis for dictionary, and a set of samples,
* finalize dictionary by adding headers and statistics.
*
* Samples must be stored concatenated in a flat buffer `samplesBuffer`,
* supplied with an array of sizes `samplesSizes`, providing the size of each sample in order.
*
* dictContentSize must be >= ZDICT_CONTENTSIZE_MIN bytes.
* maxDictSize must be >= dictContentSize, and must be >= ZDICT_DICTSIZE_MIN bytes.
*
* @return: size of dictionary stored into `dictBuffer` (<= `dictBufferCapacity`),
* or an error code, which can be tested by ZDICT_isError().
* Note: ZDICT_finalizeDictionary() will push notifications into stderr if instructed to, using notificationLevel>0.
* Note 2: dictBuffer and dictContent can overlap
*/
#define ZDICT_CONTENTSIZE_MIN 128
#define ZDICT_DICTSIZE_MIN 256
ZDICTLIB_API size_t ZDICT_finalizeDictionary(void* dictBuffer, size_t dictBufferCapacity,
const void* dictContent, size_t dictContentSize,
const void* samplesBuffer, const size_t* samplesSizes, unsigned nbSamples,
ZDICT_params_t parameters);
typedef struct {
unsigned selectivityLevel; /* 0 means default; larger => select more => larger dictionary */
ZDICT_params_t zParams;
@@ -256,7 +279,7 @@ ZDICTLIB_API size_t ZDICT_trainFromBuffer_legacy(
# define ZDICT_GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
# if defined (__cplusplus) && (__cplusplus >= 201402) /* C++14 or greater */
# define ZDICT_DEPRECATED(message) [[deprecated(message)]] ZDICTLIB_API
# elif (ZDICT_GCC_VERSION >= 405) || defined(__clang__)
# elif defined(__clang__) || (ZDICT_GCC_VERSION >= 405)
# define ZDICT_DEPRECATED(message) ZDICTLIB_API __attribute__((deprecated(message)))
# elif (ZDICT_GCC_VERSION >= 301)
# define ZDICT_DEPRECATED(message) ZDICTLIB_API __attribute__((deprecated))

View File

@@ -3,27 +3,12 @@ package zstd
/*
#define ZSTD_STATIC_LINKING_ONLY
#include "zstd.h"
#include "stdint.h" // for uintptr_t
// The following *_wrapper function are used for removing superflouos
// memory allocations when calling the wrapped functions from Go code.
// See https://github.com/golang/go/issues/24450 for details.
static size_t ZSTD_compress_wrapper(uintptr_t dst, size_t maxDstSize, const uintptr_t src, size_t srcSize, int compressionLevel) {
return ZSTD_compress((void*)dst, maxDstSize, (const void*)src, srcSize, compressionLevel);
}
static size_t ZSTD_decompress_wrapper(uintptr_t dst, size_t maxDstSize, uintptr_t src, size_t srcSize) {
return ZSTD_decompress((void*)dst, maxDstSize, (const void *)src, srcSize);
}
*/
import "C"
import (
"bytes"
"errors"
"io/ioutil"
"runtime"
"unsafe"
)
@@ -73,19 +58,18 @@ func CompressLevel(dst, src []byte, level int) ([]byte, error) {
dst = make([]byte, bound)
}
srcPtr := C.uintptr_t(uintptr(0)) // Do not point anywhere, if src is empty
var srcPtr *byte // Do not point anywhere, if src is empty
if len(src) > 0 {
srcPtr = C.uintptr_t(uintptr(unsafe.Pointer(&src[0])))
srcPtr = &src[0]
}
cWritten := C.ZSTD_compress_wrapper(
C.uintptr_t(uintptr(unsafe.Pointer(&dst[0]))),
cWritten := C.ZSTD_compress(
unsafe.Pointer(&dst[0]),
C.size_t(len(dst)),
srcPtr,
unsafe.Pointer(srcPtr),
C.size_t(len(src)),
C.int(level))
runtime.KeepAlive(src)
written := int(cWritten)
// Check if the return is an Error code
if err := getError(written); err != nil {
@@ -103,13 +87,12 @@ func Decompress(dst, src []byte) ([]byte, error) {
}
decompress := func(dst, src []byte) ([]byte, error) {
cWritten := C.ZSTD_decompress_wrapper(
C.uintptr_t(uintptr(unsafe.Pointer(&dst[0]))),
cWritten := C.ZSTD_decompress(
unsafe.Pointer(&dst[0]),
C.size_t(len(dst)),
C.uintptr_t(uintptr(unsafe.Pointer(&src[0]))),
unsafe.Pointer(&src[0]),
C.size_t(len(src)))
runtime.KeepAlive(src)
written := int(cWritten)
// Check error
if err := getError(written); err != nil {
@@ -120,8 +103,7 @@ func Decompress(dst, src []byte) ([]byte, error) {
if len(dst) == 0 {
// Attempt to use zStd to determine decompressed size (may result in error or 0)
size := int(C.size_t(C.ZSTD_getDecompressedSize(unsafe.Pointer(&src[0]), C.size_t(len(src)))))
size := int(C.ZSTD_getDecompressedSize(unsafe.Pointer(&src[0]), C.size_t(len(src))))
if err := getError(size); err != nil {
return nil, err
}

474
vendor/github.com/DataDog/zstd/zstd.h generated vendored
View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -72,16 +72,21 @@ extern "C" {
/*------ Version ------*/
#define ZSTD_VERSION_MAJOR 1
#define ZSTD_VERSION_MINOR 4
#define ZSTD_VERSION_RELEASE 4
#define ZSTD_VERSION_RELEASE 8
#define ZSTD_VERSION_NUMBER (ZSTD_VERSION_MAJOR *100*100 + ZSTD_VERSION_MINOR *100 + ZSTD_VERSION_RELEASE)
ZSTDLIB_API unsigned ZSTD_versionNumber(void); /**< to check runtime library version */
/*! ZSTD_versionNumber() :
* Return runtime library version, the value is (MAJOR*100*100 + MINOR*100 + RELEASE). */
ZSTDLIB_API unsigned ZSTD_versionNumber(void);
#define ZSTD_LIB_VERSION ZSTD_VERSION_MAJOR.ZSTD_VERSION_MINOR.ZSTD_VERSION_RELEASE
#define ZSTD_QUOTE(str) #str
#define ZSTD_EXPAND_AND_QUOTE(str) ZSTD_QUOTE(str)
#define ZSTD_VERSION_STRING ZSTD_EXPAND_AND_QUOTE(ZSTD_LIB_VERSION)
ZSTDLIB_API const char* ZSTD_versionString(void); /* requires v1.3.0+ */
/*! ZSTD_versionString() :
* Return runtime library version, like "1.4.5". Requires v1.3.0+. */
ZSTDLIB_API const char* ZSTD_versionString(void);
/* *************************************
* Default constant
@@ -274,7 +279,10 @@ typedef enum {
* Default level is ZSTD_CLEVEL_DEFAULT==3.
* Special: value 0 means default, which is controlled by ZSTD_CLEVEL_DEFAULT.
* Note 1 : it's possible to pass a negative compression level.
* Note 2 : setting a level resets all other compression parameters to default */
* Note 2 : setting a level does not automatically set all other compression parameters
* to default. Setting this will however eventually dynamically impact the compression
* parameters which have not been manually set. The manually set
* ones will 'stick'. */
/* Advanced compression parameters :
* It's possible to pin down compression parameters to some specific values.
* In which case, these values are no longer dynamically selected by the compressor */
@@ -331,7 +339,9 @@ typedef enum {
* for large inputs, by finding large matches at long distance.
* It increases memory usage and window size.
* Note: enabling this parameter increases default ZSTD_c_windowLog to 128 MB
* except when expressly set to a different value. */
* except when expressly set to a different value.
* Note: will be enabled by default if ZSTD_c_windowLog >= 128 MB and
* compression strategy >= ZSTD_btopt (== compression level 16+) */
ZSTD_c_ldmHashLog=161, /* Size of the table for long distance matching, as a power of 2.
* Larger values increase memory usage and compression ratio,
* but decrease compression speed.
@@ -362,16 +372,20 @@ typedef enum {
ZSTD_c_dictIDFlag=202, /* When applicable, dictionary's ID is written into frame header (default:1) */
/* multi-threading parameters */
/* These parameters are only useful if multi-threading is enabled (compiled with build macro ZSTD_MULTITHREAD).
* They return an error otherwise. */
/* These parameters are only active if multi-threading is enabled (compiled with build macro ZSTD_MULTITHREAD).
* Otherwise, trying to set any other value than default (0) will be a no-op and return an error.
* In a situation where it's unknown if the linked library supports multi-threading or not,
* setting ZSTD_c_nbWorkers to any value >= 1 and consulting the return value provides a quick way to check this property.
*/
ZSTD_c_nbWorkers=400, /* Select how many threads will be spawned to compress in parallel.
* When nbWorkers >= 1, triggers asynchronous mode when used with ZSTD_compressStream*() :
* When nbWorkers >= 1, triggers asynchronous mode when invoking ZSTD_compressStream*() :
* ZSTD_compressStream*() consumes input and flush output if possible, but immediately gives back control to caller,
* while compression work is performed in parallel, within worker threads.
* while compression is performed in parallel, within worker thread(s).
* (note : a strong exception to this rule is when first invocation of ZSTD_compressStream2() sets ZSTD_e_end :
* in which case, ZSTD_compressStream2() delegates to ZSTD_compress2(), which is always a blocking call).
* More workers improve speed, but also increase memory usage.
* Default value is `0`, aka "single-threaded mode" : no worker is spawned, compression is performed inside Caller's thread, all invocations are blocking */
* Default value is `0`, aka "single-threaded mode" : no worker is spawned,
* compression is performed inside Caller's thread, and all invocations are blocking */
ZSTD_c_jobSize=401, /* Size of a compression job. This value is enforced only when nbWorkers >= 1.
* Each compression job is completed in parallel, so this value can indirectly impact the nb of active threads.
* 0 means default, which is dynamically determined based on compression parameters.
@@ -400,6 +414,11 @@ typedef enum {
* ZSTD_c_literalCompressionMode
* ZSTD_c_targetCBlockSize
* ZSTD_c_srcSizeHint
* ZSTD_c_enableDedicatedDictSearch
* ZSTD_c_stableInBuffer
* ZSTD_c_stableOutBuffer
* ZSTD_c_blockDelimiters
* ZSTD_c_validateSequences
* Because they are not stable, it's necessary to define ZSTD_STATIC_LINKING_ONLY to access them.
* note : never ever use experimentalParam? names directly;
* also, the enums values themselves are unstable and can still change.
@@ -410,7 +429,12 @@ typedef enum {
ZSTD_c_experimentalParam4=1001,
ZSTD_c_experimentalParam5=1002,
ZSTD_c_experimentalParam6=1003,
ZSTD_c_experimentalParam7=1004
ZSTD_c_experimentalParam7=1004,
ZSTD_c_experimentalParam8=1005,
ZSTD_c_experimentalParam9=1006,
ZSTD_c_experimentalParam10=1007,
ZSTD_c_experimentalParam11=1008,
ZSTD_c_experimentalParam12=1009
} ZSTD_cParameter;
typedef struct {
@@ -519,11 +543,15 @@ typedef enum {
/* note : additional experimental parameters are also available
* within the experimental section of the API.
* At the time of this writing, they include :
* ZSTD_c_format
* ZSTD_d_format
* ZSTD_d_stableOutBuffer
* ZSTD_d_forceIgnoreChecksum
* Because they are not stable, it's necessary to define ZSTD_STATIC_LINKING_ONLY to access them.
* note : never ever use experimentalParam? names directly
*/
ZSTD_d_experimentalParam1=1000
ZSTD_d_experimentalParam1=1000,
ZSTD_d_experimentalParam2=1001,
ZSTD_d_experimentalParam3=1002
} ZSTD_dParameter;
@@ -659,8 +687,9 @@ typedef enum {
* - Compression parameters cannot be changed once compression is started (save a list of exceptions in multi-threading mode)
* - output->pos must be <= dstCapacity, input->pos must be <= srcSize
* - output->pos and input->pos will be updated. They are guaranteed to remain below their respective limit.
* - endOp must be a valid directive
* - When nbWorkers==0 (default), function is blocking : it completes its job before returning to caller.
* - When nbWorkers>=1, function is non-blocking : it just acquires a copy of input, and distributes jobs to internal worker threads, flush whatever is available,
* - When nbWorkers>=1, function is non-blocking : it copies a portion of input, distributes jobs to internal worker threads, flush to output whatever is available,
* and then immediately returns, just indicating that there is some data remaining to be flushed.
* The function nonetheless guarantees forward progress : it will return only after it reads or write at least 1+ byte.
* - Exception : if the first call requests a ZSTD_e_end directive and provides enough dstCapacity, the function delegates to ZSTD_compress2() which is always blocking.
@@ -763,7 +792,7 @@ ZSTDLIB_API size_t ZSTD_freeDStream(ZSTD_DStream* zds);
/* This function is redundant with the advanced API and equivalent to:
*
* ZSTD_DCtx_reset(zds);
* ZSTD_DCtx_reset(zds, ZSTD_reset_session_only);
* ZSTD_DCtx_refDDict(zds, NULL);
*/
ZSTDLIB_API size_t ZSTD_initDStream(ZSTD_DStream* zds);
@@ -1095,21 +1124,40 @@ ZSTDLIB_API size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict);
typedef struct ZSTD_CCtx_params_s ZSTD_CCtx_params;
typedef struct {
unsigned int matchPos; /* Match pos in dst */
/* If seqDef.offset > 3, then this is seqDef.offset - 3
* If seqDef.offset < 3, then this is the corresponding repeat offset
* But if seqDef.offset < 3 and litLength == 0, this is the
* repeat offset before the corresponding repeat offset
* And if seqDef.offset == 3 and litLength == 0, this is the
* most recent repeat offset - 1
*/
unsigned int offset;
unsigned int litLength; /* Literal length */
unsigned int matchLength; /* Match length */
/* 0 when seq not rep and seqDef.offset otherwise
* when litLength == 0 this will be <= 4, otherwise <= 3 like normal
*/
unsigned int rep;
unsigned int offset; /* The offset of the match. (NOT the same as the offset code)
* If offset == 0 and matchLength == 0, this sequence represents the last
* literals in the block of litLength size.
*/
unsigned int litLength; /* Literal length of the sequence. */
unsigned int matchLength; /* Match length of the sequence. */
/* Note: Users of this API may provide a sequence with matchLength == litLength == offset == 0.
* In this case, we will treat the sequence as a marker for a block boundary.
*/
unsigned int rep; /* Represents which repeat offset is represented by the field 'offset'.
* Ranges from [0, 3].
*
* Repeat offsets are essentially previous offsets from previous sequences sorted in
* recency order. For more detail, see doc/zstd_compression_format.md
*
* If rep == 0, then 'offset' does not contain a repeat offset.
* If rep > 0:
* If litLength != 0:
* rep == 1 --> offset == repeat_offset_1
* rep == 2 --> offset == repeat_offset_2
* rep == 3 --> offset == repeat_offset_3
* If litLength == 0:
* rep == 1 --> offset == repeat_offset_2
* rep == 2 --> offset == repeat_offset_3
* rep == 3 --> offset == repeat_offset_1 - 1
*
* Note: This field is optional. ZSTD_generateSequences() will calculate the value of
* 'rep', but repeat offsets do not necessarily need to be calculated from an external
* sequence provider's perspective. For example, ZSTD_compressSequences() does not
* use this 'rep' field at all (as of now).
*/
} ZSTD_Sequence;
typedef struct {
@@ -1151,6 +1199,12 @@ typedef enum {
* Decoder cannot recognise automatically this format, requiring this instruction. */
} ZSTD_format_e;
typedef enum {
/* Note: this enum controls ZSTD_d_forceIgnoreChecksum */
ZSTD_d_validateChecksum = 0,
ZSTD_d_ignoreChecksum = 1
} ZSTD_forceIgnoreChecksum_e;
typedef enum {
/* Note: this enum and the behavior it controls are effectively internal
* implementation details of the compressor. They are expected to continue
@@ -1248,14 +1302,74 @@ ZSTDLIB_API unsigned long long ZSTD_decompressBound(const void* src, size_t srcS
* or an error code (if srcSize is too small) */
ZSTDLIB_API size_t ZSTD_frameHeaderSize(const void* src, size_t srcSize);
/*! ZSTD_getSequences() :
* Extract sequences from the sequence store
typedef enum {
ZSTD_sf_noBlockDelimiters = 0, /* Representation of ZSTD_Sequence has no block delimiters, sequences only */
ZSTD_sf_explicitBlockDelimiters = 1 /* Representation of ZSTD_Sequence contains explicit block delimiters */
} ZSTD_sequenceFormat_e;
/*! ZSTD_generateSequences() :
* Generate sequences using ZSTD_compress2, given a source buffer.
*
* Each block will end with a dummy sequence
* with offset == 0, matchLength == 0, and litLength == length of last literals.
* litLength may be == 0, and if so, then the sequence of (of: 0 ml: 0 ll: 0)
* simply acts as a block delimiter.
*
* zc can be used to insert custom compression params.
* This function invokes ZSTD_compress2
* @return : number of sequences extracted
*
* The output of this function can be fed into ZSTD_compressSequences() with CCtx
* setting of ZSTD_c_blockDelimiters as ZSTD_sf_explicitBlockDelimiters
* @return : number of sequences generated
*/
ZSTDLIB_API size_t ZSTD_getSequences(ZSTD_CCtx* zc, ZSTD_Sequence* outSeqs,
size_t outSeqsSize, const void* src, size_t srcSize);
ZSTDLIB_API size_t ZSTD_generateSequences(ZSTD_CCtx* zc, ZSTD_Sequence* outSeqs,
size_t outSeqsSize, const void* src, size_t srcSize);
/*! ZSTD_mergeBlockDelimiters() :
* Given an array of ZSTD_Sequence, remove all sequences that represent block delimiters/last literals
* by merging them into into the literals of the next sequence.
*
* As such, the final generated result has no explicit representation of block boundaries,
* and the final last literals segment is not represented in the sequences.
*
* The output of this function can be fed into ZSTD_compressSequences() with CCtx
* setting of ZSTD_c_blockDelimiters as ZSTD_sf_noBlockDelimiters
* @return : number of sequences left after merging
*/
ZSTDLIB_API size_t ZSTD_mergeBlockDelimiters(ZSTD_Sequence* sequences, size_t seqsSize);
/*! ZSTD_compressSequences() :
* Compress an array of ZSTD_Sequence, generated from the original source buffer, into dst.
* If a dictionary is included, then the cctx should reference the dict. (see: ZSTD_CCtx_refCDict(), ZSTD_CCtx_loadDictionary(), etc.)
* The entire source is compressed into a single frame.
*
* The compression behavior changes based on cctx params. In particular:
* If ZSTD_c_blockDelimiters == ZSTD_sf_noBlockDelimiters, the array of ZSTD_Sequence is expected to contain
* no block delimiters (defined in ZSTD_Sequence). Block boundaries are roughly determined based on
* the block size derived from the cctx, and sequences may be split. This is the default setting.
*
* If ZSTD_c_blockDelimiters == ZSTD_sf_explicitBlockDelimiters, the array of ZSTD_Sequence is expected to contain
* block delimiters (defined in ZSTD_Sequence). Behavior is undefined if no block delimiters are provided.
*
* If ZSTD_c_validateSequences == 0, this function will blindly accept the sequences provided. Invalid sequences cause undefined
* behavior. If ZSTD_c_validateSequences == 1, then if sequence is invalid (see doc/zstd_compression_format.md for
* specifics regarding offset/matchlength requirements) then the function will bail out and return an error.
*
* In addition to the two adjustable experimental params, there are other important cctx params.
* - ZSTD_c_minMatch MUST be set as less than or equal to the smallest match generated by the match finder. It has a minimum value of ZSTD_MINMATCH_MIN.
* - ZSTD_c_compressionLevel accordingly adjusts the strength of the entropy coder, as it would in typical compression.
* - ZSTD_c_windowLog affects offset validation: this function will return an error at higher debug levels if a provided offset
* is larger than what the spec allows for a given window log and dictionary (if present). See: doc/zstd_compression_format.md
*
* Note: Repcodes are, as of now, always re-calculated within this function, so ZSTD_Sequence::rep is unused.
* Note 2: Once we integrate ability to ingest repcodes, the explicit block delims mode must respect those repcodes exactly,
* and cannot emit an RLE block that disagrees with the repcode history
* @return : final compressed size or a ZSTD error.
*/
ZSTDLIB_API size_t ZSTD_compressSequences(ZSTD_CCtx* const cctx, void* dst, size_t dstSize,
const ZSTD_Sequence* inSeqs, size_t inSeqsSize,
const void* src, size_t srcSize);
/***************************************
@@ -1263,23 +1377,28 @@ ZSTDLIB_API size_t ZSTD_getSequences(ZSTD_CCtx* zc, ZSTD_Sequence* outSeqs,
***************************************/
/*! ZSTD_estimate*() :
* These functions make it possible to estimate memory usage of a future
* {D,C}Ctx, before its creation.
* These functions make it possible to estimate memory usage
* of a future {D,C}Ctx, before its creation.
*
* ZSTD_estimateCCtxSize() will provide a budget large enough for any
* compression level up to selected one. Unlike ZSTD_estimateCStreamSize*(),
* this estimate does not include space for a window buffer, so this estimate
* is guaranteed to be enough for single-shot compressions, but not streaming
* compressions. It will however assume the input may be arbitrarily large,
* which is the worst case. If srcSize is known to always be small,
* ZSTD_estimateCCtxSize_usingCParams() can provide a tighter estimation.
* ZSTD_estimateCCtxSize_usingCParams() can be used in tandem with
* ZSTD_getCParams() to create cParams from compressionLevel.
* ZSTD_estimateCCtxSize_usingCCtxParams() can be used in tandem with
* ZSTD_CCtxParams_setParameter().
* ZSTD_estimateCCtxSize() will provide a memory budget large enough
* for any compression level up to selected one.
* Note : Unlike ZSTD_estimateCStreamSize*(), this estimate
* does not include space for a window buffer.
* Therefore, the estimation is only guaranteed for single-shot compressions, not streaming.
* The estimate will assume the input may be arbitrarily large,
* which is the worst case.
*
* Note: only single-threaded compression is supported. This function will
* return an error code if ZSTD_c_nbWorkers is >= 1. */
* When srcSize can be bound by a known and rather "small" value,
* this fact can be used to provide a tighter estimation
* because the CCtx compression context will need less memory.
* This tighter estimation can be provided by more advanced functions
* ZSTD_estimateCCtxSize_usingCParams(), which can be used in tandem with ZSTD_getCParams(),
* and ZSTD_estimateCCtxSize_usingCCtxParams(), which can be used in tandem with ZSTD_CCtxParams_setParameter().
* Both can be used to estimate memory using custom compression parameters and arbitrary srcSize limits.
*
* Note 2 : only single-threaded compression is supported.
* ZSTD_estimateCCtxSize_usingCCtxParams() will return an error code if ZSTD_c_nbWorkers is >= 1.
*/
ZSTDLIB_API size_t ZSTD_estimateCCtxSize(int compressionLevel);
ZSTDLIB_API size_t ZSTD_estimateCCtxSize_usingCParams(ZSTD_compressionParameters cParams);
ZSTDLIB_API size_t ZSTD_estimateCCtxSize_usingCCtxParams(const ZSTD_CCtx_params* params);
@@ -1362,7 +1481,11 @@ ZSTDLIB_API const ZSTD_DDict* ZSTD_initStaticDDict(
typedef void* (*ZSTD_allocFunction) (void* opaque, size_t size);
typedef void (*ZSTD_freeFunction) (void* opaque, void* address);
typedef struct { ZSTD_allocFunction customAlloc; ZSTD_freeFunction customFree; void* opaque; } ZSTD_customMem;
static ZSTD_customMem const ZSTD_defaultCMem = { NULL, NULL, NULL }; /**< this constant defers to stdlib's functions */
static
#ifdef __GNUC__
__attribute__((__unused__))
#endif
ZSTD_customMem const ZSTD_defaultCMem = { NULL, NULL, NULL }; /**< this constant defers to stdlib's functions */
ZSTDLIB_API ZSTD_CCtx* ZSTD_createCCtx_advanced(ZSTD_customMem customMem);
ZSTDLIB_API ZSTD_CStream* ZSTD_createCStream_advanced(ZSTD_customMem customMem);
@@ -1375,13 +1498,36 @@ ZSTDLIB_API ZSTD_CDict* ZSTD_createCDict_advanced(const void* dict, size_t dictS
ZSTD_compressionParameters cParams,
ZSTD_customMem customMem);
/* ! Thread pool :
* These prototypes make it possible to share a thread pool among multiple compression contexts.
* This can limit resources for applications with multiple threads where each one uses
* a threaded compression mode (via ZSTD_c_nbWorkers parameter).
* ZSTD_createThreadPool creates a new thread pool with a given number of threads.
* Note that the lifetime of such pool must exist while being used.
* ZSTD_CCtx_refThreadPool assigns a thread pool to a context (use NULL argument value
* to use an internal thread pool).
* ZSTD_freeThreadPool frees a thread pool.
*/
typedef struct POOL_ctx_s ZSTD_threadPool;
ZSTDLIB_API ZSTD_threadPool* ZSTD_createThreadPool(size_t numThreads);
ZSTDLIB_API void ZSTD_freeThreadPool (ZSTD_threadPool* pool);
ZSTDLIB_API size_t ZSTD_CCtx_refThreadPool(ZSTD_CCtx* cctx, ZSTD_threadPool* pool);
/*
* This API is temporary and is expected to change or disappear in the future!
*/
ZSTDLIB_API ZSTD_CDict* ZSTD_createCDict_advanced2(
const void* dict, size_t dictSize,
ZSTD_dictLoadMethod_e dictLoadMethod,
ZSTD_dictContentType_e dictContentType,
const ZSTD_CCtx_params* cctxParams,
ZSTD_customMem customMem);
ZSTDLIB_API ZSTD_DDict* ZSTD_createDDict_advanced(const void* dict, size_t dictSize,
ZSTD_dictLoadMethod_e dictLoadMethod,
ZSTD_dictContentType_e dictContentType,
ZSTD_customMem customMem);
/***************************************
* Advanced compression functions
***************************************/
@@ -1394,6 +1540,12 @@ ZSTDLIB_API ZSTD_DDict* ZSTD_createDDict_advanced(const void* dict, size_t dictS
* note: equivalent to ZSTD_createCDict_advanced(), with dictLoadMethod==ZSTD_dlm_byRef */
ZSTDLIB_API ZSTD_CDict* ZSTD_createCDict_byReference(const void* dictBuffer, size_t dictSize, int compressionLevel);
/*! ZSTD_getDictID_fromCDict() :
* Provides the dictID of the dictionary loaded into `cdict`.
* If @return == 0, the dictionary is not conformant to Zstandard specification, or empty.
* Non-conformant dictionaries can still be loaded, but as content-only dictionaries. */
ZSTDLIB_API unsigned ZSTD_getDictID_fromCDict(const ZSTD_CDict* cdict);
/*! ZSTD_getCParams() :
* @return ZSTD_compressionParameters structure for a selected compression level and estimated srcSize.
* `estimatedSrcSize` value is optional, select 0 if not known */
@@ -1508,6 +1660,143 @@ ZSTDLIB_API size_t ZSTD_CCtx_refPrefix_advanced(ZSTD_CCtx* cctx, const void* pre
* but compression ratio may regress significantly if guess considerably underestimates */
#define ZSTD_c_srcSizeHint ZSTD_c_experimentalParam7
/* Controls whether the new and experimental "dedicated dictionary search
* structure" can be used. This feature is still rough around the edges, be
* prepared for surprising behavior!
*
* How to use it:
*
* When using a CDict, whether to use this feature or not is controlled at
* CDict creation, and it must be set in a CCtxParams set passed into that
* construction (via ZSTD_createCDict_advanced2()). A compression will then
* use the feature or not based on how the CDict was constructed; the value of
* this param, set in the CCtx, will have no effect.
*
* However, when a dictionary buffer is passed into a CCtx, such as via
* ZSTD_CCtx_loadDictionary(), this param can be set on the CCtx to control
* whether the CDict that is created internally can use the feature or not.
*
* What it does:
*
* Normally, the internal data structures of the CDict are analogous to what
* would be stored in a CCtx after compressing the contents of a dictionary.
* To an approximation, a compression using a dictionary can then use those
* data structures to simply continue what is effectively a streaming
* compression where the simulated compression of the dictionary left off.
* Which is to say, the search structures in the CDict are normally the same
* format as in the CCtx.
*
* It is possible to do better, since the CDict is not like a CCtx: the search
* structures are written once during CDict creation, and then are only read
* after that, while the search structures in the CCtx are both read and
* written as the compression goes along. This means we can choose a search
* structure for the dictionary that is read-optimized.
*
* This feature enables the use of that different structure.
*
* Note that some of the members of the ZSTD_compressionParameters struct have
* different semantics and constraints in the dedicated search structure. It is
* highly recommended that you simply set a compression level in the CCtxParams
* you pass into the CDict creation call, and avoid messing with the cParams
* directly.
*
* Effects:
*
* This will only have any effect when the selected ZSTD_strategy
* implementation supports this feature. Currently, that's limited to
* ZSTD_greedy, ZSTD_lazy, and ZSTD_lazy2.
*
* Note that this means that the CDict tables can no longer be copied into the
* CCtx, so the dict attachment mode ZSTD_dictForceCopy will no longer be
* useable. The dictionary can only be attached or reloaded.
*
* In general, you should expect compression to be faster--sometimes very much
* so--and CDict creation to be slightly slower. Eventually, we will probably
* make this mode the default.
*/
#define ZSTD_c_enableDedicatedDictSearch ZSTD_c_experimentalParam8
/* ZSTD_c_stableInBuffer
* Experimental parameter.
* Default is 0 == disabled. Set to 1 to enable.
*
* Tells the compressor that the ZSTD_inBuffer will ALWAYS be the same
* between calls, except for the modifications that zstd makes to pos (the
* caller must not modify pos). This is checked by the compressor, and
* compression will fail if it ever changes. This means the only flush
* mode that makes sense is ZSTD_e_end, so zstd will error if ZSTD_e_end
* is not used. The data in the ZSTD_inBuffer in the range [src, src + pos)
* MUST not be modified during compression or you will get data corruption.
*
* When this flag is enabled zstd won't allocate an input window buffer,
* because the user guarantees it can reference the ZSTD_inBuffer until
* the frame is complete. But, it will still allocate an output buffer
* large enough to fit a block (see ZSTD_c_stableOutBuffer). This will also
* avoid the memcpy() from the input buffer to the input window buffer.
*
* NOTE: ZSTD_compressStream2() will error if ZSTD_e_end is not used.
* That means this flag cannot be used with ZSTD_compressStream().
*
* NOTE: So long as the ZSTD_inBuffer always points to valid memory, using
* this flag is ALWAYS memory safe, and will never access out-of-bounds
* memory. However, compression WILL fail if you violate the preconditions.
*
* WARNING: The data in the ZSTD_inBuffer in the range [dst, dst + pos) MUST
* not be modified during compression or you will get data corruption. This
* is because zstd needs to reference data in the ZSTD_inBuffer to find
* matches. Normally zstd maintains its own window buffer for this purpose,
* but passing this flag tells zstd to use the user provided buffer.
*/
#define ZSTD_c_stableInBuffer ZSTD_c_experimentalParam9
/* ZSTD_c_stableOutBuffer
* Experimental parameter.
* Default is 0 == disabled. Set to 1 to enable.
*
* Tells he compressor that the ZSTD_outBuffer will not be resized between
* calls. Specifically: (out.size - out.pos) will never grow. This gives the
* compressor the freedom to say: If the compressed data doesn't fit in the
* output buffer then return ZSTD_error_dstSizeTooSmall. This allows us to
* always decompress directly into the output buffer, instead of decompressing
* into an internal buffer and copying to the output buffer.
*
* When this flag is enabled zstd won't allocate an output buffer, because
* it can write directly to the ZSTD_outBuffer. It will still allocate the
* input window buffer (see ZSTD_c_stableInBuffer).
*
* Zstd will check that (out.size - out.pos) never grows and return an error
* if it does. While not strictly necessary, this should prevent surprises.
*/
#define ZSTD_c_stableOutBuffer ZSTD_c_experimentalParam10
/* ZSTD_c_blockDelimiters
* Default is 0 == ZSTD_sf_noBlockDelimiters.
*
* For use with sequence compression API: ZSTD_compressSequences().
*
* Designates whether or not the given array of ZSTD_Sequence contains block delimiters
* and last literals, which are defined as sequences with offset == 0 and matchLength == 0.
* See the definition of ZSTD_Sequence for more specifics.
*/
#define ZSTD_c_blockDelimiters ZSTD_c_experimentalParam11
/* ZSTD_c_validateSequences
* Default is 0 == disabled. Set to 1 to enable sequence validation.
*
* For use with sequence compression API: ZSTD_compressSequences().
* Designates whether or not we validate sequences provided to ZSTD_compressSequences()
* during function execution.
*
* Without validation, providing a sequence that does not conform to the zstd spec will cause
* undefined behavior, and may produce a corrupted block.
*
* With validation enabled, a if sequence is invalid (see doc/zstd_compression_format.md for
* specifics regarding offset/matchlength requirements) then the function will bail out and
* return an error.
*
*/
#define ZSTD_c_validateSequences ZSTD_c_experimentalParam12
/*! ZSTD_CCtx_getParameter() :
* Get the requested compression parameter value, selected by enum ZSTD_cParameter,
* and store it into int* value.
@@ -1556,8 +1845,10 @@ ZSTDLIB_API size_t ZSTD_CCtxParams_init_advanced(ZSTD_CCtx_params* cctxParams, Z
/*! ZSTD_CCtxParams_setParameter() :
* Similar to ZSTD_CCtx_setParameter.
* Set one compression parameter, selected by enum ZSTD_cParameter.
* Parameters must be applied to a ZSTD_CCtx using ZSTD_CCtx_setParametersUsingCCtxParams().
* @result : 0, or an error code (which can be tested with ZSTD_isError()).
* Parameters must be applied to a ZSTD_CCtx using
* ZSTD_CCtx_setParametersUsingCCtxParams().
* @result : a code representing success or failure (which can be tested with
* ZSTD_isError()).
*/
ZSTDLIB_API size_t ZSTD_CCtxParams_setParameter(ZSTD_CCtx_params* params, ZSTD_cParameter param, int value);
@@ -1637,11 +1928,60 @@ ZSTDLIB_API size_t ZSTD_DCtx_refPrefix_advanced(ZSTD_DCtx* dctx, const void* pre
*/
ZSTDLIB_API size_t ZSTD_DCtx_setMaxWindowSize(ZSTD_DCtx* dctx, size_t maxWindowSize);
/*! ZSTD_DCtx_getParameter() :
* Get the requested decompression parameter value, selected by enum ZSTD_dParameter,
* and store it into int* value.
* @return : 0, or an error code (which can be tested with ZSTD_isError()).
*/
ZSTDLIB_API size_t ZSTD_DCtx_getParameter(ZSTD_DCtx* dctx, ZSTD_dParameter param, int* value);
/* ZSTD_d_format
* experimental parameter,
* allowing selection between ZSTD_format_e input compression formats
*/
#define ZSTD_d_format ZSTD_d_experimentalParam1
/* ZSTD_d_stableOutBuffer
* Experimental parameter.
* Default is 0 == disabled. Set to 1 to enable.
*
* Tells the decompressor that the ZSTD_outBuffer will ALWAYS be the same
* between calls, except for the modifications that zstd makes to pos (the
* caller must not modify pos). This is checked by the decompressor, and
* decompression will fail if it ever changes. Therefore the ZSTD_outBuffer
* MUST be large enough to fit the entire decompressed frame. This will be
* checked when the frame content size is known. The data in the ZSTD_outBuffer
* in the range [dst, dst + pos) MUST not be modified during decompression
* or you will get data corruption.
*
* When this flags is enabled zstd won't allocate an output buffer, because
* it can write directly to the ZSTD_outBuffer, but it will still allocate
* an input buffer large enough to fit any compressed block. This will also
* avoid the memcpy() from the internal output buffer to the ZSTD_outBuffer.
* If you need to avoid the input buffer allocation use the buffer-less
* streaming API.
*
* NOTE: So long as the ZSTD_outBuffer always points to valid memory, using
* this flag is ALWAYS memory safe, and will never access out-of-bounds
* memory. However, decompression WILL fail if you violate the preconditions.
*
* WARNING: The data in the ZSTD_outBuffer in the range [dst, dst + pos) MUST
* not be modified during decompression or you will get data corruption. This
* is because zstd needs to reference data in the ZSTD_outBuffer to regenerate
* matches. Normally zstd maintains its own buffer for this purpose, but passing
* this flag tells zstd to use the user provided buffer.
*/
#define ZSTD_d_stableOutBuffer ZSTD_d_experimentalParam2
/* ZSTD_d_forceIgnoreChecksum
* Experimental parameter.
* Default is 0 == disabled. Set to 1 to enable
*
* Tells the decompressor to skip checksum validation during decompression, regardless
* of whether checksumming was specified during compression. This offers some
* slight performance benefits, and may be useful for debugging.
* Param has values of type ZSTD_forceIgnoreChecksum_e
*/
#define ZSTD_d_forceIgnoreChecksum ZSTD_d_experimentalParam3
/*! ZSTD_DCtx_setFormat() :
* Instruct the decoder context about what kind of data to decode next.
@@ -1670,7 +2010,8 @@ ZSTDLIB_API size_t ZSTD_decompressStream_simpleArgs (
********************************************************************/
/*===== Advanced Streaming compression functions =====*/
/**! ZSTD_initCStream_srcSize() :
/*! ZSTD_initCStream_srcSize() :
* This function is deprecated, and equivalent to:
* ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
* ZSTD_CCtx_refCDict(zcs, NULL); // clear the dictionary (if any)
@@ -1687,7 +2028,7 @@ ZSTD_initCStream_srcSize(ZSTD_CStream* zcs,
int compressionLevel,
unsigned long long pledgedSrcSize);
/**! ZSTD_initCStream_usingDict() :
/*! ZSTD_initCStream_usingDict() :
* This function is deprecated, and is equivalent to:
* ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
* ZSTD_CCtx_setParameter(zcs, ZSTD_c_compressionLevel, compressionLevel);
@@ -1704,7 +2045,7 @@ ZSTD_initCStream_usingDict(ZSTD_CStream* zcs,
const void* dict, size_t dictSize,
int compressionLevel);
/**! ZSTD_initCStream_advanced() :
/*! ZSTD_initCStream_advanced() :
* This function is deprecated, and is approximately equivalent to:
* ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
* // Pseudocode: Set each zstd parameter and leave the rest as-is.
@@ -1725,7 +2066,7 @@ ZSTD_initCStream_advanced(ZSTD_CStream* zcs,
ZSTD_parameters params,
unsigned long long pledgedSrcSize);
/**! ZSTD_initCStream_usingCDict() :
/*! ZSTD_initCStream_usingCDict() :
* This function is deprecated, and equivalent to:
* ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
* ZSTD_CCtx_refCDict(zcs, cdict);
@@ -1735,7 +2076,7 @@ ZSTD_initCStream_advanced(ZSTD_CStream* zcs,
*/
ZSTDLIB_API size_t ZSTD_initCStream_usingCDict(ZSTD_CStream* zcs, const ZSTD_CDict* cdict);
/**! ZSTD_initCStream_usingCDict_advanced() :
/*! ZSTD_initCStream_usingCDict_advanced() :
* This function is DEPRECATED, and is approximately equivalent to:
* ZSTD_CCtx_reset(zcs, ZSTD_reset_session_only);
* // Pseudocode: Set each zstd frame parameter and leave the rest as-is.
@@ -1808,7 +2149,8 @@ ZSTDLIB_API size_t ZSTD_toFlushNow(ZSTD_CCtx* cctx);
/*===== Advanced Streaming decompression functions =====*/
/**
/*!
* This function is deprecated, and is equivalent to:
*
* ZSTD_DCtx_reset(zds, ZSTD_reset_session_only);
@@ -1819,7 +2161,7 @@ ZSTDLIB_API size_t ZSTD_toFlushNow(ZSTD_CCtx* cctx);
*/
ZSTDLIB_API size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize);
/**
/*!
* This function is deprecated, and is equivalent to:
*
* ZSTD_DCtx_reset(zds, ZSTD_reset_session_only);
@@ -1830,7 +2172,7 @@ ZSTDLIB_API size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dic
*/
ZSTDLIB_API size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* zds, const ZSTD_DDict* ddict);
/**
/*!
* This function is deprecated, and is equivalent to:
*
* ZSTD_DCtx_reset(zds, ZSTD_reset_session_only);
@@ -1892,7 +2234,7 @@ ZSTDLIB_API size_t ZSTD_compressContinue(ZSTD_CCtx* cctx, void* dst, size_t dstC
ZSTDLIB_API size_t ZSTD_compressEnd(ZSTD_CCtx* cctx, void* dst, size_t dstCapacity, const void* src, size_t srcSize);
/*-
/**
Buffer-less streaming decompression (synchronous mode)
A ZSTD_DCtx object is required to track streaming operations.

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -13,8 +13,8 @@
/*-*************************************
* Dependencies
***************************************/
#include <stdlib.h> /* malloc, calloc, free */
#include <string.h> /* memset */
#define ZSTD_DEPS_NEED_MALLOC
#include "zstd_deps.h" /* ZSTD_malloc, ZSTD_calloc, ZSTD_free, ZSTD_memset */
#include "error_private.h"
#include "zstd_internal.h"
@@ -53,31 +53,31 @@ const char* ZSTD_getErrorString(ZSTD_ErrorCode code) { return ERR_getErrorString
/*=**************************************************************
* Custom allocator
****************************************************************/
void* ZSTD_malloc(size_t size, ZSTD_customMem customMem)
void* ZSTD_customMalloc(size_t size, ZSTD_customMem customMem)
{
if (customMem.customAlloc)
return customMem.customAlloc(customMem.opaque, size);
return malloc(size);
return ZSTD_malloc(size);
}
void* ZSTD_calloc(size_t size, ZSTD_customMem customMem)
void* ZSTD_customCalloc(size_t size, ZSTD_customMem customMem)
{
if (customMem.customAlloc) {
/* calloc implemented as malloc+memset;
* not as efficient as calloc, but next best guess for custom malloc */
void* const ptr = customMem.customAlloc(customMem.opaque, size);
memset(ptr, 0, size);
ZSTD_memset(ptr, 0, size);
return ptr;
}
return calloc(1, size);
return ZSTD_calloc(1, size);
}
void ZSTD_free(void* ptr, ZSTD_customMem customMem)
void ZSTD_customFree(void* ptr, ZSTD_customMem customMem)
{
if (ptr!=NULL) {
if (customMem.customFree)
customMem.customFree(customMem.opaque, ptr);
else
free(ptr);
ZSTD_free(ptr);
}
}

File diff suppressed because it is too large Load Diff

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -28,7 +28,6 @@
extern "C" {
#endif
/*-*************************************
* Constants
***************************************/
@@ -64,7 +63,7 @@ typedef struct {
} ZSTD_localDict;
typedef struct {
U32 CTable[HUF_CTABLE_SIZE_U32(255)];
HUF_CElt CTable[HUF_CTABLE_SIZE_U32(255)];
HUF_repeat repeatMode;
} ZSTD_hufCTables_t;
@@ -83,10 +82,27 @@ typedef struct {
} ZSTD_entropyCTables_t;
typedef struct {
U32 off;
U32 len;
U32 off; /* Offset code (offset + ZSTD_REP_MOVE) for the match */
U32 len; /* Raw length of match */
} ZSTD_match_t;
typedef struct {
U32 offset; /* Offset of sequence */
U32 litLength; /* Length of literals prior to match */
U32 matchLength; /* Raw length of match */
} rawSeq;
typedef struct {
rawSeq* seq; /* The start of the sequences */
size_t pos; /* The index in seq where reading stopped. pos <= size. */
size_t posInSequence; /* The position within the sequence at seq[pos] where reading
stopped. posInSequence <= seq[pos].litLength + seq[pos].matchLength */
size_t size; /* The number of sequences. <= capacity. */
size_t capacity; /* The capacity starting from `seq` pointer */
} rawSeqStore_t;
UNUSED_ATTR static const rawSeqStore_t kNullRawSeqStore = {NULL, 0, 0, 0, 0};
typedef struct {
int price;
U32 off;
@@ -147,9 +163,13 @@ struct ZSTD_matchState_t {
U32* hashTable;
U32* hashTable3;
U32* chainTable;
int dedicatedDictSearch; /* Indicates whether this matchState is using the
* dedicated dictionary search structure.
*/
optState_t opt; /* optimal parser state */
const ZSTD_matchState_t* dictMatchState;
ZSTD_compressionParameters cParams;
const rawSeqStore_t* ldmSeqStore;
};
typedef struct {
@@ -166,6 +186,7 @@ typedef struct {
typedef struct {
ZSTD_window_t window; /* State for the window round buffer management */
ldmEntry_t* hashTable;
U32 loadedDictEnd;
BYTE* bucketOffsets; /* Next position in bucket to insert entry */
U64 hashPower; /* Used to compute the rolling hash.
* Depends on ldmParams.minMatchLength */
@@ -180,19 +201,6 @@ typedef struct {
U32 windowLog; /* Window log for the LDM */
} ldmParams_t;
typedef struct {
U32 offset;
U32 litLength;
U32 matchLength;
} rawSeq;
typedef struct {
rawSeq* seq; /* The start of the sequences */
size_t pos; /* The position where reading stopped. <= size. */
size_t size; /* The number of sequences. <= capacity. */
size_t capacity; /* The capacity starting from `seq` pointer */
} rawSeqStore_t;
typedef struct {
int collectSequences;
ZSTD_Sequence* seqStart;
@@ -227,10 +235,34 @@ struct ZSTD_CCtx_params_s {
/* Long distance matching parameters */
ldmParams_t ldmParams;
/* Dedicated dict search algorithm trigger */
int enableDedicatedDictSearch;
/* Input/output buffer modes */
ZSTD_bufferMode_e inBufferMode;
ZSTD_bufferMode_e outBufferMode;
/* Sequence compression API */
ZSTD_sequenceFormat_e blockDelimiters;
int validateSequences;
/* Internal use, for createCCtxParams() and freeCCtxParams() only */
ZSTD_customMem customMem;
}; /* typedef'd to ZSTD_CCtx_params within "zstd.h" */
#define COMPRESS_SEQUENCES_WORKSPACE_SIZE (sizeof(unsigned) * (MaxSeq + 2))
#define ENTROPY_WORKSPACE_SIZE (HUF_WORKSPACE_SIZE + COMPRESS_SEQUENCES_WORKSPACE_SIZE)
/**
* Indicates whether this compression proceeds directly from user-provided
* source buffer to user-provided destination buffer (ZSTDb_not_buffered), or
* whether the context needs to buffer the input/output (ZSTDb_buffered).
*/
typedef enum {
ZSTDb_not_buffered,
ZSTDb_buffered
} ZSTD_buffered_policy_e;
struct ZSTD_CCtx_s {
ZSTD_compressionStage_e stage;
int cParamsChanged; /* == 1 if cParams(except wlog) or compression level are changed in requestedParams. Triggers transmission of new params to ZSTDMT (if available) then reset to 0. */
@@ -246,9 +278,11 @@ struct ZSTD_CCtx_s {
unsigned long long producedCSize;
XXH64_state_t xxhState;
ZSTD_customMem customMem;
ZSTD_threadPool* pool;
size_t staticSize;
SeqCollector seqCollector;
int isFirstBlock;
int initialized;
seqStore_t seqStore; /* sequences storage ptrs */
ldmState_t ldmState; /* long distance matching state */
@@ -256,7 +290,10 @@ struct ZSTD_CCtx_s {
size_t maxNbLdmSequences;
rawSeqStore_t externSeqStore; /* Mutable reference to external sequences */
ZSTD_blockState_t blockState;
U32* entropyWorkspace; /* entropy workspace of HUF_WORKSPACE_SIZE bytes */
U32* entropyWorkspace; /* entropy workspace of ENTROPY_WORKSPACE_SIZE bytes */
/* Wether we are streaming or not */
ZSTD_buffered_policy_e bufferedPolicy;
/* streaming */
char* inBuff;
@@ -271,6 +308,10 @@ struct ZSTD_CCtx_s {
ZSTD_cStreamStage streamStage;
U32 frameEnded;
/* Stable in/out buffer verification */
ZSTD_inBuffer expectedInBuffer;
size_t expectedOutBufferSize;
/* Dictionary */
ZSTD_localDict localDict;
const ZSTD_CDict* cdict;
@@ -284,8 +325,32 @@ struct ZSTD_CCtx_s {
typedef enum { ZSTD_dtlm_fast, ZSTD_dtlm_full } ZSTD_dictTableLoadMethod_e;
typedef enum { ZSTD_noDict = 0, ZSTD_extDict = 1, ZSTD_dictMatchState = 2 } ZSTD_dictMode_e;
typedef enum {
ZSTD_noDict = 0,
ZSTD_extDict = 1,
ZSTD_dictMatchState = 2,
ZSTD_dedicatedDictSearch = 3
} ZSTD_dictMode_e;
typedef enum {
ZSTD_cpm_noAttachDict = 0, /* Compression with ZSTD_noDict or ZSTD_extDict.
* In this mode we use both the srcSize and the dictSize
* when selecting and adjusting parameters.
*/
ZSTD_cpm_attachDict = 1, /* Compression with ZSTD_dictMatchState or ZSTD_dedicatedDictSearch.
* In this mode we only take the srcSize into account when selecting
* and adjusting parameters.
*/
ZSTD_cpm_createCDict = 2, /* Creating a CDict.
* In this mode we take both the source size and the dictionary size
* into account when selecting and adjusting the parameters.
*/
ZSTD_cpm_unknown = 3, /* ZSTD_getCParams, ZSTD_getParams, ZSTD_adjustParams.
* We don't know what these parameters are for. We default to the legacy
* behavior of taking both the source size and the dict size into account
* when selecting and adjusting parameters.
*/
} ZSTD_cParamMode_e;
typedef size_t (*ZSTD_blockCompressor) (
ZSTD_matchState_t* bs, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
@@ -324,6 +389,31 @@ MEM_STATIC U32 ZSTD_MLcode(U32 mlBase)
return (mlBase > 127) ? ZSTD_highbit32(mlBase) + ML_deltaCode : ML_Code[mlBase];
}
typedef struct repcodes_s {
U32 rep[3];
} repcodes_t;
MEM_STATIC repcodes_t ZSTD_updateRep(U32 const rep[3], U32 const offset, U32 const ll0)
{
repcodes_t newReps;
if (offset >= ZSTD_REP_NUM) { /* full offset */
newReps.rep[2] = rep[1];
newReps.rep[1] = rep[0];
newReps.rep[0] = offset - ZSTD_REP_MOVE;
} else { /* repcode */
U32 const repCode = offset + ll0;
if (repCode > 0) { /* note : if repCode==0, no change */
U32 const currentOffset = (repCode==ZSTD_REP_NUM) ? (rep[0] - 1) : rep[repCode];
newReps.rep[2] = (repCode >= 2) ? rep[1] : rep[2];
newReps.rep[1] = rep[0];
newReps.rep[0] = currentOffset;
} else { /* repCode == 0 */
ZSTD_memcpy(&newReps, rep, sizeof(newReps));
}
}
return newReps;
}
/* ZSTD_cParam_withinBounds:
* @return 1 if value is within cParam bounds,
* 0 otherwise */
@@ -336,6 +426,30 @@ MEM_STATIC int ZSTD_cParam_withinBounds(ZSTD_cParameter cParam, int value)
return 1;
}
/* ZSTD_noCompressBlock() :
* Writes uncompressed block to dst buffer from given src.
* Returns the size of the block */
MEM_STATIC size_t ZSTD_noCompressBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize, U32 lastBlock)
{
U32 const cBlockHeader24 = lastBlock + (((U32)bt_raw)<<1) + (U32)(srcSize << 3);
RETURN_ERROR_IF(srcSize + ZSTD_blockHeaderSize > dstCapacity,
dstSize_tooSmall, "dst buf too small for uncompressed block");
MEM_writeLE24(dst, cBlockHeader24);
ZSTD_memcpy((BYTE*)dst + ZSTD_blockHeaderSize, src, srcSize);
return ZSTD_blockHeaderSize + srcSize;
}
MEM_STATIC size_t ZSTD_rleCompressBlock (void* dst, size_t dstCapacity, BYTE src, size_t srcSize, U32 lastBlock)
{
BYTE* const op = (BYTE*)dst;
U32 const cBlockHeader = lastBlock + (((U32)bt_rle)<<1) + (U32)(srcSize << 3);
RETURN_ERROR_IF(dstCapacity < 4, dstSize_tooSmall, "");
MEM_writeLE24(op, cBlockHeader);
op[3] = src;
return 4;
}
/* ZSTD_minGain() :
* minimum compression required
* to generate a compress block or a compressed literals section.
@@ -348,6 +462,21 @@ MEM_STATIC size_t ZSTD_minGain(size_t srcSize, ZSTD_strategy strat)
return (srcSize >> minlog) + 2;
}
MEM_STATIC int ZSTD_disableLiteralsCompression(const ZSTD_CCtx_params* cctxParams)
{
switch (cctxParams->literalCompressionMode) {
case ZSTD_lcm_huffman:
return 0;
case ZSTD_lcm_uncompressed:
return 1;
default:
assert(0 /* impossible: pre-validated */);
/* fall-through */
case ZSTD_lcm_auto:
return (cctxParams->cParams.strategy == ZSTD_fast) && (cctxParams->cParams.targetLength > 0);
}
}
/*! ZSTD_safecopyLiterals() :
* memcpy() function that won't read beyond more than WILDCOPY_OVERLENGTH bytes past ilimit_w.
* Only called when the sequence ends past ilimit_w, so it only needs to be optimized for single
@@ -432,9 +561,12 @@ static unsigned ZSTD_NbCommonBytes (size_t val)
if (MEM_isLittleEndian()) {
if (MEM_64bits()) {
# if defined(_MSC_VER) && defined(_WIN64)
unsigned long r = 0;
_BitScanForward64( &r, (U64)val );
return (unsigned)(r>>3);
# if STATIC_BMI2
return _tzcnt_u64(val) >> 3;
# else
unsigned long r = 0;
return _BitScanForward64( &r, (U64)val ) ? (unsigned)(r >> 3) : 0;
# endif
# elif defined(__GNUC__) && (__GNUC__ >= 4)
return (__builtin_ctzll((U64)val) >> 3);
# else
@@ -451,8 +583,7 @@ static unsigned ZSTD_NbCommonBytes (size_t val)
} else { /* 32 bits */
# if defined(_MSC_VER)
unsigned long r=0;
_BitScanForward( &r, (U32)val );
return (unsigned)(r>>3);
return _BitScanForward( &r, (U32)val ) ? (unsigned)(r >> 3) : 0;
# elif defined(__GNUC__) && (__GNUC__ >= 3)
return (__builtin_ctz((U32)val) >> 3);
# else
@@ -466,9 +597,12 @@ static unsigned ZSTD_NbCommonBytes (size_t val)
} else { /* Big Endian CPU */
if (MEM_64bits()) {
# if defined(_MSC_VER) && defined(_WIN64)
unsigned long r = 0;
_BitScanReverse64( &r, val );
return (unsigned)(r>>3);
# if STATIC_BMI2
return _lzcnt_u64(val) >> 3;
# else
unsigned long r = 0;
return _BitScanReverse64(&r, (U64)val) ? (unsigned)(r >> 3) : 0;
# endif
# elif defined(__GNUC__) && (__GNUC__ >= 4)
return (__builtin_clzll(val) >> 3);
# else
@@ -482,8 +616,7 @@ static unsigned ZSTD_NbCommonBytes (size_t val)
} else { /* 32 bits */
# if defined(_MSC_VER)
unsigned long r = 0;
_BitScanReverse( &r, (unsigned long)val );
return (unsigned)(r>>3);
return _BitScanReverse( &r, (unsigned long)val ) ? (unsigned)(r >> 3) : 0;
# elif defined(__GNUC__) && (__GNUC__ >= 3)
return (__builtin_clz((U32)val) >> 3);
# else
@@ -564,7 +697,8 @@ static const U64 prime8bytes = 0xCF1BBCDCB7A56463ULL;
static size_t ZSTD_hash8(U64 u, U32 h) { return (size_t)(((u) * prime8bytes) >> (64-h)) ; }
static size_t ZSTD_hash8Ptr(const void* p, U32 h) { return ZSTD_hash8(MEM_readLE64(p), h); }
MEM_STATIC size_t ZSTD_hashPtr(const void* p, U32 hBits, U32 mls)
MEM_STATIC FORCE_INLINE_ATTR
size_t ZSTD_hashPtr(const void* p, U32 hBits, U32 mls)
{
switch(mls)
{
@@ -680,7 +814,7 @@ MEM_STATIC ZSTD_dictMode_e ZSTD_matchState_dictMode(const ZSTD_matchState_t *ms)
return ZSTD_window_hasExtDict(ms->window) ?
ZSTD_extDict :
ms->dictMatchState != NULL ?
ZSTD_dictMatchState :
(ms->dictMatchState->dedicatedDictSearch ? ZSTD_dedicatedDictSearch : ZSTD_dictMatchState) :
ZSTD_noDict;
}
@@ -692,8 +826,8 @@ MEM_STATIC ZSTD_dictMode_e ZSTD_matchState_dictMode(const ZSTD_matchState_t *ms)
MEM_STATIC U32 ZSTD_window_needOverflowCorrection(ZSTD_window_t const window,
void const* srcEnd)
{
U32 const current = (U32)((BYTE const*)srcEnd - window.base);
return current > ZSTD_CURRENT_MAX;
U32 const curr = (U32)((BYTE const*)srcEnd - window.base);
return curr > ZSTD_CURRENT_MAX;
}
/**
@@ -729,18 +863,30 @@ MEM_STATIC U32 ZSTD_window_correctOverflow(ZSTD_window_t* window, U32 cycleLog,
* windowLog <= 31 ==> 3<<29 + 1<<windowLog < 7<<29 < 1<<32.
*/
U32 const cycleMask = (1U << cycleLog) - 1;
U32 const current = (U32)((BYTE const*)src - window->base);
U32 const newCurrent = (current & cycleMask) + maxDist;
U32 const correction = current - newCurrent;
U32 const curr = (U32)((BYTE const*)src - window->base);
U32 const currentCycle0 = curr & cycleMask;
/* Exclude zero so that newCurrent - maxDist >= 1. */
U32 const currentCycle1 = currentCycle0 == 0 ? (1U << cycleLog) : currentCycle0;
U32 const newCurrent = currentCycle1 + maxDist;
U32 const correction = curr - newCurrent;
assert((maxDist & cycleMask) == 0);
assert(current > newCurrent);
assert(curr > newCurrent);
/* Loose bound, should be around 1<<29 (see above) */
assert(correction > 1<<28);
window->base += correction;
window->dictBase += correction;
window->lowLimit -= correction;
window->dictLimit -= correction;
if (window->lowLimit <= correction) window->lowLimit = 1;
else window->lowLimit -= correction;
if (window->dictLimit <= correction) window->dictLimit = 1;
else window->dictLimit -= correction;
/* Ensure we can still reference the full window. */
assert(newCurrent >= maxDist);
assert(newCurrent - maxDist >= 1);
/* Ensure that lowLimit and dictLimit didn't underflow. */
assert(window->lowLimit <= newCurrent);
assert(window->dictLimit <= newCurrent);
DEBUGLOG(4, "Correction of 0x%x bytes to lowLimit=0x%x", correction,
window->lowLimit);
@@ -844,6 +990,15 @@ ZSTD_checkDictValidity(const ZSTD_window_t* window,
} } }
}
MEM_STATIC void ZSTD_window_init(ZSTD_window_t* window) {
ZSTD_memset(window, 0, sizeof(*window));
window->base = (BYTE const*)"";
window->dictBase = (BYTE const*)"";
window->dictLimit = 1; /* start from 1, so that 1st position is valid */
window->lowLimit = 1; /* it ensures first and later CCtx usages compress the same */
window->nextSrc = window->base + 1; /* see issue #1241 */
}
/**
* ZSTD_window_update():
* Updates the window by appending [src, src + srcSize) to the window.
@@ -857,6 +1012,10 @@ MEM_STATIC U32 ZSTD_window_update(ZSTD_window_t* window,
BYTE const* const ip = (BYTE const*)src;
U32 contiguous = 1;
DEBUGLOG(5, "ZSTD_window_update");
if (srcSize == 0)
return contiguous;
assert(window->base != NULL);
assert(window->dictBase != NULL);
/* Check if blocks follow each other */
if (src != window->nextSrc) {
/* not contiguous */
@@ -867,7 +1026,7 @@ MEM_STATIC U32 ZSTD_window_update(ZSTD_window_t* window,
window->dictLimit = (U32)distanceFromBase;
window->dictBase = window->base;
window->base = ip - distanceFromBase;
// ms->nextToUpdate = window->dictLimit;
/* ms->nextToUpdate = window->dictLimit; */
if (window->dictLimit - window->lowLimit < HASH_READ_SIZE) window->lowLimit = window->dictLimit; /* too small extDict */
contiguous = 0;
}
@@ -883,12 +1042,35 @@ MEM_STATIC U32 ZSTD_window_update(ZSTD_window_t* window,
return contiguous;
}
MEM_STATIC U32 ZSTD_getLowestMatchIndex(const ZSTD_matchState_t* ms, U32 current, unsigned windowLog)
/**
* Returns the lowest allowed match index. It may either be in the ext-dict or the prefix.
*/
MEM_STATIC U32 ZSTD_getLowestMatchIndex(const ZSTD_matchState_t* ms, U32 curr, unsigned windowLog)
{
U32 const maxDistance = 1U << windowLog;
U32 const lowestValid = ms->window.lowLimit;
U32 const withinWindow = (current - lowestValid > maxDistance) ? current - maxDistance : lowestValid;
U32 const withinWindow = (curr - lowestValid > maxDistance) ? curr - maxDistance : lowestValid;
U32 const isDictionary = (ms->loadedDictEnd != 0);
/* When using a dictionary the entire dictionary is valid if a single byte of the dictionary
* is within the window. We invalidate the dictionary (and set loadedDictEnd to 0) when it isn't
* valid for the entire block. So this check is sufficient to find the lowest valid match index.
*/
U32 const matchLowest = isDictionary ? lowestValid : withinWindow;
return matchLowest;
}
/**
* Returns the lowest allowed match index in the prefix.
*/
MEM_STATIC U32 ZSTD_getLowestPrefixIndex(const ZSTD_matchState_t* ms, U32 curr, unsigned windowLog)
{
U32 const maxDistance = 1U << windowLog;
U32 const lowestValid = ms->window.dictLimit;
U32 const withinWindow = (curr - lowestValid > maxDistance) ? curr - maxDistance : lowestValid;
U32 const isDictionary = (ms->loadedDictEnd != 0);
/* When computing the lowest prefix index we need to take the dictionary into account to handle
* the edge case where the dictionary and the source are contiguous in memory.
*/
U32 const matchLowest = isDictionary ? lowestValid : withinWindow;
return matchLowest;
}
@@ -931,6 +1113,20 @@ MEM_STATIC void ZSTD_debugTable(const U32* table, U32 max)
}
#endif
/* ===============================================================
* Shared internal declarations
* These prototypes may be called from sources not in lib/compress
* =============================================================== */
/* ZSTD_loadCEntropy() :
* dict : must point at beginning of a valid zstd dictionary.
* return : size of dictionary header (size of magic number + dict ID + entropy tables)
* assumptions : magic number supposed already checked
* and dictSize >= 8 */
size_t ZSTD_loadCEntropy(ZSTD_compressedBlockState_t* bs, void* workspace,
const void* const dict, size_t dictSize);
void ZSTD_reset_compressedBlockState(ZSTD_compressedBlockState_t* bs);
/* ==============================================================
* Private declarations
@@ -940,9 +1136,10 @@ MEM_STATIC void ZSTD_debugTable(const U32* table, U32 max)
/* ZSTD_getCParamsFromCCtxParams() :
* cParams are built depending on compressionLevel, src size hints,
* LDM and manually set compression parameters.
* Note: srcSizeHint == 0 means 0!
*/
ZSTD_compressionParameters ZSTD_getCParamsFromCCtxParams(
const ZSTD_CCtx_params* CCtxParams, U64 srcSizeHint, size_t dictSize);
const ZSTD_CCtx_params* CCtxParams, U64 srcSizeHint, size_t dictSize, ZSTD_cParamMode_e mode);
/*! ZSTD_initCStream_internal() :
* Private use only. Init streaming operation.
@@ -999,5 +1196,8 @@ size_t ZSTD_writeLastEmptyBlock(void* dst, size_t dstCapacity);
*/
size_t ZSTD_referenceExternalSequences(ZSTD_CCtx* cctx, rawSeq* seq, size_t nbSeq);
/** ZSTD_cycleLog() :
* condition for correct operation : hashLog > 1 */
U32 ZSTD_cycleLog(U32 hashLog, ZSTD_strategy strat);
#endif /* ZSTD_COMPRESS_H */

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -18,7 +18,7 @@ size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src,
BYTE* const ostart = (BYTE* const)dst;
U32 const flSize = 1 + (srcSize>31) + (srcSize>4095);
RETURN_ERROR_IF(srcSize + flSize > dstCapacity, dstSize_tooSmall);
RETURN_ERROR_IF(srcSize + flSize > dstCapacity, dstSize_tooSmall, "");
switch(flSize)
{
@@ -35,7 +35,8 @@ size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src,
assert(0);
}
memcpy(ostart + flSize, src, srcSize);
ZSTD_memcpy(ostart + flSize, src, srcSize);
DEBUGLOG(5, "Raw literals: %u -> %u", (U32)srcSize, (U32)(srcSize + flSize));
return srcSize + flSize;
}
@@ -62,6 +63,7 @@ size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, const void*
}
ostart[flSize] = *(const BYTE*)src;
DEBUGLOG(5, "RLE literals: %u -> %u", (U32)srcSize, (U32)flSize + 1);
return flSize+1;
}
@@ -80,11 +82,11 @@ size_t ZSTD_compressLiterals (ZSTD_hufCTables_t const* prevHuf,
symbolEncodingType_e hType = set_compressed;
size_t cLitSize;
DEBUGLOG(5,"ZSTD_compressLiterals (disableLiteralCompression=%i)",
disableLiteralCompression);
DEBUGLOG(5,"ZSTD_compressLiterals (disableLiteralCompression=%i srcSize=%u)",
disableLiteralCompression, (U32)srcSize);
/* Prepare nextEntropy assuming reusing the existing table */
memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
if (disableLiteralCompression)
return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
@@ -102,24 +104,25 @@ size_t ZSTD_compressLiterals (ZSTD_hufCTables_t const* prevHuf,
cLitSize = singleStream ?
HUF_compress1X_repeat(
ostart+lhSize, dstCapacity-lhSize, src, srcSize,
255, 11, entropyWorkspace, entropyWorkspaceSize,
HUF_SYMBOLVALUE_MAX, HUF_TABLELOG_DEFAULT, entropyWorkspace, entropyWorkspaceSize,
(HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2) :
HUF_compress4X_repeat(
ostart+lhSize, dstCapacity-lhSize, src, srcSize,
255, 11, entropyWorkspace, entropyWorkspaceSize,
HUF_SYMBOLVALUE_MAX, HUF_TABLELOG_DEFAULT, entropyWorkspace, entropyWorkspaceSize,
(HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2);
if (repeat != HUF_repeat_none) {
/* reused the existing table */
DEBUGLOG(5, "Reusing previous huffman table");
hType = set_repeat;
}
}
if ((cLitSize==0) | (cLitSize >= srcSize - minGain) | ERR_isError(cLitSize)) {
memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
}
if (cLitSize==1) {
memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
return ZSTD_compressRleLiteralsBlock(dst, dstCapacity, src, srcSize);
}
@@ -150,5 +153,6 @@ size_t ZSTD_compressLiterals (ZSTD_hufCTables_t const* prevHuf,
default: /* not possible : lhSize is {3,4,5} */
assert(0);
}
DEBUGLOG(5, "Compressed literals: %u -> %u", (U32)srcSize, (U32)(lhSize+cLitSize));
return lhSize+cLitSize;
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -50,6 +50,19 @@ static unsigned ZSTD_getFSEMaxSymbolValue(FSE_CTable const* ctable) {
return maxSymbolValue;
}
/**
* Returns true if we should use ncount=-1 else we should
* use ncount=1 for low probability symbols instead.
*/
static unsigned ZSTD_useLowProbCount(size_t const nbSeq)
{
/* Heuristic: This should cover most blocks <= 16K and
* start to fade out after 16K to about 32K depending on
* comprssibility.
*/
return nbSeq >= 2048;
}
/**
* Returns the cost in bytes of encoding the normalized count header.
* Returns an error if any of the helper functions return an error.
@@ -60,7 +73,7 @@ static size_t ZSTD_NCountCost(unsigned const* count, unsigned const max,
BYTE wksp[FSE_NCOUNTBOUND];
S16 norm[MaxSeq + 1];
const U32 tableLog = FSE_optimalTableLog(FSELog, nbSeq, max);
FORWARD_IF_ERROR(FSE_normalizeCount(norm, tableLog, count, nbSeq, max));
FORWARD_IF_ERROR(FSE_normalizeCount(norm, tableLog, count, nbSeq, max, ZSTD_useLowProbCount(nbSeq)), "");
return FSE_writeNCount(wksp, sizeof(wksp), norm, max, tableLog);
}
@@ -86,7 +99,7 @@ static size_t ZSTD_entropyCost(unsigned const* count, unsigned const max, size_t
* Returns the cost in bits of encoding the distribution in count using ctable.
* Returns an error if ctable cannot represent all the symbols in count.
*/
static size_t ZSTD_fseBitCost(
size_t ZSTD_fseBitCost(
FSE_CTable const* ctable,
unsigned const* count,
unsigned const max)
@@ -96,18 +109,22 @@ static size_t ZSTD_fseBitCost(
unsigned s;
FSE_CState_t cstate;
FSE_initCState(&cstate, ctable);
RETURN_ERROR_IF(ZSTD_getFSEMaxSymbolValue(ctable) < max, GENERIC,
"Repeat FSE_CTable has maxSymbolValue %u < %u",
if (ZSTD_getFSEMaxSymbolValue(ctable) < max) {
DEBUGLOG(5, "Repeat FSE_CTable has maxSymbolValue %u < %u",
ZSTD_getFSEMaxSymbolValue(ctable), max);
return ERROR(GENERIC);
}
for (s = 0; s <= max; ++s) {
unsigned const tableLog = cstate.stateLog;
unsigned const badCost = (tableLog + 1) << kAccuracyLog;
unsigned const bitCost = FSE_bitCost(cstate.symbolTT, tableLog, s, kAccuracyLog);
if (count[s] == 0)
continue;
RETURN_ERROR_IF(bitCost >= badCost, GENERIC,
"Repeat FSE_CTable has Prob[%u] == 0", s);
cost += count[s] * bitCost;
if (bitCost >= badCost) {
DEBUGLOG(5, "Repeat FSE_CTable has Prob[%u] == 0", s);
return ERROR(GENERIC);
}
cost += (size_t)count[s] * bitCost;
}
return cost >> kAccuracyLog;
}
@@ -117,15 +134,15 @@ static size_t ZSTD_fseBitCost(
* table described by norm. The max symbol support by norm is assumed >= max.
* norm must be valid for every symbol with non-zero probability in count.
*/
static size_t ZSTD_crossEntropyCost(short const* norm, unsigned accuracyLog,
unsigned const* count, unsigned const max)
size_t ZSTD_crossEntropyCost(short const* norm, unsigned accuracyLog,
unsigned const* count, unsigned const max)
{
unsigned const shift = 8 - accuracyLog;
size_t cost = 0;
unsigned s;
assert(accuracyLog <= 8);
for (s = 0; s <= max; ++s) {
unsigned const normAcc = norm[s] != -1 ? norm[s] : 1;
unsigned const normAcc = (norm[s] != -1) ? (unsigned)norm[s] : 1;
unsigned const norm256 = normAcc << shift;
assert(norm256 > 0);
assert(norm256 < 256);
@@ -230,15 +247,15 @@ ZSTD_buildCTable(void* dst, size_t dstCapacity,
switch (type) {
case set_rle:
FORWARD_IF_ERROR(FSE_buildCTable_rle(nextCTable, (BYTE)max));
RETURN_ERROR_IF(dstCapacity==0, dstSize_tooSmall);
FORWARD_IF_ERROR(FSE_buildCTable_rle(nextCTable, (BYTE)max), "");
RETURN_ERROR_IF(dstCapacity==0, dstSize_tooSmall, "not enough space");
*op = codeTable[0];
return 1;
case set_repeat:
memcpy(nextCTable, prevCTable, prevCTableSize);
ZSTD_memcpy(nextCTable, prevCTable, prevCTableSize);
return 0;
case set_basic:
FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, defaultNorm, defaultMax, defaultNormLog, entropyWorkspace, entropyWorkspaceSize)); /* note : could be pre-calculated */
FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, defaultNorm, defaultMax, defaultNormLog, entropyWorkspace, entropyWorkspaceSize), ""); /* note : could be pre-calculated */
return 0;
case set_compressed: {
S16 norm[MaxSeq + 1];
@@ -249,14 +266,15 @@ ZSTD_buildCTable(void* dst, size_t dstCapacity,
nbSeq_1--;
}
assert(nbSeq_1 > 1);
FORWARD_IF_ERROR(FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max));
assert(entropyWorkspaceSize >= FSE_BUILD_CTABLE_WORKSPACE_SIZE(MaxSeq, MaxFSELog));
FORWARD_IF_ERROR(FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max, ZSTD_useLowProbCount(nbSeq_1)), "");
{ size_t const NCountSize = FSE_writeNCount(op, oend - op, norm, max, tableLog); /* overflow protected */
FORWARD_IF_ERROR(NCountSize);
FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, norm, max, tableLog, entropyWorkspace, entropyWorkspaceSize));
FORWARD_IF_ERROR(NCountSize, "FSE_writeNCount failed");
FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, norm, max, tableLog, entropyWorkspace, entropyWorkspaceSize), "");
return NCountSize;
}
}
default: assert(0); RETURN_ERROR(GENERIC);
default: assert(0); RETURN_ERROR(GENERIC, "impossible to reach");
}
}
@@ -290,7 +308,7 @@ ZSTD_encodeSequences_body(
if (MEM_32bits()) BIT_flushBits(&blockStream);
if (longOffsets) {
U32 const ofBits = ofCodeTable[nbSeq-1];
int const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1);
unsigned const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1);
if (extraBits) {
BIT_addBits(&blockStream, sequences[nbSeq-1].offset, extraBits);
BIT_flushBits(&blockStream);
@@ -327,7 +345,7 @@ ZSTD_encodeSequences_body(
BIT_addBits(&blockStream, sequences[n].matchLength, mlBits);
if (MEM_32bits() || (ofBits+mlBits+llBits > 56)) BIT_flushBits(&blockStream);
if (longOffsets) {
int const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1);
unsigned const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1);
if (extraBits) {
BIT_addBits(&blockStream, sequences[n].offset, extraBits);
BIT_flushBits(&blockStream); /* (7)*/

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -44,4 +44,11 @@ size_t ZSTD_encodeSequences(
FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable,
seqDef const* sequences, size_t nbSeq, int longOffsets, int bmi2);
size_t ZSTD_fseBitCost(
FSE_CTable const* ctable,
unsigned const* count,
unsigned const max);
size_t ZSTD_crossEntropyCost(short const* norm, unsigned accuracyLog,
unsigned const* count, unsigned const max);
#endif /* ZSTD_COMPRESS_SEQUENCES_H */

View File

@@ -0,0 +1,849 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
/*-*************************************
* Dependencies
***************************************/
#include "zstd_compress_superblock.h"
#include "zstd_internal.h" /* ZSTD_getSequenceLength */
#include "hist.h" /* HIST_countFast_wksp */
#include "zstd_compress_internal.h"
#include "zstd_compress_sequences.h"
#include "zstd_compress_literals.h"
/*-*************************************
* Superblock entropy buffer structs
***************************************/
/** ZSTD_hufCTablesMetadata_t :
* Stores Literals Block Type for a super-block in hType, and
* huffman tree description in hufDesBuffer.
* hufDesSize refers to the size of huffman tree description in bytes.
* This metadata is populated in ZSTD_buildSuperBlockEntropy_literal() */
typedef struct {
symbolEncodingType_e hType;
BYTE hufDesBuffer[ZSTD_MAX_HUF_HEADER_SIZE];
size_t hufDesSize;
} ZSTD_hufCTablesMetadata_t;
/** ZSTD_fseCTablesMetadata_t :
* Stores symbol compression modes for a super-block in {ll, ol, ml}Type, and
* fse tables in fseTablesBuffer.
* fseTablesSize refers to the size of fse tables in bytes.
* This metadata is populated in ZSTD_buildSuperBlockEntropy_sequences() */
typedef struct {
symbolEncodingType_e llType;
symbolEncodingType_e ofType;
symbolEncodingType_e mlType;
BYTE fseTablesBuffer[ZSTD_MAX_FSE_HEADERS_SIZE];
size_t fseTablesSize;
size_t lastCountSize; /* This is to account for bug in 1.3.4. More detail in ZSTD_compressSubBlock_sequences() */
} ZSTD_fseCTablesMetadata_t;
typedef struct {
ZSTD_hufCTablesMetadata_t hufMetadata;
ZSTD_fseCTablesMetadata_t fseMetadata;
} ZSTD_entropyCTablesMetadata_t;
/** ZSTD_buildSuperBlockEntropy_literal() :
* Builds entropy for the super-block literals.
* Stores literals block type (raw, rle, compressed, repeat) and
* huffman description table to hufMetadata.
* @return : size of huffman description table or error code */
static size_t ZSTD_buildSuperBlockEntropy_literal(void* const src, size_t srcSize,
const ZSTD_hufCTables_t* prevHuf,
ZSTD_hufCTables_t* nextHuf,
ZSTD_hufCTablesMetadata_t* hufMetadata,
const int disableLiteralsCompression,
void* workspace, size_t wkspSize)
{
BYTE* const wkspStart = (BYTE*)workspace;
BYTE* const wkspEnd = wkspStart + wkspSize;
BYTE* const countWkspStart = wkspStart;
unsigned* const countWksp = (unsigned*)workspace;
const size_t countWkspSize = (HUF_SYMBOLVALUE_MAX + 1) * sizeof(unsigned);
BYTE* const nodeWksp = countWkspStart + countWkspSize;
const size_t nodeWkspSize = wkspEnd-nodeWksp;
unsigned maxSymbolValue = 255;
unsigned huffLog = HUF_TABLELOG_DEFAULT;
HUF_repeat repeat = prevHuf->repeatMode;
DEBUGLOG(5, "ZSTD_buildSuperBlockEntropy_literal (srcSize=%zu)", srcSize);
/* Prepare nextEntropy assuming reusing the existing table */
ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
if (disableLiteralsCompression) {
DEBUGLOG(5, "set_basic - disabled");
hufMetadata->hType = set_basic;
return 0;
}
/* small ? don't even attempt compression (speed opt) */
# define COMPRESS_LITERALS_SIZE_MIN 63
{ size_t const minLitSize = (prevHuf->repeatMode == HUF_repeat_valid) ? 6 : COMPRESS_LITERALS_SIZE_MIN;
if (srcSize <= minLitSize) {
DEBUGLOG(5, "set_basic - too small");
hufMetadata->hType = set_basic;
return 0;
}
}
/* Scan input and build symbol stats */
{ size_t const largest = HIST_count_wksp (countWksp, &maxSymbolValue, (const BYTE*)src, srcSize, workspace, wkspSize);
FORWARD_IF_ERROR(largest, "HIST_count_wksp failed");
if (largest == srcSize) {
DEBUGLOG(5, "set_rle");
hufMetadata->hType = set_rle;
return 0;
}
if (largest <= (srcSize >> 7)+4) {
DEBUGLOG(5, "set_basic - no gain");
hufMetadata->hType = set_basic;
return 0;
}
}
/* Validate the previous Huffman table */
if (repeat == HUF_repeat_check && !HUF_validateCTable((HUF_CElt const*)prevHuf->CTable, countWksp, maxSymbolValue)) {
repeat = HUF_repeat_none;
}
/* Build Huffman Tree */
ZSTD_memset(nextHuf->CTable, 0, sizeof(nextHuf->CTable));
huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue);
{ size_t const maxBits = HUF_buildCTable_wksp((HUF_CElt*)nextHuf->CTable, countWksp,
maxSymbolValue, huffLog,
nodeWksp, nodeWkspSize);
FORWARD_IF_ERROR(maxBits, "HUF_buildCTable_wksp");
huffLog = (U32)maxBits;
{ /* Build and write the CTable */
size_t const newCSize = HUF_estimateCompressedSize(
(HUF_CElt*)nextHuf->CTable, countWksp, maxSymbolValue);
size_t const hSize = HUF_writeCTable(
hufMetadata->hufDesBuffer, sizeof(hufMetadata->hufDesBuffer),
(HUF_CElt*)nextHuf->CTable, maxSymbolValue, huffLog);
/* Check against repeating the previous CTable */
if (repeat != HUF_repeat_none) {
size_t const oldCSize = HUF_estimateCompressedSize(
(HUF_CElt const*)prevHuf->CTable, countWksp, maxSymbolValue);
if (oldCSize < srcSize && (oldCSize <= hSize + newCSize || hSize + 12 >= srcSize)) {
DEBUGLOG(5, "set_repeat - smaller");
ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
hufMetadata->hType = set_repeat;
return 0;
}
}
if (newCSize + hSize >= srcSize) {
DEBUGLOG(5, "set_basic - no gains");
ZSTD_memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
hufMetadata->hType = set_basic;
return 0;
}
DEBUGLOG(5, "set_compressed (hSize=%u)", (U32)hSize);
hufMetadata->hType = set_compressed;
nextHuf->repeatMode = HUF_repeat_check;
return hSize;
}
}
}
/** ZSTD_buildSuperBlockEntropy_sequences() :
* Builds entropy for the super-block sequences.
* Stores symbol compression modes and fse table to fseMetadata.
* @return : size of fse tables or error code */
static size_t ZSTD_buildSuperBlockEntropy_sequences(seqStore_t* seqStorePtr,
const ZSTD_fseCTables_t* prevEntropy,
ZSTD_fseCTables_t* nextEntropy,
const ZSTD_CCtx_params* cctxParams,
ZSTD_fseCTablesMetadata_t* fseMetadata,
void* workspace, size_t wkspSize)
{
BYTE* const wkspStart = (BYTE*)workspace;
BYTE* const wkspEnd = wkspStart + wkspSize;
BYTE* const countWkspStart = wkspStart;
unsigned* const countWksp = (unsigned*)workspace;
const size_t countWkspSize = (MaxSeq + 1) * sizeof(unsigned);
BYTE* const cTableWksp = countWkspStart + countWkspSize;
const size_t cTableWkspSize = wkspEnd-cTableWksp;
ZSTD_strategy const strategy = cctxParams->cParams.strategy;
FSE_CTable* CTable_LitLength = nextEntropy->litlengthCTable;
FSE_CTable* CTable_OffsetBits = nextEntropy->offcodeCTable;
FSE_CTable* CTable_MatchLength = nextEntropy->matchlengthCTable;
const BYTE* const ofCodeTable = seqStorePtr->ofCode;
const BYTE* const llCodeTable = seqStorePtr->llCode;
const BYTE* const mlCodeTable = seqStorePtr->mlCode;
size_t const nbSeq = seqStorePtr->sequences - seqStorePtr->sequencesStart;
BYTE* const ostart = fseMetadata->fseTablesBuffer;
BYTE* const oend = ostart + sizeof(fseMetadata->fseTablesBuffer);
BYTE* op = ostart;
assert(cTableWkspSize >= (1 << MaxFSELog) * sizeof(FSE_FUNCTION_TYPE));
DEBUGLOG(5, "ZSTD_buildSuperBlockEntropy_sequences (nbSeq=%zu)", nbSeq);
ZSTD_memset(workspace, 0, wkspSize);
fseMetadata->lastCountSize = 0;
/* convert length/distances into codes */
ZSTD_seqToCodes(seqStorePtr);
/* build CTable for Literal Lengths */
{ U32 LLtype;
unsigned max = MaxLL;
size_t const mostFrequent = HIST_countFast_wksp(countWksp, &max, llCodeTable, nbSeq, workspace, wkspSize); /* can't fail */
DEBUGLOG(5, "Building LL table");
nextEntropy->litlength_repeatMode = prevEntropy->litlength_repeatMode;
LLtype = ZSTD_selectEncodingType(&nextEntropy->litlength_repeatMode,
countWksp, max, mostFrequent, nbSeq,
LLFSELog, prevEntropy->litlengthCTable,
LL_defaultNorm, LL_defaultNormLog,
ZSTD_defaultAllowed, strategy);
assert(set_basic < set_compressed && set_rle < set_compressed);
assert(!(LLtype < set_compressed && nextEntropy->litlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */
{ size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_LitLength, LLFSELog, (symbolEncodingType_e)LLtype,
countWksp, max, llCodeTable, nbSeq, LL_defaultNorm, LL_defaultNormLog, MaxLL,
prevEntropy->litlengthCTable, sizeof(prevEntropy->litlengthCTable),
cTableWksp, cTableWkspSize);
FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for LitLens failed");
if (LLtype == set_compressed)
fseMetadata->lastCountSize = countSize;
op += countSize;
fseMetadata->llType = (symbolEncodingType_e) LLtype;
} }
/* build CTable for Offsets */
{ U32 Offtype;
unsigned max = MaxOff;
size_t const mostFrequent = HIST_countFast_wksp(countWksp, &max, ofCodeTable, nbSeq, workspace, wkspSize); /* can't fail */
/* We can only use the basic table if max <= DefaultMaxOff, otherwise the offsets are too large */
ZSTD_defaultPolicy_e const defaultPolicy = (max <= DefaultMaxOff) ? ZSTD_defaultAllowed : ZSTD_defaultDisallowed;
DEBUGLOG(5, "Building OF table");
nextEntropy->offcode_repeatMode = prevEntropy->offcode_repeatMode;
Offtype = ZSTD_selectEncodingType(&nextEntropy->offcode_repeatMode,
countWksp, max, mostFrequent, nbSeq,
OffFSELog, prevEntropy->offcodeCTable,
OF_defaultNorm, OF_defaultNormLog,
defaultPolicy, strategy);
assert(!(Offtype < set_compressed && nextEntropy->offcode_repeatMode != FSE_repeat_none)); /* We don't copy tables */
{ size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_OffsetBits, OffFSELog, (symbolEncodingType_e)Offtype,
countWksp, max, ofCodeTable, nbSeq, OF_defaultNorm, OF_defaultNormLog, DefaultMaxOff,
prevEntropy->offcodeCTable, sizeof(prevEntropy->offcodeCTable),
cTableWksp, cTableWkspSize);
FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for Offsets failed");
if (Offtype == set_compressed)
fseMetadata->lastCountSize = countSize;
op += countSize;
fseMetadata->ofType = (symbolEncodingType_e) Offtype;
} }
/* build CTable for MatchLengths */
{ U32 MLtype;
unsigned max = MaxML;
size_t const mostFrequent = HIST_countFast_wksp(countWksp, &max, mlCodeTable, nbSeq, workspace, wkspSize); /* can't fail */
DEBUGLOG(5, "Building ML table (remaining space : %i)", (int)(oend-op));
nextEntropy->matchlength_repeatMode = prevEntropy->matchlength_repeatMode;
MLtype = ZSTD_selectEncodingType(&nextEntropy->matchlength_repeatMode,
countWksp, max, mostFrequent, nbSeq,
MLFSELog, prevEntropy->matchlengthCTable,
ML_defaultNorm, ML_defaultNormLog,
ZSTD_defaultAllowed, strategy);
assert(!(MLtype < set_compressed && nextEntropy->matchlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */
{ size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_MatchLength, MLFSELog, (symbolEncodingType_e)MLtype,
countWksp, max, mlCodeTable, nbSeq, ML_defaultNorm, ML_defaultNormLog, MaxML,
prevEntropy->matchlengthCTable, sizeof(prevEntropy->matchlengthCTable),
cTableWksp, cTableWkspSize);
FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for MatchLengths failed");
if (MLtype == set_compressed)
fseMetadata->lastCountSize = countSize;
op += countSize;
fseMetadata->mlType = (symbolEncodingType_e) MLtype;
} }
assert((size_t) (op-ostart) <= sizeof(fseMetadata->fseTablesBuffer));
return op-ostart;
}
/** ZSTD_buildSuperBlockEntropy() :
* Builds entropy for the super-block.
* @return : 0 on success or error code */
static size_t
ZSTD_buildSuperBlockEntropy(seqStore_t* seqStorePtr,
const ZSTD_entropyCTables_t* prevEntropy,
ZSTD_entropyCTables_t* nextEntropy,
const ZSTD_CCtx_params* cctxParams,
ZSTD_entropyCTablesMetadata_t* entropyMetadata,
void* workspace, size_t wkspSize)
{
size_t const litSize = seqStorePtr->lit - seqStorePtr->litStart;
DEBUGLOG(5, "ZSTD_buildSuperBlockEntropy");
entropyMetadata->hufMetadata.hufDesSize =
ZSTD_buildSuperBlockEntropy_literal(seqStorePtr->litStart, litSize,
&prevEntropy->huf, &nextEntropy->huf,
&entropyMetadata->hufMetadata,
ZSTD_disableLiteralsCompression(cctxParams),
workspace, wkspSize);
FORWARD_IF_ERROR(entropyMetadata->hufMetadata.hufDesSize, "ZSTD_buildSuperBlockEntropy_literal failed");
entropyMetadata->fseMetadata.fseTablesSize =
ZSTD_buildSuperBlockEntropy_sequences(seqStorePtr,
&prevEntropy->fse, &nextEntropy->fse,
cctxParams,
&entropyMetadata->fseMetadata,
workspace, wkspSize);
FORWARD_IF_ERROR(entropyMetadata->fseMetadata.fseTablesSize, "ZSTD_buildSuperBlockEntropy_sequences failed");
return 0;
}
/** ZSTD_compressSubBlock_literal() :
* Compresses literals section for a sub-block.
* When we have to write the Huffman table we will sometimes choose a header
* size larger than necessary. This is because we have to pick the header size
* before we know the table size + compressed size, so we have a bound on the
* table size. If we guessed incorrectly, we fall back to uncompressed literals.
*
* We write the header when writeEntropy=1 and set entropyWrriten=1 when we succeeded
* in writing the header, otherwise it is set to 0.
*
* hufMetadata->hType has literals block type info.
* If it is set_basic, all sub-blocks literals section will be Raw_Literals_Block.
* If it is set_rle, all sub-blocks literals section will be RLE_Literals_Block.
* If it is set_compressed, first sub-block's literals section will be Compressed_Literals_Block
* If it is set_compressed, first sub-block's literals section will be Treeless_Literals_Block
* and the following sub-blocks' literals sections will be Treeless_Literals_Block.
* @return : compressed size of literals section of a sub-block
* Or 0 if it unable to compress.
* Or error code */
static size_t ZSTD_compressSubBlock_literal(const HUF_CElt* hufTable,
const ZSTD_hufCTablesMetadata_t* hufMetadata,
const BYTE* literals, size_t litSize,
void* dst, size_t dstSize,
const int bmi2, int writeEntropy, int* entropyWritten)
{
size_t const header = writeEntropy ? 200 : 0;
size_t const lhSize = 3 + (litSize >= (1 KB - header)) + (litSize >= (16 KB - header));
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstSize;
BYTE* op = ostart + lhSize;
U32 const singleStream = lhSize == 3;
symbolEncodingType_e hType = writeEntropy ? hufMetadata->hType : set_repeat;
size_t cLitSize = 0;
(void)bmi2; /* TODO bmi2... */
DEBUGLOG(5, "ZSTD_compressSubBlock_literal (litSize=%zu, lhSize=%zu, writeEntropy=%d)", litSize, lhSize, writeEntropy);
*entropyWritten = 0;
if (litSize == 0 || hufMetadata->hType == set_basic) {
DEBUGLOG(5, "ZSTD_compressSubBlock_literal using raw literal");
return ZSTD_noCompressLiterals(dst, dstSize, literals, litSize);
} else if (hufMetadata->hType == set_rle) {
DEBUGLOG(5, "ZSTD_compressSubBlock_literal using rle literal");
return ZSTD_compressRleLiteralsBlock(dst, dstSize, literals, litSize);
}
assert(litSize > 0);
assert(hufMetadata->hType == set_compressed || hufMetadata->hType == set_repeat);
if (writeEntropy && hufMetadata->hType == set_compressed) {
ZSTD_memcpy(op, hufMetadata->hufDesBuffer, hufMetadata->hufDesSize);
op += hufMetadata->hufDesSize;
cLitSize += hufMetadata->hufDesSize;
DEBUGLOG(5, "ZSTD_compressSubBlock_literal (hSize=%zu)", hufMetadata->hufDesSize);
}
/* TODO bmi2 */
{ const size_t cSize = singleStream ? HUF_compress1X_usingCTable(op, oend-op, literals, litSize, hufTable)
: HUF_compress4X_usingCTable(op, oend-op, literals, litSize, hufTable);
op += cSize;
cLitSize += cSize;
if (cSize == 0 || ERR_isError(cSize)) {
DEBUGLOG(5, "Failed to write entropy tables %s", ZSTD_getErrorName(cSize));
return 0;
}
/* If we expand and we aren't writing a header then emit uncompressed */
if (!writeEntropy && cLitSize >= litSize) {
DEBUGLOG(5, "ZSTD_compressSubBlock_literal using raw literal because uncompressible");
return ZSTD_noCompressLiterals(dst, dstSize, literals, litSize);
}
/* If we are writing headers then allow expansion that doesn't change our header size. */
if (lhSize < (size_t)(3 + (cLitSize >= 1 KB) + (cLitSize >= 16 KB))) {
assert(cLitSize > litSize);
DEBUGLOG(5, "Literals expanded beyond allowed header size");
return ZSTD_noCompressLiterals(dst, dstSize, literals, litSize);
}
DEBUGLOG(5, "ZSTD_compressSubBlock_literal (cSize=%zu)", cSize);
}
/* Build header */
switch(lhSize)
{
case 3: /* 2 - 2 - 10 - 10 */
{ U32 const lhc = hType + ((!singleStream) << 2) + ((U32)litSize<<4) + ((U32)cLitSize<<14);
MEM_writeLE24(ostart, lhc);
break;
}
case 4: /* 2 - 2 - 14 - 14 */
{ U32 const lhc = hType + (2 << 2) + ((U32)litSize<<4) + ((U32)cLitSize<<18);
MEM_writeLE32(ostart, lhc);
break;
}
case 5: /* 2 - 2 - 18 - 18 */
{ U32 const lhc = hType + (3 << 2) + ((U32)litSize<<4) + ((U32)cLitSize<<22);
MEM_writeLE32(ostart, lhc);
ostart[4] = (BYTE)(cLitSize >> 10);
break;
}
default: /* not possible : lhSize is {3,4,5} */
assert(0);
}
*entropyWritten = 1;
DEBUGLOG(5, "Compressed literals: %u -> %u", (U32)litSize, (U32)(op-ostart));
return op-ostart;
}
static size_t ZSTD_seqDecompressedSize(seqStore_t const* seqStore, const seqDef* sequences, size_t nbSeq, size_t litSize, int lastSequence) {
const seqDef* const sstart = sequences;
const seqDef* const send = sequences + nbSeq;
const seqDef* sp = sstart;
size_t matchLengthSum = 0;
size_t litLengthSum = 0;
while (send-sp > 0) {
ZSTD_sequenceLength const seqLen = ZSTD_getSequenceLength(seqStore, sp);
litLengthSum += seqLen.litLength;
matchLengthSum += seqLen.matchLength;
sp++;
}
assert(litLengthSum <= litSize);
if (!lastSequence) {
assert(litLengthSum == litSize);
}
return matchLengthSum + litSize;
}
/** ZSTD_compressSubBlock_sequences() :
* Compresses sequences section for a sub-block.
* fseMetadata->llType, fseMetadata->ofType, and fseMetadata->mlType have
* symbol compression modes for the super-block.
* The first successfully compressed block will have these in its header.
* We set entropyWritten=1 when we succeed in compressing the sequences.
* The following sub-blocks will always have repeat mode.
* @return : compressed size of sequences section of a sub-block
* Or 0 if it is unable to compress
* Or error code. */
static size_t ZSTD_compressSubBlock_sequences(const ZSTD_fseCTables_t* fseTables,
const ZSTD_fseCTablesMetadata_t* fseMetadata,
const seqDef* sequences, size_t nbSeq,
const BYTE* llCode, const BYTE* mlCode, const BYTE* ofCode,
const ZSTD_CCtx_params* cctxParams,
void* dst, size_t dstCapacity,
const int bmi2, int writeEntropy, int* entropyWritten)
{
const int longOffsets = cctxParams->cParams.windowLog > STREAM_ACCUMULATOR_MIN;
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstCapacity;
BYTE* op = ostart;
BYTE* seqHead;
DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (nbSeq=%zu, writeEntropy=%d, longOffsets=%d)", nbSeq, writeEntropy, longOffsets);
*entropyWritten = 0;
/* Sequences Header */
RETURN_ERROR_IF((oend-op) < 3 /*max nbSeq Size*/ + 1 /*seqHead*/,
dstSize_tooSmall, "");
if (nbSeq < 0x7F)
*op++ = (BYTE)nbSeq;
else if (nbSeq < LONGNBSEQ)
op[0] = (BYTE)((nbSeq>>8) + 0x80), op[1] = (BYTE)nbSeq, op+=2;
else
op[0]=0xFF, MEM_writeLE16(op+1, (U16)(nbSeq - LONGNBSEQ)), op+=3;
if (nbSeq==0) {
return op - ostart;
}
/* seqHead : flags for FSE encoding type */
seqHead = op++;
DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (seqHeadSize=%u)", (unsigned)(op-ostart));
if (writeEntropy) {
const U32 LLtype = fseMetadata->llType;
const U32 Offtype = fseMetadata->ofType;
const U32 MLtype = fseMetadata->mlType;
DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (fseTablesSize=%zu)", fseMetadata->fseTablesSize);
*seqHead = (BYTE)((LLtype<<6) + (Offtype<<4) + (MLtype<<2));
ZSTD_memcpy(op, fseMetadata->fseTablesBuffer, fseMetadata->fseTablesSize);
op += fseMetadata->fseTablesSize;
} else {
const U32 repeat = set_repeat;
*seqHead = (BYTE)((repeat<<6) + (repeat<<4) + (repeat<<2));
}
{ size_t const bitstreamSize = ZSTD_encodeSequences(
op, oend - op,
fseTables->matchlengthCTable, mlCode,
fseTables->offcodeCTable, ofCode,
fseTables->litlengthCTable, llCode,
sequences, nbSeq,
longOffsets, bmi2);
FORWARD_IF_ERROR(bitstreamSize, "ZSTD_encodeSequences failed");
op += bitstreamSize;
/* zstd versions <= 1.3.4 mistakenly report corruption when
* FSE_readNCount() receives a buffer < 4 bytes.
* Fixed by https://github.com/facebook/zstd/pull/1146.
* This can happen when the last set_compressed table present is 2
* bytes and the bitstream is only one byte.
* In this exceedingly rare case, we will simply emit an uncompressed
* block, since it isn't worth optimizing.
*/
#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
if (writeEntropy && fseMetadata->lastCountSize && fseMetadata->lastCountSize + bitstreamSize < 4) {
/* NCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */
assert(fseMetadata->lastCountSize + bitstreamSize == 3);
DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.3.4 by "
"emitting an uncompressed block.");
return 0;
}
#endif
DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (bitstreamSize=%zu)", bitstreamSize);
}
/* zstd versions <= 1.4.0 mistakenly report error when
* sequences section body size is less than 3 bytes.
* Fixed by https://github.com/facebook/zstd/pull/1664.
* This can happen when the previous sequences section block is compressed
* with rle mode and the current block's sequences section is compressed
* with repeat mode where sequences section body size can be 1 byte.
*/
#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
if (op-seqHead < 4) {
DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.4.0 by emitting "
"an uncompressed block when sequences are < 4 bytes");
return 0;
}
#endif
*entropyWritten = 1;
return op - ostart;
}
/** ZSTD_compressSubBlock() :
* Compresses a single sub-block.
* @return : compressed size of the sub-block
* Or 0 if it failed to compress. */
static size_t ZSTD_compressSubBlock(const ZSTD_entropyCTables_t* entropy,
const ZSTD_entropyCTablesMetadata_t* entropyMetadata,
const seqDef* sequences, size_t nbSeq,
const BYTE* literals, size_t litSize,
const BYTE* llCode, const BYTE* mlCode, const BYTE* ofCode,
const ZSTD_CCtx_params* cctxParams,
void* dst, size_t dstCapacity,
const int bmi2,
int writeLitEntropy, int writeSeqEntropy,
int* litEntropyWritten, int* seqEntropyWritten,
U32 lastBlock)
{
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstCapacity;
BYTE* op = ostart + ZSTD_blockHeaderSize;
DEBUGLOG(5, "ZSTD_compressSubBlock (litSize=%zu, nbSeq=%zu, writeLitEntropy=%d, writeSeqEntropy=%d, lastBlock=%d)",
litSize, nbSeq, writeLitEntropy, writeSeqEntropy, lastBlock);
{ size_t cLitSize = ZSTD_compressSubBlock_literal((const HUF_CElt*)entropy->huf.CTable,
&entropyMetadata->hufMetadata, literals, litSize,
op, oend-op, bmi2, writeLitEntropy, litEntropyWritten);
FORWARD_IF_ERROR(cLitSize, "ZSTD_compressSubBlock_literal failed");
if (cLitSize == 0) return 0;
op += cLitSize;
}
{ size_t cSeqSize = ZSTD_compressSubBlock_sequences(&entropy->fse,
&entropyMetadata->fseMetadata,
sequences, nbSeq,
llCode, mlCode, ofCode,
cctxParams,
op, oend-op,
bmi2, writeSeqEntropy, seqEntropyWritten);
FORWARD_IF_ERROR(cSeqSize, "ZSTD_compressSubBlock_sequences failed");
if (cSeqSize == 0) return 0;
op += cSeqSize;
}
/* Write block header */
{ size_t cSize = (op-ostart)-ZSTD_blockHeaderSize;
U32 const cBlockHeader24 = lastBlock + (((U32)bt_compressed)<<1) + (U32)(cSize << 3);
MEM_writeLE24(ostart, cBlockHeader24);
}
return op-ostart;
}
static size_t ZSTD_estimateSubBlockSize_literal(const BYTE* literals, size_t litSize,
const ZSTD_hufCTables_t* huf,
const ZSTD_hufCTablesMetadata_t* hufMetadata,
void* workspace, size_t wkspSize,
int writeEntropy)
{
unsigned* const countWksp = (unsigned*)workspace;
unsigned maxSymbolValue = 255;
size_t literalSectionHeaderSize = 3; /* Use hard coded size of 3 bytes */
if (hufMetadata->hType == set_basic) return litSize;
else if (hufMetadata->hType == set_rle) return 1;
else if (hufMetadata->hType == set_compressed || hufMetadata->hType == set_repeat) {
size_t const largest = HIST_count_wksp (countWksp, &maxSymbolValue, (const BYTE*)literals, litSize, workspace, wkspSize);
if (ZSTD_isError(largest)) return litSize;
{ size_t cLitSizeEstimate = HUF_estimateCompressedSize((const HUF_CElt*)huf->CTable, countWksp, maxSymbolValue);
if (writeEntropy) cLitSizeEstimate += hufMetadata->hufDesSize;
return cLitSizeEstimate + literalSectionHeaderSize;
} }
assert(0); /* impossible */
return 0;
}
static size_t ZSTD_estimateSubBlockSize_symbolType(symbolEncodingType_e type,
const BYTE* codeTable, unsigned maxCode,
size_t nbSeq, const FSE_CTable* fseCTable,
const U32* additionalBits,
short const* defaultNorm, U32 defaultNormLog, U32 defaultMax,
void* workspace, size_t wkspSize)
{
unsigned* const countWksp = (unsigned*)workspace;
const BYTE* ctp = codeTable;
const BYTE* const ctStart = ctp;
const BYTE* const ctEnd = ctStart + nbSeq;
size_t cSymbolTypeSizeEstimateInBits = 0;
unsigned max = maxCode;
HIST_countFast_wksp(countWksp, &max, codeTable, nbSeq, workspace, wkspSize); /* can't fail */
if (type == set_basic) {
/* We selected this encoding type, so it must be valid. */
assert(max <= defaultMax);
cSymbolTypeSizeEstimateInBits = max <= defaultMax
? ZSTD_crossEntropyCost(defaultNorm, defaultNormLog, countWksp, max)
: ERROR(GENERIC);
} else if (type == set_rle) {
cSymbolTypeSizeEstimateInBits = 0;
} else if (type == set_compressed || type == set_repeat) {
cSymbolTypeSizeEstimateInBits = ZSTD_fseBitCost(fseCTable, countWksp, max);
}
if (ZSTD_isError(cSymbolTypeSizeEstimateInBits)) return nbSeq * 10;
while (ctp < ctEnd) {
if (additionalBits) cSymbolTypeSizeEstimateInBits += additionalBits[*ctp];
else cSymbolTypeSizeEstimateInBits += *ctp; /* for offset, offset code is also the number of additional bits */
ctp++;
}
return cSymbolTypeSizeEstimateInBits / 8;
}
static size_t ZSTD_estimateSubBlockSize_sequences(const BYTE* ofCodeTable,
const BYTE* llCodeTable,
const BYTE* mlCodeTable,
size_t nbSeq,
const ZSTD_fseCTables_t* fseTables,
const ZSTD_fseCTablesMetadata_t* fseMetadata,
void* workspace, size_t wkspSize,
int writeEntropy)
{
size_t sequencesSectionHeaderSize = 3; /* Use hard coded size of 3 bytes */
size_t cSeqSizeEstimate = 0;
cSeqSizeEstimate += ZSTD_estimateSubBlockSize_symbolType(fseMetadata->ofType, ofCodeTable, MaxOff,
nbSeq, fseTables->offcodeCTable, NULL,
OF_defaultNorm, OF_defaultNormLog, DefaultMaxOff,
workspace, wkspSize);
cSeqSizeEstimate += ZSTD_estimateSubBlockSize_symbolType(fseMetadata->llType, llCodeTable, MaxLL,
nbSeq, fseTables->litlengthCTable, LL_bits,
LL_defaultNorm, LL_defaultNormLog, MaxLL,
workspace, wkspSize);
cSeqSizeEstimate += ZSTD_estimateSubBlockSize_symbolType(fseMetadata->mlType, mlCodeTable, MaxML,
nbSeq, fseTables->matchlengthCTable, ML_bits,
ML_defaultNorm, ML_defaultNormLog, MaxML,
workspace, wkspSize);
if (writeEntropy) cSeqSizeEstimate += fseMetadata->fseTablesSize;
return cSeqSizeEstimate + sequencesSectionHeaderSize;
}
static size_t ZSTD_estimateSubBlockSize(const BYTE* literals, size_t litSize,
const BYTE* ofCodeTable,
const BYTE* llCodeTable,
const BYTE* mlCodeTable,
size_t nbSeq,
const ZSTD_entropyCTables_t* entropy,
const ZSTD_entropyCTablesMetadata_t* entropyMetadata,
void* workspace, size_t wkspSize,
int writeLitEntropy, int writeSeqEntropy) {
size_t cSizeEstimate = 0;
cSizeEstimate += ZSTD_estimateSubBlockSize_literal(literals, litSize,
&entropy->huf, &entropyMetadata->hufMetadata,
workspace, wkspSize, writeLitEntropy);
cSizeEstimate += ZSTD_estimateSubBlockSize_sequences(ofCodeTable, llCodeTable, mlCodeTable,
nbSeq, &entropy->fse, &entropyMetadata->fseMetadata,
workspace, wkspSize, writeSeqEntropy);
return cSizeEstimate + ZSTD_blockHeaderSize;
}
static int ZSTD_needSequenceEntropyTables(ZSTD_fseCTablesMetadata_t const* fseMetadata)
{
if (fseMetadata->llType == set_compressed || fseMetadata->llType == set_rle)
return 1;
if (fseMetadata->mlType == set_compressed || fseMetadata->mlType == set_rle)
return 1;
if (fseMetadata->ofType == set_compressed || fseMetadata->ofType == set_rle)
return 1;
return 0;
}
/** ZSTD_compressSubBlock_multi() :
* Breaks super-block into multiple sub-blocks and compresses them.
* Entropy will be written to the first block.
* The following blocks will use repeat mode to compress.
* All sub-blocks are compressed blocks (no raw or rle blocks).
* @return : compressed size of the super block (which is multiple ZSTD blocks)
* Or 0 if it failed to compress. */
static size_t ZSTD_compressSubBlock_multi(const seqStore_t* seqStorePtr,
const ZSTD_compressedBlockState_t* prevCBlock,
ZSTD_compressedBlockState_t* nextCBlock,
const ZSTD_entropyCTablesMetadata_t* entropyMetadata,
const ZSTD_CCtx_params* cctxParams,
void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
const int bmi2, U32 lastBlock,
void* workspace, size_t wkspSize)
{
const seqDef* const sstart = seqStorePtr->sequencesStart;
const seqDef* const send = seqStorePtr->sequences;
const seqDef* sp = sstart;
const BYTE* const lstart = seqStorePtr->litStart;
const BYTE* const lend = seqStorePtr->lit;
const BYTE* lp = lstart;
BYTE const* ip = (BYTE const*)src;
BYTE const* const iend = ip + srcSize;
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstCapacity;
BYTE* op = ostart;
const BYTE* llCodePtr = seqStorePtr->llCode;
const BYTE* mlCodePtr = seqStorePtr->mlCode;
const BYTE* ofCodePtr = seqStorePtr->ofCode;
size_t targetCBlockSize = cctxParams->targetCBlockSize;
size_t litSize, seqCount;
int writeLitEntropy = entropyMetadata->hufMetadata.hType == set_compressed;
int writeSeqEntropy = 1;
int lastSequence = 0;
DEBUGLOG(5, "ZSTD_compressSubBlock_multi (litSize=%u, nbSeq=%u)",
(unsigned)(lend-lp), (unsigned)(send-sstart));
litSize = 0;
seqCount = 0;
do {
size_t cBlockSizeEstimate = 0;
if (sstart == send) {
lastSequence = 1;
} else {
const seqDef* const sequence = sp + seqCount;
lastSequence = sequence == send - 1;
litSize += ZSTD_getSequenceLength(seqStorePtr, sequence).litLength;
seqCount++;
}
if (lastSequence) {
assert(lp <= lend);
assert(litSize <= (size_t)(lend - lp));
litSize = (size_t)(lend - lp);
}
/* I think there is an optimization opportunity here.
* Calling ZSTD_estimateSubBlockSize for every sequence can be wasteful
* since it recalculates estimate from scratch.
* For example, it would recount literal distribution and symbol codes everytime.
*/
cBlockSizeEstimate = ZSTD_estimateSubBlockSize(lp, litSize, ofCodePtr, llCodePtr, mlCodePtr, seqCount,
&nextCBlock->entropy, entropyMetadata,
workspace, wkspSize, writeLitEntropy, writeSeqEntropy);
if (cBlockSizeEstimate > targetCBlockSize || lastSequence) {
int litEntropyWritten = 0;
int seqEntropyWritten = 0;
const size_t decompressedSize = ZSTD_seqDecompressedSize(seqStorePtr, sp, seqCount, litSize, lastSequence);
const size_t cSize = ZSTD_compressSubBlock(&nextCBlock->entropy, entropyMetadata,
sp, seqCount,
lp, litSize,
llCodePtr, mlCodePtr, ofCodePtr,
cctxParams,
op, oend-op,
bmi2, writeLitEntropy, writeSeqEntropy,
&litEntropyWritten, &seqEntropyWritten,
lastBlock && lastSequence);
FORWARD_IF_ERROR(cSize, "ZSTD_compressSubBlock failed");
if (cSize > 0 && cSize < decompressedSize) {
DEBUGLOG(5, "Committed the sub-block");
assert(ip + decompressedSize <= iend);
ip += decompressedSize;
sp += seqCount;
lp += litSize;
op += cSize;
llCodePtr += seqCount;
mlCodePtr += seqCount;
ofCodePtr += seqCount;
litSize = 0;
seqCount = 0;
/* Entropy only needs to be written once */
if (litEntropyWritten) {
writeLitEntropy = 0;
}
if (seqEntropyWritten) {
writeSeqEntropy = 0;
}
}
}
} while (!lastSequence);
if (writeLitEntropy) {
DEBUGLOG(5, "ZSTD_compressSubBlock_multi has literal entropy tables unwritten");
ZSTD_memcpy(&nextCBlock->entropy.huf, &prevCBlock->entropy.huf, sizeof(prevCBlock->entropy.huf));
}
if (writeSeqEntropy && ZSTD_needSequenceEntropyTables(&entropyMetadata->fseMetadata)) {
/* If we haven't written our entropy tables, then we've violated our contract and
* must emit an uncompressed block.
*/
DEBUGLOG(5, "ZSTD_compressSubBlock_multi has sequence entropy tables unwritten");
return 0;
}
if (ip < iend) {
size_t const cSize = ZSTD_noCompressBlock(op, oend - op, ip, iend - ip, lastBlock);
DEBUGLOG(5, "ZSTD_compressSubBlock_multi last sub-block uncompressed, %zu bytes", (size_t)(iend - ip));
FORWARD_IF_ERROR(cSize, "ZSTD_noCompressBlock failed");
assert(cSize != 0);
op += cSize;
/* We have to regenerate the repcodes because we've skipped some sequences */
if (sp < send) {
seqDef const* seq;
repcodes_t rep;
ZSTD_memcpy(&rep, prevCBlock->rep, sizeof(rep));
for (seq = sstart; seq < sp; ++seq) {
rep = ZSTD_updateRep(rep.rep, seq->offset - 1, ZSTD_getSequenceLength(seqStorePtr, seq).litLength == 0);
}
ZSTD_memcpy(nextCBlock->rep, &rep, sizeof(rep));
}
}
DEBUGLOG(5, "ZSTD_compressSubBlock_multi compressed");
return op-ostart;
}
size_t ZSTD_compressSuperBlock(ZSTD_CCtx* zc,
void* dst, size_t dstCapacity,
void const* src, size_t srcSize,
unsigned lastBlock) {
ZSTD_entropyCTablesMetadata_t entropyMetadata;
FORWARD_IF_ERROR(ZSTD_buildSuperBlockEntropy(&zc->seqStore,
&zc->blockState.prevCBlock->entropy,
&zc->blockState.nextCBlock->entropy,
&zc->appliedParams,
&entropyMetadata,
zc->entropyWorkspace, ENTROPY_WORKSPACE_SIZE /* statically allocated in resetCCtx */), "");
return ZSTD_compressSubBlock_multi(&zc->seqStore,
zc->blockState.prevCBlock,
zc->blockState.nextCBlock,
&entropyMetadata,
&zc->appliedParams,
dst, dstCapacity,
src, srcSize,
zc->bmi2, lastBlock,
zc->entropyWorkspace, ENTROPY_WORKSPACE_SIZE /* statically allocated in resetCCtx */);
}

View File

@@ -0,0 +1,32 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef ZSTD_COMPRESS_ADVANCED_H
#define ZSTD_COMPRESS_ADVANCED_H
/*-*************************************
* Dependencies
***************************************/
#include "zstd.h" /* ZSTD_CCtx */
/*-*************************************
* Target Compressed Block Size
***************************************/
/* ZSTD_compressSuperBlock() :
* Used to compress a super block when targetCBlockSize is being used.
* The given block will be compressed into multiple sub blocks that are around targetCBlockSize. */
size_t ZSTD_compressSuperBlock(ZSTD_CCtx* zc,
void* dst, size_t dstCapacity,
void const* src, size_t srcSize,
unsigned lastBlock);
#endif /* ZSTD_COMPRESS_ADVANCED_H */

139
vendor/github.com/DataDog/zstd/zstd_ctx.go generated vendored Normal file
View File

@@ -0,0 +1,139 @@
package zstd
/*
#define ZSTD_STATIC_LINKING_ONLY
#include "zstd.h"
*/
import "C"
import (
"bytes"
"io/ioutil"
"runtime"
"unsafe"
)
type Ctx interface {
// Compress src into dst. If you have a buffer to use, you can pass it to
// prevent allocation. If it is too small, or if nil is passed, a new buffer
// will be allocated and returned.
Compress(dst, src []byte) ([]byte, error)
// CompressLevel is the same as Compress but you can pass a compression level
CompressLevel(dst, src []byte, level int) ([]byte, error)
// Decompress src into dst. If you have a buffer to use, you can pass it to
// prevent allocation. If it is too small, or if nil is passed, a new buffer
// will be allocated and returned.
Decompress(dst, src []byte) ([]byte, error)
}
type ctx struct {
cctx *C.ZSTD_CCtx
dctx *C.ZSTD_DCtx
}
// Create a new ZStd Context.
// When compressing/decompressing many times, it is recommended to allocate a
// context just once, and re-use it for each successive compression operation.
// This will make workload friendlier for system's memory.
// Note : re-using context is just a speed / resource optimization.
// It doesn't change the compression ratio, which remains identical.
// Note 2 : In multi-threaded environments,
// use one different context per thread for parallel execution.
//
func NewCtx() Ctx {
c := &ctx{
cctx: C.ZSTD_createCCtx(),
dctx: C.ZSTD_createDCtx(),
}
runtime.SetFinalizer(c, finalizeCtx)
return c
}
func (c *ctx) Compress(dst, src []byte) ([]byte, error) {
return c.CompressLevel(dst, src, DefaultCompression)
}
func (c *ctx) CompressLevel(dst, src []byte, level int) ([]byte, error) {
bound := CompressBound(len(src))
if cap(dst) >= bound {
dst = dst[0:bound] // Reuse dst buffer
} else {
dst = make([]byte, bound)
}
var srcPtr *byte // Do not point anywhere, if src is empty
if len(src) > 0 {
srcPtr = &src[0]
}
cWritten := C.ZSTD_compressCCtx(
c.cctx,
unsafe.Pointer(&dst[0]),
C.size_t(len(dst)),
unsafe.Pointer(srcPtr),
C.size_t(len(src)),
C.int(level))
written := int(cWritten)
// Check if the return is an Error code
if err := getError(written); err != nil {
return nil, err
}
return dst[:written], nil
}
func (c *ctx) Decompress(dst, src []byte) ([]byte, error) {
if len(src) == 0 {
return []byte{}, ErrEmptySlice
}
decompress := func(dst, src []byte) ([]byte, error) {
cWritten := C.ZSTD_decompressDCtx(
c.dctx,
unsafe.Pointer(&dst[0]),
C.size_t(len(dst)),
unsafe.Pointer(&src[0]),
C.size_t(len(src)))
written := int(cWritten)
// Check error
if err := getError(written); err != nil {
return nil, err
}
return dst[:written], nil
}
if len(dst) == 0 {
// Attempt to use zStd to determine decompressed size (may result in error or 0)
size := int(C.size_t(C.ZSTD_getDecompressedSize(unsafe.Pointer(&src[0]), C.size_t(len(src)))))
if err := getError(size); err != nil {
return nil, err
}
if size > 0 {
dst = make([]byte, size)
} else {
dst = make([]byte, len(src)*3) // starting guess
}
}
for i := 0; i < 3; i++ { // 3 tries to allocate a bigger buffer
result, err := decompress(dst, src)
if !IsDstSizeTooSmallError(err) {
return result, err
}
dst = make([]byte, len(dst)*2) // Grow buffer by 2
}
// We failed getting a dst buffer of correct size, use stream API
r := NewReader(bytes.NewReader(src))
defer r.Close()
return ioutil.ReadAll(r)
}
func finalizeCtx(c *ctx) {
C.ZSTD_freeCCtx(c.cctx)
C.ZSTD_freeDCtx(c.dctx)
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -24,16 +24,6 @@ extern "C" {
* Constants
***************************************/
/* define "workspace is too large" as this number of times larger than needed */
#define ZSTD_WORKSPACETOOLARGE_FACTOR 3
/* when workspace is continuously too large
* during at least this number of times,
* context's memory usage is considered wasteful,
* because it's sized to handle a worst case scenario which rarely happens.
* In which case, resize it down to free some memory */
#define ZSTD_WORKSPACETOOLARGE_MAXDURATION 128
/* Since the workspace is effectively its own little malloc implementation /
* arena, when we run under ASAN, we should similarly insert redzones between
* each internal element of the workspace, so ASAN will catch overruns that
@@ -54,6 +44,16 @@ typedef enum {
ZSTD_cwksp_alloc_aligned
} ZSTD_cwksp_alloc_phase_e;
/**
* Used to describe whether the workspace is statically allocated (and will not
* necessarily ever be freed), or if it's dynamically allocated and we can
* expect a well-formed caller to free this.
*/
typedef enum {
ZSTD_cwksp_dynamic_alloc,
ZSTD_cwksp_static_alloc
} ZSTD_cwksp_static_alloc_e;
/**
* Zstd fits all its internal datastructures into a single continuous buffer,
* so that it only needs to perform a single OS allocation (or so that a buffer
@@ -102,7 +102,7 @@ typedef enum {
*
* - Static objects: this is optionally the enclosing ZSTD_CCtx or ZSTD_CDict,
* so that literally everything fits in a single buffer. Note: if present,
* this must be the first object in the workspace, since ZSTD_free{CCtx,
* this must be the first object in the workspace, since ZSTD_customFree{CCtx,
* CDict}() rely on a pointer comparison to see whether one or two frees are
* required.
*
@@ -147,9 +147,10 @@ typedef struct {
void* tableValidEnd;
void* allocStart;
int allocFailed;
BYTE allocFailed;
int workspaceOversizedDuration;
ZSTD_cwksp_alloc_phase_e phase;
ZSTD_cwksp_static_alloc_e isStatic;
} ZSTD_cwksp;
/*-*************************************
@@ -188,7 +189,9 @@ MEM_STATIC size_t ZSTD_cwksp_align(size_t size, size_t const align) {
* else is though.
*/
MEM_STATIC size_t ZSTD_cwksp_alloc_size(size_t size) {
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
if (size == 0)
return 0;
#if ZSTD_ADDRESS_SANITIZER && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
return size + 2 * ZSTD_CWKSP_ASAN_REDZONE_SIZE;
#else
return size;
@@ -238,7 +241,10 @@ MEM_STATIC void* ZSTD_cwksp_reserve_internal(
ZSTD_cwksp_internal_advance_phase(ws, phase);
alloc = (BYTE *)ws->allocStart - bytes;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
if (bytes == 0)
return NULL;
#if ZSTD_ADDRESS_SANITIZER && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* over-reserve space */
alloc = (BYTE *)alloc - 2 * ZSTD_CWKSP_ASAN_REDZONE_SIZE;
#endif
@@ -257,11 +263,13 @@ MEM_STATIC void* ZSTD_cwksp_reserve_internal(
}
ws->allocStart = alloc;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
#if ZSTD_ADDRESS_SANITIZER && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* Move alloc so there's ZSTD_CWKSP_ASAN_REDZONE_SIZE unused space on
* either size. */
alloc = (BYTE *)alloc + ZSTD_CWKSP_ASAN_REDZONE_SIZE;
__asan_unpoison_memory_region(alloc, bytes);
if (ws->isStatic == ZSTD_cwksp_dynamic_alloc) {
__asan_unpoison_memory_region(alloc, bytes);
}
#endif
return alloc;
@@ -306,8 +314,10 @@ MEM_STATIC void* ZSTD_cwksp_reserve_table(ZSTD_cwksp* ws, size_t bytes) {
}
ws->tableEnd = end;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
__asan_unpoison_memory_region(alloc, bytes);
#if ZSTD_ADDRESS_SANITIZER && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
if (ws->isStatic == ZSTD_cwksp_dynamic_alloc) {
__asan_unpoison_memory_region(alloc, bytes);
}
#endif
return alloc;
@@ -321,7 +331,7 @@ MEM_STATIC void* ZSTD_cwksp_reserve_object(ZSTD_cwksp* ws, size_t bytes) {
void* alloc = ws->objectEnd;
void* end = (BYTE*)alloc + roundedBytes;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
#if ZSTD_ADDRESS_SANITIZER && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* over-reserve space */
end = (BYTE *)end + 2 * ZSTD_CWKSP_ASAN_REDZONE_SIZE;
#endif
@@ -342,11 +352,13 @@ MEM_STATIC void* ZSTD_cwksp_reserve_object(ZSTD_cwksp* ws, size_t bytes) {
ws->tableEnd = end;
ws->tableValidEnd = end;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
#if ZSTD_ADDRESS_SANITIZER && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* Move alloc so there's ZSTD_CWKSP_ASAN_REDZONE_SIZE unused space on
* either size. */
alloc = (BYTE *)alloc + ZSTD_CWKSP_ASAN_REDZONE_SIZE;
__asan_unpoison_memory_region(alloc, bytes);
if (ws->isStatic == ZSTD_cwksp_dynamic_alloc) {
__asan_unpoison_memory_region(alloc, bytes);
}
#endif
return alloc;
@@ -355,7 +367,7 @@ MEM_STATIC void* ZSTD_cwksp_reserve_object(ZSTD_cwksp* ws, size_t bytes) {
MEM_STATIC void ZSTD_cwksp_mark_tables_dirty(ZSTD_cwksp* ws) {
DEBUGLOG(4, "cwksp: ZSTD_cwksp_mark_tables_dirty");
#if defined (MEMORY_SANITIZER) && !defined (ZSTD_MSAN_DONT_POISON_WORKSPACE)
#if ZSTD_MEMORY_SANITIZER && !defined (ZSTD_MSAN_DONT_POISON_WORKSPACE)
/* To validate that the table re-use logic is sound, and that we don't
* access table space that we haven't cleaned, we re-"poison" the table
* space every time we mark it dirty. */
@@ -390,7 +402,7 @@ MEM_STATIC void ZSTD_cwksp_clean_tables(ZSTD_cwksp* ws) {
assert(ws->tableValidEnd >= ws->objectEnd);
assert(ws->tableValidEnd <= ws->allocStart);
if (ws->tableValidEnd < ws->tableEnd) {
memset(ws->tableValidEnd, 0, (BYTE*)ws->tableEnd - (BYTE*)ws->tableValidEnd);
ZSTD_memset(ws->tableValidEnd, 0, (BYTE*)ws->tableEnd - (BYTE*)ws->tableValidEnd);
}
ZSTD_cwksp_mark_tables_clean(ws);
}
@@ -402,8 +414,12 @@ MEM_STATIC void ZSTD_cwksp_clean_tables(ZSTD_cwksp* ws) {
MEM_STATIC void ZSTD_cwksp_clear_tables(ZSTD_cwksp* ws) {
DEBUGLOG(4, "cwksp: clearing tables!");
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
{
#if ZSTD_ADDRESS_SANITIZER && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* We don't do this when the workspace is statically allocated, because
* when that is the case, we have no capability to hook into the end of the
* workspace's lifecycle to unpoison the memory.
*/
if (ws->isStatic == ZSTD_cwksp_dynamic_alloc) {
size_t size = (BYTE*)ws->tableValidEnd - (BYTE*)ws->objectEnd;
__asan_poison_memory_region(ws->objectEnd, size);
}
@@ -420,7 +436,7 @@ MEM_STATIC void ZSTD_cwksp_clear_tables(ZSTD_cwksp* ws) {
MEM_STATIC void ZSTD_cwksp_clear(ZSTD_cwksp* ws) {
DEBUGLOG(4, "cwksp: clearing!");
#if defined (MEMORY_SANITIZER) && !defined (ZSTD_MSAN_DONT_POISON_WORKSPACE)
#if ZSTD_MEMORY_SANITIZER && !defined (ZSTD_MSAN_DONT_POISON_WORKSPACE)
/* To validate that the context re-use logic is sound, and that we don't
* access stuff that this compression hasn't initialized, we re-"poison"
* the workspace (or at least the non-static, non-table parts of it)
@@ -431,8 +447,12 @@ MEM_STATIC void ZSTD_cwksp_clear(ZSTD_cwksp* ws) {
}
#endif
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
{
#if ZSTD_ADDRESS_SANITIZER && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* We don't do this when the workspace is statically allocated, because
* when that is the case, we have no capability to hook into the end of the
* workspace's lifecycle to unpoison the memory.
*/
if (ws->isStatic == ZSTD_cwksp_dynamic_alloc) {
size_t size = (BYTE*)ws->workspaceEnd - (BYTE*)ws->objectEnd;
__asan_poison_memory_region(ws->objectEnd, size);
}
@@ -452,7 +472,7 @@ MEM_STATIC void ZSTD_cwksp_clear(ZSTD_cwksp* ws) {
* Any existing values in the workspace are ignored (the previously managed
* buffer, if present, must be separately freed).
*/
MEM_STATIC void ZSTD_cwksp_init(ZSTD_cwksp* ws, void* start, size_t size) {
MEM_STATIC void ZSTD_cwksp_init(ZSTD_cwksp* ws, void* start, size_t size, ZSTD_cwksp_static_alloc_e isStatic) {
DEBUGLOG(4, "cwksp: init'ing workspace with %zd bytes", size);
assert(((size_t)start & (sizeof(void*)-1)) == 0); /* ensure correct alignment */
ws->workspace = start;
@@ -460,24 +480,25 @@ MEM_STATIC void ZSTD_cwksp_init(ZSTD_cwksp* ws, void* start, size_t size) {
ws->objectEnd = ws->workspace;
ws->tableValidEnd = ws->objectEnd;
ws->phase = ZSTD_cwksp_alloc_objects;
ws->isStatic = isStatic;
ZSTD_cwksp_clear(ws);
ws->workspaceOversizedDuration = 0;
ZSTD_cwksp_assert_internal_consistency(ws);
}
MEM_STATIC size_t ZSTD_cwksp_create(ZSTD_cwksp* ws, size_t size, ZSTD_customMem customMem) {
void* workspace = ZSTD_malloc(size, customMem);
void* workspace = ZSTD_customMalloc(size, customMem);
DEBUGLOG(4, "cwksp: creating new workspace with %zd bytes", size);
RETURN_ERROR_IF(workspace == NULL, memory_allocation);
ZSTD_cwksp_init(ws, workspace, size);
RETURN_ERROR_IF(workspace == NULL, memory_allocation, "NULL pointer!");
ZSTD_cwksp_init(ws, workspace, size, ZSTD_cwksp_dynamic_alloc);
return 0;
}
MEM_STATIC void ZSTD_cwksp_free(ZSTD_cwksp* ws, ZSTD_customMem customMem) {
void *ptr = ws->workspace;
DEBUGLOG(4, "cwksp: freeing workspace");
memset(ws, 0, sizeof(ZSTD_cwksp));
ZSTD_free(ptr, customMem);
ZSTD_memset(ws, 0, sizeof(ZSTD_cwksp));
ZSTD_customFree(ptr, customMem);
}
/**
@@ -486,13 +507,18 @@ MEM_STATIC void ZSTD_cwksp_free(ZSTD_cwksp* ws, ZSTD_customMem customMem) {
*/
MEM_STATIC void ZSTD_cwksp_move(ZSTD_cwksp* dst, ZSTD_cwksp* src) {
*dst = *src;
memset(src, 0, sizeof(ZSTD_cwksp));
ZSTD_memset(src, 0, sizeof(ZSTD_cwksp));
}
MEM_STATIC size_t ZSTD_cwksp_sizeof(const ZSTD_cwksp* ws) {
return (size_t)((BYTE*)ws->workspaceEnd - (BYTE*)ws->workspace);
}
MEM_STATIC size_t ZSTD_cwksp_used(const ZSTD_cwksp* ws) {
return (size_t)((BYTE*)ws->tableEnd - (BYTE*)ws->workspace)
+ (size_t)((BYTE*)ws->workspaceEnd - (BYTE*)ws->allocStart);
}
MEM_STATIC int ZSTD_cwksp_reserve_failed(const ZSTD_cwksp* ws) {
return ws->allocFailed;
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -14,7 +14,7 @@
/*-*******************************************************
* Dependencies
*********************************************************/
#include <string.h> /* memcpy, memmove, memset */
#include "zstd_deps.h" /* ZSTD_memcpy, ZSTD_memmove, ZSTD_memset */
#include "cpu.h" /* bmi2 */
#include "mem.h" /* low level memory routines */
#define FSE_STATIC_LINKING_ONLY
@@ -25,7 +25,7 @@
#include "zstd_ddict.h"
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
# include "zstd_legacy.h"
# include "../legacy/zstd_legacy.h"
#endif
@@ -65,6 +65,10 @@ void ZSTD_copyDDictParameters(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict)
dctx->virtualStart = ddict->dictContent;
dctx->dictEnd = (const BYTE*)ddict->dictContent + ddict->dictSize;
dctx->previousDstEnd = dctx->dictEnd;
#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
dctx->dictContentBeginForFuzzing = dctx->prefixStart;
dctx->dictContentEndForFuzzing = dctx->previousDstEnd;
#endif
if (ddict->entropyPresent) {
dctx->litEntropy = 1;
dctx->fseEntropy = 1;
@@ -107,7 +111,7 @@ ZSTD_loadEntropy_intoDDict(ZSTD_DDict* ddict,
/* load entropy tables */
RETURN_ERROR_IF(ZSTD_isError(ZSTD_loadDEntropy(
&ddict->entropy, ddict->dictContent, ddict->dictSize)),
dictionary_corrupted);
dictionary_corrupted, "");
ddict->entropyPresent = 1;
return 0;
}
@@ -123,17 +127,17 @@ static size_t ZSTD_initDDict_internal(ZSTD_DDict* ddict,
ddict->dictContent = dict;
if (!dict) dictSize = 0;
} else {
void* const internalBuffer = ZSTD_malloc(dictSize, ddict->cMem);
void* const internalBuffer = ZSTD_customMalloc(dictSize, ddict->cMem);
ddict->dictBuffer = internalBuffer;
ddict->dictContent = internalBuffer;
if (!internalBuffer) return ERROR(memory_allocation);
memcpy(internalBuffer, dict, dictSize);
ZSTD_memcpy(internalBuffer, dict, dictSize);
}
ddict->dictSize = dictSize;
ddict->entropy.hufTable[0] = (HUF_DTable)((HufLog)*0x1000001); /* cover both little and big endian */
/* parse dictionary content */
FORWARD_IF_ERROR( ZSTD_loadEntropy_intoDDict(ddict, dictContentType) );
FORWARD_IF_ERROR( ZSTD_loadEntropy_intoDDict(ddict, dictContentType) , "");
return 0;
}
@@ -143,9 +147,9 @@ ZSTD_DDict* ZSTD_createDDict_advanced(const void* dict, size_t dictSize,
ZSTD_dictContentType_e dictContentType,
ZSTD_customMem customMem)
{
if (!customMem.customAlloc ^ !customMem.customFree) return NULL;
if ((!customMem.customAlloc) ^ (!customMem.customFree)) return NULL;
{ ZSTD_DDict* const ddict = (ZSTD_DDict*) ZSTD_malloc(sizeof(ZSTD_DDict), customMem);
{ ZSTD_DDict* const ddict = (ZSTD_DDict*) ZSTD_customMalloc(sizeof(ZSTD_DDict), customMem);
if (ddict == NULL) return NULL;
ddict->cMem = customMem;
{ size_t const initResult = ZSTD_initDDict_internal(ddict,
@@ -194,7 +198,7 @@ const ZSTD_DDict* ZSTD_initStaticDDict(
if ((size_t)sBuffer & 7) return NULL; /* 8-aligned */
if (sBufferSize < neededSpace) return NULL;
if (dictLoadMethod == ZSTD_dlm_byCopy) {
memcpy(ddict+1, dict, dictSize); /* local copy */
ZSTD_memcpy(ddict+1, dict, dictSize); /* local copy */
dict = ddict+1;
}
if (ZSTD_isError( ZSTD_initDDict_internal(ddict,
@@ -209,8 +213,8 @@ size_t ZSTD_freeDDict(ZSTD_DDict* ddict)
{
if (ddict==NULL) return 0; /* support free on NULL */
{ ZSTD_customMem const cMem = ddict->cMem;
ZSTD_free(ddict->dictBuffer, cMem);
ZSTD_free(ddict, cMem);
ZSTD_customFree(ddict->dictBuffer, cMem);
ZSTD_customFree(ddict, cMem);
return 0;
}
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -15,7 +15,7 @@
/*-*******************************************************
* Dependencies
*********************************************************/
#include <stddef.h> /* size_t */
#include "zstd_deps.h" /* size_t */
#include "zstd.h" /* ZSTD_DDict, and several public functions */

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -15,7 +15,7 @@
/*-*******************************************************
* Dependencies
*********************************************************/
#include <stddef.h> /* size_t */
#include "zstd_deps.h" /* size_t */
#include "zstd.h" /* DCtx, and some public functions */
#include "zstd_internal.h" /* blockProperties_t, and some public functions */
#include "zstd_decompress_internal.h" /* ZSTD_seqSymbol */
@@ -48,12 +48,15 @@ size_t ZSTD_decompressBlock_internal(ZSTD_DCtx* dctx,
* this function must be called with valid parameters only
* (dt is large enough, normalizedCounter distribution total is a power of 2, max is within range, etc.)
* in which case it cannot fail.
* The workspace must be 4-byte aligned and at least ZSTD_BUILD_FSE_TABLE_WKSP_SIZE bytes, which is
* defined in zstd_decompress_internal.h.
* Internal use only.
*/
void ZSTD_buildFSETable(ZSTD_seqSymbol* dt,
const short* normalizedCounter, unsigned maxSymbolValue,
const U32* baseValue, const U32* nbAdditionalBits,
unsigned tableLog);
unsigned tableLog, void* wksp, size_t wkspSize,
int bmi2);
#endif /* ZSTD_DEC_BLOCK_H */

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -27,26 +27,26 @@
/*-*******************************************************
* Constants
*********************************************************/
static const U32 LL_base[MaxLL+1] = {
static UNUSED_ATTR const U32 LL_base[MaxLL+1] = {
0, 1, 2, 3, 4, 5, 6, 7,
8, 9, 10, 11, 12, 13, 14, 15,
16, 18, 20, 22, 24, 28, 32, 40,
48, 64, 0x80, 0x100, 0x200, 0x400, 0x800, 0x1000,
0x2000, 0x4000, 0x8000, 0x10000 };
static const U32 OF_base[MaxOff+1] = {
static UNUSED_ATTR const U32 OF_base[MaxOff+1] = {
0, 1, 1, 5, 0xD, 0x1D, 0x3D, 0x7D,
0xFD, 0x1FD, 0x3FD, 0x7FD, 0xFFD, 0x1FFD, 0x3FFD, 0x7FFD,
0xFFFD, 0x1FFFD, 0x3FFFD, 0x7FFFD, 0xFFFFD, 0x1FFFFD, 0x3FFFFD, 0x7FFFFD,
0xFFFFFD, 0x1FFFFFD, 0x3FFFFFD, 0x7FFFFFD, 0xFFFFFFD, 0x1FFFFFFD, 0x3FFFFFFD, 0x7FFFFFFD };
static const U32 OF_bits[MaxOff+1] = {
static UNUSED_ATTR const U32 OF_bits[MaxOff+1] = {
0, 1, 2, 3, 4, 5, 6, 7,
8, 9, 10, 11, 12, 13, 14, 15,
16, 17, 18, 19, 20, 21, 22, 23,
24, 25, 26, 27, 28, 29, 30, 31 };
static const U32 ML_base[MaxML+1] = {
static UNUSED_ATTR const U32 ML_base[MaxML+1] = {
3, 4, 5, 6, 7, 8, 9, 10,
11, 12, 13, 14, 15, 16, 17, 18,
19, 20, 21, 22, 23, 24, 25, 26,
@@ -73,12 +73,16 @@ static const U32 ML_base[MaxML+1] = {
#define SEQSYMBOL_TABLE_SIZE(log) (1 + (1 << (log)))
#define ZSTD_BUILD_FSE_TABLE_WKSP_SIZE (sizeof(S16) * (MaxSeq + 1) + (1u << MaxFSELog) + sizeof(U64))
#define ZSTD_BUILD_FSE_TABLE_WKSP_SIZE_U32 ((ZSTD_BUILD_FSE_TABLE_WKSP_SIZE + sizeof(U32) - 1) / sizeof(U32))
typedef struct {
ZSTD_seqSymbol LLTable[SEQSYMBOL_TABLE_SIZE(LLFSELog)]; /* Note : Space reserved for FSE Tables */
ZSTD_seqSymbol OFTable[SEQSYMBOL_TABLE_SIZE(OffFSELog)]; /* is also used as temporary workspace while building hufTable during DDict creation */
ZSTD_seqSymbol MLTable[SEQSYMBOL_TABLE_SIZE(MLFSELog)]; /* and therefore must be at least HUF_DECOMPRESS_WORKSPACE_SIZE large */
HUF_DTable hufTable[HUF_DTABLE_SIZE(HufLog)]; /* can accommodate HUF_decompress4X */
U32 rep[ZSTD_REP_NUM];
U32 workspace[ZSTD_BUILD_FSE_TABLE_WKSP_SIZE_U32];
} ZSTD_entropyDTables_t;
typedef enum { ZSTDds_getFrameHeaderSize, ZSTDds_decodeFrameHeader,
@@ -117,6 +121,8 @@ struct ZSTD_DCtx_s
XXH64_state_t xxhState;
size_t headerSize;
ZSTD_format_e format;
ZSTD_forceIgnoreChecksum_e forceIgnoreChecksum; /* User specified: if == 1, will ignore checksums in compressed frame. Default == 0 */
U32 validateChecksum; /* if == 1, will validate checksum. Is == 1 if (fParams.checksumFlag == 1) and (forceIgnoreChecksum == 0). */
const BYTE* litPtr;
ZSTD_customMem customMem;
size_t litSize;
@@ -147,10 +153,19 @@ struct ZSTD_DCtx_s
U32 legacyVersion;
U32 hostageByte;
int noForwardProgress;
ZSTD_bufferMode_e outBufferMode;
ZSTD_outBuffer expectedOutBuffer;
/* workspace */
BYTE litBuffer[ZSTD_BLOCKSIZE_MAX + WILDCOPY_OVERLENGTH];
BYTE headerBuffer[ZSTD_FRAMEHEADERSIZE_MAX];
size_t oversizedDuration;
#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
void const* dictContentBeginForFuzzing;
void const* dictContentEndForFuzzing;
#endif
}; /* typedef'd to ZSTD_DCtx within "zstd.h" */
@@ -160,7 +175,7 @@ struct ZSTD_DCtx_s
/*! ZSTD_loadDEntropy() :
* dict : must point at beginning of a valid zstd dictionary.
* @return : size of entropy tables read */
* @return : size of dictionary header (size of magic number + dict ID + entropy tables) */
size_t ZSTD_loadDEntropy(ZSTD_entropyDTables_t* entropy,
const void* const dict, size_t const dictSize);

111
vendor/github.com/DataDog/zstd/zstd_deps.h generated vendored Normal file
View File

@@ -0,0 +1,111 @@
/*
* Copyright (c) 2016-2020, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
/* This file provides common libc dependencies that zstd requires.
* The purpose is to allow replacing this file with a custom implementation
* to compile zstd without libc support.
*/
/* Need:
* NULL
* INT_MAX
* UINT_MAX
* ZSTD_memcpy()
* ZSTD_memset()
* ZSTD_memmove()
*/
#ifndef ZSTD_DEPS_COMMON
#define ZSTD_DEPS_COMMON
#include <limits.h>
#include <stddef.h>
#include <string.h>
#if defined(__GNUC__) && __GNUC__ >= 4
# define ZSTD_memcpy(d,s,l) __builtin_memcpy((d),(s),(l))
# define ZSTD_memmove(d,s,l) __builtin_memmove((d),(s),(l))
# define ZSTD_memset(p,v,l) __builtin_memset((p),(v),(l))
#else
# define ZSTD_memcpy(d,s,l) memcpy((d),(s),(l))
# define ZSTD_memmove(d,s,l) memmove((d),(s),(l))
# define ZSTD_memset(p,v,l) memset((p),(v),(l))
#endif
#endif /* ZSTD_DEPS_COMMON */
/* Need:
* ZSTD_malloc()
* ZSTD_free()
* ZSTD_calloc()
*/
#ifdef ZSTD_DEPS_NEED_MALLOC
#ifndef ZSTD_DEPS_MALLOC
#define ZSTD_DEPS_MALLOC
#include <stdlib.h>
#define ZSTD_malloc(s) malloc(s)
#define ZSTD_calloc(n,s) calloc((n), (s))
#define ZSTD_free(p) free((p))
#endif /* ZSTD_DEPS_MALLOC */
#endif /* ZSTD_DEPS_NEED_MALLOC */
/*
* Provides 64-bit math support.
* Need:
* U64 ZSTD_div64(U64 dividend, U32 divisor)
*/
#ifdef ZSTD_DEPS_NEED_MATH64
#ifndef ZSTD_DEPS_MATH64
#define ZSTD_DEPS_MATH64
#define ZSTD_div64(dividend, divisor) ((dividend) / (divisor))
#endif /* ZSTD_DEPS_MATH64 */
#endif /* ZSTD_DEPS_NEED_MATH64 */
/* Need:
* assert()
*/
#ifdef ZSTD_DEPS_NEED_ASSERT
#ifndef ZSTD_DEPS_ASSERT
#define ZSTD_DEPS_ASSERT
#include <assert.h>
#endif /* ZSTD_DEPS_ASSERT */
#endif /* ZSTD_DEPS_NEED_ASSERT */
/* Need:
* ZSTD_DEBUG_PRINT()
*/
#ifdef ZSTD_DEPS_NEED_IO
#ifndef ZSTD_DEPS_IO
#define ZSTD_DEPS_IO
#include <stdio.h>
#define ZSTD_DEBUG_PRINT(...) fprintf(stderr, __VA_ARGS__)
#endif /* ZSTD_DEPS_IO */
#endif /* ZSTD_DEPS_NEED_IO */
/* Only requested when <stdint.h> is known to be present.
* Need:
* intptr_t
*/
#ifdef ZSTD_DEPS_NEED_STDINT
#ifndef ZSTD_DEPS_STDINT
#define ZSTD_DEPS_STDINT
#include <stdint.h>
#endif /* ZSTD_DEPS_STDINT */
#endif /* ZSTD_DEPS_NEED_STDINT */

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -31,15 +31,15 @@ void ZSTD_fillDoubleHashTable(ZSTD_matchState_t* ms,
* is empty.
*/
for (; ip + fastHashFillStep - 1 <= iend; ip += fastHashFillStep) {
U32 const current = (U32)(ip - base);
U32 const curr = (U32)(ip - base);
U32 i;
for (i = 0; i < fastHashFillStep; ++i) {
size_t const smHash = ZSTD_hashPtr(ip + i, hBitsS, mls);
size_t const lgHash = ZSTD_hashPtr(ip + i, hBitsL, 8);
if (i == 0)
hashSmall[smHash] = current + i;
hashSmall[smHash] = curr + i;
if (i == 0 || hashLarge[lgHash] == 0)
hashLarge[lgHash] = current + i;
hashLarge[lgHash] = curr + i;
/* Only load extra positions for ZSTD_dtlm_full */
if (dtlm == ZSTD_dtlm_fast)
break;
@@ -63,10 +63,8 @@ size_t ZSTD_compressBlock_doubleFast_generic(
const BYTE* ip = istart;
const BYTE* anchor = istart;
const U32 endIndex = (U32)((size_t)(istart - base) + srcSize);
const U32 lowestValid = ms->window.dictLimit;
const U32 maxDistance = 1U << cParams->windowLog;
/* presumes that, if there is a dictionary, it must be using Attach mode */
const U32 prefixLowestIndex = (endIndex - lowestValid > maxDistance) ? endIndex - maxDistance : lowestValid;
const U32 prefixLowestIndex = ZSTD_getLowestPrefixIndex(ms, endIndex, cParams->windowLog);
const BYTE* const prefixLowest = base + prefixLowestIndex;
const BYTE* const iend = istart + srcSize;
const BYTE* const ilimit = iend - HASH_READ_SIZE;
@@ -96,7 +94,7 @@ size_t ZSTD_compressBlock_doubleFast_generic(
dictCParams->hashLog : hBitsL;
const U32 dictHBitsS = dictMode == ZSTD_dictMatchState ?
dictCParams->chainLog : hBitsS;
const U32 dictAndPrefixLength = (U32)(ip - prefixLowest + dictEnd - dictStart);
const U32 dictAndPrefixLength = (U32)((ip - prefixLowest) + (dictEnd - dictStart));
DEBUGLOG(5, "ZSTD_compressBlock_doubleFast_generic");
@@ -104,13 +102,15 @@ size_t ZSTD_compressBlock_doubleFast_generic(
/* if a dictionary is attached, it must be within window range */
if (dictMode == ZSTD_dictMatchState) {
assert(lowestValid + maxDistance >= endIndex);
assert(ms->window.dictLimit + (1U << cParams->windowLog) >= endIndex);
}
/* init */
ip += (dictAndPrefixLength == 0);
if (dictMode == ZSTD_noDict) {
U32 const maxRep = (U32)(ip - prefixLowest);
U32 const curr = (U32)(ip - base);
U32 const windowLow = ZSTD_getLowestPrefixIndex(ms, curr, cParams->windowLog);
U32 const maxRep = curr - windowLow;
if (offset_2 > maxRep) offsetSaved = offset_2, offset_2 = 0;
if (offset_1 > maxRep) offsetSaved = offset_1, offset_1 = 0;
}
@@ -129,17 +129,17 @@ size_t ZSTD_compressBlock_doubleFast_generic(
size_t const h = ZSTD_hashPtr(ip, hBitsS, mls);
size_t const dictHL = ZSTD_hashPtr(ip, dictHBitsL, 8);
size_t const dictHS = ZSTD_hashPtr(ip, dictHBitsS, mls);
U32 const current = (U32)(ip-base);
U32 const curr = (U32)(ip-base);
U32 const matchIndexL = hashLong[h2];
U32 matchIndexS = hashSmall[h];
const BYTE* matchLong = base + matchIndexL;
const BYTE* match = base + matchIndexS;
const U32 repIndex = current + 1 - offset_1;
const U32 repIndex = curr + 1 - offset_1;
const BYTE* repMatch = (dictMode == ZSTD_dictMatchState
&& repIndex < prefixLowestIndex) ?
dictBase + (repIndex - dictIndexDelta) :
base + repIndex;
hashLong[h2] = hashSmall[h] = current; /* update hash tables */
hashLong[h2] = hashSmall[h] = curr; /* update hash tables */
/* check dictMatchState repcode */
if (dictMode == ZSTD_dictMatchState
@@ -177,7 +177,7 @@ size_t ZSTD_compressBlock_doubleFast_generic(
if (dictMatchL > dictStart && MEM_read64(dictMatchL) == MEM_read64(ip)) {
mLength = ZSTD_count_2segments(ip+8, dictMatchL+8, iend, dictEnd, prefixLowest) + 8;
offset = (U32)(current - dictMatchIndexL - dictIndexDelta);
offset = (U32)(curr - dictMatchIndexL - dictIndexDelta);
while (((ip>anchor) & (dictMatchL>dictStart)) && (ip[-1] == dictMatchL[-1])) { ip--; dictMatchL--; mLength++; } /* catch up */
goto _match_found;
} }
@@ -198,6 +198,9 @@ size_t ZSTD_compressBlock_doubleFast_generic(
} }
ip += ((ip-anchor) >> kSearchStrength) + 1;
#if defined(__aarch64__)
PREFETCH_L1(ip+256);
#endif
continue;
_search_next_long:
@@ -206,7 +209,7 @@ _search_next_long:
size_t const dictHLNext = ZSTD_hashPtr(ip+1, dictHBitsL, 8);
U32 const matchIndexL3 = hashLong[hl3];
const BYTE* matchL3 = base + matchIndexL3;
hashLong[hl3] = current + 1;
hashLong[hl3] = curr + 1;
/* check prefix long +1 match */
if (matchIndexL3 > prefixLowestIndex) {
@@ -225,7 +228,7 @@ _search_next_long:
if (dictMatchL3 > dictStart && MEM_read64(dictMatchL3) == MEM_read64(ip+1)) {
mLength = ZSTD_count_2segments(ip+1+8, dictMatchL3+8, iend, dictEnd, prefixLowest) + 8;
ip++;
offset = (U32)(current + 1 - dictMatchIndexL3 - dictIndexDelta);
offset = (U32)(curr + 1 - dictMatchIndexL3 - dictIndexDelta);
while (((ip>anchor) & (dictMatchL3>dictStart)) && (ip[-1] == dictMatchL3[-1])) { ip--; dictMatchL3--; mLength++; } /* catch up */
goto _match_found;
} } }
@@ -233,7 +236,7 @@ _search_next_long:
/* if no long +1 match, explore the short match we found */
if (dictMode == ZSTD_dictMatchState && matchIndexS < prefixLowestIndex) {
mLength = ZSTD_count_2segments(ip+4, match+4, iend, dictEnd, prefixLowest) + 4;
offset = (U32)(current - matchIndexS);
offset = (U32)(curr - matchIndexS);
while (((ip>anchor) & (match>dictStart)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */
} else {
mLength = ZSTD_count(ip+4, match+4, iend) + 4;
@@ -257,7 +260,7 @@ _match_stored:
if (ip <= ilimit) {
/* Complementary insertion */
/* done after iLimit test, as candidates could be > iend-8 */
{ U32 const indexToInsert = current+2;
{ U32 const indexToInsert = curr+2;
hashLong[ZSTD_hashPtr(base+indexToInsert, hBitsL, 8)] = indexToInsert;
hashLong[ZSTD_hashPtr(ip-2, hBitsL, 8)] = (U32)(ip-2-base);
hashSmall[ZSTD_hashPtr(base+indexToInsert, hBitsS, mls)] = indexToInsert;
@@ -271,7 +274,7 @@ _match_stored:
U32 const repIndex2 = current2 - offset_2;
const BYTE* repMatch2 = dictMode == ZSTD_dictMatchState
&& repIndex2 < prefixLowestIndex ?
dictBase - dictIndexDelta + repIndex2 :
dictBase + repIndex2 - dictIndexDelta :
base + repIndex2;
if ( ((U32)((prefixLowestIndex-1) - (U32)repIndex2) >= 3 /* intentional overflow */)
&& (MEM_read32(repMatch2) == MEM_read32(ip)) ) {
@@ -398,12 +401,12 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
const BYTE* const matchLongBase = matchLongIndex < prefixStartIndex ? dictBase : base;
const BYTE* matchLong = matchLongBase + matchLongIndex;
const U32 current = (U32)(ip-base);
const U32 repIndex = current + 1 - offset_1; /* offset_1 expected <= current +1 */
const U32 curr = (U32)(ip-base);
const U32 repIndex = curr + 1 - offset_1; /* offset_1 expected <= curr +1 */
const BYTE* const repBase = repIndex < prefixStartIndex ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
size_t mLength;
hashSmall[hSmall] = hashLong[hLong] = current; /* update hash table */
hashSmall[hSmall] = hashLong[hLong] = curr; /* update hash table */
if ((((U32)((prefixStartIndex-1) - repIndex) >= 3) /* intentional underflow : ensure repIndex doesn't overlap dict + prefix */
& (repIndex > dictStartIndex))
@@ -418,7 +421,7 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
const BYTE* const lowMatchPtr = matchLongIndex < prefixStartIndex ? dictStart : prefixStart;
U32 offset;
mLength = ZSTD_count_2segments(ip+8, matchLong+8, iend, matchEnd, prefixStart) + 8;
offset = current - matchLongIndex;
offset = curr - matchLongIndex;
while (((ip>anchor) & (matchLong>lowMatchPtr)) && (ip[-1] == matchLong[-1])) { ip--; matchLong--; mLength++; } /* catch up */
offset_2 = offset_1;
offset_1 = offset;
@@ -430,19 +433,19 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
const BYTE* const match3Base = matchIndex3 < prefixStartIndex ? dictBase : base;
const BYTE* match3 = match3Base + matchIndex3;
U32 offset;
hashLong[h3] = current + 1;
hashLong[h3] = curr + 1;
if ( (matchIndex3 > dictStartIndex) && (MEM_read64(match3) == MEM_read64(ip+1)) ) {
const BYTE* const matchEnd = matchIndex3 < prefixStartIndex ? dictEnd : iend;
const BYTE* const lowMatchPtr = matchIndex3 < prefixStartIndex ? dictStart : prefixStart;
mLength = ZSTD_count_2segments(ip+9, match3+8, iend, matchEnd, prefixStart) + 8;
ip++;
offset = current+1 - matchIndex3;
offset = curr+1 - matchIndex3;
while (((ip>anchor) & (match3>lowMatchPtr)) && (ip[-1] == match3[-1])) { ip--; match3--; mLength++; } /* catch up */
} else {
const BYTE* const matchEnd = matchIndex < prefixStartIndex ? dictEnd : iend;
const BYTE* const lowMatchPtr = matchIndex < prefixStartIndex ? dictStart : prefixStart;
mLength = ZSTD_count_2segments(ip+4, match+4, iend, matchEnd, prefixStart) + 4;
offset = current - matchIndex;
offset = curr - matchIndex;
while (((ip>anchor) & (match>lowMatchPtr)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */
}
offset_2 = offset_1;
@@ -461,7 +464,7 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
if (ip <= ilimit) {
/* Complementary insertion */
/* done after iLimit test, as candidates could be > iend-8 */
{ U32 const indexToInsert = current+2;
{ U32 const indexToInsert = curr+2;
hashLong[ZSTD_hashPtr(base+indexToInsert, hBitsL, 8)] = indexToInsert;
hashLong[ZSTD_hashPtr(ip-2, hBitsL, 8)] = (U32)(ip-2-base);
hashSmall[ZSTD_hashPtr(base+indexToInsert, hBitsS, mls)] = indexToInsert;

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -76,6 +76,8 @@ typedef enum {
/* following error codes are __NOT STABLE__, they can be removed or changed in future versions */
ZSTD_error_frameIndex_tooLarge = 100,
ZSTD_error_seekableIO = 102,
ZSTD_error_dstBuffer_wrong = 104,
ZSTD_error_srcBuffer_wrong = 105,
ZSTD_error_maxCode = 120 /* never EVER use this value directly, it can change in future versions! Use ZSTD_isError() instead */
} ZSTD_ErrorCode;

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -29,16 +29,16 @@ void ZSTD_fillHashTable(ZSTD_matchState_t* ms,
* Insert the other positions if their hash entry is empty.
*/
for ( ; ip + fastHashFillStep < iend + 2; ip += fastHashFillStep) {
U32 const current = (U32)(ip - base);
U32 const curr = (U32)(ip - base);
size_t const hash0 = ZSTD_hashPtr(ip, hBits, mls);
hashTable[hash0] = current;
hashTable[hash0] = curr;
if (dtlm == ZSTD_dtlm_fast) continue;
/* Only load extra positions for ZSTD_dtlm_full */
{ U32 p;
for (p = 1; p < fastHashFillStep; ++p) {
size_t const hash = ZSTD_hashPtr(ip + p, hBits, mls);
if (hashTable[hash] == 0) { /* not yet filled */
hashTable[hash] = current + p;
hashTable[hash] = curr + p;
} } } }
}
@@ -61,9 +61,7 @@ ZSTD_compressBlock_fast_generic(
const BYTE* ip1;
const BYTE* anchor = istart;
const U32 endIndex = (U32)((size_t)(istart - base) + srcSize);
const U32 maxDistance = 1U << cParams->windowLog;
const U32 validStartIndex = ms->window.dictLimit;
const U32 prefixStartIndex = (endIndex - validStartIndex > maxDistance) ? endIndex - maxDistance : validStartIndex;
const U32 prefixStartIndex = ZSTD_getLowestPrefixIndex(ms, endIndex, cParams->windowLog);
const BYTE* const prefixStart = base + prefixStartIndex;
const BYTE* const iend = istart + srcSize;
const BYTE* const ilimit = iend - HASH_READ_SIZE;
@@ -74,12 +72,21 @@ ZSTD_compressBlock_fast_generic(
DEBUGLOG(5, "ZSTD_compressBlock_fast_generic");
ip0 += (ip0 == prefixStart);
ip1 = ip0 + 1;
{ U32 const maxRep = (U32)(ip0 - prefixStart);
{ U32 const curr = (U32)(ip0 - base);
U32 const windowLow = ZSTD_getLowestPrefixIndex(ms, curr, cParams->windowLog);
U32 const maxRep = curr - windowLow;
if (offset_2 > maxRep) offsetSaved = offset_2, offset_2 = 0;
if (offset_1 > maxRep) offsetSaved = offset_1, offset_1 = 0;
}
/* Main Search Loop */
#ifdef __INTEL_COMPILER
/* From intel 'The vector pragma indicates that the loop should be
* vectorized if it is legal to do so'. Can be used together with
* #pragma ivdep (but have opted to exclude that because intel
* warns against using it).*/
#pragma vector always
#endif
while (ip1 < ilimit) { /* < instead of <=, because check at ip0+2 */
size_t mLength;
BYTE const* ip2 = ip0 + 2;
@@ -91,19 +98,25 @@ ZSTD_compressBlock_fast_generic(
U32 const current1 = (U32)(ip1-base);
U32 const matchIndex0 = hashTable[h0];
U32 const matchIndex1 = hashTable[h1];
BYTE const* repMatch = ip2-offset_1;
BYTE const* repMatch = ip2 - offset_1;
const BYTE* match0 = base + matchIndex0;
const BYTE* match1 = base + matchIndex1;
U32 offcode;
#if defined(__aarch64__)
PREFETCH_L1(ip0+256);
#endif
hashTable[h0] = current0; /* update hash table */
hashTable[h1] = current1; /* update hash table */
assert(ip0 + 1 == ip1);
if ((offset_1 > 0) & (MEM_read32(repMatch) == MEM_read32(ip2))) {
mLength = ip2[-1] == repMatch[-1] ? 1 : 0;
mLength = (ip2[-1] == repMatch[-1]) ? 1 : 0;
ip0 = ip2 - mLength;
match0 = repMatch - mLength;
mLength += 4;
offcode = 0;
goto _match;
}
@@ -128,19 +141,18 @@ _offset: /* Requires: ip0, match0 */
offset_2 = offset_1;
offset_1 = (U32)(ip0-match0);
offcode = offset_1 + ZSTD_REP_MOVE;
mLength = 0;
mLength = 4;
/* Count the backwards match length */
while (((ip0>anchor) & (match0>prefixStart))
&& (ip0[-1] == match0[-1])) { ip0--; match0--; mLength++; } /* catch up */
_match: /* Requires: ip0, match0, offcode */
/* Count the forward length */
mLength += ZSTD_count(ip0+mLength+4, match0+mLength+4, iend) + 4;
mLength += ZSTD_count(ip0+mLength, match0+mLength, iend);
ZSTD_storeSeq(seqStore, (size_t)(ip0-anchor), anchor, iend, offcode, mLength-MINMATCH);
/* match found */
ip0 += mLength;
anchor = ip0;
ip1 = ip0 + 1;
if (ip0 <= ilimit) {
/* Fill Table */
@@ -148,19 +160,18 @@ _match: /* Requires: ip0, match0, offcode */
hashTable[ZSTD_hashPtr(base+current0+2, hlog, mls)] = current0+2; /* here because current+2 could be > iend-8 */
hashTable[ZSTD_hashPtr(ip0-2, hlog, mls)] = (U32)(ip0-2-base);
while ( ((ip0 <= ilimit) & (offset_2>0)) /* offset_2==0 means offset_2 is invalidated */
&& (MEM_read32(ip0) == MEM_read32(ip0 - offset_2)) ) {
/* store sequence */
size_t const rLength = ZSTD_count(ip0+4, ip0+4-offset_2, iend) + 4;
{ U32 const tmpOff = offset_2; offset_2 = offset_1; offset_1 = tmpOff; } /* swap offset_2 <=> offset_1 */
hashTable[ZSTD_hashPtr(ip0, hlog, mls)] = (U32)(ip0-base);
ip0 += rLength;
ip1 = ip0 + 1;
ZSTD_storeSeq(seqStore, 0 /*litLen*/, anchor, iend, 0 /*offCode*/, rLength-MINMATCH);
anchor = ip0;
continue; /* faster when present (confirmed on gcc-8) ... (?) */
}
}
if (offset_2 > 0) { /* offset_2==0 means offset_2 is invalidated */
while ( (ip0 <= ilimit) && (MEM_read32(ip0) == MEM_read32(ip0 - offset_2)) ) {
/* store sequence */
size_t const rLength = ZSTD_count(ip0+4, ip0+4-offset_2, iend) + 4;
{ U32 const tmpOff = offset_2; offset_2 = offset_1; offset_1 = tmpOff; } /* swap offset_2 <=> offset_1 */
hashTable[ZSTD_hashPtr(ip0, hlog, mls)] = (U32)(ip0-base);
ip0 += rLength;
ZSTD_storeSeq(seqStore, 0 /*litLen*/, anchor, iend, 0 /*offCode*/, rLength-MINMATCH);
anchor = ip0;
continue; /* faster when present (confirmed on gcc-8) ... (?) */
} } }
ip1 = ip0 + 1;
}
/* save reps for next block */
@@ -247,14 +258,14 @@ size_t ZSTD_compressBlock_fast_dictMatchState_generic(
while (ip < ilimit) { /* < instead of <=, because repcode check at (ip+1) */
size_t mLength;
size_t const h = ZSTD_hashPtr(ip, hlog, mls);
U32 const current = (U32)(ip-base);
U32 const curr = (U32)(ip-base);
U32 const matchIndex = hashTable[h];
const BYTE* match = base + matchIndex;
const U32 repIndex = current + 1 - offset_1;
const U32 repIndex = curr + 1 - offset_1;
const BYTE* repMatch = (repIndex < prefixStartIndex) ?
dictBase + (repIndex - dictIndexDelta) :
base + repIndex;
hashTable[h] = current; /* update hash table */
hashTable[h] = curr; /* update hash table */
if ( ((U32)((prefixStartIndex-1) - repIndex) >= 3) /* intentional underflow : ensure repIndex isn't overlapping dict + prefix */
&& (MEM_read32(repMatch) == MEM_read32(ip+1)) ) {
@@ -273,7 +284,7 @@ size_t ZSTD_compressBlock_fast_dictMatchState_generic(
continue;
} else {
/* found a dict match */
U32 const offset = (U32)(current-dictMatchIndex-dictIndexDelta);
U32 const offset = (U32)(curr-dictMatchIndex-dictIndexDelta);
mLength = ZSTD_count_2segments(ip+4, dictMatch+4, iend, dictEnd, prefixStart) + 4;
while (((ip>anchor) & (dictMatch>dictStart))
&& (ip[-1] == dictMatch[-1])) {
@@ -305,8 +316,8 @@ size_t ZSTD_compressBlock_fast_dictMatchState_generic(
if (ip <= ilimit) {
/* Fill Table */
assert(base+current+2 > istart); /* check base overflow */
hashTable[ZSTD_hashPtr(base+current+2, hlog, mls)] = current+2; /* here because current+2 could be > iend-8 */
assert(base+curr+2 > istart); /* check base overflow */
hashTable[ZSTD_hashPtr(base+curr+2, hlog, mls)] = curr+2; /* here because curr+2 could be > iend-8 */
hashTable[ZSTD_hashPtr(ip-2, hlog, mls)] = (U32)(ip-2-base);
/* check immediate repcode */
@@ -387,7 +398,7 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(
const BYTE* const ilimit = iend - 8;
U32 offset_1=rep[0], offset_2=rep[1];
DEBUGLOG(5, "ZSTD_compressBlock_fast_extDict_generic");
DEBUGLOG(5, "ZSTD_compressBlock_fast_extDict_generic (offset_1=%u)", offset_1);
/* switch to "regular" variant if extDict is invalidated due to maxDistance */
if (prefixStartIndex == dictStartIndex)
@@ -399,12 +410,13 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(
const U32 matchIndex = hashTable[h];
const BYTE* const matchBase = matchIndex < prefixStartIndex ? dictBase : base;
const BYTE* match = matchBase + matchIndex;
const U32 current = (U32)(ip-base);
const U32 repIndex = current + 1 - offset_1;
const U32 curr = (U32)(ip-base);
const U32 repIndex = curr + 1 - offset_1;
const BYTE* const repBase = repIndex < prefixStartIndex ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
hashTable[h] = current; /* update hash table */
assert(offset_1 <= current +1); /* check repIndex */
hashTable[h] = curr; /* update hash table */
DEBUGLOG(7, "offset_1 = %u , curr = %u", offset_1, curr);
assert(offset_1 <= curr +1); /* check repIndex */
if ( (((U32)((prefixStartIndex-1) - repIndex) >= 3) /* intentional underflow */ & (repIndex > dictStartIndex))
&& (MEM_read32(repMatch) == MEM_read32(ip+1)) ) {
@@ -423,7 +435,7 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(
}
{ const BYTE* const matchEnd = matchIndex < prefixStartIndex ? dictEnd : iend;
const BYTE* const lowMatchPtr = matchIndex < prefixStartIndex ? dictStart : prefixStart;
U32 const offset = current - matchIndex;
U32 const offset = curr - matchIndex;
size_t mLength = ZSTD_count_2segments(ip+4, match+4, iend, matchEnd, prefixStart) + 4;
while (((ip>anchor) & (match>lowMatchPtr)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */
offset_2 = offset_1; offset_1 = offset; /* update offset history */
@@ -434,7 +446,7 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(
if (ip <= ilimit) {
/* Fill Table */
hashTable[ZSTD_hashPtr(base+current+2, hlog, mls)] = current+2;
hashTable[ZSTD_hashPtr(base+curr+2, hlog, mls)] = curr+2;
hashTable[ZSTD_hashPtr(ip-2, hlog, mls)] = (U32)(ip-2-base);
/* check immediate repcode */
while (ip <= ilimit) {

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -19,6 +19,9 @@
/*-*************************************
* Dependencies
***************************************/
#if !defined(ZSTD_NO_INTRINSICS) && defined(__ARM_NEON)
#include <arm_neon.h>
#endif
#include "compiler.h"
#include "mem.h"
#include "debug.h" /* assert, DEBUGLOG, RAWLOG, g_debuglevel */
@@ -53,6 +56,31 @@ extern "C" {
#define MIN(a,b) ((a)<(b) ? (a) : (b))
#define MAX(a,b) ((a)>(b) ? (a) : (b))
/**
* Ignore: this is an internal helper.
*
* This is a helper function to help force C99-correctness during compilation.
* Under strict compilation modes, variadic macro arguments can't be empty.
* However, variadic function arguments can be. Using a function therefore lets
* us statically check that at least one (string) argument was passed,
* independent of the compilation flags.
*/
static INLINE_KEYWORD UNUSED_ATTR
void _force_has_format_string(const char *format, ...) {
(void)format;
}
/**
* Ignore: this is an internal helper.
*
* We want to force this function invocation to be syntactically correct, but
* we don't want to force runtime evaluation of its arguments.
*/
#define _FORCE_HAS_FORMAT_STRING(...) \
if (0) { \
_force_has_format_string(__VA_ARGS__); \
}
/**
* Return the specified error if the condition evaluates to true.
*
@@ -62,7 +90,9 @@ extern "C" {
*/
#define RETURN_ERROR_IF(cond, err, ...) \
if (cond) { \
RAWLOG(3, "%s:%d: ERROR!: check %s failed, returning %s", __FILE__, __LINE__, ZSTD_QUOTE(cond), ZSTD_QUOTE(ERROR(err))); \
RAWLOG(3, "%s:%d: ERROR!: check %s failed, returning %s", \
__FILE__, __LINE__, ZSTD_QUOTE(cond), ZSTD_QUOTE(ERROR(err))); \
_FORCE_HAS_FORMAT_STRING(__VA_ARGS__); \
RAWLOG(3, ": " __VA_ARGS__); \
RAWLOG(3, "\n"); \
return ERROR(err); \
@@ -75,7 +105,9 @@ extern "C" {
*/
#define RETURN_ERROR(err, ...) \
do { \
RAWLOG(3, "%s:%d: ERROR!: unconditional check failed, returning %s", __FILE__, __LINE__, ZSTD_QUOTE(ERROR(err))); \
RAWLOG(3, "%s:%d: ERROR!: unconditional check failed, returning %s", \
__FILE__, __LINE__, ZSTD_QUOTE(ERROR(err))); \
_FORCE_HAS_FORMAT_STRING(__VA_ARGS__); \
RAWLOG(3, ": " __VA_ARGS__); \
RAWLOG(3, "\n"); \
return ERROR(err); \
@@ -90,7 +122,9 @@ extern "C" {
do { \
size_t const err_code = (err); \
if (ERR_isError(err_code)) { \
RAWLOG(3, "%s:%d: ERROR!: forwarding error in %s: %s", __FILE__, __LINE__, ZSTD_QUOTE(err), ERR_getErrorName(err_code)); \
RAWLOG(3, "%s:%d: ERROR!: forwarding error in %s: %s", \
__FILE__, __LINE__, ZSTD_QUOTE(err), ERR_getErrorName(err_code)); \
_FORCE_HAS_FORMAT_STRING(__VA_ARGS__); \
RAWLOG(3, ": " __VA_ARGS__); \
RAWLOG(3, "\n"); \
return err_code; \
@@ -105,7 +139,7 @@ extern "C" {
#define ZSTD_REP_NUM 3 /* number of repcodes */
#define ZSTD_REP_MOVE (ZSTD_REP_NUM-1)
static const U32 repStartValue[ZSTD_REP_NUM] = { 1, 4, 8 };
static UNUSED_ATTR const U32 repStartValue[ZSTD_REP_NUM] = { 1, 4, 8 };
#define KB *(1 <<10)
#define MB *(1 <<20)
@@ -119,15 +153,17 @@ static const U32 repStartValue[ZSTD_REP_NUM] = { 1, 4, 8 };
#define BIT0 1
#define ZSTD_WINDOWLOG_ABSOLUTEMIN 10
static const size_t ZSTD_fcs_fieldSize[4] = { 0, 2, 4, 8 };
static const size_t ZSTD_did_fieldSize[4] = { 0, 1, 2, 4 };
static UNUSED_ATTR const size_t ZSTD_fcs_fieldSize[4] = { 0, 2, 4, 8 };
static UNUSED_ATTR const size_t ZSTD_did_fieldSize[4] = { 0, 1, 2, 4 };
#define ZSTD_FRAMEIDSIZE 4 /* magic number size */
#define ZSTD_BLOCKHEADERSIZE 3 /* C standard doesn't allow `static const` variable to be init using another `static const` variable */
static const size_t ZSTD_blockHeaderSize = ZSTD_BLOCKHEADERSIZE;
static UNUSED_ATTR const size_t ZSTD_blockHeaderSize = ZSTD_BLOCKHEADERSIZE;
typedef enum { bt_raw, bt_rle, bt_compressed, bt_reserved } blockType_e;
#define ZSTD_FRAMECHECKSUMSIZE 4
#define MIN_SEQUENCES_SIZE 1 /* nbSeq==0 */
#define MIN_CBLOCK_SIZE (1 /*litCSize*/ + 1 /* RLE or RAW */ + MIN_SEQUENCES_SIZE /* nbSeq==0 */) /* for a non-null block */
@@ -150,51 +186,77 @@ typedef enum { set_basic, set_rle, set_compressed, set_repeat } symbolEncodingTy
#define OffFSELog 8
#define MaxFSELog MAX(MAX(MLFSELog, LLFSELog), OffFSELog)
static const U32 LL_bits[MaxLL+1] = { 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
1, 1, 1, 1, 2, 2, 3, 3,
4, 6, 7, 8, 9,10,11,12,
13,14,15,16 };
static const S16 LL_defaultNorm[MaxLL+1] = { 4, 3, 2, 2, 2, 2, 2, 2,
2, 2, 2, 2, 2, 1, 1, 1,
2, 2, 2, 2, 2, 2, 2, 2,
2, 3, 2, 1, 1, 1, 1, 1,
-1,-1,-1,-1 };
#define ZSTD_MAX_HUF_HEADER_SIZE 128 /* header + <= 127 byte tree description */
/* Each table cannot take more than #symbols * FSELog bits */
#define ZSTD_MAX_FSE_HEADERS_SIZE (((MaxML + 1) * MLFSELog + (MaxLL + 1) * LLFSELog + (MaxOff + 1) * OffFSELog + 7) / 8)
static UNUSED_ATTR const U32 LL_bits[MaxLL+1] = {
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
1, 1, 1, 1, 2, 2, 3, 3,
4, 6, 7, 8, 9,10,11,12,
13,14,15,16
};
static UNUSED_ATTR const S16 LL_defaultNorm[MaxLL+1] = {
4, 3, 2, 2, 2, 2, 2, 2,
2, 2, 2, 2, 2, 1, 1, 1,
2, 2, 2, 2, 2, 2, 2, 2,
2, 3, 2, 1, 1, 1, 1, 1,
-1,-1,-1,-1
};
#define LL_DEFAULTNORMLOG 6 /* for static allocation */
static const U32 LL_defaultNormLog = LL_DEFAULTNORMLOG;
static UNUSED_ATTR const U32 LL_defaultNormLog = LL_DEFAULTNORMLOG;
static const U32 ML_bits[MaxML+1] = { 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
1, 1, 1, 1, 2, 2, 3, 3,
4, 4, 5, 7, 8, 9,10,11,
12,13,14,15,16 };
static const S16 ML_defaultNorm[MaxML+1] = { 1, 4, 3, 2, 2, 2, 2, 2,
2, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1,-1,-1,
-1,-1,-1,-1,-1 };
static UNUSED_ATTR const U32 ML_bits[MaxML+1] = {
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
1, 1, 1, 1, 2, 2, 3, 3,
4, 4, 5, 7, 8, 9,10,11,
12,13,14,15,16
};
static UNUSED_ATTR const S16 ML_defaultNorm[MaxML+1] = {
1, 4, 3, 2, 2, 2, 2, 2,
2, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1,-1,-1,
-1,-1,-1,-1,-1
};
#define ML_DEFAULTNORMLOG 6 /* for static allocation */
static const U32 ML_defaultNormLog = ML_DEFAULTNORMLOG;
static UNUSED_ATTR const U32 ML_defaultNormLog = ML_DEFAULTNORMLOG;
static const S16 OF_defaultNorm[DefaultMaxOff+1] = { 1, 1, 1, 1, 1, 1, 2, 2,
2, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1,
-1,-1,-1,-1,-1 };
static UNUSED_ATTR const S16 OF_defaultNorm[DefaultMaxOff+1] = {
1, 1, 1, 1, 1, 1, 2, 2,
2, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1,
-1,-1,-1,-1,-1
};
#define OF_DEFAULTNORMLOG 5 /* for static allocation */
static const U32 OF_defaultNormLog = OF_DEFAULTNORMLOG;
static UNUSED_ATTR const U32 OF_defaultNormLog = OF_DEFAULTNORMLOG;
/*-*******************************************
* Shared functions to include for inlining
*********************************************/
static void ZSTD_copy8(void* dst, const void* src) { memcpy(dst, src, 8); }
static void ZSTD_copy8(void* dst, const void* src) {
#if !defined(ZSTD_NO_INTRINSICS) && defined(__ARM_NEON)
vst1_u8((uint8_t*)dst, vld1_u8((const uint8_t*)src));
#else
ZSTD_memcpy(dst, src, 8);
#endif
}
#define COPY8(d,s) { ZSTD_copy8(d,s); d+=8; s+=8; }
static void ZSTD_copy16(void* dst, const void* src) { memcpy(dst, src, 16); }
static void ZSTD_copy16(void* dst, const void* src) {
#if !defined(ZSTD_NO_INTRINSICS) && defined(__ARM_NEON)
vst1q_u8((uint8_t*)dst, vld1q_u8((const uint8_t*)src));
#else
ZSTD_memcpy(dst, src, 16);
#endif
}
#define COPY16(d,s) { ZSTD_copy16(d,s); d+=16; s+=16; }
#define WILDCOPY_OVERLENGTH 32
@@ -207,13 +269,13 @@ typedef enum {
} ZSTD_overlap_e;
/*! ZSTD_wildcopy() :
* Custom version of memcpy(), can over read/write up to WILDCOPY_OVERLENGTH bytes (if length==0)
* Custom version of ZSTD_memcpy(), can over read/write up to WILDCOPY_OVERLENGTH bytes (if length==0)
* @param ovtype controls the overlap detection
* - ZSTD_no_overlap: The source and destination are guaranteed to be at least WILDCOPY_VECLEN bytes apart.
* - ZSTD_overlap_src_before_dst: The src and dst may overlap, but they MUST be at least 8 bytes apart.
* The src buffer must be before the dst buffer.
*/
MEM_STATIC FORCE_INLINE_ATTR DONT_VECTORIZE
MEM_STATIC FORCE_INLINE_ATTR
void ZSTD_wildcopy(void* dst, const void* src, ptrdiff_t length, ZSTD_overlap_e const ovtype)
{
ptrdiff_t diff = (BYTE*)dst - (const BYTE*)src;
@@ -230,47 +292,110 @@ void ZSTD_wildcopy(void* dst, const void* src, ptrdiff_t length, ZSTD_overlap_e
} while (op < oend);
} else {
assert(diff >= WILDCOPY_VECLEN || diff <= -WILDCOPY_VECLEN);
/* Separate out the first two COPY16() calls because the copy length is
/* Separate out the first COPY16() call because the copy length is
* almost certain to be short, so the branches have different
* probabilities.
* On gcc-9 unrolling once is +1.6%, twice is +2%, thrice is +1.8%.
* On clang-8 unrolling once is +1.4%, twice is +3.3%, thrice is +3%.
* probabilities. Since it is almost certain to be short, only do
* one COPY16() in the first call. Then, do two calls per loop since
* at that point it is more likely to have a high trip count.
*/
COPY16(op, ip);
COPY16(op, ip);
if (op >= oend) return;
#ifdef __aarch64__
do {
COPY16(op, ip);
}
while (op < oend);
#else
ZSTD_copy16(op, ip);
if (16 >= length) return;
op += 16;
ip += 16;
do {
COPY16(op, ip);
COPY16(op, ip);
}
while (op < oend);
#endif
}
}
MEM_STATIC size_t ZSTD_limitCopy(void* dst, size_t dstCapacity, const void* src, size_t srcSize)
{
size_t const length = MIN(dstCapacity, srcSize);
if (length > 0) {
ZSTD_memcpy(dst, src, length);
}
return length;
}
/* define "workspace is too large" as this number of times larger than needed */
#define ZSTD_WORKSPACETOOLARGE_FACTOR 3
/* when workspace is continuously too large
* during at least this number of times,
* context's memory usage is considered wasteful,
* because it's sized to handle a worst case scenario which rarely happens.
* In which case, resize it down to free some memory */
#define ZSTD_WORKSPACETOOLARGE_MAXDURATION 128
/* Controls whether the input/output buffer is buffered or stable. */
typedef enum {
ZSTD_bm_buffered = 0, /* Buffer the input/output */
ZSTD_bm_stable = 1 /* ZSTD_inBuffer/ZSTD_outBuffer is stable */
} ZSTD_bufferMode_e;
/*-*******************************************
* Private declarations
*********************************************/
typedef struct seqDef_s {
U32 offset;
U32 offset; /* Offset code of the sequence */
U16 litLength;
U16 matchLength;
} seqDef;
typedef struct {
seqDef* sequencesStart;
seqDef* sequences;
seqDef* sequences; /* ptr to end of sequences */
BYTE* litStart;
BYTE* lit;
BYTE* lit; /* ptr to end of literals */
BYTE* llCode;
BYTE* mlCode;
BYTE* ofCode;
size_t maxNbSeq;
size_t maxNbLit;
U32 longLengthID; /* 0 == no longLength; 1 == Lit.longLength; 2 == Match.longLength; */
U32 longLengthPos;
/* longLengthPos and longLengthID to allow us to represent either a single litLength or matchLength
* in the seqStore that has a value larger than U16 (if it exists). To do so, we increment
* the existing value of the litLength or matchLength by 0x10000.
*/
U32 longLengthID; /* 0 == no longLength; 1 == Represent the long literal; 2 == Represent the long match; */
U32 longLengthPos; /* Index of the sequence to apply long length modification to */
} seqStore_t;
typedef struct {
U32 litLength;
U32 matchLength;
} ZSTD_sequenceLength;
/**
* Returns the ZSTD_sequenceLength for the given sequences. It handles the decoding of long sequences
* indicated by longLengthPos and longLengthID, and adds MINMATCH back to matchLength.
*/
MEM_STATIC ZSTD_sequenceLength ZSTD_getSequenceLength(seqStore_t const* seqStore, seqDef const* seq)
{
ZSTD_sequenceLength seqLen;
seqLen.litLength = seq->litLength;
seqLen.matchLength = seq->matchLength + MINMATCH;
if (seqStore->longLengthPos == (U32)(seq - seqStore->sequencesStart)) {
if (seqStore->longLengthID == 1) {
seqLen.litLength += 0xFFFF;
}
if (seqStore->longLengthID == 2) {
seqLen.matchLength += 0xFFFF;
}
}
return seqLen;
}
/**
* Contains the compressed frame size and an upper-bound for the decompressed frame size.
* Note: before using `compressedSize`, check for errors using ZSTD_isError().
@@ -286,9 +411,9 @@ const seqStore_t* ZSTD_getSeqStore(const ZSTD_CCtx* ctx); /* compress & dictBu
void ZSTD_seqToCodes(const seqStore_t* seqStorePtr); /* compress, dictBuilder, decodeCorpus (shouldn't get its definition from here) */
/* custom memory allocation functions */
void* ZSTD_malloc(size_t size, ZSTD_customMem customMem);
void* ZSTD_calloc(size_t size, ZSTD_customMem customMem);
void ZSTD_free(void* ptr, ZSTD_customMem customMem);
void* ZSTD_customMalloc(size_t size, ZSTD_customMem customMem);
void* ZSTD_customCalloc(size_t size, ZSTD_customMem customMem);
void ZSTD_customFree(void* ptr, ZSTD_customMem customMem);
MEM_STATIC U32 ZSTD_highbit32(U32 val) /* compress, dictBuilder, decodeCorpus */
@@ -296,9 +421,12 @@ MEM_STATIC U32 ZSTD_highbit32(U32 val) /* compress, dictBuilder, decodeCorpus
assert(val != 0);
{
# if defined(_MSC_VER) /* Visual */
unsigned long r=0;
_BitScanReverse(&r, val);
return (unsigned)r;
# if STATIC_BMI2 == 1
return _lzcnt_u32(val)^31;
# else
unsigned long r=0;
return _BitScanReverse(&r, val) ? (unsigned)r : 0;
# endif
# elif defined(__GNUC__) && (__GNUC__ >= 3) /* GCC Intrinsic */
return __builtin_clz (val) ^ 31;
# elif defined(__ICCARM__) /* IAR Intrinsic */

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -58,11 +58,11 @@ ZSTD_updateDUBT(ZSTD_matchState_t* ms,
/** ZSTD_insertDUBT1() :
* sort one already inserted but unsorted position
* assumption : current >= btlow == (current - btmask)
* assumption : curr >= btlow == (curr - btmask)
* doesn't fail */
static void
ZSTD_insertDUBT1(ZSTD_matchState_t* ms,
U32 current, const BYTE* inputEnd,
U32 curr, const BYTE* inputEnd,
U32 nbCompares, U32 btLow,
const ZSTD_dictMode_e dictMode)
{
@@ -74,41 +74,41 @@ ZSTD_insertDUBT1(ZSTD_matchState_t* ms,
const BYTE* const base = ms->window.base;
const BYTE* const dictBase = ms->window.dictBase;
const U32 dictLimit = ms->window.dictLimit;
const BYTE* const ip = (current>=dictLimit) ? base + current : dictBase + current;
const BYTE* const iend = (current>=dictLimit) ? inputEnd : dictBase + dictLimit;
const BYTE* const ip = (curr>=dictLimit) ? base + curr : dictBase + curr;
const BYTE* const iend = (curr>=dictLimit) ? inputEnd : dictBase + dictLimit;
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const prefixStart = base + dictLimit;
const BYTE* match;
U32* smallerPtr = bt + 2*(current&btMask);
U32* smallerPtr = bt + 2*(curr&btMask);
U32* largerPtr = smallerPtr + 1;
U32 matchIndex = *smallerPtr; /* this candidate is unsorted : next sorted candidate is reached through *smallerPtr, while *largerPtr contains previous unsorted candidate (which is already saved and can be overwritten) */
U32 dummy32; /* to be nullified at the end */
U32 const windowValid = ms->window.lowLimit;
U32 const maxDistance = 1U << cParams->windowLog;
U32 const windowLow = (current - windowValid > maxDistance) ? current - maxDistance : windowValid;
U32 const windowLow = (curr - windowValid > maxDistance) ? curr - maxDistance : windowValid;
DEBUGLOG(8, "ZSTD_insertDUBT1(%u) (dictLimit=%u, lowLimit=%u)",
current, dictLimit, windowLow);
assert(current >= btLow);
curr, dictLimit, windowLow);
assert(curr >= btLow);
assert(ip < iend); /* condition for ZSTD_count */
while (nbCompares-- && (matchIndex > windowLow)) {
U32* const nextPtr = bt + 2*(matchIndex & btMask);
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
assert(matchIndex < current);
assert(matchIndex < curr);
/* note : all candidates are now supposed sorted,
* but it's still possible to have nextPtr[1] == ZSTD_DUBT_UNSORTED_MARK
* when a real index has the same value as ZSTD_DUBT_UNSORTED_MARK */
if ( (dictMode != ZSTD_extDict)
|| (matchIndex+matchLength >= dictLimit) /* both in current segment*/
|| (current < dictLimit) /* both in extDict */) {
|| (curr < dictLimit) /* both in extDict */) {
const BYTE* const mBase = ( (dictMode != ZSTD_extDict)
|| (matchIndex+matchLength >= dictLimit)) ?
base : dictBase;
assert( (matchIndex+matchLength >= dictLimit) /* might be wrong if extDict is incorrectly set to 0 */
|| (current < dictLimit) );
|| (curr < dictLimit) );
match = mBase + matchIndex;
matchLength += ZSTD_count(ip+matchLength, match+matchLength, iend);
} else {
@@ -119,7 +119,7 @@ ZSTD_insertDUBT1(ZSTD_matchState_t* ms,
}
DEBUGLOG(8, "ZSTD_insertDUBT1: comparing %u with %u : found %u common bytes ",
current, matchIndex, (U32)matchLength);
curr, matchIndex, (U32)matchLength);
if (ip+matchLength == iend) { /* equal : no way to know if inf or sup */
break; /* drop , to guarantee consistency ; miss a bit of compression, but other solutions can corrupt tree */
@@ -168,7 +168,7 @@ ZSTD_DUBT_findBetterDictMatch (
const BYTE* const base = ms->window.base;
const BYTE* const prefixStart = base + ms->window.dictLimit;
U32 const current = (U32)(ip-base);
U32 const curr = (U32)(ip-base);
const BYTE* const dictBase = dms->window.base;
const BYTE* const dictEnd = dms->window.nextSrc;
U32 const dictHighLimit = (U32)(dms->window.nextSrc - dms->window.base);
@@ -195,10 +195,10 @@ ZSTD_DUBT_findBetterDictMatch (
if (matchLength > bestLength) {
U32 matchIndex = dictMatchIndex + dictIndexDelta;
if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(current-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) ) {
if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(curr-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) ) {
DEBUGLOG(9, "ZSTD_DUBT_findBetterDictMatch(%u) : found better match length %u -> %u and offsetCode %u -> %u (dictMatchIndex %u, matchIndex %u)",
current, (U32)bestLength, (U32)matchLength, (U32)*offsetPtr, ZSTD_REP_MOVE + current - matchIndex, dictMatchIndex, matchIndex);
bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + current - matchIndex;
curr, (U32)bestLength, (U32)matchLength, (U32)*offsetPtr, ZSTD_REP_MOVE + curr - matchIndex, dictMatchIndex, matchIndex);
bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + curr - matchIndex;
}
if (ip+matchLength == iend) { /* reached end of input : ip[matchLength] is not valid, no way to know if it's larger or smaller than match */
break; /* drop, to guarantee consistency (miss a little bit of compression) */
@@ -218,9 +218,9 @@ ZSTD_DUBT_findBetterDictMatch (
}
if (bestLength >= MINMATCH) {
U32 const mIndex = current - ((U32)*offsetPtr - ZSTD_REP_MOVE); (void)mIndex;
U32 const mIndex = curr - ((U32)*offsetPtr - ZSTD_REP_MOVE); (void)mIndex;
DEBUGLOG(8, "ZSTD_DUBT_findBetterDictMatch(%u) : found match of length %u and offsetCode %u (pos %u)",
current, (U32)bestLength, (U32)*offsetPtr, mIndex);
curr, (U32)bestLength, (U32)*offsetPtr, mIndex);
}
return bestLength;
@@ -241,13 +241,13 @@ ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
U32 matchIndex = hashTable[h];
const BYTE* const base = ms->window.base;
U32 const current = (U32)(ip-base);
U32 const windowLow = ZSTD_getLowestMatchIndex(ms, current, cParams->windowLog);
U32 const curr = (U32)(ip-base);
U32 const windowLow = ZSTD_getLowestMatchIndex(ms, curr, cParams->windowLog);
U32* const bt = ms->chainTable;
U32 const btLog = cParams->chainLog - 1;
U32 const btMask = (1 << btLog) - 1;
U32 const btLow = (btMask >= current) ? 0 : current - btMask;
U32 const btLow = (btMask >= curr) ? 0 : curr - btMask;
U32 const unsortLimit = MAX(btLow, windowLow);
U32* nextCandidate = bt + 2*(matchIndex&btMask);
@@ -256,8 +256,9 @@ ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
U32 nbCandidates = nbCompares;
U32 previousCandidate = 0;
DEBUGLOG(7, "ZSTD_DUBT_findBestMatch (%u) ", current);
DEBUGLOG(7, "ZSTD_DUBT_findBestMatch (%u) ", curr);
assert(ip <= iend-8); /* required for h calculation */
assert(dictMode != ZSTD_dedicatedDictSearch);
/* reach end of unsorted candidates list */
while ( (matchIndex > unsortLimit)
@@ -299,14 +300,14 @@ ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
const U32 dictLimit = ms->window.dictLimit;
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const prefixStart = base + dictLimit;
U32* smallerPtr = bt + 2*(current&btMask);
U32* largerPtr = bt + 2*(current&btMask) + 1;
U32 matchEndIdx = current + 8 + 1;
U32* smallerPtr = bt + 2*(curr&btMask);
U32* largerPtr = bt + 2*(curr&btMask) + 1;
U32 matchEndIdx = curr + 8 + 1;
U32 dummy32; /* to be nullified at the end */
size_t bestLength = 0;
matchIndex = hashTable[h];
hashTable[h] = current; /* Update Hash Table */
hashTable[h] = curr; /* Update Hash Table */
while (nbCompares-- && (matchIndex > windowLow)) {
U32* const nextPtr = bt + 2*(matchIndex & btMask);
@@ -326,8 +327,8 @@ ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
if (matchLength > bestLength) {
if (matchLength > matchEndIdx - matchIndex)
matchEndIdx = matchIndex + (U32)matchLength;
if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(current-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) )
bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + current - matchIndex;
if ( (4*(int)(matchLength-bestLength)) > (int)(ZSTD_highbit32(curr-matchIndex+1) - ZSTD_highbit32((U32)offsetPtr[0]+1)) )
bestLength = matchLength, *offsetPtr = ZSTD_REP_MOVE + curr - matchIndex;
if (ip+matchLength == iend) { /* equal : no way to know if inf or sup */
if (dictMode == ZSTD_dictMatchState) {
nbCompares = 0; /* in addition to avoiding checking any
@@ -363,12 +364,12 @@ ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
mls, dictMode);
}
assert(matchEndIdx > current+8); /* ensure nextToUpdate is increased */
assert(matchEndIdx > curr+8); /* ensure nextToUpdate is increased */
ms->nextToUpdate = matchEndIdx - 8; /* skip repetitive patterns */
if (bestLength >= MINMATCH) {
U32 const mIndex = current - ((U32)*offsetPtr - ZSTD_REP_MOVE); (void)mIndex;
U32 const mIndex = curr - ((U32)*offsetPtr - ZSTD_REP_MOVE); (void)mIndex;
DEBUGLOG(8, "ZSTD_DUBT_findBestMatch(%u) : found match of length %u and offsetCode %u (pos %u)",
current, (U32)bestLength, (U32)*offsetPtr, mIndex);
curr, (U32)bestLength, (U32)*offsetPtr, mIndex);
}
return bestLength;
}
@@ -446,7 +447,7 @@ static size_t ZSTD_BtFindBestMatch_extDict_selectMLS (
/* Update chains up to ip (excluded)
Assumption : always within prefix (i.e. not within extDict) */
static U32 ZSTD_insertAndFindFirstIndex_internal(
FORCE_INLINE_TEMPLATE U32 ZSTD_insertAndFindFirstIndex_internal(
ZSTD_matchState_t* ms,
const ZSTD_compressionParameters* const cParams,
const BYTE* ip, U32 const mls)
@@ -475,6 +476,121 @@ U32 ZSTD_insertAndFindFirstIndex(ZSTD_matchState_t* ms, const BYTE* ip) {
return ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, ms->cParams.minMatch);
}
void ZSTD_dedicatedDictSearch_lazy_loadDictionary(ZSTD_matchState_t* ms, const BYTE* const ip)
{
const BYTE* const base = ms->window.base;
U32 const target = (U32)(ip - base);
U32* const hashTable = ms->hashTable;
U32* const chainTable = ms->chainTable;
U32 const chainSize = 1 << ms->cParams.chainLog;
U32 idx = ms->nextToUpdate;
U32 const minChain = chainSize < target ? target - chainSize : idx;
U32 const bucketSize = 1 << ZSTD_LAZY_DDSS_BUCKET_LOG;
U32 const cacheSize = bucketSize - 1;
U32 const chainAttempts = (1 << ms->cParams.searchLog) - cacheSize;
U32 const chainLimit = chainAttempts > 255 ? 255 : chainAttempts;
/* We know the hashtable is oversized by a factor of `bucketSize`.
* We are going to temporarily pretend `bucketSize == 1`, keeping only a
* single entry. We will use the rest of the space to construct a temporary
* chaintable.
*/
U32 const hashLog = ms->cParams.hashLog - ZSTD_LAZY_DDSS_BUCKET_LOG;
U32* const tmpHashTable = hashTable;
U32* const tmpChainTable = hashTable + ((size_t)1 << hashLog);
U32 const tmpChainSize = ((1 << ZSTD_LAZY_DDSS_BUCKET_LOG) - 1) << hashLog;
U32 const tmpMinChain = tmpChainSize < target ? target - tmpChainSize : idx;
U32 hashIdx;
assert(ms->cParams.chainLog <= 24);
assert(ms->cParams.hashLog >= ms->cParams.chainLog);
assert(idx != 0);
assert(tmpMinChain <= minChain);
/* fill conventional hash table and conventional chain table */
for ( ; idx < target; idx++) {
U32 const h = (U32)ZSTD_hashPtr(base + idx, hashLog, ms->cParams.minMatch);
if (idx >= tmpMinChain) {
tmpChainTable[idx - tmpMinChain] = hashTable[h];
}
tmpHashTable[h] = idx;
}
/* sort chains into ddss chain table */
{
U32 chainPos = 0;
for (hashIdx = 0; hashIdx < (1U << hashLog); hashIdx++) {
U32 count;
U32 countBeyondMinChain = 0;
U32 i = tmpHashTable[hashIdx];
for (count = 0; i >= tmpMinChain && count < cacheSize; count++) {
/* skip through the chain to the first position that won't be
* in the hash cache bucket */
if (i < minChain) {
countBeyondMinChain++;
}
i = tmpChainTable[i - tmpMinChain];
}
if (count == cacheSize) {
for (count = 0; count < chainLimit;) {
if (i < minChain) {
if (!i || countBeyondMinChain++ > cacheSize) {
/* only allow pulling `cacheSize` number of entries
* into the cache or chainTable beyond `minChain`,
* to replace the entries pulled out of the
* chainTable into the cache. This lets us reach
* back further without increasing the total number
* of entries in the chainTable, guaranteeing the
* DDSS chain table will fit into the space
* allocated for the regular one. */
break;
}
}
chainTable[chainPos++] = i;
count++;
if (i < tmpMinChain) {
break;
}
i = tmpChainTable[i - tmpMinChain];
}
} else {
count = 0;
}
if (count) {
tmpHashTable[hashIdx] = ((chainPos - count) << 8) + count;
} else {
tmpHashTable[hashIdx] = 0;
}
}
assert(chainPos <= chainSize); /* I believe this is guaranteed... */
}
/* move chain pointers into the last entry of each hash bucket */
for (hashIdx = (1 << hashLog); hashIdx; ) {
U32 const bucketIdx = --hashIdx << ZSTD_LAZY_DDSS_BUCKET_LOG;
U32 const chainPackedPointer = tmpHashTable[hashIdx];
U32 i;
for (i = 0; i < cacheSize; i++) {
hashTable[bucketIdx + i] = 0;
}
hashTable[bucketIdx + bucketSize - 1] = chainPackedPointer;
}
/* fill the buckets of the hash table */
for (idx = ms->nextToUpdate; idx < target; idx++) {
U32 const h = (U32)ZSTD_hashPtr(base + idx, hashLog, ms->cParams.minMatch)
<< ZSTD_LAZY_DDSS_BUCKET_LOG;
U32 i;
/* Shift hash cache down 1. */
for (i = cacheSize - 1; i; i--)
hashTable[h + i] = hashTable[h + i - 1];
hashTable[h] = idx;
}
ms->nextToUpdate = target;
}
/* inlining is important to hardwire a hot branch (template emulation) */
FORCE_INLINE_TEMPLATE
@@ -493,20 +609,33 @@ size_t ZSTD_HcFindBestMatch_generic (
const U32 dictLimit = ms->window.dictLimit;
const BYTE* const prefixStart = base + dictLimit;
const BYTE* const dictEnd = dictBase + dictLimit;
const U32 current = (U32)(ip-base);
const U32 curr = (U32)(ip-base);
const U32 maxDistance = 1U << cParams->windowLog;
const U32 lowestValid = ms->window.lowLimit;
const U32 withinMaxDistance = (current - lowestValid > maxDistance) ? current - maxDistance : lowestValid;
const U32 withinMaxDistance = (curr - lowestValid > maxDistance) ? curr - maxDistance : lowestValid;
const U32 isDictionary = (ms->loadedDictEnd != 0);
const U32 lowLimit = isDictionary ? lowestValid : withinMaxDistance;
const U32 minChain = current > chainSize ? current - chainSize : 0;
const U32 minChain = curr > chainSize ? curr - chainSize : 0;
U32 nbAttempts = 1U << cParams->searchLog;
size_t ml=4-1;
/* HC4 match finder */
U32 matchIndex = ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, mls);
const ZSTD_matchState_t* const dms = ms->dictMatchState;
const U32 ddsHashLog = dictMode == ZSTD_dedicatedDictSearch
? dms->cParams.hashLog - ZSTD_LAZY_DDSS_BUCKET_LOG : 0;
const size_t ddsIdx = dictMode == ZSTD_dedicatedDictSearch
? ZSTD_hashPtr(ip, ddsHashLog, mls) << ZSTD_LAZY_DDSS_BUCKET_LOG : 0;
for ( ; (matchIndex>lowLimit) & (nbAttempts>0) ; nbAttempts--) {
U32 matchIndex;
if (dictMode == ZSTD_dedicatedDictSearch) {
const U32* entry = &dms->hashTable[ddsIdx];
PREFETCH_L1(entry);
}
/* HC4 match finder */
matchIndex = ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, mls);
for ( ; (matchIndex>=lowLimit) & (nbAttempts>0) ; nbAttempts--) {
size_t currentMl=0;
if ((dictMode != ZSTD_extDict) || matchIndex >= dictLimit) {
const BYTE* const match = base + matchIndex;
@@ -523,7 +652,7 @@ size_t ZSTD_HcFindBestMatch_generic (
/* save best solution */
if (currentMl > ml) {
ml = currentMl;
*offsetPtr = current - matchIndex + ZSTD_REP_MOVE;
*offsetPtr = curr - matchIndex + ZSTD_REP_MOVE;
if (ip+currentMl == iLimit) break; /* best possible, avoids read overflow on next attempt */
}
@@ -531,8 +660,92 @@ size_t ZSTD_HcFindBestMatch_generic (
matchIndex = NEXT_IN_CHAIN(matchIndex, chainMask);
}
if (dictMode == ZSTD_dictMatchState) {
const ZSTD_matchState_t* const dms = ms->dictMatchState;
if (dictMode == ZSTD_dedicatedDictSearch) {
const U32 ddsLowestIndex = dms->window.dictLimit;
const BYTE* const ddsBase = dms->window.base;
const BYTE* const ddsEnd = dms->window.nextSrc;
const U32 ddsSize = (U32)(ddsEnd - ddsBase);
const U32 ddsIndexDelta = dictLimit - ddsSize;
const U32 bucketSize = (1 << ZSTD_LAZY_DDSS_BUCKET_LOG);
const U32 bucketLimit = nbAttempts < bucketSize - 1 ? nbAttempts : bucketSize - 1;
U32 ddsAttempt;
for (ddsAttempt = 0; ddsAttempt < bucketSize - 1; ddsAttempt++) {
PREFETCH_L1(ddsBase + dms->hashTable[ddsIdx + ddsAttempt]);
}
{
U32 const chainPackedPointer = dms->hashTable[ddsIdx + bucketSize - 1];
U32 const chainIndex = chainPackedPointer >> 8;
PREFETCH_L1(&dms->chainTable[chainIndex]);
}
for (ddsAttempt = 0; ddsAttempt < bucketLimit; ddsAttempt++) {
size_t currentMl=0;
const BYTE* match;
matchIndex = dms->hashTable[ddsIdx + ddsAttempt];
match = ddsBase + matchIndex;
if (!matchIndex) {
return ml;
}
/* guaranteed by table construction */
(void)ddsLowestIndex;
assert(matchIndex >= ddsLowestIndex);
assert(match+4 <= ddsEnd);
if (MEM_read32(match) == MEM_read32(ip)) {
/* assumption : matchIndex <= dictLimit-4 (by table construction) */
currentMl = ZSTD_count_2segments(ip+4, match+4, iLimit, ddsEnd, prefixStart) + 4;
}
/* save best solution */
if (currentMl > ml) {
ml = currentMl;
*offsetPtr = curr - (matchIndex + ddsIndexDelta) + ZSTD_REP_MOVE;
if (ip+currentMl == iLimit) {
/* best possible, avoids read overflow on next attempt */
return ml;
}
}
}
{
U32 const chainPackedPointer = dms->hashTable[ddsIdx + bucketSize - 1];
U32 chainIndex = chainPackedPointer >> 8;
U32 const chainLength = chainPackedPointer & 0xFF;
U32 const chainAttempts = nbAttempts - ddsAttempt;
U32 const chainLimit = chainAttempts > chainLength ? chainLength : chainAttempts;
U32 chainAttempt;
for (chainAttempt = 0 ; chainAttempt < chainLimit; chainAttempt++) {
PREFETCH_L1(ddsBase + dms->chainTable[chainIndex + chainAttempt]);
}
for (chainAttempt = 0 ; chainAttempt < chainLimit; chainAttempt++, chainIndex++) {
size_t currentMl=0;
const BYTE* match;
matchIndex = dms->chainTable[chainIndex];
match = ddsBase + matchIndex;
/* guaranteed by table construction */
assert(matchIndex >= ddsLowestIndex);
assert(match+4 <= ddsEnd);
if (MEM_read32(match) == MEM_read32(ip)) {
/* assumption : matchIndex <= dictLimit-4 (by table construction) */
currentMl = ZSTD_count_2segments(ip+4, match+4, iLimit, ddsEnd, prefixStart) + 4;
}
/* save best solution */
if (currentMl > ml) {
ml = currentMl;
*offsetPtr = curr - (matchIndex + ddsIndexDelta) + ZSTD_REP_MOVE;
if (ip+currentMl == iLimit) break; /* best possible, avoids read overflow on next attempt */
}
}
}
} else if (dictMode == ZSTD_dictMatchState) {
const U32* const dmsChainTable = dms->chainTable;
const U32 dmsChainSize = (1 << dms->cParams.chainLog);
const U32 dmsChainMask = dmsChainSize - 1;
@@ -545,7 +758,7 @@ size_t ZSTD_HcFindBestMatch_generic (
matchIndex = dms->hashTable[ZSTD_hashPtr(ip, dms->cParams.hashLog, mls)];
for ( ; (matchIndex>dmsLowestIndex) & (nbAttempts>0) ; nbAttempts--) {
for ( ; (matchIndex>=dmsLowestIndex) & (nbAttempts>0) ; nbAttempts--) {
size_t currentMl=0;
const BYTE* const match = dmsBase + matchIndex;
assert(match+4 <= dmsEnd);
@@ -555,11 +768,12 @@ size_t ZSTD_HcFindBestMatch_generic (
/* save best solution */
if (currentMl > ml) {
ml = currentMl;
*offsetPtr = current - (matchIndex + dmsIndexDelta) + ZSTD_REP_MOVE;
*offsetPtr = curr - (matchIndex + dmsIndexDelta) + ZSTD_REP_MOVE;
if (ip+currentMl == iLimit) break; /* best possible, avoids read overflow on next attempt */
}
if (matchIndex <= dmsMinChain) break;
matchIndex = dmsChainTable[matchIndex & dmsChainMask];
}
}
@@ -600,6 +814,22 @@ static size_t ZSTD_HcFindBestMatch_dictMatchState_selectMLS (
}
static size_t ZSTD_HcFindBestMatch_dedicatedDictSearch_selectMLS (
ZSTD_matchState_t* ms,
const BYTE* ip, const BYTE* const iLimit,
size_t* offsetPtr)
{
switch(ms->cParams.minMatch)
{
default : /* includes case 3 */
case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_dedicatedDictSearch);
case 5 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 5, ZSTD_dedicatedDictSearch);
case 7 :
case 6 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 6, ZSTD_dedicatedDictSearch);
}
}
FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_extDict_selectMLS (
ZSTD_matchState_t* ms,
const BYTE* ip, const BYTE* const iLimit,
@@ -641,35 +871,62 @@ ZSTD_compressBlock_lazy_generic(
typedef size_t (*searchMax_f)(
ZSTD_matchState_t* ms,
const BYTE* ip, const BYTE* iLimit, size_t* offsetPtr);
searchMax_f const searchMax = dictMode == ZSTD_dictMatchState ?
(searchMethod==search_binaryTree ? ZSTD_BtFindBestMatch_dictMatchState_selectMLS
: ZSTD_HcFindBestMatch_dictMatchState_selectMLS) :
(searchMethod==search_binaryTree ? ZSTD_BtFindBestMatch_selectMLS
: ZSTD_HcFindBestMatch_selectMLS);
/**
* This table is indexed first by the four ZSTD_dictMode_e values, and then
* by the two searchMethod_e values. NULLs are placed for configurations
* that should never occur (extDict modes go to the other implementation
* below and there is no DDSS for binary tree search yet).
*/
const searchMax_f searchFuncs[4][2] = {
{
ZSTD_HcFindBestMatch_selectMLS,
ZSTD_BtFindBestMatch_selectMLS
},
{
NULL,
NULL
},
{
ZSTD_HcFindBestMatch_dictMatchState_selectMLS,
ZSTD_BtFindBestMatch_dictMatchState_selectMLS
},
{
ZSTD_HcFindBestMatch_dedicatedDictSearch_selectMLS,
NULL
}
};
searchMax_f const searchMax = searchFuncs[dictMode][searchMethod == search_binaryTree];
U32 offset_1 = rep[0], offset_2 = rep[1], savedOffset=0;
const int isDMS = dictMode == ZSTD_dictMatchState;
const int isDDS = dictMode == ZSTD_dedicatedDictSearch;
const int isDxS = isDMS || isDDS;
const ZSTD_matchState_t* const dms = ms->dictMatchState;
const U32 dictLowestIndex = dictMode == ZSTD_dictMatchState ?
dms->window.dictLimit : 0;
const BYTE* const dictBase = dictMode == ZSTD_dictMatchState ?
dms->window.base : NULL;
const BYTE* const dictLowest = dictMode == ZSTD_dictMatchState ?
dictBase + dictLowestIndex : NULL;
const BYTE* const dictEnd = dictMode == ZSTD_dictMatchState ?
dms->window.nextSrc : NULL;
const U32 dictIndexDelta = dictMode == ZSTD_dictMatchState ?
const U32 dictLowestIndex = isDxS ? dms->window.dictLimit : 0;
const BYTE* const dictBase = isDxS ? dms->window.base : NULL;
const BYTE* const dictLowest = isDxS ? dictBase + dictLowestIndex : NULL;
const BYTE* const dictEnd = isDxS ? dms->window.nextSrc : NULL;
const U32 dictIndexDelta = isDxS ?
prefixLowestIndex - (U32)(dictEnd - dictBase) :
0;
const U32 dictAndPrefixLength = (U32)(ip - prefixLowest + dictEnd - dictLowest);
const U32 dictAndPrefixLength = (U32)((ip - prefixLowest) + (dictEnd - dictLowest));
assert(searchMax != NULL);
DEBUGLOG(5, "ZSTD_compressBlock_lazy_generic (dictMode=%u)", (U32)dictMode);
/* init */
ip += (dictAndPrefixLength == 0);
if (dictMode == ZSTD_noDict) {
U32 const maxRep = (U32)(ip - prefixLowest);
U32 const curr = (U32)(ip - base);
U32 const windowLow = ZSTD_getLowestPrefixIndex(ms, curr, ms->cParams.windowLog);
U32 const maxRep = curr - windowLow;
if (offset_2 > maxRep) savedOffset = offset_2, offset_2 = 0;
if (offset_1 > maxRep) savedOffset = offset_1, offset_1 = 0;
}
if (dictMode == ZSTD_dictMatchState) {
if (isDxS) {
/* dictMatchState repCode checks don't currently handle repCode == 0
* disabling. */
assert(offset_1 <= dictAndPrefixLength);
@@ -677,15 +934,21 @@ ZSTD_compressBlock_lazy_generic(
}
/* Match Loop */
#if defined(__GNUC__) && defined(__x86_64__)
/* I've measured random a 5% speed loss on levels 5 & 6 (greedy) when the
* code alignment is perturbed. To fix the instability align the loop on 32-bytes.
*/
__asm__(".p2align 5");
#endif
while (ip < ilimit) {
size_t matchLength=0;
size_t offset=0;
const BYTE* start=ip+1;
/* check repCode */
if (dictMode == ZSTD_dictMatchState) {
if (isDxS) {
const U32 repIndex = (U32)(ip - base) + 1 - offset_1;
const BYTE* repMatch = (dictMode == ZSTD_dictMatchState
const BYTE* repMatch = ((dictMode == ZSTD_dictMatchState || dictMode == ZSTD_dedicatedDictSearch)
&& repIndex < prefixLowestIndex) ?
dictBase + (repIndex - dictIndexDelta) :
base + repIndex;
@@ -726,7 +989,7 @@ ZSTD_compressBlock_lazy_generic(
if ((mlRep >= 4) && (gain2 > gain1))
matchLength = mlRep, offset = 0, start = ip;
}
if (dictMode == ZSTD_dictMatchState) {
if (isDxS) {
const U32 repIndex = (U32)(ip - base) - offset_1;
const BYTE* repMatch = repIndex < prefixLowestIndex ?
dictBase + (repIndex - dictIndexDelta) :
@@ -761,7 +1024,7 @@ ZSTD_compressBlock_lazy_generic(
if ((mlRep >= 4) && (gain2 > gain1))
matchLength = mlRep, offset = 0, start = ip;
}
if (dictMode == ZSTD_dictMatchState) {
if (isDxS) {
const U32 repIndex = (U32)(ip - base) - offset_1;
const BYTE* repMatch = repIndex < prefixLowestIndex ?
dictBase + (repIndex - dictIndexDelta) :
@@ -799,7 +1062,7 @@ ZSTD_compressBlock_lazy_generic(
&& (start[-1] == (start-(offset-ZSTD_REP_MOVE))[-1]) ) /* only search for offset within prefix */
{ start--; matchLength++; }
}
if (dictMode == ZSTD_dictMatchState) {
if (isDxS) {
U32 const matchIndex = (U32)((start-base) - (offset - ZSTD_REP_MOVE));
const BYTE* match = (matchIndex < prefixLowestIndex) ? dictBase + matchIndex - dictIndexDelta : base + matchIndex;
const BYTE* const mStart = (matchIndex < prefixLowestIndex) ? dictLowest : prefixLowest;
@@ -815,12 +1078,11 @@ _storeSequence:
}
/* check immediate repcode */
if (dictMode == ZSTD_dictMatchState) {
if (isDxS) {
while (ip <= ilimit) {
U32 const current2 = (U32)(ip-base);
U32 const repIndex = current2 - offset_2;
const BYTE* repMatch = dictMode == ZSTD_dictMatchState
&& repIndex < prefixLowestIndex ?
const BYTE* repMatch = repIndex < prefixLowestIndex ?
dictBase - dictIndexDelta + repIndex :
base + repIndex;
if ( ((U32)((prefixLowestIndex-1) - (U32)repIndex) >= 3 /* intentional overflow */)
@@ -915,6 +1177,28 @@ size_t ZSTD_compressBlock_greedy_dictMatchState(
}
size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_dedicatedDictSearch);
}
size_t ZSTD_compressBlock_lazy_dedicatedDictSearch(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_dedicatedDictSearch);
}
size_t ZSTD_compressBlock_greedy_dedicatedDictSearch(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0, ZSTD_dedicatedDictSearch);
}
FORCE_INLINE_TEMPLATE
size_t ZSTD_compressBlock_lazy_extDict_generic(
ZSTD_matchState_t* ms, seqStore_t* seqStore,
@@ -929,11 +1213,11 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
const BYTE* const ilimit = iend - 8;
const BYTE* const base = ms->window.base;
const U32 dictLimit = ms->window.dictLimit;
const U32 lowestIndex = ms->window.lowLimit;
const BYTE* const prefixStart = base + dictLimit;
const BYTE* const dictBase = ms->window.dictBase;
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const dictStart = dictBase + lowestIndex;
const BYTE* const dictStart = dictBase + ms->window.lowLimit;
const U32 windowLog = ms->cParams.windowLog;
typedef size_t (*searchMax_f)(
ZSTD_matchState_t* ms,
@@ -942,21 +1226,30 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
U32 offset_1 = rep[0], offset_2 = rep[1];
DEBUGLOG(5, "ZSTD_compressBlock_lazy_extDict_generic");
/* init */
ip += (ip == prefixStart);
/* Match Loop */
#if defined(__GNUC__) && defined(__x86_64__)
/* I've measured random a 5% speed loss on levels 5 & 6 (greedy) when the
* code alignment is perturbed. To fix the instability align the loop on 32-bytes.
*/
__asm__(".p2align 5");
#endif
while (ip < ilimit) {
size_t matchLength=0;
size_t offset=0;
const BYTE* start=ip+1;
U32 current = (U32)(ip-base);
U32 curr = (U32)(ip-base);
/* check repCode */
{ const U32 repIndex = (U32)(current+1 - offset_1);
{ const U32 windowLow = ZSTD_getLowestMatchIndex(ms, curr+1, windowLog);
const U32 repIndex = (U32)(curr+1 - offset_1);
const BYTE* const repBase = repIndex < dictLimit ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */
if (MEM_read32(ip+1) == MEM_read32(repMatch)) {
/* repcode detected we should take it */
const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend;
@@ -980,13 +1273,14 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
if (depth>=1)
while (ip<ilimit) {
ip ++;
current++;
curr++;
/* check repCode */
if (offset) {
const U32 repIndex = (U32)(current - offset_1);
const U32 windowLow = ZSTD_getLowestMatchIndex(ms, curr, windowLog);
const U32 repIndex = (U32)(curr - offset_1);
const BYTE* const repBase = repIndex < dictLimit ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */
if (MEM_read32(ip) == MEM_read32(repMatch)) {
/* repcode detected */
const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend;
@@ -1010,13 +1304,14 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
/* let's find an even better one */
if ((depth==2) && (ip<ilimit)) {
ip ++;
current++;
curr++;
/* check repCode */
if (offset) {
const U32 repIndex = (U32)(current - offset_1);
const U32 windowLow = ZSTD_getLowestMatchIndex(ms, curr, windowLog);
const U32 repIndex = (U32)(curr - offset_1);
const BYTE* const repBase = repIndex < dictLimit ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */
if (MEM_read32(ip) == MEM_read32(repMatch)) {
/* repcode detected */
const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend;
@@ -1057,10 +1352,12 @@ _storeSequence:
/* check immediate repcode */
while (ip <= ilimit) {
const U32 repIndex = (U32)((ip-base) - offset_2);
const U32 repCurrent = (U32)(ip-base);
const U32 windowLow = ZSTD_getLowestMatchIndex(ms, repCurrent, windowLog);
const U32 repIndex = repCurrent - offset_2;
const BYTE* const repBase = repIndex < dictLimit ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */
if (MEM_read32(ip) == MEM_read32(repMatch)) {
/* repcode detected we should take it */
const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend;

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -17,8 +17,18 @@ extern "C" {
#include "zstd_compress_internal.h"
/**
* Dedicated Dictionary Search Structure bucket log. In the
* ZSTD_dedicatedDictSearch mode, the hashTable has
* 2 ** ZSTD_LAZY_DDSS_BUCKET_LOG entries in each bucket, rather than just
* one.
*/
#define ZSTD_LAZY_DDSS_BUCKET_LOG 2
U32 ZSTD_insertAndFindFirstIndex(ZSTD_matchState_t* ms, const BYTE* ip);
void ZSTD_dedicatedDictSearch_lazy_loadDictionary(ZSTD_matchState_t* ms, const BYTE* const ip);
void ZSTD_preserveUnsortedMark (U32* const table, U32 const size, U32 const reducerValue); /*! used in ZSTD_reduceIndex(). preemptively increase value of ZSTD_DUBT_UNSORTED_MARK */
size_t ZSTD_compressBlock_btlazy2(
@@ -47,6 +57,16 @@ size_t ZSTD_compressBlock_greedy_dictMatchState(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize);
size_t ZSTD_compressBlock_lazy2_dedicatedDictSearch(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize);
size_t ZSTD_compressBlock_lazy_dedicatedDictSearch(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize);
size_t ZSTD_compressBlock_greedy_dedicatedDictSearch(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize);
size_t ZSTD_compressBlock_greedy_extDict(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize);

View File

@@ -1,10 +1,11 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#include "zstd_ldm.h"
@@ -26,13 +27,6 @@ void ZSTD_ldm_adjustParameters(ldmParams_t* params,
DEBUGLOG(4, "ZSTD_ldm_adjustParameters");
if (!params->bucketSizeLog) params->bucketSizeLog = LDM_BUCKET_SIZE_LOG;
if (!params->minMatchLength) params->minMatchLength = LDM_MIN_MATCH_LENGTH;
if (cParams->strategy >= ZSTD_btopt) {
/* Get out of the way of the optimal parser */
U32 const minMatch = MAX(cParams->targetLength, params->minMatchLength);
assert(minMatch >= ZSTD_LDM_MINMATCH_MIN);
assert(minMatch <= ZSTD_LDM_MINMATCH_MAX);
params->minMatchLength = minMatch;
}
if (params->hashLog == 0) {
params->hashLog = MAX(ZSTD_HASHLOG_MIN, params->windowLog - LDM_HASH_RLOG);
assert(params->hashLog <= ZSTD_HASHLOG_MAX);
@@ -149,10 +143,10 @@ static void ZSTD_ldm_makeEntryAndInsertByTag(ldmState_t* ldmState,
* We count only bytes where pMatch >= pBase and pIn >= pAnchor. */
static size_t ZSTD_ldm_countBackwardsMatch(
const BYTE* pIn, const BYTE* pAnchor,
const BYTE* pMatch, const BYTE* pBase)
const BYTE* pMatch, const BYTE* pMatchBase)
{
size_t matchLength = 0;
while (pIn > pAnchor && pMatch > pBase && pIn[-1] == pMatch[-1]) {
while (pIn > pAnchor && pMatch > pMatchBase && pIn[-1] == pMatch[-1]) {
pIn--;
pMatch--;
matchLength++;
@@ -160,6 +154,27 @@ static size_t ZSTD_ldm_countBackwardsMatch(
return matchLength;
}
/** ZSTD_ldm_countBackwardsMatch_2segments() :
* Returns the number of bytes that match backwards from pMatch,
* even with the backwards match spanning 2 different segments.
*
* On reaching `pMatchBase`, start counting from mEnd */
static size_t ZSTD_ldm_countBackwardsMatch_2segments(
const BYTE* pIn, const BYTE* pAnchor,
const BYTE* pMatch, const BYTE* pMatchBase,
const BYTE* pExtDictStart, const BYTE* pExtDictEnd)
{
size_t matchLength = ZSTD_ldm_countBackwardsMatch(pIn, pAnchor, pMatch, pMatchBase);
if (pMatch - matchLength != pMatchBase || pMatchBase == pExtDictStart) {
/* If backwards match is entirely in the extDict or prefix, immediately return */
return matchLength;
}
DEBUGLOG(7, "ZSTD_ldm_countBackwardsMatch_2segments: found 2-parts backwards match (length in prefix==%zu)", matchLength);
matchLength += ZSTD_ldm_countBackwardsMatch(pIn - matchLength, pAnchor, pExtDictEnd, pExtDictStart);
DEBUGLOG(7, "final backwards match length = %zu", matchLength);
return matchLength;
}
/** ZSTD_ldm_fillFastTables() :
*
* Fills the relevant tables for the ZSTD_fast and ZSTD_dfast strategies.
@@ -223,6 +238,20 @@ static U64 ZSTD_ldm_fillLdmHashTable(ldmState_t* state,
return rollingHash;
}
void ZSTD_ldm_fillHashTable(
ldmState_t* state, const BYTE* ip,
const BYTE* iend, ldmParams_t const* params)
{
DEBUGLOG(5, "ZSTD_ldm_fillHashTable");
if ((size_t)(iend - ip) >= params->minMatchLength) {
U64 startingHash = ZSTD_rollingHash_compute(ip, params->minMatchLength);
ZSTD_ldm_fillLdmHashTable(
state, startingHash, ip, iend - params->minMatchLength, state->window.base,
params->hashLog - params->bucketSizeLog,
*params);
}
}
/** ZSTD_ldm_limitTableUpdate() :
*
@@ -231,10 +260,10 @@ static U64 ZSTD_ldm_fillLdmHashTable(ldmState_t* state,
* (after a long match, only update tables a limited amount). */
static void ZSTD_ldm_limitTableUpdate(ZSTD_matchState_t* ms, const BYTE* anchor)
{
U32 const current = (U32)(anchor - ms->window.base);
if (current > ms->nextToUpdate + 1024) {
U32 const curr = (U32)(anchor - ms->window.base);
if (curr > ms->nextToUpdate + 1024) {
ms->nextToUpdate =
current - MIN(512, current - ms->nextToUpdate - 1024);
curr - MIN(512, curr - ms->nextToUpdate - 1024);
}
}
@@ -271,7 +300,7 @@ static size_t ZSTD_ldm_generateSequences_internal(
while (ip <= ilimit) {
size_t mLength;
U32 const current = (U32)(ip - base);
U32 const curr = (U32)(ip - base);
size_t forwardMatchLength = 0, backwardMatchLength = 0;
ldmEntry_t* bestEntry = NULL;
if (ip != istart) {
@@ -321,8 +350,9 @@ static size_t ZSTD_ldm_generateSequences_internal(
continue;
}
curBackwardMatchLength =
ZSTD_ldm_countBackwardsMatch(ip, anchor, pMatch,
lowMatchPtr);
ZSTD_ldm_countBackwardsMatch_2segments(ip, anchor,
pMatch, lowMatchPtr,
dictStart, dictEnd);
curTotalMatchLength = curForwardMatchLength +
curBackwardMatchLength;
} else { /* !extDict */
@@ -350,7 +380,7 @@ static size_t ZSTD_ldm_generateSequences_internal(
/* No match found -- continue searching */
if (bestEntry == NULL) {
ZSTD_ldm_makeEntryAndInsertByTag(ldmState, rollingHash,
hBits, current,
hBits, curr,
*params);
ip++;
continue;
@@ -362,11 +392,11 @@ static size_t ZSTD_ldm_generateSequences_internal(
{
/* Store the sequence:
* ip = current - backwardMatchLength
* ip = curr - backwardMatchLength
* The match is at (bestEntry->offset - backwardMatchLength)
*/
U32 const matchIndex = bestEntry->offset;
U32 const offset = current - matchIndex;
U32 const offset = curr - matchIndex;
rawSeq* const seq = rawSeqStore->seq + rawSeqStore->size;
/* Out of sequence storage */
@@ -449,6 +479,8 @@ size_t ZSTD_ldm_generateSequences(
U32 const correction = ZSTD_window_correctOverflow(
&ldmState->window, /* cycleLog */ 0, maxDist, chunkStart);
ZSTD_ldm_reduceTable(ldmState->hashTable, ldmHSize, correction);
/* invalidate dictionaries on overflow correction */
ldmState->loadedDictEnd = 0;
}
/* 2. We enforce the maximum offset allowed.
*
@@ -457,8 +489,14 @@ size_t ZSTD_ldm_generateSequences(
* TODO: * Test the chunk size.
* * Try invalidation after the sequence generation and test the
* the offset against maxDist directly.
*
* NOTE: Because of dictionaries + sequence splitting we MUST make sure
* that any offset used is valid at the END of the sequence, since it may
* be split into two sequences. This condition holds when using
* ZSTD_window_enforceMaxDist(), but if we move to checking offsets
* against maxDist directly, we'll have to carefully handle that case.
*/
ZSTD_window_enforceMaxDist(&ldmState->window, chunkEnd, maxDist, NULL, NULL);
ZSTD_window_enforceMaxDist(&ldmState->window, chunkEnd, maxDist, &ldmState->loadedDictEnd, NULL);
/* 3. Generate the sequences for the chunk, and get newLeftoverSize. */
newLeftoverSize = ZSTD_ldm_generateSequences_internal(
ldmState, sequences, params, chunkStart, chunkSize);
@@ -539,6 +577,23 @@ static rawSeq maybeSplitSequence(rawSeqStore_t* rawSeqStore,
return sequence;
}
void ZSTD_ldm_skipRawSeqStoreBytes(rawSeqStore_t* rawSeqStore, size_t nbBytes) {
U32 currPos = (U32)(rawSeqStore->posInSequence + nbBytes);
while (currPos && rawSeqStore->pos < rawSeqStore->size) {
rawSeq currSeq = rawSeqStore->seq[rawSeqStore->pos];
if (currPos >= currSeq.litLength + currSeq.matchLength) {
currPos -= currSeq.litLength + currSeq.matchLength;
rawSeqStore->pos++;
} else {
rawSeqStore->posInSequence = currPos;
break;
}
}
if (currPos == 0 || rawSeqStore->pos == rawSeqStore->size) {
rawSeqStore->posInSequence = 0;
}
}
size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore,
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
@@ -554,6 +609,15 @@ size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore,
BYTE const* ip = istart;
DEBUGLOG(5, "ZSTD_ldm_blockCompress: srcSize=%zu", srcSize);
/* If using opt parser, use LDMs only as candidates rather than always accepting them */
if (cParams->strategy >= ZSTD_btopt) {
size_t lastLLSize;
ms->ldmSeqStore = rawSeqStore;
lastLLSize = blockCompressor(ms, seqStore, rep, src, srcSize);
ZSTD_ldm_skipRawSeqStoreBytes(rawSeqStore, srcSize);
return lastLLSize;
}
assert(rawSeqStore->pos <= rawSeqStore->size);
assert(rawSeqStore->size <= rawSeqStore->capacity);
/* Loop through each sequence and apply the block compressor to the lits */
@@ -566,14 +630,13 @@ size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore,
if (sequence.offset == 0)
break;
assert(sequence.offset <= (1U << cParams->windowLog));
assert(ip + sequence.litLength + sequence.matchLength <= iend);
/* Fill tables for block compressor */
ZSTD_ldm_limitTableUpdate(ms, ip);
ZSTD_ldm_fillFastTables(ms, ip);
/* Run the block compressor */
DEBUGLOG(5, "calling block compressor on segment of size %u", sequence.litLength);
DEBUGLOG(5, "pos %u : calling block compressor on segment of size %u", (unsigned)(ip-istart), sequence.litLength);
{
size_t const newLitLength =
blockCompressor(ms, seqStore, rep, ip, sequence.litLength);

View File

@@ -1,10 +1,11 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef ZSTD_LDM_H
@@ -23,6 +24,10 @@ extern "C" {
#define ZSTD_LDM_DEFAULT_WINDOW_LOG ZSTD_WINDOWLOG_LIMIT_DEFAULT
void ZSTD_ldm_fillHashTable(
ldmState_t* state, const BYTE* ip,
const BYTE* iend, ldmParams_t const* params);
/**
* ZSTD_ldm_generateSequences():
*
@@ -73,6 +78,12 @@ size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore,
void ZSTD_ldm_skipSequences(rawSeqStore_t* rawSeqStore, size_t srcSize,
U32 const minMatch);
/* ZSTD_ldm_skipRawSeqStoreBytes():
* Moves forward in rawSeqStore by nbBytes, updating fields 'pos' and 'posInSequence'.
* Not to be used in conjunction with ZSTD_ldm_skipSequences().
* Must be called for data with is not passed to ZSTD_ldm_blockCompress().
*/
void ZSTD_ldm_skipRawSeqStoreBytes(rawSeqStore_t* rawSeqStore, size_t nbBytes);
/** ZSTD_ldm_getTableSize() :
* Estimate the space needed for long distance matching tables or 0 if LDM is

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Przemyslaw Skibinski, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Przemyslaw Skibinski, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -249,40 +249,6 @@ static U32 ZSTD_litLengthPrice(U32 const litLength, const optState_t* const optP
}
}
/* ZSTD_litLengthContribution() :
* @return ( cost(litlength) - cost(0) )
* this value can then be added to rawLiteralsCost()
* to provide a cost which is directly comparable to a match ending at same position */
static int ZSTD_litLengthContribution(U32 const litLength, const optState_t* const optPtr, int optLevel)
{
if (optPtr->priceType >= zop_predef) return (int)WEIGHT(litLength, optLevel);
/* dynamic statistics */
{ U32 const llCode = ZSTD_LLcode(litLength);
int const contribution = (int)(LL_bits[llCode] * BITCOST_MULTIPLIER)
+ (int)WEIGHT(optPtr->litLengthFreq[0], optLevel) /* note: log2litLengthSum cancel out */
- (int)WEIGHT(optPtr->litLengthFreq[llCode], optLevel);
#if 1
return contribution;
#else
return MAX(0, contribution); /* sometimes better, sometimes not ... */
#endif
}
}
/* ZSTD_literalsContribution() :
* creates a fake cost for the literals part of a sequence
* which can be compared to the ending cost of a match
* should a new match start at this position */
static int ZSTD_literalsContribution(const BYTE* const literals, U32 const litLength,
const optState_t* const optPtr,
int optLevel)
{
int const contribution = (int)ZSTD_rawLiteralsCost(literals, litLength, optPtr, optLevel)
+ ZSTD_litLengthContribution(litLength, optPtr, optLevel);
return contribution;
}
/* ZSTD_getMatchPrice() :
* Provides the cost of the match part (offset + matchLength) of a sequence
* Must be combined with ZSTD_fullLiteralsCost() to get the full cost of a sequence.
@@ -420,32 +386,32 @@ static U32 ZSTD_insertBt1(
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const prefixStart = base + dictLimit;
const BYTE* match;
const U32 current = (U32)(ip-base);
const U32 btLow = btMask >= current ? 0 : current - btMask;
U32* smallerPtr = bt + 2*(current&btMask);
const U32 curr = (U32)(ip-base);
const U32 btLow = btMask >= curr ? 0 : curr - btMask;
U32* smallerPtr = bt + 2*(curr&btMask);
U32* largerPtr = smallerPtr + 1;
U32 dummy32; /* to be nullified at the end */
U32 const windowLow = ms->window.lowLimit;
U32 matchEndIdx = current+8+1;
U32 matchEndIdx = curr+8+1;
size_t bestLength = 8;
U32 nbCompares = 1U << cParams->searchLog;
#ifdef ZSTD_C_PREDICT
U32 predictedSmall = *(bt + 2*((current-1)&btMask) + 0);
U32 predictedLarge = *(bt + 2*((current-1)&btMask) + 1);
U32 predictedSmall = *(bt + 2*((curr-1)&btMask) + 0);
U32 predictedLarge = *(bt + 2*((curr-1)&btMask) + 1);
predictedSmall += (predictedSmall>0);
predictedLarge += (predictedLarge>0);
#endif /* ZSTD_C_PREDICT */
DEBUGLOG(8, "ZSTD_insertBt1 (%u)", current);
DEBUGLOG(8, "ZSTD_insertBt1 (%u)", curr);
assert(ip <= iend-8); /* required for h calculation */
hashTable[h] = current; /* Update Hash Table */
hashTable[h] = curr; /* Update Hash Table */
assert(windowLow > 0);
while (nbCompares-- && (matchIndex >= windowLow)) {
U32* const nextPtr = bt + 2*(matchIndex & btMask);
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
assert(matchIndex < current);
assert(matchIndex < curr);
#ifdef ZSTD_C_PREDICT /* note : can create issues when hlog small <= 11 */
const U32* predictPtr = bt + 2*((matchIndex-1) & btMask); /* written this way, as bt is a roll buffer */
@@ -508,8 +474,8 @@ static U32 ZSTD_insertBt1(
*smallerPtr = *largerPtr = 0;
{ U32 positions = 0;
if (bestLength > 384) positions = MIN(192, (U32)(bestLength - 384)); /* speed optimization */
assert(matchEndIdx > current + 8);
return MAX(positions, matchEndIdx - (current + 8));
assert(matchEndIdx > curr + 8);
return MAX(positions, matchEndIdx - (curr + 8));
}
}
@@ -553,7 +519,7 @@ U32 ZSTD_insertBtAndGetAllMatches (
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32 const sufficient_len = MIN(cParams->targetLength, ZSTD_OPT_NUM -1);
const BYTE* const base = ms->window.base;
U32 const current = (U32)(ip-base);
U32 const curr = (U32)(ip-base);
U32 const hashLog = cParams->hashLog;
U32 const minMatch = (mls==3) ? 3 : 4;
U32* const hashTable = ms->hashTable;
@@ -567,12 +533,12 @@ U32 ZSTD_insertBtAndGetAllMatches (
U32 const dictLimit = ms->window.dictLimit;
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const prefixStart = base + dictLimit;
U32 const btLow = (btMask >= current) ? 0 : current - btMask;
U32 const windowLow = ZSTD_getLowestMatchIndex(ms, current, cParams->windowLog);
U32 const btLow = (btMask >= curr) ? 0 : curr - btMask;
U32 const windowLow = ZSTD_getLowestMatchIndex(ms, curr, cParams->windowLog);
U32 const matchLow = windowLow ? windowLow : 1;
U32* smallerPtr = bt + 2*(current&btMask);
U32* largerPtr = bt + 2*(current&btMask) + 1;
U32 matchEndIdx = current+8+1; /* farthest referenced position of any match => detects repetitive patterns */
U32* smallerPtr = bt + 2*(curr&btMask);
U32* largerPtr = bt + 2*(curr&btMask) + 1;
U32 matchEndIdx = curr+8+1; /* farthest referenced position of any match => detects repetitive patterns */
U32 dummy32; /* to be nullified at the end */
U32 mnum = 0;
U32 nbCompares = 1U << cParams->searchLog;
@@ -591,7 +557,7 @@ U32 ZSTD_insertBtAndGetAllMatches (
U32 const dmsBtLow = dictMode == ZSTD_dictMatchState && dmsBtMask < dmsHighLimit - dmsLowLimit ? dmsHighLimit - dmsBtMask : dmsLowLimit;
size_t bestLength = lengthToBeat-1;
DEBUGLOG(8, "ZSTD_insertBtAndGetAllMatches: current=%u", current);
DEBUGLOG(8, "ZSTD_insertBtAndGetAllMatches: current=%u", curr);
/* check repCode */
assert(ll0 <= 1); /* necessarily 1 or 0 */
@@ -599,26 +565,29 @@ U32 ZSTD_insertBtAndGetAllMatches (
U32 repCode;
for (repCode = ll0; repCode < lastR; repCode++) {
U32 const repOffset = (repCode==ZSTD_REP_NUM) ? (rep[0] - 1) : rep[repCode];
U32 const repIndex = current - repOffset;
U32 const repIndex = curr - repOffset;
U32 repLen = 0;
assert(current >= dictLimit);
if (repOffset-1 /* intentional overflow, discards 0 and -1 */ < current-dictLimit) { /* equivalent to `current > repIndex >= dictLimit` */
if (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(ip - repOffset, minMatch)) {
assert(curr >= dictLimit);
if (repOffset-1 /* intentional overflow, discards 0 and -1 */ < curr-dictLimit) { /* equivalent to `curr > repIndex >= dictLimit` */
/* We must validate the repcode offset because when we're using a dictionary the
* valid offset range shrinks when the dictionary goes out of bounds.
*/
if ((repIndex >= windowLow) & (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(ip - repOffset, minMatch))) {
repLen = (U32)ZSTD_count(ip+minMatch, ip+minMatch-repOffset, iLimit) + minMatch;
}
} else { /* repIndex < dictLimit || repIndex >= current */
} else { /* repIndex < dictLimit || repIndex >= curr */
const BYTE* const repMatch = dictMode == ZSTD_dictMatchState ?
dmsBase + repIndex - dmsIndexDelta :
dictBase + repIndex;
assert(current >= windowLow);
assert(curr >= windowLow);
if ( dictMode == ZSTD_extDict
&& ( ((repOffset-1) /*intentional overflow*/ < current - windowLow) /* equivalent to `current > repIndex >= windowLow` */
&& ( ((repOffset-1) /*intentional overflow*/ < curr - windowLow) /* equivalent to `curr > repIndex >= windowLow` */
& (((U32)((dictLimit-1) - repIndex) >= 3) ) /* intentional overflow : do not test positions overlapping 2 memory segments */)
&& (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(repMatch, minMatch)) ) {
repLen = (U32)ZSTD_count_2segments(ip+minMatch, repMatch+minMatch, iLimit, dictEnd, prefixStart) + minMatch;
}
if (dictMode == ZSTD_dictMatchState
&& ( ((repOffset-1) /*intentional overflow*/ < current - (dmsLowLimit + dmsIndexDelta)) /* equivalent to `current > repIndex >= dmsLowLimit` */
&& ( ((repOffset-1) /*intentional overflow*/ < curr - (dmsLowLimit + dmsIndexDelta)) /* equivalent to `curr > repIndex >= dmsLowLimit` */
& ((U32)((dictLimit-1) - repIndex) >= 3) ) /* intentional overflow : do not test positions overlapping 2 memory segments */
&& (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(repMatch, minMatch)) ) {
repLen = (U32)ZSTD_count_2segments(ip+minMatch, repMatch+minMatch, iLimit, dmsEnd, prefixStart) + minMatch;
@@ -640,7 +609,7 @@ U32 ZSTD_insertBtAndGetAllMatches (
if ((mls == 3) /*static*/ && (bestLength < mls)) {
U32 const matchIndex3 = ZSTD_insertAndFindFirstIndexHash3(ms, nextToUpdate3, ip);
if ((matchIndex3 >= matchLow)
& (current - matchIndex3 < (1<<18)) /*heuristic : longer distance likely too expensive*/ ) {
& (curr - matchIndex3 < (1<<18)) /*heuristic : longer distance likely too expensive*/ ) {
size_t mlen;
if ((dictMode == ZSTD_noDict) /*static*/ || (dictMode == ZSTD_dictMatchState) /*static*/ || (matchIndex3 >= dictLimit)) {
const BYTE* const match = base + matchIndex3;
@@ -655,26 +624,26 @@ U32 ZSTD_insertBtAndGetAllMatches (
DEBUGLOG(8, "found small match with hlog3, of length %u",
(U32)mlen);
bestLength = mlen;
assert(current > matchIndex3);
assert(curr > matchIndex3);
assert(mnum==0); /* no prior solution */
matches[0].off = (current - matchIndex3) + ZSTD_REP_MOVE;
matches[0].off = (curr - matchIndex3) + ZSTD_REP_MOVE;
matches[0].len = (U32)mlen;
mnum = 1;
if ( (mlen > sufficient_len) |
(ip+mlen == iLimit) ) { /* best possible length */
ms->nextToUpdate = current+1; /* skip insertion */
ms->nextToUpdate = curr+1; /* skip insertion */
return 1;
} } }
/* no dictMatchState lookup: dicts don't have a populated HC3 table */
}
hashTable[h] = current; /* Update Hash Table */
hashTable[h] = curr; /* Update Hash Table */
while (nbCompares-- && (matchIndex >= matchLow)) {
U32* const nextPtr = bt + 2*(matchIndex & btMask);
const BYTE* match;
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
assert(current > matchIndex);
assert(curr > matchIndex);
if ((dictMode == ZSTD_noDict) || (dictMode == ZSTD_dictMatchState) || (matchIndex+matchLength >= dictLimit)) {
assert(matchIndex+matchLength >= dictLimit); /* ensure the condition is correct when !extDict */
@@ -691,12 +660,12 @@ U32 ZSTD_insertBtAndGetAllMatches (
if (matchLength > bestLength) {
DEBUGLOG(8, "found match of length %u at distance %u (offCode=%u)",
(U32)matchLength, current - matchIndex, current - matchIndex + ZSTD_REP_MOVE);
(U32)matchLength, curr - matchIndex, curr - matchIndex + ZSTD_REP_MOVE);
assert(matchEndIdx > matchIndex);
if (matchLength > matchEndIdx - matchIndex)
matchEndIdx = matchIndex + (U32)matchLength;
bestLength = matchLength;
matches[mnum].off = (current - matchIndex) + ZSTD_REP_MOVE;
matches[mnum].off = (curr - matchIndex) + ZSTD_REP_MOVE;
matches[mnum].len = (U32)matchLength;
mnum++;
if ( (matchLength > ZSTD_OPT_NUM)
@@ -739,11 +708,11 @@ U32 ZSTD_insertBtAndGetAllMatches (
if (matchLength > bestLength) {
matchIndex = dictMatchIndex + dmsIndexDelta;
DEBUGLOG(8, "found dms match of length %u at distance %u (offCode=%u)",
(U32)matchLength, current - matchIndex, current - matchIndex + ZSTD_REP_MOVE);
(U32)matchLength, curr - matchIndex, curr - matchIndex + ZSTD_REP_MOVE);
if (matchLength > matchEndIdx - matchIndex)
matchEndIdx = matchIndex + (U32)matchLength;
bestLength = matchLength;
matches[mnum].off = (current - matchIndex) + ZSTD_REP_MOVE;
matches[mnum].off = (curr - matchIndex) + ZSTD_REP_MOVE;
matches[mnum].len = (U32)matchLength;
mnum++;
if ( (matchLength > ZSTD_OPT_NUM)
@@ -764,7 +733,7 @@ U32 ZSTD_insertBtAndGetAllMatches (
}
}
assert(matchEndIdx > current+8);
assert(matchEndIdx > curr+8);
ms->nextToUpdate = matchEndIdx - 8; /* skip repetitive patterns */
return mnum;
}
@@ -795,34 +764,144 @@ FORCE_INLINE_TEMPLATE U32 ZSTD_BtGetAllMatches (
}
}
/*************************
* LDM helper functions *
*************************/
/* Struct containing info needed to make decision about ldm inclusion */
typedef struct {
rawSeqStore_t seqStore; /* External match candidates store for this block */
U32 startPosInBlock; /* Start position of the current match candidate */
U32 endPosInBlock; /* End position of the current match candidate */
U32 offset; /* Offset of the match candidate */
} ZSTD_optLdm_t;
/* ZSTD_optLdm_skipRawSeqStoreBytes():
* Moves forward in rawSeqStore by nbBytes, which will update the fields 'pos' and 'posInSequence'.
*/
static void ZSTD_optLdm_skipRawSeqStoreBytes(rawSeqStore_t* rawSeqStore, size_t nbBytes) {
U32 currPos = (U32)(rawSeqStore->posInSequence + nbBytes);
while (currPos && rawSeqStore->pos < rawSeqStore->size) {
rawSeq currSeq = rawSeqStore->seq[rawSeqStore->pos];
if (currPos >= currSeq.litLength + currSeq.matchLength) {
currPos -= currSeq.litLength + currSeq.matchLength;
rawSeqStore->pos++;
} else {
rawSeqStore->posInSequence = currPos;
break;
}
}
if (currPos == 0 || rawSeqStore->pos == rawSeqStore->size) {
rawSeqStore->posInSequence = 0;
}
}
/* ZSTD_opt_getNextMatchAndUpdateSeqStore():
* Calculates the beginning and end of the next match in the current block.
* Updates 'pos' and 'posInSequence' of the ldmSeqStore.
*/
static void ZSTD_opt_getNextMatchAndUpdateSeqStore(ZSTD_optLdm_t* optLdm, U32 currPosInBlock,
U32 blockBytesRemaining) {
rawSeq currSeq;
U32 currBlockEndPos;
U32 literalsBytesRemaining;
U32 matchBytesRemaining;
/* Setting match end position to MAX to ensure we never use an LDM during this block */
if (optLdm->seqStore.size == 0 || optLdm->seqStore.pos >= optLdm->seqStore.size) {
optLdm->startPosInBlock = UINT_MAX;
optLdm->endPosInBlock = UINT_MAX;
return;
}
/* Calculate appropriate bytes left in matchLength and litLength after adjusting
based on ldmSeqStore->posInSequence */
currSeq = optLdm->seqStore.seq[optLdm->seqStore.pos];
assert(optLdm->seqStore.posInSequence <= currSeq.litLength + currSeq.matchLength);
currBlockEndPos = currPosInBlock + blockBytesRemaining;
literalsBytesRemaining = (optLdm->seqStore.posInSequence < currSeq.litLength) ?
currSeq.litLength - (U32)optLdm->seqStore.posInSequence :
0;
matchBytesRemaining = (literalsBytesRemaining == 0) ?
currSeq.matchLength - ((U32)optLdm->seqStore.posInSequence - currSeq.litLength) :
currSeq.matchLength;
/* If there are more literal bytes than bytes remaining in block, no ldm is possible */
if (literalsBytesRemaining >= blockBytesRemaining) {
optLdm->startPosInBlock = UINT_MAX;
optLdm->endPosInBlock = UINT_MAX;
ZSTD_optLdm_skipRawSeqStoreBytes(&optLdm->seqStore, blockBytesRemaining);
return;
}
/* Matches may be < MINMATCH by this process. In that case, we will reject them
when we are deciding whether or not to add the ldm */
optLdm->startPosInBlock = currPosInBlock + literalsBytesRemaining;
optLdm->endPosInBlock = optLdm->startPosInBlock + matchBytesRemaining;
optLdm->offset = currSeq.offset;
if (optLdm->endPosInBlock > currBlockEndPos) {
/* Match ends after the block ends, we can't use the whole match */
optLdm->endPosInBlock = currBlockEndPos;
ZSTD_optLdm_skipRawSeqStoreBytes(&optLdm->seqStore, currBlockEndPos - currPosInBlock);
} else {
/* Consume nb of bytes equal to size of sequence left */
ZSTD_optLdm_skipRawSeqStoreBytes(&optLdm->seqStore, literalsBytesRemaining + matchBytesRemaining);
}
}
/* ZSTD_optLdm_maybeAddMatch():
* Adds a match if it's long enough, based on it's 'matchStartPosInBlock'
* and 'matchEndPosInBlock', into 'matches'. Maintains the correct ordering of 'matches'
*/
static void ZSTD_optLdm_maybeAddMatch(ZSTD_match_t* matches, U32* nbMatches,
ZSTD_optLdm_t* optLdm, U32 currPosInBlock) {
U32 posDiff = currPosInBlock - optLdm->startPosInBlock;
/* Note: ZSTD_match_t actually contains offCode and matchLength (before subtracting MINMATCH) */
U32 candidateMatchLength = optLdm->endPosInBlock - optLdm->startPosInBlock - posDiff;
U32 candidateOffCode = optLdm->offset + ZSTD_REP_MOVE;
/* Ensure that current block position is not outside of the match */
if (currPosInBlock < optLdm->startPosInBlock
|| currPosInBlock >= optLdm->endPosInBlock
|| candidateMatchLength < MINMATCH) {
return;
}
if (*nbMatches == 0 || ((candidateMatchLength > matches[*nbMatches-1].len) && *nbMatches < ZSTD_OPT_NUM)) {
DEBUGLOG(6, "ZSTD_optLdm_maybeAddMatch(): Adding ldm candidate match (offCode: %u matchLength %u) at block position=%u",
candidateOffCode, candidateMatchLength, currPosInBlock);
matches[*nbMatches].len = candidateMatchLength;
matches[*nbMatches].off = candidateOffCode;
(*nbMatches)++;
}
}
/* ZSTD_optLdm_processMatchCandidate():
* Wrapper function to update ldm seq store and call ldm functions as necessary.
*/
static void ZSTD_optLdm_processMatchCandidate(ZSTD_optLdm_t* optLdm, ZSTD_match_t* matches, U32* nbMatches,
U32 currPosInBlock, U32 remainingBytes) {
if (optLdm->seqStore.size == 0 || optLdm->seqStore.pos >= optLdm->seqStore.size) {
return;
}
if (currPosInBlock >= optLdm->endPosInBlock) {
if (currPosInBlock > optLdm->endPosInBlock) {
/* The position at which ZSTD_optLdm_processMatchCandidate() is called is not necessarily
* at the end of a match from the ldm seq store, and will often be some bytes
* over beyond matchEndPosInBlock. As such, we need to correct for these "overshoots"
*/
U32 posOvershoot = currPosInBlock - optLdm->endPosInBlock;
ZSTD_optLdm_skipRawSeqStoreBytes(&optLdm->seqStore, posOvershoot);
}
ZSTD_opt_getNextMatchAndUpdateSeqStore(optLdm, currPosInBlock, remainingBytes);
}
ZSTD_optLdm_maybeAddMatch(matches, nbMatches, optLdm, currPosInBlock);
}
/*-*******************************
* Optimal parser
*********************************/
typedef struct repcodes_s {
U32 rep[3];
} repcodes_t;
static repcodes_t ZSTD_updateRep(U32 const rep[3], U32 const offset, U32 const ll0)
{
repcodes_t newReps;
if (offset >= ZSTD_REP_NUM) { /* full offset */
newReps.rep[2] = rep[1];
newReps.rep[1] = rep[0];
newReps.rep[0] = offset - ZSTD_REP_MOVE;
} else { /* repcode */
U32 const repCode = offset + ll0;
if (repCode > 0) { /* note : if repCode==0, no change */
U32 const currentOffset = (repCode==ZSTD_REP_NUM) ? (rep[0] - 1) : rep[repCode];
newReps.rep[2] = (repCode >= 2) ? rep[1] : rep[2];
newReps.rep[1] = rep[0];
newReps.rep[0] = currentOffset;
} else { /* repCode == 0 */
memcpy(&newReps, rep, sizeof(newReps));
}
}
return newReps;
}
static U32 ZSTD_totalLen(ZSTD_optimal_t sol)
@@ -839,7 +918,7 @@ listStats(const U32* table, int lastEltID)
int enb;
for (enb=0; enb < nbElts; enb++) {
(void)table;
//RAWLOG(2, "%3i:%3i, ", enb, table[enb]);
/* RAWLOG(2, "%3i:%3i, ", enb, table[enb]); */
RAWLOG(2, "%4i,", table[enb]);
}
RAWLOG(2, " \n");
@@ -872,6 +951,11 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
ZSTD_optimal_t* const opt = optStatePtr->priceTable;
ZSTD_match_t* const matches = optStatePtr->matchTable;
ZSTD_optimal_t lastSequence;
ZSTD_optLdm_t optLdm;
optLdm.seqStore = ms->ldmSeqStore ? *ms->ldmSeqStore : kNullRawSeqStore;
optLdm.endPosInBlock = optLdm.startPosInBlock = optLdm.offset = 0;
ZSTD_opt_getNextMatchAndUpdateSeqStore(&optLdm, (U32)(ip-istart), (U32)(iend-ip));
/* init */
DEBUGLOG(5, "ZSTD_compressBlock_opt_generic: current=%u, prefix=%u, nextToUpdate=%u",
@@ -887,14 +971,21 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
/* find first match */
{ U32 const litlen = (U32)(ip - anchor);
U32 const ll0 = !litlen;
U32 const nbMatches = ZSTD_BtGetAllMatches(matches, ms, &nextToUpdate3, ip, iend, dictMode, rep, ll0, minMatch);
U32 nbMatches = ZSTD_BtGetAllMatches(matches, ms, &nextToUpdate3, ip, iend, dictMode, rep, ll0, minMatch);
ZSTD_optLdm_processMatchCandidate(&optLdm, matches, &nbMatches,
(U32)(ip-istart), (U32)(iend - ip));
if (!nbMatches) { ip++; continue; }
/* initialize opt[0] */
{ U32 i ; for (i=0; i<ZSTD_REP_NUM; i++) opt[0].rep[i] = rep[i]; }
opt[0].mlen = 0; /* means is_a_literal */
opt[0].litlen = litlen;
opt[0].price = ZSTD_literalsContribution(anchor, litlen, optStatePtr, optLevel);
/* We don't need to include the actual price of the literals because
* it is static for the duration of the forward pass, and is included
* in every price. We include the literal length to avoid negative
* prices when we subtract the previous literal length.
*/
opt[0].price = ZSTD_litLengthPrice(litlen, optStatePtr, optLevel);
/* large match -> immediate encoding */
{ U32 const maxML = matches[nbMatches-1].len;
@@ -923,7 +1014,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
for (matchNb = 0; matchNb < nbMatches; matchNb++) {
U32 const offset = matches[matchNb].off;
U32 const end = matches[matchNb].len;
repcodes_t const repHistory = ZSTD_updateRep(rep, offset, ll0);
for ( ; pos <= end ; pos++ ) {
U32 const matchPrice = ZSTD_getMatchPrice(offset, pos, optStatePtr, optLevel);
U32 const sequencePrice = literalsPrice + matchPrice;
@@ -933,8 +1023,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
opt[pos].off = offset;
opt[pos].litlen = litlen;
opt[pos].price = sequencePrice;
ZSTD_STATIC_ASSERT(sizeof(opt[pos].rep) == sizeof(repHistory));
memcpy(opt[pos].rep, &repHistory, sizeof(repHistory));
} }
last_pos = pos-1;
}
@@ -961,7 +1049,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
opt[cur].off = 0;
opt[cur].litlen = litlen;
opt[cur].price = price;
memcpy(opt[cur].rep, opt[cur-1].rep, sizeof(opt[cur].rep));
} else {
DEBUGLOG(7, "cPos:%zi==rPos:%u : literal would cost more (%.2f>%.2f) (hist:%u,%u,%u)",
inr-istart, cur, ZSTD_fCost(price), ZSTD_fCost(opt[cur].price),
@@ -969,6 +1056,21 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
}
}
/* Set the repcodes of the current position. We must do it here
* because we rely on the repcodes of the 2nd to last sequence being
* correct to set the next chunks repcodes during the backward
* traversal.
*/
ZSTD_STATIC_ASSERT(sizeof(opt[cur].rep) == sizeof(repcodes_t));
assert(cur >= opt[cur].mlen);
if (opt[cur].mlen != 0) {
U32 const prev = cur - opt[cur].mlen;
repcodes_t newReps = ZSTD_updateRep(opt[prev].rep, opt[cur].off, opt[cur].litlen==0);
ZSTD_memcpy(opt[cur].rep, &newReps, sizeof(repcodes_t));
} else {
ZSTD_memcpy(opt[cur].rep, opt[cur - 1].rep, sizeof(repcodes_t));
}
/* last match must start at a minimum distance of 8 from oend */
if (inr > ilimit) continue;
@@ -984,8 +1086,12 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
U32 const litlen = (opt[cur].mlen == 0) ? opt[cur].litlen : 0;
U32 const previousPrice = opt[cur].price;
U32 const basePrice = previousPrice + ZSTD_litLengthPrice(0, optStatePtr, optLevel);
U32 const nbMatches = ZSTD_BtGetAllMatches(matches, ms, &nextToUpdate3, inr, iend, dictMode, opt[cur].rep, ll0, minMatch);
U32 nbMatches = ZSTD_BtGetAllMatches(matches, ms, &nextToUpdate3, inr, iend, dictMode, opt[cur].rep, ll0, minMatch);
U32 matchNb;
ZSTD_optLdm_processMatchCandidate(&optLdm, matches, &nbMatches,
(U32)(inr-istart), (U32)(iend-inr));
if (!nbMatches) {
DEBUGLOG(7, "rPos:%u : no match found", cur);
continue;
@@ -1009,7 +1115,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
/* set prices using matches found at position == cur */
for (matchNb = 0; matchNb < nbMatches; matchNb++) {
U32 const offset = matches[matchNb].off;
repcodes_t const repHistory = ZSTD_updateRep(opt[cur].rep, offset, ll0);
U32 const lastML = matches[matchNb].len;
U32 const startML = (matchNb>0) ? matches[matchNb-1].len+1 : minMatch;
U32 mlen;
@@ -1029,8 +1134,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
opt[pos].off = offset;
opt[pos].litlen = litlen;
opt[pos].price = price;
ZSTD_STATIC_ASSERT(sizeof(opt[pos].rep) == sizeof(repHistory));
memcpy(opt[pos].rep, &repHistory, sizeof(repHistory));
} else {
DEBUGLOG(7, "rPos:%u (ml=%2u) => new price is worse (%.2f>=%.2f)",
pos, mlen, ZSTD_fCost(price), ZSTD_fCost(opt[pos].price));
@@ -1046,6 +1149,17 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
_shortestPath: /* cur, last_pos, best_mlen, best_off have to be set */
assert(opt[0].mlen == 0);
/* Set the next chunk's repcodes based on the repcodes of the beginning
* of the last match, and the last sequence. This avoids us having to
* update them while traversing the sequences.
*/
if (lastSequence.mlen != 0) {
repcodes_t reps = ZSTD_updateRep(opt[cur].rep, lastSequence.off, lastSequence.litlen==0);
ZSTD_memcpy(rep, &reps, sizeof(reps));
} else {
ZSTD_memcpy(rep, opt[cur].rep, sizeof(repcodes_t));
}
{ U32 const storeEnd = cur + 1;
U32 storeStart = storeEnd;
U32 seqPos = cur;
@@ -1082,20 +1196,6 @@ _shortestPath: /* cur, last_pos, best_mlen, best_off have to be set */
continue; /* will finish */
}
/* repcodes update : like ZSTD_updateRep(), but update in place */
if (offCode >= ZSTD_REP_NUM) { /* full offset */
rep[2] = rep[1];
rep[1] = rep[0];
rep[0] = offCode - ZSTD_REP_MOVE;
} else { /* repcode */
U32 const repCode = offCode + (llen==0);
if (repCode) { /* note : if repCode==0, no change */
U32 const currentOffset = (repCode==ZSTD_REP_NUM) ? (rep[0] - 1) : rep[repCode];
if (repCode >= 2) rep[2] = rep[1];
rep[1] = rep[0];
rep[0] = currentOffset;
} }
assert(anchor + llen <= iend);
ZSTD_updateStats(optStatePtr, llen, anchor, offCode, mlen);
ZSTD_storeSeq(seqStore, llen, anchor, iend, offCode, mlen-MINMATCH);
@@ -1104,7 +1204,6 @@ _shortestPath: /* cur, last_pos, best_mlen, best_off have to be set */
} }
ZSTD_setBasePrices(optStatePtr, optLevel);
}
} /* while (ip < ilimit) */
/* Return the last literals size */
@@ -1156,7 +1255,7 @@ ZSTD_initStats_ultra(ZSTD_matchState_t* ms,
const void* src, size_t srcSize)
{
U32 tmpRep[ZSTD_REP_NUM]; /* updated rep codes will sink here */
memcpy(tmpRep, rep, sizeof(tmpRep));
ZSTD_memcpy(tmpRep, rep, sizeof(tmpRep));
DEBUGLOG(4, "ZSTD_initStats_ultra (srcSize=%zu)", srcSize);
assert(ms->opt.litLengthSum == 0); /* first block */
@@ -1189,7 +1288,7 @@ size_t ZSTD_compressBlock_btultra2(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize)
{
U32 const current = (U32)((const BYTE*)src - ms->window.base);
U32 const curr = (U32)((const BYTE*)src - ms->window.base);
DEBUGLOG(5, "ZSTD_compressBlock_btultra2 (srcSize=%zu)", srcSize);
/* 2-pass strategy:
@@ -1204,7 +1303,7 @@ size_t ZSTD_compressBlock_btultra2(
if ( (ms->opt.litLengthSum==0) /* first block */
&& (seqStore->sequences == seqStore->sequencesStart) /* no ldm */
&& (ms->window.dictLimit == ms->window.lowLimit) /* no dictionary */
&& (current == ms->window.dictLimit) /* start of frame, nothing already loaded nor skipped */
&& (curr == ms->window.dictLimit) /* start of frame, nothing already loaded nor skipped */
&& (srcSize > ZSTD_PREDEF_THRESHOLD)
) {
ZSTD_initStats_ultra(ms, seqStore, rep, src, srcSize);

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -2,7 +2,6 @@ package zstd
/*
#define ZSTD_STATIC_LINKING_ONLY
#include "stdint.h" // for uintptr_t
#include "zstd.h"
typedef struct compressStream2_result_s {
@@ -11,9 +10,10 @@ typedef struct compressStream2_result_s {
size_t bytes_written;
} compressStream2_result;
static void ZSTD_compressStream2_wrapper(compressStream2_result* result, ZSTD_CCtx* ctx, uintptr_t dst, size_t maxDstSize, const uintptr_t src, size_t srcSize) {
ZSTD_outBuffer outBuffer = { (void*)dst, maxDstSize, 0 };
ZSTD_inBuffer inBuffer = { (void*)src, srcSize, 0 };
static void ZSTD_compressStream2_wrapper(compressStream2_result* result, ZSTD_CCtx* ctx,
void* dst, size_t maxDstSize, const void* src, size_t srcSize) {
ZSTD_outBuffer outBuffer = { dst, maxDstSize, 0 };
ZSTD_inBuffer inBuffer = { src, srcSize, 0 };
size_t retCode = ZSTD_compressStream2(ctx, &outBuffer, &inBuffer, ZSTD_e_continue);
result->return_code = retCode;
@@ -21,9 +21,21 @@ static void ZSTD_compressStream2_wrapper(compressStream2_result* result, ZSTD_CC
result->bytes_written = outBuffer.pos;
}
static void ZSTD_compressStream2_finish(compressStream2_result* result, ZSTD_CCtx* ctx, uintptr_t dst, size_t maxDstSize, const uintptr_t src, size_t srcSize) {
ZSTD_outBuffer outBuffer = { (void*)dst, maxDstSize, 0 };
ZSTD_inBuffer inBuffer = { (void*)src, srcSize, 0 };
static void ZSTD_compressStream2_flush(compressStream2_result* result, ZSTD_CCtx* ctx,
void* dst, size_t maxDstSize, const void* src, size_t srcSize) {
ZSTD_outBuffer outBuffer = { dst, maxDstSize, 0 };
ZSTD_inBuffer inBuffer = { src, srcSize, 0 };
size_t retCode = ZSTD_compressStream2(ctx, &outBuffer, &inBuffer, ZSTD_e_flush);
result->return_code = retCode;
result->bytes_consumed = inBuffer.pos;
result->bytes_written = outBuffer.pos;
}
static void ZSTD_compressStream2_finish(compressStream2_result* result, ZSTD_CCtx* ctx,
void* dst, size_t maxDstSize, const void* src, size_t srcSize) {
ZSTD_outBuffer outBuffer = { dst, maxDstSize, 0 };
ZSTD_inBuffer inBuffer = { src, srcSize, 0 };
size_t retCode = ZSTD_compressStream2(ctx, &outBuffer, &inBuffer, ZSTD_e_end);
result->return_code = retCode;
@@ -38,9 +50,10 @@ typedef struct decompressStream2_result_s {
size_t bytes_written;
} decompressStream2_result;
static void ZSTD_decompressStream_wrapper(decompressStream2_result* result, ZSTD_DCtx* ctx, uintptr_t dst, size_t maxDstSize, const uintptr_t src, size_t srcSize) {
ZSTD_outBuffer outBuffer = { (void*)dst, maxDstSize, 0 };
ZSTD_inBuffer inBuffer = { (void*)src, srcSize, 0 };
static void ZSTD_decompressStream_wrapper(decompressStream2_result* result, ZSTD_DCtx* ctx,
void* dst, size_t maxDstSize, const void* src, size_t srcSize) {
ZSTD_outBuffer outBuffer = { dst, maxDstSize, 0 };
ZSTD_inBuffer inBuffer = { src, srcSize, 0 };
size_t retCode = ZSTD_decompressStream(ctx, &outBuffer, &inBuffer);
result->return_code = retCode;
@@ -155,20 +168,19 @@ func (w *Writer) Write(p []byte) (int, error) {
srcData = w.srcBuffer
}
srcPtr := C.uintptr_t(uintptr(0)) // Do not point anywhere, if src is empty
var srcPtr *byte // Do not point anywhere, if src is empty
if len(srcData) > 0 {
srcPtr = C.uintptr_t(uintptr(unsafe.Pointer(&srcData[0])))
srcPtr = &srcData[0]
}
C.ZSTD_compressStream2_wrapper(
w.resultBuffer,
w.ctx,
C.uintptr_t(uintptr(unsafe.Pointer(&w.dstBuffer[0]))),
unsafe.Pointer(&w.dstBuffer[0]),
C.size_t(len(w.dstBuffer)),
srcPtr,
unsafe.Pointer(srcPtr),
C.size_t(len(srcData)),
)
runtime.KeepAlive(p) // Ensure p is kept until here so pointer doesn't disappear during C call
ret := int(w.resultBuffer.return_code)
if err := getError(ret); err != nil {
return 0, err
@@ -203,6 +215,49 @@ func (w *Writer) Write(p []byte) (int, error) {
return len(p), err
}
// Flush writes any unwritten data to the underlying io.Writer.
func (w *Writer) Flush() error {
if w.firstError != nil {
return w.firstError
}
ret := 1 // So we loop at least once
for ret > 0 {
var srcPtr *byte // Do not point anywhere, if src is empty
if len(w.srcBuffer) > 0 {
srcPtr = &w.srcBuffer[0]
}
C.ZSTD_compressStream2_flush(
w.resultBuffer,
w.ctx,
unsafe.Pointer(&w.dstBuffer[0]),
C.size_t(len(w.dstBuffer)),
unsafe.Pointer(srcPtr),
C.size_t(len(w.srcBuffer)),
)
ret = int(w.resultBuffer.return_code)
if err := getError(ret); err != nil {
return err
}
w.srcBuffer = w.srcBuffer[w.resultBuffer.bytes_consumed:]
written := int(w.resultBuffer.bytes_written)
_, err := w.underlyingWriter.Write(w.dstBuffer[:written])
if err != nil {
return err
}
if ret > 0 { // We have a hint if we need to resize the dstBuffer
w.dstBuffer = w.dstBuffer[:cap(w.dstBuffer)]
if len(w.dstBuffer) < ret {
w.dstBuffer = make([]byte, ret)
}
}
}
return nil
}
// Close closes the Writer, flushing any unwritten data to the underlying
// io.Writer and freeing objects, but does not close the underlying io.Writer.
func (w *Writer) Close() error {
@@ -212,17 +267,17 @@ func (w *Writer) Close() error {
ret := 1 // So we loop at least once
for ret > 0 {
srcPtr := C.uintptr_t(uintptr(0)) // Do not point anywhere, if src is empty
var srcPtr *byte // Do not point anywhere, if src is empty
if len(w.srcBuffer) > 0 {
srcPtr = C.uintptr_t(uintptr(unsafe.Pointer(&w.srcBuffer[0])))
srcPtr = &w.srcBuffer[0]
}
C.ZSTD_compressStream2_finish(
w.resultBuffer,
w.ctx,
C.uintptr_t(uintptr(unsafe.Pointer(&w.dstBuffer[0]))),
unsafe.Pointer(&w.dstBuffer[0]),
C.size_t(len(w.dstBuffer)),
srcPtr,
unsafe.Pointer(srcPtr),
C.size_t(len(w.srcBuffer)),
)
ret = int(w.resultBuffer.return_code)
@@ -231,7 +286,11 @@ func (w *Writer) Close() error {
}
w.srcBuffer = w.srcBuffer[w.resultBuffer.bytes_consumed:]
written := int(w.resultBuffer.bytes_written)
w.underlyingWriter.Write(w.dstBuffer[:written])
_, err := w.underlyingWriter.Write(w.dstBuffer[:written])
if err != nil {
C.ZSTD_freeCStream(w.ctx)
return err
}
if ret > 0 { // We have a hint if we need to resize the dstBuffer
w.dstBuffer = w.dstBuffer[:cap(w.dstBuffer)]
@@ -388,17 +447,17 @@ func (r *reader) Read(p []byte) (int, error) {
src = src[:r.compressionLeft+n]
// C code
srcPtr := C.uintptr_t(uintptr(0)) // Do not point anywhere, if src is empty
var srcPtr *byte // Do not point anywhere, if src is empty
if len(src) > 0 {
srcPtr = C.uintptr_t(uintptr(unsafe.Pointer(&src[0])))
srcPtr = &src[0]
}
C.ZSTD_decompressStream_wrapper(
r.resultBuffer,
r.ctx,
C.uintptr_t(uintptr(unsafe.Pointer(&r.decompressionBuffer[0]))),
unsafe.Pointer(&r.decompressionBuffer[0]),
C.size_t(len(r.decompressionBuffer)),
srcPtr,
unsafe.Pointer(srcPtr),
C.size_t(len(src)),
)
retCode := int(r.resultBuffer.return_code)

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -257,7 +257,7 @@ static U64 FSE_read64(const void* memPtr)
U64 val; memcpy(&val, memPtr, sizeof(val)); return val;
}
#endif // FSE_FORCE_MEMORY_ACCESS
#endif /* FSE_FORCE_MEMORY_ACCESS */
static U16 FSE_readLE16(const void* memPtr)
{
@@ -1078,7 +1078,7 @@ static size_t HUF_decompress_usingDTable( /* -3% slower when non static */
BYTE* const ostart = (BYTE*) dst;
BYTE* op = ostart;
BYTE* const omax = op + maxDstSize;
BYTE* const olimit = omax-15;
BYTE* const olimit = maxDstSize < 15 ? op : omax-15;
const void* ptr = DTable;
const HUF_DElt* const dt = (const HUF_DElt*)(ptr)+1;
@@ -1092,7 +1092,7 @@ static size_t HUF_decompress_usingDTable( /* -3% slower when non static */
const size_t length1 = FSE_readLE16(jumpTable);
const size_t length2 = FSE_readLE16(jumpTable+1);
const size_t length3 = FSE_readLE16(jumpTable+2);
const size_t length4 = cSrcSize - 6 - length1 - length2 - length3; // check coherency !!
const size_t length4 = cSrcSize - 6 - length1 - length2 - length3; /* check coherency !! */
const char* const start1 = (const char*)(cSrc) + 6;
const char* const start2 = start1 + length1;
const char* const start3 = start2 + length2;
@@ -1150,11 +1150,11 @@ static size_t HUF_decompress_usingDTable( /* -3% slower when non static */
/* tail */
{
// bitTail = bitD1; // *much* slower : -20% !??!
/* bitTail = bitD1; */ /* *much* slower : -20% !??! */
FSE_DStream_t bitTail;
bitTail.ptr = bitD1.ptr;
bitTail.bitsConsumed = bitD1.bitsConsumed;
bitTail.bitContainer = bitD1.bitContainer; // required in case of FSE_DStream_endOfBuffer
bitTail.bitContainer = bitD1.bitContainer; /* required in case of FSE_DStream_endOfBuffer */
bitTail.start = start1;
for ( ; (FSE_reloadDStream(&bitTail) < FSE_DStream_completed) && (op<omax) ; op++)
{
@@ -1280,7 +1280,11 @@ static size_t HUF_decompress (void* dst, size_t maxDstSize, const void* cSrc, si
* Basic Types
*********************************************************/
#if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */
# include <stdint.h>
# if defined(_AIX)
# include <inttypes.h>
# else
# include <stdint.h> /* intptr_t */
# endif
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef int16_t S16;
@@ -1483,7 +1487,9 @@ static size_t ZSTDv01_getcBlockSize(const void* src, size_t srcSize, blockProper
static size_t ZSTD_copyUncompressedBlock(void* dst, size_t maxDstSize, const void* src, size_t srcSize)
{
if (srcSize > maxDstSize) return ERROR(dstSize_tooSmall);
memcpy(dst, src, srcSize);
if (srcSize > 0) {
memcpy(dst, src, srcSize);
}
return srcSize;
}
@@ -1502,7 +1508,7 @@ static size_t ZSTD_decompressLiterals(void* ctx,
if (srcSize <= 3) return ERROR(corruption_detected);
litSize = ip[1] + (ip[0]<<8);
litSize += ((ip[-3] >> 3) & 7) << 16; // mmmmh....
litSize += ((ip[-3] >> 3) & 7) << 16; /* mmmmh.... */
op = oend - litSize;
(void)ctx;
@@ -1541,7 +1547,9 @@ static size_t ZSTDv01_decodeLiteralsBlock(void* ctx,
size_t rleSize = litbp.origSize;
if (rleSize>maxDstSize) return ERROR(dstSize_tooSmall);
if (!srcSize) return ERROR(srcSize_wrong);
memset(oend - rleSize, *ip, rleSize);
if (rleSize > 0) {
memset(oend - rleSize, *ip, rleSize);
}
*litStart = oend - rleSize;
*litSize = rleSize;
ip++;
@@ -1901,8 +1909,10 @@ static size_t ZSTD_decompressSequences(
{
size_t lastLLSize = litEnd - litPtr;
if (op+lastLLSize > oend) return ERROR(dstSize_tooSmall);
if (op != litPtr) memmove(op, litPtr, lastLLSize);
op += lastLLSize;
if (lastLLSize > 0) {
if (op != litPtr) memmove(op, litPtr, lastLLSize);
op += lastLLSize;
}
}
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -89,7 +89,11 @@ extern "C" {
* Basic Types
*****************************************************************/
#if defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
# include <stdint.h>
# if defined(_AIX)
# include <inttypes.h>
# else
# include <stdint.h> /* intptr_t */
# endif
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef int16_t S16;
@@ -189,7 +193,7 @@ MEM_STATIC void MEM_write16(void* memPtr, U16 value)
memcpy(memPtr, &value, sizeof(value));
}
#endif // MEM_FORCE_MEMORY_ACCESS
#endif /* MEM_FORCE_MEMORY_ACCESS */
MEM_STATIC U16 MEM_readLE16(const void* memPtr)
@@ -2836,7 +2840,9 @@ static size_t ZSTD_getcBlockSize(const void* src, size_t srcSize, blockPropertie
static size_t ZSTD_copyUncompressedBlock(void* dst, size_t maxDstSize, const void* src, size_t srcSize)
{
if (srcSize > maxDstSize) return ERROR(dstSize_tooSmall);
memcpy(dst, src, srcSize);
if (srcSize > 0) {
memcpy(dst, src, srcSize);
}
return srcSize;
}
@@ -3229,8 +3235,10 @@ static size_t ZSTD_decompressSequences(
size_t lastLLSize = litEnd - litPtr;
if (litPtr > litEnd) return ERROR(corruption_detected);
if (op+lastLLSize > oend) return ERROR(dstSize_tooSmall);
if (op != litPtr) memmove(op, litPtr, lastLLSize);
op += lastLLSize;
if (lastLLSize > 0) {
if (op != litPtr) memmove(op, litPtr, lastLLSize);
op += lastLLSize;
}
}
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -90,7 +90,11 @@ extern "C" {
* Basic Types
*****************************************************************/
#if defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
# include <stdint.h>
# if defined(_AIX)
# include <inttypes.h>
# else
# include <stdint.h> /* intptr_t */
# endif
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef int16_t S16;
@@ -191,7 +195,7 @@ MEM_STATIC void MEM_write16(void* memPtr, U16 value)
}
#endif // MEM_FORCE_MEMORY_ACCESS
#endif /* MEM_FORCE_MEMORY_ACCESS */
MEM_STATIC U16 MEM_readLE16(const void* memPtr)
@@ -2477,7 +2481,9 @@ static size_t ZSTD_getcBlockSize(const void* src, size_t srcSize, blockPropertie
static size_t ZSTD_copyUncompressedBlock(void* dst, size_t maxDstSize, const void* src, size_t srcSize)
{
if (srcSize > maxDstSize) return ERROR(dstSize_tooSmall);
memcpy(dst, src, srcSize);
if (srcSize > 0) {
memcpy(dst, src, srcSize);
}
return srcSize;
}
@@ -2870,8 +2876,10 @@ static size_t ZSTD_decompressSequences(
size_t lastLLSize = litEnd - litPtr;
if (litPtr > litEnd) return ERROR(corruption_detected);
if (op+lastLLSize > oend) return ERROR(dstSize_tooSmall);
if (op != litPtr) memmove(op, litPtr, lastLLSize);
op += lastLLSize;
if (lastLLSize > 0) {
if (op != litPtr) memmove(op, litPtr, lastLLSize);
op += lastLLSize;
}
}
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -52,7 +52,11 @@ extern "C" {
* Basic Types
*****************************************************************/
#if defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
# include <stdint.h>
# if defined(_AIX)
# include <inttypes.h>
# else
# include <stdint.h> /* intptr_t */
# endif
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef int16_t S16;
@@ -161,7 +165,7 @@ MEM_STATIC void MEM_write16(void* memPtr, U16 value)
memcpy(memPtr, &value, sizeof(value));
}
#endif // MEM_FORCE_MEMORY_ACCESS
#endif /* MEM_FORCE_MEMORY_ACCESS */
MEM_STATIC U16 MEM_readLE16(const void* memPtr)
@@ -2603,7 +2607,9 @@ static size_t ZSTD_getcBlockSize(const void* src, size_t srcSize, blockPropertie
static size_t ZSTD_copyRawBlock(void* dst, size_t maxDstSize, const void* src, size_t srcSize)
{
if (srcSize > maxDstSize) return ERROR(dstSize_tooSmall);
memcpy(dst, src, srcSize);
if (srcSize > 0) {
memcpy(dst, src, srcSize);
}
return srcSize;
}
@@ -3008,8 +3014,10 @@ static size_t ZSTD_decompressSequences(
size_t lastLLSize = litEnd - litPtr;
if (litPtr > litEnd) return ERROR(corruption_detected);
if (op+lastLLSize > oend) return ERROR(dstSize_tooSmall);
if (op != litPtr) memcpy(op, litPtr, lastLLSize);
op += lastLLSize;
if (lastLLSize > 0) {
if (op != litPtr) memcpy(op, litPtr, lastLLSize);
op += lastLLSize;
}
}
}
@@ -3407,7 +3415,9 @@ static size_t ZBUFF_decompressWithDictionary(ZBUFF_DCtx* zbc, const void* src, s
static size_t ZBUFF_limitCopy(void* dst, size_t maxDstSize, const void* src, size_t srcSize)
{
size_t length = MIN(maxDstSize, srcSize);
memcpy(dst, src, length);
if (length > 0) {
memcpy(dst, src, length);
}
return length;
}

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

View File

@@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@@ -80,7 +80,11 @@ extern "C" {
* Basic Types
*****************************************************************/
#if defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
# include <stdint.h>
# if defined(_AIX)
# include <inttypes.h>
# else
# include <stdint.h> /* intptr_t */
# endif
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef int16_t S16;
@@ -1804,7 +1808,7 @@ static size_t HUFv05_readStats(BYTE* huffWeight, size_t hwSize, U32* rankStats,
if (!srcSize) return ERROR(srcSize_wrong);
iSize = ip[0];
//memset(huffWeight, 0, hwSize); /* is not necessary, even though some analyzer complain ... */
/* memset(huffWeight, 0, hwSize); */ /* is not necessary, even though some analyzer complain ... */
if (iSize >= 128) { /* special header */
if (iSize >= (242)) { /* RLE */
@@ -1879,7 +1883,7 @@ size_t HUFv05_readDTableX2 (U16* DTable, const void* src, size_t srcSize)
HUFv05_DEltX2* const dt = (HUFv05_DEltX2*)dtPtr;
HUFv05_STATIC_ASSERT(sizeof(HUFv05_DEltX2) == sizeof(U16)); /* if compilation fails here, assertion is false */
//memset(huffWeight, 0, sizeof(huffWeight)); /* is not necessary, even though some analyzer complain ... */
/* memset(huffWeight, 0, sizeof(huffWeight)); */ /* is not necessary, even though some analyzer complain ... */
iSize = HUFv05_readStats(huffWeight, HUFv05_MAX_SYMBOL_VALUE + 1, rankVal, &nbSymbols, &tableLog, src, srcSize);
if (HUFv05_isError(iSize)) return iSize;
@@ -2210,7 +2214,7 @@ size_t HUFv05_readDTableX4 (unsigned* DTable, const void* src, size_t srcSize)
HUFv05_STATIC_ASSERT(sizeof(HUFv05_DEltX4) == sizeof(unsigned)); /* if compilation fails here, assertion is false */
if (memLog > HUFv05_ABSOLUTEMAX_TABLELOG) return ERROR(tableLog_tooLarge);
//memset(weightList, 0, sizeof(weightList)); /* is not necessary, even though some analyzer complain ... */
/* memset(weightList, 0, sizeof(weightList)); */ /* is not necessary, even though some analyzer complain ... */
iSize = HUFv05_readStats(weightList, HUFv05_MAX_SYMBOL_VALUE + 1, rankStats, &nbSymbols, &tableLog, src, srcSize);
if (HUFv05_isError(iSize)) return iSize;
@@ -2539,9 +2543,9 @@ size_t HUFv05_decompress (void* dst, size_t dstSize, const void* cSrc, size_t cS
return decompress[algoNb](dst, dstSize, cSrc, cSrcSize);
//return HUFv05_decompress4X2(dst, dstSize, cSrc, cSrcSize); /* multi-streams single-symbol decoding */
//return HUFv05_decompress4X4(dst, dstSize, cSrc, cSrcSize); /* multi-streams double-symbols decoding */
//return HUFv05_decompress4X6(dst, dstSize, cSrc, cSrcSize); /* multi-streams quad-symbols decoding */
/* return HUFv05_decompress4X2(dst, dstSize, cSrc, cSrcSize); */ /* multi-streams single-symbol decoding */
/* return HUFv05_decompress4X4(dst, dstSize, cSrc, cSrcSize); */ /* multi-streams double-symbols decoding */
/* return HUFv05_decompress4X6(dst, dstSize, cSrc, cSrcSize); */ /* multi-streams quad-symbols decoding */
}
/*
zstd - standard compression library
@@ -3362,8 +3366,10 @@ static size_t ZSTDv05_decompressSequences(
size_t lastLLSize = litEnd - litPtr;
if (litPtr > litEnd) return ERROR(corruption_detected); /* too many literals already used */
if (op+lastLLSize > oend) return ERROR(dstSize_tooSmall);
memcpy(op, litPtr, lastLLSize);
op += lastLLSize;
if (lastLLSize > 0) {
memcpy(op, litPtr, lastLLSize);
op += lastLLSize;
}
}
return op-ostart;
@@ -3791,7 +3797,9 @@ static size_t ZBUFFv05_blockHeaderSize = 3;
static size_t ZBUFFv05_limitCopy(void* dst, size_t maxDstSize, const void* src, size_t srcSize)
{
size_t length = MIN(maxDstSize, srcSize);
memcpy(dst, src, length);
if (length > 0) {
memcpy(dst, src, length);
}
return length;
}
@@ -3928,7 +3936,7 @@ size_t ZBUFFv05_decompressContinue(ZBUFFv05_DCtx* zbc, void* dst, size_t* maxDst
*maxDstSizePtr = 0;
return headerSize - zbc->hPos;
}
// zbc->stage = ZBUFFv05ds_decodeHeader; break; /* useless : stage follows */
/* zbc->stage = ZBUFFv05ds_decodeHeader; break; */ /* useless : stage follows */
}
/* fall-through */
case ZBUFFv05ds_decodeHeader:
@@ -4001,7 +4009,7 @@ size_t ZBUFFv05_decompressContinue(ZBUFFv05_DCtx* zbc, void* dst, size_t* maxDst
if (!decodedSize) { zbc->stage = ZBUFFv05ds_read; break; } /* this was just a header */
zbc->outEnd = zbc->outStart + decodedSize;
zbc->stage = ZBUFFv05ds_flush;
// break; /* ZBUFFv05ds_flush follows */
/* break; */ /* ZBUFFv05ds_flush follows */
}
}
/* fall-through */

Some files were not shown because too many files have changed in this diff Show More