use glide and refactor (#281)

.travis.yml

@@ -1,7 +1,7 @@
 language: go
 go: 
-  - 1.5
   - 1.6
+  - 1.7
 
 script:
-  - make travis
+  - make test

Makefile

@@ -11,44 +11,39 @@ export LD_LIBRARY_PATH
 export DYLD_LIBRARY_PATH
 export GO_BUILD_TAGS
 
-GO=GO15VENDOREXPERIMENT="1" go
-
-PKGS=$(shell $(GO) list ./... | grep -v "cmd")
-
-all: build  
+all: build
 
 build:
-	$(GO) build -o bin/ledis-server -tags '$(GO_BUILD_TAGS)' cmd/ledis-server/*
-	$(GO) build -o bin/ledis-cli -tags '$(GO_BUILD_TAGS)' cmd/ledis-cli/*
+	rm -rf vendor && ln -s _vendor/vendor vendor
+	go build -o bin/ledis-benchmark -tags '$(GO_BUILD_TAGS)' cmd/ledis-benchmark/*
+	go build -o bin/ledis-dump -tags '$(GO_BUILD_TAGS)' cmd/ledis-dump/*
+	go build -o bin/ledis-load -tags '$(GO_BUILD_TAGS)' cmd/ledis-load/*
+	go build -o bin/ledis-repair -tags '$(GO_BUILD_TAGS)' cmd/ledis-repair/*
+	rm -rf vendor
 
-build_all: build
-	$(GO) build -o bin/ledis-benchmark -tags '$(GO_BUILD_TAGS)' cmd/ledis-benchmark/*
-	$(GO) build -o bin/ledis-dump -tags '$(GO_BUILD_TAGS)' cmd/ledis-dump/*
-	$(GO) build -o bin/ledis-load -tags '$(GO_BUILD_TAGS)' cmd/ledis-load/*
-	$(GO) build -o bin/ledis-repair -tags '$(GO_BUILD_TAGS)' cmd/ledis-repair/*
-	
 test:
-	# use vendor for test
-	@rm -rf vendor && ln -s cmd/vendor vendor
-	@$(GO) test --race -tags '$(GO_BUILD_TAGS)' $(PKGS)
-	@rm -rf vendor
+	rm -rf vendor && ln -s _vendor/vendor vendor
+	go test --race -tags '$(GO_BUILD_TAGS)' -timeout 2m ./...
+	rm -rf vendor
+
 
 clean:
-	$(GO) clean -i ./...
+	go clean -i ./...
 
 fmt:
 	gofmt -w -s  . 2>&1 | grep -vE 'vendor' | awk '{print} END{if(NR>0) {exit 1}}' 
 
-deps:
-	# see https://github.com/coreos/etcd/blob/master/scripts/updatedep.sh
-	rm -rf Godeps vendor cmd/vendor
-	mkdir -p cmd/vendor
-	ln -s cmd/vendor vendor
-	godep save ./...
-	rm -rf cmd/Godeps
-	rm vendor
-	mv Godeps cmd/
-
-travis:
-	@rm -rf vendor && ln -s cmd/vendor vendor
-	@$(GO) test --race -tags '$(GO_BUILD_TAGS)' $(PKGS)
+update_vendor:
+	which glide >/dev/null || curl https://glide.sh/get | sh
+	which glide-vc || go get -v -u github.com/sgotti/glide-vc
+	rm -r vendor && mv _vendor/vendor vendor || true
+	rm -rf _vendor
+ifdef PKG
+	glide get --strip-vendor --skip-test ${PKG}
+else
+	glide update --strip-vendor --skip-test
+endif
+	@echo "removing test files"
+	glide vc --only-code --no-tests
+	mkdir -p _vendor
+	mv vendor _vendor/vendor

README.md

@@ -34,9 +34,6 @@ Create a workspace and checkout ledisdb source
 
     cd src/github.com/siddontang/ledisdb
 
-    #install godep and be sure godep command can be found in $PATH
-    go get github.com/tools/godep
-
     #set build and run environment 
     source dev.sh
 
@@ -180,7 +177,7 @@ See [Clients](https://github.com/siddontang/ledisdb/wiki/Clients) to find or con
 
 ## Requirement
 
-+ Go version >= 1.5
++ Go version >= 1.6
 
 ## Related Repos
 

_vendor/vendor/github.com/BurntSushi/toml/decode.go

@@ -225,6 +225,9 @@ func (md *MetaData) unify(data interface{}, rv reflect.Value) error {
 func (md *MetaData) unifyStruct(mapping interface{}, rv reflect.Value) error {
 	tmap, ok := mapping.(map[string]interface{})
 	if !ok {
+		if mapping == nil {
+			return nil
+		}
 		return mismatch(rv, "map", mapping)
 	}
 
@@ -267,6 +270,9 @@ func (md *MetaData) unifyStruct(mapping interface{}, rv reflect.Value) error {
 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() {
@@ -292,6 +298,9 @@ func (md *MetaData) unifyMap(mapping interface{}, rv reflect.Value) error {
 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()
@@ -305,12 +314,16 @@ func (md *MetaData) unifyArray(data interface{}, rv reflect.Value) error {
 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)
 	}
-	sliceLen := datav.Len()
-	if rv.IsNil() {
-		rv.Set(reflect.MakeSlice(rv.Type(), sliceLen, sliceLen))
+	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)
 }
 

_vendor/vendor/github.com/BurntSushi/toml/encode.go

@@ -306,19 +306,30 @@ func (enc *Encoder) eStruct(key Key, rv reflect.Value) {
 	addFields = func(rt reflect.Type, rv reflect.Value, start []int) {
 		for i := 0; i < rt.NumField(); i++ {
 			f := rt.Field(i)
-			// skip unexporded fields
-			if f.PkgPath != "" {
+			// skip unexported fields
+			if f.PkgPath != "" && !f.Anonymous {
 				continue
 			}
 			frv := rv.Field(i)
 			if f.Anonymous {
-				frv := eindirect(frv)
-				t := frv.Type()
-				if t.Kind() != reflect.Struct {
-					encPanic(errAnonNonStruct)
+				t := f.Type
+				switch t.Kind() {
+				case reflect.Struct:
+					addFields(t, frv, f.Index)
+					continue
+				case reflect.Ptr:
+					if t.Elem().Kind() == reflect.Struct {
+						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.
 				}
-				addFields(t, frv, f.Index)
-			} else if typeIsHash(tomlTypeOfGo(frv)) {
+			}
+
+			if typeIsHash(tomlTypeOfGo(frv)) {
 				fieldsSub = append(fieldsSub, append(start, f.Index...))
 			} else {
 				fieldsDirect = append(fieldsDirect, append(start, f.Index...))
@@ -336,15 +347,14 @@ func (enc *Encoder) eStruct(key Key, rv reflect.Value) {
 				continue
 			}
 
-			keyName := sft.Tag.Get("toml")
-			if keyName == "-" {
+			tag := sft.Tag.Get("toml")
+			if tag == "-" {
 				continue
 			}
+			keyName, opts := getOptions(tag)
 			if keyName == "" {
 				keyName = sft.Name
 			}
-
-			keyName, opts := getOptions(keyName)
 			if _, ok := opts["omitempty"]; ok && isEmpty(sf) {
 				continue
 			} else if _, ok := opts["omitzero"]; ok && isZero(sf) {
@@ -457,34 +467,22 @@ func getOptions(keyName string) (string, map[string]struct{}) {
 func isZero(rv reflect.Value) bool {
 	switch rv.Kind() {
 	case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
-		if rv.Int() == 0 {
-			return true
-		}
+		return rv.Int() == 0
 	case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
-		if rv.Uint() == 0 {
-			return true
-		}
+		return rv.Uint() == 0
 	case reflect.Float32, reflect.Float64:
-		if rv.Float() == 0.0 {
-			return true
-		}
+		return rv.Float() == 0.0
 	}
-
 	return false
 }
 
 func isEmpty(rv reflect.Value) bool {
 	switch rv.Kind() {
-	case reflect.String:
-		if len(strings.TrimSpace(rv.String())) == 0 {
-			return true
-		}
-	case reflect.Array, reflect.Slice, reflect.Map:
-		if rv.Len() == 0 {
-			return true
-		}
+	case reflect.Array, reflect.Slice, reflect.Map, reflect.String:
+		return rv.Len() == 0
+	case reflect.Bool:
+		return !rv.Bool()
 	}
-
 	return false
 }
 

_vendor/vendor/github.com/BurntSushi/toml/lex.go

@@ -272,8 +272,6 @@ func lexTableNameStart(lx *lexer) stateFn {
 		lx.ignore()
 		lx.push(lexTableNameEnd)
 		return lexValue // reuse string lexing
-	case isWhitespace(r):
-		return lexTableNameStart
 	default:
 		return lexBareTableName
 	}
@@ -560,7 +558,6 @@ func lexMultilineRawString(lx *lexer) stateFn {
 func lexMultilineStringEscape(lx *lexer) stateFn {
 	// Handle the special case first:
 	if isNL(lx.next()) {
-		lx.next()
 		return lexMultilineString
 	} else {
 		lx.backup()

_vendor/vendor/github.com/BurntSushi/toml/parse.go

@@ -81,7 +81,7 @@ func (p *parser) next() item {
 }
 
 func (p *parser) bug(format string, v ...interface{}) {
-	log.Fatalf("BUG: %s\n\n", fmt.Sprintf(format, v...))
+	log.Panicf("BUG: %s\n\n", fmt.Sprintf(format, v...))
 }
 
 func (p *parser) expect(typ itemType) item {
@@ -216,7 +216,7 @@ func (p *parser) value(it item) (interface{}, tomlType) {
 	case itemDatetime:
 		t, err := time.Parse("2006-01-02T15:04:05Z", it.val)
 		if err != nil {
-			p.bug("Expected Zulu formatted DateTime, but got '%s'.", it.val)
+			p.panicf("Invalid RFC3339 Zulu DateTime: '%s'.", it.val)
 		}
 		return t, p.typeOfPrimitive(it)
 	case itemArray:
@@ -401,7 +401,7 @@ func stripFirstNewline(s string) string {
 	if len(s) == 0 || s[0] != '\n' {
 		return s
 	}
-	return s[1:len(s)]
+	return s[1:]
 }
 
 func stripEscapedWhitespace(s string) string {
@@ -481,12 +481,7 @@ func (p *parser) asciiEscapeToUnicode(bs []byte) rune {
 		p.bug("Could not parse '%s' as a hexadecimal number, but the "+
 			"lexer claims it's OK: %s", s, err)
 	}
-
-	// BUG(burntsushi)
-	// I honestly don't understand how this works. I can't seem
-	// to find a way to make this fail. I figured this would fail on invalid
-	// UTF-8 characters like U+DCFF, but it doesn't.
-	if !utf8.ValidString(string(rune(hex))) {
+	if !utf8.ValidRune(rune(hex)) {
 		p.panicf("Escaped character '\\u%s' is not valid UTF-8.", s)
 	}
 	return rune(hex)

_vendor/vendor/github.com/BurntSushi/toml/type_fields.go

@@ -92,10 +92,10 @@ func typeFields(t reflect.Type) []field {
 			// Scan f.typ for fields to include.
 			for i := 0; i < f.typ.NumField(); i++ {
 				sf := f.typ.Field(i)
-				if sf.PkgPath != "" { // unexported
+				if sf.PkgPath != "" && !sf.Anonymous { // unexported
 					continue
 				}
-				name := sf.Tag.Get("toml")
+				name, _ := getOptions(sf.Tag.Get("toml"))
 				if name == "-" {
 					continue
 				}

_vendor/vendor/github.com/cupcake/rdb/crc64/crc64.go

@@ -35,7 +35,7 @@ type digest struct {
 	crc uint64
 }
 
-func New() hash.Hash {
+func New() hash.Hash64 {
 	return &digest{}
 }
 

_vendor/vendor/github.com/cupcake/rdb/decoder.go

@@ -20,6 +20,10 @@ type Decoder interface {
 	// StartDatabase is called when database n starts.
 	// Once a database starts, another database will not start until EndDatabase is called.
 	StartDatabase(n int)
+	// AUX field
+	Aux(key, value []byte)
+	// ResizeDB hint
+	ResizeDatabase(dbSize, expiresSize uint32)
 	// Set is called once for each string key.
 	Set(key, value []byte, expiry int64)
 	// StartHash is called at the beginning of a hash.
@@ -38,6 +42,7 @@ type Decoder interface {
 	EndSet(key []byte)
 	// StartList is called at the beginning of a list.
 	// Rpush will be called exactly length times before EndList.
+	// If length of the list is not known, then length is -1
 	StartList(key []byte, length, expiry int64)
 	// Rpush is called once for each value in a list.
 	Rpush(key, value []byte)
@@ -102,11 +107,12 @@ const (
 	TypeZSet   ValueType = 3
 	TypeHash   ValueType = 4
 
-	TypeHashZipmap  ValueType = 9
-	TypeListZiplist ValueType = 10
-	TypeSetIntset   ValueType = 11
-	TypeZSetZiplist ValueType = 12
-	TypeHashZiplist ValueType = 13
+	TypeHashZipmap    ValueType = 9
+	TypeListZiplist   ValueType = 10
+	TypeSetIntset     ValueType = 11
+	TypeZSetZiplist   ValueType = 12
+	TypeHashZiplist   ValueType = 13
+	TypeListQuicklist ValueType = 14
 )
 
 const (
@@ -115,6 +121,8 @@ const (
 	rdb32bitLen = 2
 	rdbEncVal   = 3
 
+	rdbFlagAux      = 0xfa
+	rdbFlagResizeDB = 0xfb
 	rdbFlagExpiryMS = 0xfc
 	rdbFlagExpiry   = 0xfd
 	rdbFlagSelectDB = 0xfe
@@ -154,6 +162,26 @@ func (d *decode) decode() error {
 			return err
 		}
 		switch objType {
+		case rdbFlagAux:
+			auxKey, err := d.readString()
+			if err != nil {
+				return err
+			}
+			auxVal, err := d.readString()
+			if err != nil {
+				return err
+			}
+			d.event.Aux(auxKey, auxVal)
+		case rdbFlagResizeDB:
+			dbSize, _, err := d.readLength()
+			if err != nil {
+				return err
+			}
+			expiresSize, _, err := d.readLength()
+			if err != nil {
+				return err
+			}
+			d.event.ResizeDatabase(dbSize, expiresSize)
 		case rdbFlagExpiryMS:
 			_, err := io.ReadFull(d.r, d.intBuf)
 			if err != nil {
@@ -188,6 +216,7 @@ func (d *decode) decode() error {
 			if err != nil {
 				return err
 			}
+			expiry = 0
 		}
 	}
 
@@ -216,6 +245,16 @@ func (d *decode) readObject(key []byte, typ ValueType, expiry int64) error {
 			d.event.Rpush(key, value)
 		}
 		d.event.EndList(key)
+	case TypeListQuicklist:
+		length, _, err := d.readLength()
+		if err != nil {
+			return err
+		}
+		d.event.StartList(key, int64(-1), expiry)
+		for i := uint32(0); i < length; i++ {
+			d.readZiplist(key, 0, false)
+		}
+		d.event.EndList(key)
 	case TypeSet:
 		cardinality, _, err := d.readLength()
 		if err != nil {
@@ -269,7 +308,7 @@ func (d *decode) readObject(key []byte, typ ValueType, expiry int64) error {
 	case TypeHashZipmap:
 		return d.readZipmap(key, expiry)
 	case TypeListZiplist:
-		return d.readZiplist(key, expiry)
+		return d.readZiplist(key, expiry, true)
 	case TypeSetIntset:
 		return d.readIntset(key, expiry)
 	case TypeZSetZiplist:
@@ -378,7 +417,7 @@ func readZipmapItemLength(buf *sliceBuffer, readFree bool) (int, int, error) {
 	return int(b), int(free), err
 }
 
-func (d *decode) readZiplist(key []byte, expiry int64) error {
+func (d *decode) readZiplist(key []byte, expiry int64, addListEvents bool) error {
 	ziplist, err := d.readString()
 	if err != nil {
 		return err
@@ -388,7 +427,9 @@ func (d *decode) readZiplist(key []byte, expiry int64) error {
 	if err != nil {
 		return err
 	}
-	d.event.StartList(key, length, expiry)
+	if addListEvents {
+		d.event.StartList(key, length, expiry)
+	}
 	for i := int64(0); i < length; i++ {
 		entry, err := readZiplistEntry(buf)
 		if err != nil {
@@ -396,7 +437,9 @@ func (d *decode) readZiplist(key []byte, expiry int64) error {
 		}
 		d.event.Rpush(key, entry)
 	}
-	d.event.EndList(key)
+	if addListEvents {
+		d.event.EndList(key)
+	}
 	return nil
 }
 
@@ -585,7 +628,7 @@ func (d *decode) checkHeader() error {
 	}
 
 	version, _ := strconv.ParseInt(string(header[5:]), 10, 64)
-	if version < 1 || version > 6 {
+	if version < 1 || version > 7 {
 		return fmt.Errorf("rdb: invalid RDB version number %d", version)
 	}
 

_vendor/vendor/github.com/cupcake/rdb/nopdecoder/nop_decoder.go

@@ -5,6 +5,8 @@ type NopDecoder struct{}
 
 func (d NopDecoder) StartRDB()                                       {}
 func (d NopDecoder) StartDatabase(n int)                             {}
+func (d NopDecoder) Aux(key, value []byte)                           {}
+func (d NopDecoder) ResizeDatabase(dbSize, expiresSize uint32)       {}
 func (d NopDecoder) EndDatabase(n int)                               {}
 func (d NopDecoder) EndRDB()                                         {}
 func (d NopDecoder) Set(key, value []byte, expiry int64)             {}

_vendor/vendor/github.com/edsrzf/mmap-go/mmap_unix.go

@@ -2,7 +2,7 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.
 
-// +build darwin dragonfly freebsd linux openbsd solaris
+// +build darwin dragonfly freebsd linux openbsd solaris netbsd
 
 package mmap
 
@@ -35,7 +35,7 @@ func mmap(len int, inprot, inflags, fd uintptr, off int64) ([]byte, error) {
 }
 
 func flush(addr, len uintptr) error {
-	_, _, errno := syscall.Syscall(syscall.SYS_MSYNC, addr, len, syscall.MS_SYNC)
+	_, _, errno := syscall.Syscall(_SYS_MSYNC, addr, len, _MS_SYNC)
 	if errno != 0 {
 		return syscall.Errno(errno)
 	}

_vendor/vendor/github.com/edsrzf/mmap-go/mmap_windows.go

@@ -73,7 +73,20 @@ func mmap(len int, prot, flags, hfile uintptr, off int64) ([]byte, error) {
 
 func flush(addr, len uintptr) error {
 	errno := syscall.FlushViewOfFile(addr, len)
-	return os.NewSyscallError("FlushViewOfFile", errno)
+	if errno != nil {
+		return os.NewSyscallError("FlushViewOfFile", errno)
+	}
+
+	handleLock.Lock()
+	defer handleLock.Unlock()
+	handle, ok := handleMap[addr]
+	if !ok {
+		// should be impossible; we would've errored above
+		return errors.New("unknown base address")
+	}
+
+	errno = syscall.FlushFileBuffers(handle)
+	return os.NewSyscallError("FlushFileBuffers", errno)
 }
 
 func lock(addr, len uintptr) error {
@@ -88,13 +101,18 @@ func unlock(addr, len uintptr) error {
 
 func unmap(addr, len uintptr) error {
 	flush(addr, len)
+	// Lock the UnmapViewOfFile along with the handleMap deletion.
+	// As soon as we unmap the view, the OS is free to give the
+	// same addr to another new map. We don't want another goroutine
+	// to insert and remove the same addr into handleMap while
+	// we're trying to remove our old addr/handle pair.
+	handleLock.Lock()
+	defer handleLock.Unlock()
 	err := syscall.UnmapViewOfFile(addr)
 	if err != nil {
 		return err
 	}
 
-	handleLock.Lock()
-	defer handleLock.Unlock()
 	handle, ok := handleMap[addr]
 	if !ok {
 		// should be impossible; we would've errored above

_vendor/vendor/github.com/edsrzf/mmap-go/msync_netbsd.go

@@ -0,0 +1,8 @@
+// Copyright 2011 Evan Shaw. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package mmap
+
+const _SYS_MSYNC = 277
+const _MS_SYNC = 0x04

_vendor/vendor/github.com/edsrzf/mmap-go/msync_unix.go

@@ -0,0 +1,14 @@
+// Copyright 2011 Evan Shaw. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// +build darwin dragonfly freebsd linux openbsd solaris
+
+package mmap
+
+import (
+	"syscall"
+)
+
+const _SYS_MSYNC = syscall.SYS_MSYNC
+const _MS_SYNC = syscall.MS_SYNC

_vendor/vendor/github.com/golang/snappy/decode.go

@@ -17,6 +17,8 @@ var (
 	ErrTooLarge = errors.New("snappy: decoded block is too large")
 	// ErrUnsupported reports that the input isn't supported.
 	ErrUnsupported = errors.New("snappy: unsupported input")
+
+	errUnsupportedLiteralLength = errors.New("snappy: unsupported literal length")
 )
 
 // DecodedLen returns the length of the decoded block.
@@ -40,96 +42,33 @@ func decodedLen(src []byte) (blockLen, headerLen int, err error) {
 	return int(v), n, nil
 }
 
+const (
+	decodeErrCodeCorrupt                  = 1
+	decodeErrCodeUnsupportedLiteralLength = 2
+)
+
 // Decode returns the decoded form of src. The returned slice may be a sub-
 // slice of dst if dst was large enough to hold the entire decoded block.
 // Otherwise, a newly allocated slice will be returned.
-// It is valid to pass a nil dst.
+//
+// The dst and src must not overlap. It is valid to pass a nil dst.
 func Decode(dst, src []byte) ([]byte, error) {
 	dLen, s, err := decodedLen(src)
 	if err != nil {
 		return nil, err
 	}
-	if len(dst) < dLen {
+	if dLen <= len(dst) {
+		dst = dst[:dLen]
+	} else {
 		dst = make([]byte, dLen)
 	}
-
-	var d, offset, length int
-	for s < len(src) {
-		switch src[s] & 0x03 {
-		case tagLiteral:
-			x := uint(src[s] >> 2)
-			switch {
-			case x < 60:
-				s++
-			case x == 60:
-				s += 2
-				if s > len(src) {
-					return nil, ErrCorrupt
-				}
-				x = uint(src[s-1])
-			case x == 61:
-				s += 3
-				if s > len(src) {
-					return nil, ErrCorrupt
-				}
-				x = uint(src[s-2]) | uint(src[s-1])<<8
-			case x == 62:
-				s += 4
-				if s > len(src) {
-					return nil, ErrCorrupt
-				}
-				x = uint(src[s-3]) | uint(src[s-2])<<8 | uint(src[s-1])<<16
-			case x == 63:
-				s += 5
-				if s > len(src) {
-					return nil, ErrCorrupt
-				}
-				x = uint(src[s-4]) | uint(src[s-3])<<8 | uint(src[s-2])<<16 | uint(src[s-1])<<24
-			}
-			length = int(x + 1)
-			if length <= 0 {
-				return nil, errors.New("snappy: unsupported literal length")
-			}
-			if length > len(dst)-d || length > len(src)-s {
-				return nil, ErrCorrupt
-			}
-			copy(dst[d:], src[s:s+length])
-			d += length
-			s += length
-			continue
-
-		case tagCopy1:
-			s += 2
-			if s > len(src) {
-				return nil, ErrCorrupt
-			}
-			length = 4 + int(src[s-2])>>2&0x7
-			offset = int(src[s-2])&0xe0<<3 | int(src[s-1])
-
-		case tagCopy2:
-			s += 3
-			if s > len(src) {
-				return nil, ErrCorrupt
-			}
-			length = 1 + int(src[s-3])>>2
-			offset = int(src[s-2]) | int(src[s-1])<<8
-
-		case tagCopy4:
-			return nil, errors.New("snappy: unsupported COPY_4 tag")
-		}
-
-		end := d + length
-		if offset > d || end > len(dst) {
-			return nil, ErrCorrupt
-		}
-		for ; d < end; d++ {
-			dst[d] = dst[d-offset]
-		}
+	switch decode(dst, src[s:]) {
+	case 0:
+		return dst, nil
+	case decodeErrCodeUnsupportedLiteralLength:
+		return nil, errUnsupportedLiteralLength
 	}
-	if d != dLen {
-		return nil, ErrCorrupt
-	}
-	return dst[:d], nil
+	return nil, ErrCorrupt
 }
 
 // NewReader returns a new Reader that decompresses from r, using the framing
@@ -138,12 +77,12 @@ func Decode(dst, src []byte) ([]byte, error) {
 func NewReader(r io.Reader) *Reader {
 	return &Reader{
 		r:       r,
-		decoded: make([]byte, maxUncompressedChunkLen),
-		buf:     make([]byte, MaxEncodedLen(maxUncompressedChunkLen)+checksumSize),
+		decoded: make([]byte, maxBlockSize),
+		buf:     make([]byte, maxEncodedLenOfMaxBlockSize+checksumSize),
 	}
 }
 
-// Reader is an io.Reader than can read Snappy-compressed bytes.
+// Reader is an io.Reader that can read Snappy-compressed bytes.
 type Reader struct {
 	r       io.Reader
 	err     error
@@ -165,9 +104,9 @@ func (r *Reader) Reset(reader io.Reader) {
 	r.readHeader = false
 }
 
-func (r *Reader) readFull(p []byte) (ok bool) {
+func (r *Reader) readFull(p []byte, allowEOF bool) (ok bool) {
 	if _, r.err = io.ReadFull(r.r, p); r.err != nil {
-		if r.err == io.ErrUnexpectedEOF {
+		if r.err == io.ErrUnexpectedEOF || (r.err == io.EOF && !allowEOF) {
 			r.err = ErrCorrupt
 		}
 		return false
@@ -186,7 +125,7 @@ func (r *Reader) Read(p []byte) (int, error) {
 			r.i += n
 			return n, nil
 		}
-		if !r.readFull(r.buf[:4]) {
+		if !r.readFull(r.buf[:4], true) {
 			return 0, r.err
 		}
 		chunkType := r.buf[0]
@@ -213,7 +152,7 @@ func (r *Reader) Read(p []byte) (int, error) {
 				return 0, r.err
 			}
 			buf := r.buf[:chunkLen]
-			if !r.readFull(buf) {
+			if !r.readFull(buf, false) {
 				return 0, r.err
 			}
 			checksum := uint32(buf[0]) | uint32(buf[1])<<8 | uint32(buf[2])<<16 | uint32(buf[3])<<24
@@ -246,13 +185,17 @@ func (r *Reader) Read(p []byte) (int, error) {
 				return 0, r.err
 			}
 			buf := r.buf[:checksumSize]
-			if !r.readFull(buf) {
+			if !r.readFull(buf, false) {
 				return 0, r.err
 			}
 			checksum := uint32(buf[0]) | uint32(buf[1])<<8 | uint32(buf[2])<<16 | uint32(buf[3])<<24
 			// Read directly into r.decoded instead of via r.buf.
 			n := chunkLen - checksumSize
-			if !r.readFull(r.decoded[:n]) {
+			if n > len(r.decoded) {
+				r.err = ErrCorrupt
+				return 0, r.err
+			}
+			if !r.readFull(r.decoded[:n], false) {
 				return 0, r.err
 			}
 			if crc(r.decoded[:n]) != checksum {
@@ -268,7 +211,7 @@ func (r *Reader) Read(p []byte) (int, error) {
 				r.err = ErrCorrupt
 				return 0, r.err
 			}
-			if !r.readFull(r.buf[:len(magicBody)]) {
+			if !r.readFull(r.buf[:len(magicBody)], false) {
 				return 0, r.err
 			}
 			for i := 0; i < len(magicBody); i++ {
@@ -287,7 +230,7 @@ func (r *Reader) Read(p []byte) (int, error) {
 		}
 		// Section 4.4 Padding (chunk type 0xfe).
 		// Section 4.6. Reserved skippable chunks (chunk types 0x80-0xfd).
-		if !r.readFull(r.buf[:chunkLen]) {
+		if !r.readFull(r.buf[:chunkLen], false) {
 			return 0, r.err
 		}
 	}

_vendor/vendor/github.com/golang/snappy/decode_amd64.go

@@ -0,0 +1,14 @@
+// Copyright 2016 The Snappy-Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// +build !appengine
+// +build gc
+// +build !noasm
+
+package snappy
+
+// decode has the same semantics as in decode_other.go.
+//
+//go:noescape
+func decode(dst, src []byte) int

_vendor/vendor/github.com/golang/snappy/decode_amd64.s

@@ -0,0 +1,490 @@
+// Copyright 2016 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// +build !appengine
+// +build gc
+// +build !noasm
+
+#include "textflag.h"
+
+// The asm code generally follows the pure Go code in decode_other.go, except
+// where marked with a "!!!".
+
+// func decode(dst, src []byte) int
+//
+// All local variables fit into registers. The non-zero stack size is only to
+// spill registers and push args when issuing a CALL. The register allocation:
+//	- AX	scratch
+//	- BX	scratch
+//	- CX	length or x
+//	- DX	offset
+//	- SI	&src[s]
+//	- DI	&dst[d]
+//	+ R8	dst_base
+//	+ R9	dst_len
+//	+ R10	dst_base + dst_len
+//	+ R11	src_base
+//	+ R12	src_len
+//	+ R13	src_base + src_len
+//	- R14	used by doCopy
+//	- R15	used by doCopy
+//
+// The registers R8-R13 (marked with a "+") are set at the start of the
+// function, and after a CALL returns, and are not otherwise modified.
+//
+// The d variable is implicitly DI - R8,  and len(dst)-d is R10 - DI.
+// The s variable is implicitly SI - R11, and len(src)-s is R13 - SI.
+TEXT ·decode(SB), NOSPLIT, $48-56
+	// Initialize SI, DI and R8-R13.
+	MOVQ dst_base+0(FP), R8
+	MOVQ dst_len+8(FP), R9
+	MOVQ R8, DI
+	MOVQ R8, R10
+	ADDQ R9, R10
+	MOVQ src_base+24(FP), R11
+	MOVQ src_len+32(FP), R12
+	MOVQ R11, SI
+	MOVQ R11, R13
+	ADDQ R12, R13
+
+loop:
+	// for s < len(src)
+	CMPQ SI, R13
+	JEQ  end
+
+	// CX = uint32(src[s])
+	//
+	// switch src[s] & 0x03
+	MOVBLZX (SI), CX
+	MOVL    CX, BX
+	ANDL    $3, BX
+	CMPL    BX, $1
+	JAE     tagCopy
+
+	// ----------------------------------------
+	// The code below handles literal tags.
+
+	// case tagLiteral:
+	// x := uint32(src[s] >> 2)
+	// switch
+	SHRL $2, CX
+	CMPL CX, $60
+	JAE  tagLit60Plus
+
+	// case x < 60:
+	// s++
+	INCQ SI
+
+doLit:
+	// This is the end of the inner "switch", when we have a literal tag.
+	//
+	// We assume that CX == x and x fits in a uint32, where x is the variable
+	// used in the pure Go decode_other.go code.
+
+	// length = int(x) + 1
+	//
+	// Unlike the pure Go code, we don't need to check if length <= 0 because
+	// CX can hold 64 bits, so the increment cannot overflow.
+	INCQ CX
+
+	// Prepare to check if copying length bytes will run past the end of dst or
+	// src.
+	//
+	// AX = len(dst) - d
+	// BX = len(src) - s
+	MOVQ R10, AX
+	SUBQ DI, AX
+	MOVQ R13, BX
+	SUBQ SI, BX
+
+	// !!! Try a faster technique for short (16 or fewer bytes) copies.
+	//
+	// if length > 16 || len(dst)-d < 16 || len(src)-s < 16 {
+	//   goto callMemmove // Fall back on calling runtime·memmove.
+	// }
+	//
+	// The C++ snappy code calls this TryFastAppend. It also checks len(src)-s
+	// against 21 instead of 16, because it cannot assume that all of its input
+	// is contiguous in memory and so it needs to leave enough source bytes to
+	// read the next tag without refilling buffers, but Go's Decode assumes
+	// contiguousness (the src argument is a []byte).
+	CMPQ CX, $16
+	JGT  callMemmove
+	CMPQ AX, $16
+	JLT  callMemmove
+	CMPQ BX, $16
+	JLT  callMemmove
+
+	// !!! Implement the copy from src to dst as a 16-byte load and store.
+	// (Decode's documentation says that dst and src must not overlap.)
+	//
+	// This always copies 16 bytes, instead of only length bytes, but that's
+	// OK. If the input is a valid Snappy encoding then subsequent iterations
+	// will fix up the overrun. Otherwise, Decode returns a nil []byte (and a
+	// non-nil error), so the overrun will be ignored.
+	//
+	// Note that on amd64, it is legal and cheap to issue unaligned 8-byte or
+	// 16-byte loads and stores. This technique probably wouldn't be as
+	// effective on architectures that are fussier about alignment.
+	MOVOU 0(SI), X0
+	MOVOU X0, 0(DI)
+
+	// d += length
+	// s += length
+	ADDQ CX, DI
+	ADDQ CX, SI
+	JMP  loop
+
+callMemmove:
+	// if length > len(dst)-d || length > len(src)-s { etc }
+	CMPQ CX, AX
+	JGT  errCorrupt
+	CMPQ CX, BX
+	JGT  errCorrupt
+
+	// copy(dst[d:], src[s:s+length])
+	//
+	// This means calling runtime·memmove(&dst[d], &src[s], length), so we push
+	// DI, SI and CX as arguments. Coincidentally, we also need to spill those
+	// three registers to the stack, to save local variables across the CALL.
+	MOVQ DI, 0(SP)
+	MOVQ SI, 8(SP)
+	MOVQ CX, 16(SP)
+	MOVQ DI, 24(SP)
+	MOVQ SI, 32(SP)
+	MOVQ CX, 40(SP)
+	CALL runtime·memmove(SB)
+
+	// Restore local variables: unspill registers from the stack and
+	// re-calculate R8-R13.
+	MOVQ 24(SP), DI
+	MOVQ 32(SP), SI
+	MOVQ 40(SP), CX
+	MOVQ dst_base+0(FP), R8
+	MOVQ dst_len+8(FP), R9
+	MOVQ R8, R10
+	ADDQ R9, R10
+	MOVQ src_base+24(FP), R11
+	MOVQ src_len+32(FP), R12
+	MOVQ R11, R13
+	ADDQ R12, R13
+
+	// d += length
+	// s += length
+	ADDQ CX, DI
+	ADDQ CX, SI
+	JMP  loop
+
+tagLit60Plus:
+	// !!! This fragment does the
+	//
+	// s += x - 58; if uint(s) > uint(len(src)) { etc }
+	//
+	// checks. In the asm version, we code it once instead of once per switch case.
+	ADDQ CX, SI
+	SUBQ $58, SI
+	MOVQ SI, BX
+	SUBQ R11, BX
+	CMPQ BX, R12
+	JA   errCorrupt
+
+	// case x == 60:
+	CMPL CX, $61
+	JEQ  tagLit61
+	JA   tagLit62Plus
+
+	// x = uint32(src[s-1])
+	MOVBLZX -1(SI), CX
+	JMP     doLit
+
+tagLit61:
+	// case x == 61:
+	// x = uint32(src[s-2]) | uint32(src[s-1])<<8
+	MOVWLZX -2(SI), CX
+	JMP     doLit
+
+tagLit62Plus:
+	CMPL CX, $62
+	JA   tagLit63
+
+	// case x == 62:
+	// x = uint32(src[s-3]) | uint32(src[s-2])<<8 | uint32(src[s-1])<<16
+	MOVWLZX -3(SI), CX
+	MOVBLZX -1(SI), BX
+	SHLL    $16, BX
+	ORL     BX, CX
+	JMP     doLit
+
+tagLit63:
+	// case x == 63:
+	// x = uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24
+	MOVL -4(SI), CX
+	JMP  doLit
+
+// The code above handles literal tags.
+// ----------------------------------------
+// The code below handles copy tags.
+
+tagCopy4:
+	// case tagCopy4:
+	// s += 5
+	ADDQ $5, SI
+
+	// if uint(s) > uint(len(src)) { etc }
+	MOVQ SI, BX
+	SUBQ R11, BX
+	CMPQ BX, R12
+	JA   errCorrupt
+
+	// length = 1 + int(src[s-5])>>2
+	SHRQ $2, CX
+	INCQ CX
+
+	// offset = int(uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24)
+	MOVLQZX -4(SI), DX
+	JMP     doCopy
+
+tagCopy2:
+	// case tagCopy2:
+	// s += 3
+	ADDQ $3, SI
+
+	// if uint(s) > uint(len(src)) { etc }
+	MOVQ SI, BX
+	SUBQ R11, BX
+	CMPQ BX, R12
+	JA   errCorrupt
+
+	// length = 1 + int(src[s-3])>>2
+	SHRQ $2, CX
+	INCQ CX
+
+	// offset = int(uint32(src[s-2]) | uint32(src[s-1])<<8)
+	MOVWQZX -2(SI), DX
+	JMP     doCopy
+
+tagCopy:
+	// We have a copy tag. We assume that:
+	//	- BX == src[s] & 0x03
+	//	- CX == src[s]
+	CMPQ BX, $2
+	JEQ  tagCopy2
+	JA   tagCopy4
+
+	// case tagCopy1:
+	// s += 2
+	ADDQ $2, SI
+
+	// if uint(s) > uint(len(src)) { etc }
+	MOVQ SI, BX
+	SUBQ R11, BX
+	CMPQ BX, R12
+	JA   errCorrupt
+
+	// offset = int(uint32(src[s-2])&0xe0<<3 | uint32(src[s-1]))
+	MOVQ    CX, DX
+	ANDQ    $0xe0, DX
+	SHLQ    $3, DX
+	MOVBQZX -1(SI), BX
+	ORQ     BX, DX
+
+	// length = 4 + int(src[s-2])>>2&0x7
+	SHRQ $2, CX
+	ANDQ $7, CX
+	ADDQ $4, CX
+
+doCopy:
+	// This is the end of the outer "switch", when we have a copy tag.
+	//
+	// We assume that:
+	//	- CX == length && CX > 0
+	//	- DX == offset
+
+	// if offset <= 0 { etc }
+	CMPQ DX, $0
+	JLE  errCorrupt
+
+	// if d < offset { etc }
+	MOVQ DI, BX
+	SUBQ R8, BX
+	CMPQ BX, DX
+	JLT  errCorrupt
+
+	// if length > len(dst)-d { etc }
+	MOVQ R10, BX
+	SUBQ DI, BX
+	CMPQ CX, BX
+	JGT  errCorrupt
+
+	// forwardCopy(dst[d:d+length], dst[d-offset:]); d += length
+	//
+	// Set:
+	//	- R14 = len(dst)-d
+	//	- R15 = &dst[d-offset]
+	MOVQ R10, R14
+	SUBQ DI, R14
+	MOVQ DI, R15
+	SUBQ DX, R15
+
+	// !!! Try a faster technique for short (16 or fewer bytes) forward copies.
+	//
+	// First, try using two 8-byte load/stores, similar to the doLit technique
+	// above. Even if dst[d:d+length] and dst[d-offset:] can overlap, this is
+	// still OK if offset >= 8. Note that this has to be two 8-byte load/stores
+	// and not one 16-byte load/store, and the first store has to be before the
+	// second load, due to the overlap if offset is in the range [8, 16).
+	//
+	// if length > 16 || offset < 8 || len(dst)-d < 16 {
+	//   goto slowForwardCopy
+	// }
+	// copy 16 bytes
+	// d += length
+	CMPQ CX, $16
+	JGT  slowForwardCopy
+	CMPQ DX, $8
+	JLT  slowForwardCopy
+	CMPQ R14, $16
+	JLT  slowForwardCopy
+	MOVQ 0(R15), AX
+	MOVQ AX, 0(DI)
+	MOVQ 8(R15), BX
+	MOVQ BX, 8(DI)
+	ADDQ CX, DI
+	JMP  loop
+
+slowForwardCopy:
+	// !!! If the forward copy is longer than 16 bytes, or if offset < 8, we
+	// can still try 8-byte load stores, provided we can overrun up to 10 extra
+	// bytes. As above, the overrun will be fixed up by subsequent iterations
+	// of the outermost loop.
+	//
+	// The C++ snappy code calls this technique IncrementalCopyFastPath. Its
+	// commentary says:
+	//
+	// ----
+	//
+	// The main part of this loop is a simple copy of eight bytes at a time
+	// until we've copied (at least) the requested amount of bytes.  However,
+	// if d and d-offset are less than eight bytes apart (indicating a
+	// repeating pattern of length < 8), we first need to expand the pattern in
+	// order to get the correct results. For instance, if the buffer looks like
+	// this, with the eight-byte <d-offset> and <d> patterns marked as
+	// intervals:
+	//
+	//    abxxxxxxxxxxxx
+	//    [------]           d-offset
+	//      [------]         d
+	//
+	// a single eight-byte copy from <d-offset> to <d> will repeat the pattern
+	// once, after which we can move <d> two bytes without moving <d-offset>:
+	//
+	//    ababxxxxxxxxxx
+	//    [------]           d-offset
+	//        [------]       d
+	//
+	// and repeat the exercise until the two no longer overlap.
+	//
+	// This allows us to do very well in the special case of one single byte
+	// repeated many times, without taking a big hit for more general cases.
+	//
+	// The worst case of extra writing past the end of the match occurs when
+	// offset == 1 and length == 1; the last copy will read from byte positions
+	// [0..7] and write to [4..11], whereas it was only supposed to write to
+	// position 1. Thus, ten excess bytes.
+	//
+	// ----
+	//
+	// That "10 byte overrun" worst case is confirmed by Go's
+	// TestSlowForwardCopyOverrun, which also tests the fixUpSlowForwardCopy
+	// and finishSlowForwardCopy algorithm.
+	//
+	// if length > len(dst)-d-10 {
+	//   goto verySlowForwardCopy
+	// }
+	SUBQ $10, R14
+	CMPQ CX, R14
+	JGT  verySlowForwardCopy
+
+makeOffsetAtLeast8:
+	// !!! As above, expand the pattern so that offset >= 8 and we can use
+	// 8-byte load/stores.
+	//
+	// for offset < 8 {
+	//   copy 8 bytes from dst[d-offset:] to dst[d:]
+	//   length -= offset
+	//   d      += offset
+	//   offset += offset
+	//   // The two previous lines together means that d-offset, and therefore
+	//   // R15, is unchanged.
+	// }
+	CMPQ DX, $8
+	JGE  fixUpSlowForwardCopy
+	MOVQ (R15), BX
+	MOVQ BX, (DI)
+	SUBQ DX, CX
+	ADDQ DX, DI
+	ADDQ DX, DX
+	JMP  makeOffsetAtLeast8
+
+fixUpSlowForwardCopy:
+	// !!! Add length (which might be negative now) to d (implied by DI being
+	// &dst[d]) so that d ends up at the right place when we jump back to the
+	// top of the loop. Before we do that, though, we save DI to AX so that, if
+	// length is positive, copying the remaining length bytes will write to the
+	// right place.
+	MOVQ DI, AX
+	ADDQ CX, DI
+
+finishSlowForwardCopy:
+	// !!! Repeat 8-byte load/stores until length <= 0. Ending with a negative
+	// length means that we overrun, but as above, that will be fixed up by
+	// subsequent iterations of the outermost loop.
+	CMPQ CX, $0
+	JLE  loop
+	MOVQ (R15), BX
+	MOVQ BX, (AX)
+	ADDQ $8, R15
+	ADDQ $8, AX
+	SUBQ $8, CX
+	JMP  finishSlowForwardCopy
+
+verySlowForwardCopy:
+	// verySlowForwardCopy is a simple implementation of forward copy. In C
+	// parlance, this is a do/while loop instead of a while loop, since we know
+	// that length > 0. In Go syntax:
+	//
+	// for {
+	//   dst[d] = dst[d - offset]
+	//   d++
+	//   length--
+	//   if length == 0 {
+	//     break
+	//   }
+	// }
+	MOVB (R15), BX
+	MOVB BX, (DI)
+	INCQ R15
+	INCQ DI
+	DECQ CX
+	JNZ  verySlowForwardCopy
+	JMP  loop
+
+// The code above handles copy tags.
+// ----------------------------------------
+
+end:
+	// This is the end of the "for s < len(src)".
+	//
+	// if d != len(dst) { etc }
+	CMPQ DI, R10
+	JNE  errCorrupt
+
+	// return 0
+	MOVQ $0, ret+48(FP)
+	RET
+
+errCorrupt:
+	// return decodeErrCodeCorrupt
+	MOVQ $1, ret+48(FP)
+	RET

_vendor/vendor/github.com/golang/snappy/decode_other.go

@@ -0,0 +1,101 @@
+// Copyright 2016 The Snappy-Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// +build !amd64 appengine !gc noasm
+
+package snappy
+
+// decode writes the decoding of src to dst. It assumes that the varint-encoded
+// length of the decompressed bytes has already been read, and that len(dst)
+// equals that length.
+//
+// It returns 0 on success or a decodeErrCodeXxx error code on failure.
+func decode(dst, src []byte) int {
+	var d, s, offset, length int
+	for s < len(src) {
+		switch src[s] & 0x03 {
+		case tagLiteral:
+			x := uint32(src[s] >> 2)
+			switch {
+			case x < 60:
+				s++
+			case x == 60:
+				s += 2
+				if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line.
+					return decodeErrCodeCorrupt
+				}
+				x = uint32(src[s-1])
+			case x == 61:
+				s += 3
+				if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line.
+					return decodeErrCodeCorrupt
+				}
+				x = uint32(src[s-2]) | uint32(src[s-1])<<8
+			case x == 62:
+				s += 4
+				if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line.
+					return decodeErrCodeCorrupt
+				}
+				x = uint32(src[s-3]) | uint32(src[s-2])<<8 | uint32(src[s-1])<<16
+			case x == 63:
+				s += 5
+				if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line.
+					return decodeErrCodeCorrupt
+				}
+				x = uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24
+			}
+			length = int(x) + 1
+			if length <= 0 {
+				return decodeErrCodeUnsupportedLiteralLength
+			}
+			if length > len(dst)-d || length > len(src)-s {
+				return decodeErrCodeCorrupt
+			}
+			copy(dst[d:], src[s:s+length])
+			d += length
+			s += length
+			continue
+
+		case tagCopy1:
+			s += 2
+			if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line.
+				return decodeErrCodeCorrupt
+			}
+			length = 4 + int(src[s-2])>>2&0x7
+			offset = int(uint32(src[s-2])&0xe0<<3 | uint32(src[s-1]))
+
+		case tagCopy2:
+			s += 3
+			if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line.
+				return decodeErrCodeCorrupt
+			}
+			length = 1 + int(src[s-3])>>2
+			offset = int(uint32(src[s-2]) | uint32(src[s-1])<<8)
+
+		case tagCopy4:
+			s += 5
+			if uint(s) > uint(len(src)) { // The uint conversions catch overflow from the previous line.
+				return decodeErrCodeCorrupt
+			}
+			length = 1 + int(src[s-5])>>2
+			offset = int(uint32(src[s-4]) | uint32(src[s-3])<<8 | uint32(src[s-2])<<16 | uint32(src[s-1])<<24)
+		}
+
+		if offset <= 0 || d < offset || length > len(dst)-d {
+			return decodeErrCodeCorrupt
+		}
+		// Copy from an earlier sub-slice of dst to a later sub-slice. Unlike
+		// the built-in copy function, this byte-by-byte copy always runs
+		// forwards, even if the slices overlap. Conceptually, this is:
+		//
+		// d += forwardCopy(dst[d:d+length], dst[d-offset:])
+		for end := d + length; d != end; d++ {
+			dst[d] = dst[d-offset]
+		}
+	}
+	if d != len(dst) {
+		return decodeErrCodeCorrupt
+	}
+	return 0
+}

_vendor/vendor/github.com/golang/snappy/encode.go

@@ -0,0 +1,285 @@
+// Copyright 2011 The Snappy-Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package snappy
+
+import (
+	"encoding/binary"
+	"errors"
+	"io"
+)
+
+// Encode returns the encoded form of src. The returned slice may be a sub-
+// slice of dst if dst was large enough to hold the entire encoded block.
+// Otherwise, a newly allocated slice will be returned.
+//
+// The dst and src must not overlap. It is valid to pass a nil dst.
+func Encode(dst, src []byte) []byte {
+	if n := MaxEncodedLen(len(src)); n < 0 {
+		panic(ErrTooLarge)
+	} else if len(dst) < n {
+		dst = make([]byte, n)
+	}
+
+	// The block starts with the varint-encoded length of the decompressed bytes.
+	d := binary.PutUvarint(dst, uint64(len(src)))
+
+	for len(src) > 0 {
+		p := src
+		src = nil
+		if len(p) > maxBlockSize {
+			p, src = p[:maxBlockSize], p[maxBlockSize:]
+		}
+		if len(p) < minNonLiteralBlockSize {
+			d += emitLiteral(dst[d:], p)
+		} else {
+			d += encodeBlock(dst[d:], p)
+		}
+	}
+	return dst[:d]
+}
+
+// inputMargin is the minimum number of extra input bytes to keep, inside
+// encodeBlock's inner loop. On some architectures, this margin lets us
+// implement a fast path for emitLiteral, where the copy of short (<= 16 byte)
+// literals can be implemented as a single load to and store from a 16-byte
+// register. That literal's actual length can be as short as 1 byte, so this
+// can copy up to 15 bytes too much, but that's OK as subsequent iterations of
+// the encoding loop will fix up the copy overrun, and this inputMargin ensures
+// that we don't overrun the dst and src buffers.
+const inputMargin = 16 - 1
+
+// minNonLiteralBlockSize is the minimum size of the input to encodeBlock that
+// could be encoded with a copy tag. This is the minimum with respect to the
+// algorithm used by encodeBlock, not a minimum enforced by the file format.
+//
+// The encoded output must start with at least a 1 byte literal, as there are
+// no previous bytes to copy. A minimal (1 byte) copy after that, generated
+// from an emitCopy call in encodeBlock's main loop, would require at least
+// another inputMargin bytes, for the reason above: we want any emitLiteral
+// calls inside encodeBlock's main loop to use the fast path if possible, which
+// requires being able to overrun by inputMargin bytes. Thus,
+// minNonLiteralBlockSize equals 1 + 1 + inputMargin.
+//
+// The C++ code doesn't use this exact threshold, but it could, as discussed at
+// https://groups.google.com/d/topic/snappy-compression/oGbhsdIJSJ8/discussion
+// The difference between Go (2+inputMargin) and C++ (inputMargin) is purely an
+// optimization. It should not affect the encoded form. This is tested by
+// TestSameEncodingAsCppShortCopies.
+const minNonLiteralBlockSize = 1 + 1 + inputMargin
+
+// MaxEncodedLen returns the maximum length of a snappy block, given its
+// uncompressed length.
+//
+// It will return a negative value if srcLen is too large to encode.
+func MaxEncodedLen(srcLen int) int {
+	n := uint64(srcLen)
+	if n > 0xffffffff {
+		return -1
+	}
+	// Compressed data can be defined as:
+	//    compressed := item* literal*
+	//    item       := literal* copy
+	//
+	// The trailing literal sequence has a space blowup of at most 62/60
+	// since a literal of length 60 needs one tag byte + one extra byte
+	// for length information.
+	//
+	// Item blowup is trickier to measure. Suppose the "copy" op copies
+	// 4 bytes of data. Because of a special check in the encoding code,
+	// we produce a 4-byte copy only if the offset is < 65536. Therefore
+	// the copy op takes 3 bytes to encode, and this type of item leads
+	// to at most the 62/60 blowup for representing literals.
+	//
+	// Suppose the "copy" op copies 5 bytes of data. If the offset is big
+	// enough, it will take 5 bytes to encode the copy op. Therefore the
+	// worst case here is a one-byte literal followed by a five-byte copy.
+	// That is, 6 bytes of input turn into 7 bytes of "compressed" data.
+	//
+	// This last factor dominates the blowup, so the final estimate is:
+	n = 32 + n + n/6
+	if n > 0xffffffff {
+		return -1
+	}
+	return int(n)
+}
+
+var errClosed = errors.New("snappy: Writer is closed")
+
+// NewWriter returns a new Writer that compresses to w.
+//
+// The Writer returned does not buffer writes. There is no need to Flush or
+// Close such a Writer.
+//
+// Deprecated: the Writer returned is not suitable for many small writes, only
+// for few large writes. Use NewBufferedWriter instead, which is efficient
+// regardless of the frequency and shape of the writes, and remember to Close
+// that Writer when done.
+func NewWriter(w io.Writer) *Writer {
+	return &Writer{
+		w:    w,
+		obuf: make([]byte, obufLen),
+	}
+}
+
+// NewBufferedWriter returns a new Writer that compresses to w, using the
+// framing format described at
+// https://github.com/google/snappy/blob/master/framing_format.txt
+//
+// The Writer returned buffers writes. Users must call Close to guarantee all
+// data has been forwarded to the underlying io.Writer. They may also call
+// Flush zero or more times before calling Close.
+func NewBufferedWriter(w io.Writer) *Writer {
+	return &Writer{
+		w:    w,
+		ibuf: make([]byte, 0, maxBlockSize),
+		obuf: make([]byte, obufLen),
+	}
+}
+
+// Writer is an io.Writer that can write Snappy-compressed bytes.
+type Writer struct {
+	w   io.Writer
+	err error
+
+	// ibuf is a buffer for the incoming (uncompressed) bytes.
+	//
+	// Its use is optional. For backwards compatibility, Writers created by the
+	// NewWriter function have ibuf == nil, do not buffer incoming bytes, and
+	// therefore do not need to be Flush'ed or Close'd.
+	ibuf []byte
+
+	// obuf is a buffer for the outgoing (compressed) bytes.
+	obuf []byte
+
+	// wroteStreamHeader is whether we have written the stream header.
+	wroteStreamHeader bool
+}
+
+// Reset discards the writer's state and switches the Snappy writer to write to
+// w. This permits reusing a Writer rather than allocating a new one.
+func (w *Writer) Reset(writer io.Writer) {
+	w.w = writer
+	w.err = nil
+	if w.ibuf != nil {
+		w.ibuf = w.ibuf[:0]
+	}
+	w.wroteStreamHeader = false
+}
+
+// Write satisfies the io.Writer interface.
+func (w *Writer) Write(p []byte) (nRet int, errRet error) {
+	if w.ibuf == nil {
+		// Do not buffer incoming bytes. This does not perform or compress well
+		// if the caller of Writer.Write writes many small slices. This
+		// behavior is therefore deprecated, but still supported for backwards
+		// compatibility with code that doesn't explicitly Flush or Close.
+		return w.write(p)
+	}
+
+	// The remainder of this method is based on bufio.Writer.Write from the
+	// standard library.
+
+	for len(p) > (cap(w.ibuf)-len(w.ibuf)) && w.err == nil {
+		var n int
+		if len(w.ibuf) == 0 {
+			// Large write, empty buffer.
+			// Write directly from p to avoid copy.
+			n, _ = w.write(p)
+		} else {
+			n = copy(w.ibuf[len(w.ibuf):cap(w.ibuf)], p)
+			w.ibuf = w.ibuf[:len(w.ibuf)+n]
+			w.Flush()
+		}
+		nRet += n
+		p = p[n:]
+	}
+	if w.err != nil {
+		return nRet, w.err
+	}
+	n := copy(w.ibuf[len(w.ibuf):cap(w.ibuf)], p)
+	w.ibuf = w.ibuf[:len(w.ibuf)+n]
+	nRet += n
+	return nRet, nil
+}
+
+func (w *Writer) write(p []byte) (nRet int, errRet error) {
+	if w.err != nil {
+		return 0, w.err
+	}
+	for len(p) > 0 {
+		obufStart := len(magicChunk)
+		if !w.wroteStreamHeader {
+			w.wroteStreamHeader = true
+			copy(w.obuf, magicChunk)
+			obufStart = 0
+		}
+
+		var uncompressed []byte
+		if len(p) > maxBlockSize {
+			uncompressed, p = p[:maxBlockSize], p[maxBlockSize:]
+		} else {
+			uncompressed, p = p, nil
+		}
+		checksum := crc(uncompressed)
+
+		// Compress the buffer, discarding the result if the improvement
+		// isn't at least 12.5%.
+		compressed := Encode(w.obuf[obufHeaderLen:], uncompressed)
+		chunkType := uint8(chunkTypeCompressedData)
+		chunkLen := 4 + len(compressed)
+		obufEnd := obufHeaderLen + len(compressed)
+		if len(compressed) >= len(uncompressed)-len(uncompressed)/8 {
+			chunkType = chunkTypeUncompressedData
+			chunkLen = 4 + len(uncompressed)
+			obufEnd = obufHeaderLen
+		}
+
+		// Fill in the per-chunk header that comes before the body.
+		w.obuf[len(magicChunk)+0] = chunkType
+		w.obuf[len(magicChunk)+1] = uint8(chunkLen >> 0)
+		w.obuf[len(magicChunk)+2] = uint8(chunkLen >> 8)
+		w.obuf[len(magicChunk)+3] = uint8(chunkLen >> 16)
+		w.obuf[len(magicChunk)+4] = uint8(checksum >> 0)
+		w.obuf[len(magicChunk)+5] = uint8(checksum >> 8)
+		w.obuf[len(magicChunk)+6] = uint8(checksum >> 16)
+		w.obuf[len(magicChunk)+7] = uint8(checksum >> 24)
+
+		if _, err := w.w.Write(w.obuf[obufStart:obufEnd]); err != nil {
+			w.err = err
+			return nRet, err
+		}
+		if chunkType == chunkTypeUncompressedData {
+			if _, err := w.w.Write(uncompressed); err != nil {
+				w.err = err
+				return nRet, err
+			}
+		}
+		nRet += len(uncompressed)
+	}
+	return nRet, nil
+}
+
+// Flush flushes the Writer to its underlying io.Writer.
+func (w *Writer) Flush() error {
+	if w.err != nil {
+		return w.err
+	}
+	if len(w.ibuf) == 0 {
+		return nil
+	}
+	w.write(w.ibuf)
+	w.ibuf = w.ibuf[:0]
+	return w.err
+}
+
+// Close calls Flush and then closes the Writer.
+func (w *Writer) Close() error {
+	w.Flush()
+	ret := w.err
+	if w.err == nil {
+		w.err = errClosed
+	}
+	return ret
+}

_vendor/vendor/github.com/golang/snappy/encode_amd64.go

@@ -0,0 +1,29 @@
+// Copyright 2016 The Snappy-Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// +build !appengine
+// +build gc
+// +build !noasm
+
+package snappy
+
+// emitLiteral has the same semantics as in encode_other.go.
+//
+//go:noescape
+func emitLiteral(dst, lit []byte) int
+
+// emitCopy has the same semantics as in encode_other.go.
+//
+//go:noescape
+func emitCopy(dst []byte, offset, length int) int
+
+// extendMatch has the same semantics as in encode_other.go.
+//
+//go:noescape
+func extendMatch(src []byte, i, j int) int
+
+// encodeBlock has the same semantics as in encode_other.go.
+//
+//go:noescape
+func encodeBlock(dst, src []byte) (d int)

_vendor/vendor/github.com/golang/snappy/encode_amd64.s

@@ -0,0 +1,730 @@
+// Copyright 2016 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// +build !appengine
+// +build gc
+// +build !noasm
+
+#include "textflag.h"
+
+// The XXX lines assemble on Go 1.4, 1.5 and 1.7, but not 1.6, due to a
+// Go toolchain regression. See https://github.com/golang/go/issues/15426 and
+// https://github.com/golang/snappy/issues/29
+//
+// As a workaround, the package was built with a known good assembler, and
+// those instructions were disassembled by "objdump -d" to yield the
+//	4e 0f b7 7c 5c 78       movzwq 0x78(%rsp,%r11,2),%r15
+// style comments, in AT&T asm syntax. Note that rsp here is a physical
+// register, not Go/asm's SP pseudo-register (see https://golang.org/doc/asm).
+// The instructions were then encoded as "BYTE $0x.." sequences, which assemble
+// fine on Go 1.6.
+
+// The asm code generally follows the pure Go code in encode_other.go, except
+// where marked with a "!!!".
+
+// ----------------------------------------------------------------------------
+
+// func emitLiteral(dst, lit []byte) int
+//
+// All local variables fit into registers. The register allocation:
+//	- AX	len(lit)
+//	- BX	n
+//	- DX	return value
+//	- DI	&dst[i]
+//	- R10	&lit[0]
+//
+// The 24 bytes of stack space is to call runtime·memmove.
+//
+// The unusual register allocation of local variables, such as R10 for the
+// source pointer, matches the allocation used at the call site in encodeBlock,
+// which makes it easier to manually inline this function.
+TEXT ·emitLiteral(SB), NOSPLIT, $24-56
+	MOVQ dst_base+0(FP), DI
+	MOVQ lit_base+24(FP), R10
+	MOVQ lit_len+32(FP), AX
+	MOVQ AX, DX
+	MOVL AX, BX
+	SUBL $1, BX
+
+	CMPL BX, $60
+	JLT  oneByte
+	CMPL BX, $256
+	JLT  twoBytes
+
+threeBytes:
+	MOVB $0xf4, 0(DI)
+	MOVW BX, 1(DI)
+	ADDQ $3, DI
+	ADDQ $3, DX
+	JMP  memmove
+
+twoBytes:
+	MOVB $0xf0, 0(DI)
+	MOVB BX, 1(DI)
+	ADDQ $2, DI
+	ADDQ $2, DX
+	JMP  memmove
+
+oneByte:
+	SHLB $2, BX
+	MOVB BX, 0(DI)
+	ADDQ $1, DI
+	ADDQ $1, DX
+
+memmove:
+	MOVQ DX, ret+48(FP)
+
+	// copy(dst[i:], lit)
+	//
+	// This means calling runtime·memmove(&dst[i], &lit[0], len(lit)), so we push
+	// DI, R10 and AX as arguments.
+	MOVQ DI, 0(SP)
+	MOVQ R10, 8(SP)
+	MOVQ AX, 16(SP)
+	CALL runtime·memmove(SB)
+	RET
+
+// ----------------------------------------------------------------------------
+
+// func emitCopy(dst []byte, offset, length int) int
+//
+// All local variables fit into registers. The register allocation:
+//	- AX	length
+//	- SI	&dst[0]
+//	- DI	&dst[i]
+//	- R11	offset
+//
+// The unusual register allocation of local variables, such as R11 for the
+// offset, matches the allocation used at the call site in encodeBlock, which
+// makes it easier to manually inline this function.
+TEXT ·emitCopy(SB), NOSPLIT, $0-48
+	MOVQ dst_base+0(FP), DI
+	MOVQ DI, SI
+	MOVQ offset+24(FP), R11
+	MOVQ length+32(FP), AX
+
+loop0:
+	// for length >= 68 { etc }
+	CMPL AX, $68
+	JLT  step1
+
+	// Emit a length 64 copy, encoded as 3 bytes.
+	MOVB $0xfe, 0(DI)
+	MOVW R11, 1(DI)
+	ADDQ $3, DI
+	SUBL $64, AX
+	JMP  loop0
+
+step1:
+	// if length > 64 { etc }
+	CMPL AX, $64
+	JLE  step2
+
+	// Emit a length 60 copy, encoded as 3 bytes.
+	MOVB $0xee, 0(DI)
+	MOVW R11, 1(DI)
+	ADDQ $3, DI
+	SUBL $60, AX
+
+step2:
+	// if length >= 12 || offset >= 2048 { goto step3 }
+	CMPL AX, $12
+	JGE  step3
+	CMPL R11, $2048
+	JGE  step3
+
+	// Emit the remaining copy, encoded as 2 bytes.
+	MOVB R11, 1(DI)
+	SHRL $8, R11
+	SHLB $5, R11
+	SUBB $4, AX
+	SHLB $2, AX
+	ORB  AX, R11
+	ORB  $1, R11
+	MOVB R11, 0(DI)
+	ADDQ $2, DI
+
+	// Return the number of bytes written.
+	SUBQ SI, DI
+	MOVQ DI, ret+40(FP)
+	RET
+
+step3:
+	// Emit the remaining copy, encoded as 3 bytes.
+	SUBL $1, AX
+	SHLB $2, AX
+	ORB  $2, AX
+	MOVB AX, 0(DI)
+	MOVW R11, 1(DI)
+	ADDQ $3, DI
+
+	// Return the number of bytes written.
+	SUBQ SI, DI
+	MOVQ DI, ret+40(FP)
+	RET
+
+// ----------------------------------------------------------------------------
+
+// func extendMatch(src []byte, i, j int) int
+//
+// All local variables fit into registers. The register allocation:
+//	- DX	&src[0]
+//	- SI	&src[j]
+//	- R13	&src[len(src) - 8]
+//	- R14	&src[len(src)]
+//	- R15	&src[i]
+//
+// The unusual register allocation of local variables, such as R15 for a source
+// pointer, matches the allocation used at the call site in encodeBlock, which
+// makes it easier to manually inline this function.
+TEXT ·extendMatch(SB), NOSPLIT, $0-48
+	MOVQ src_base+0(FP), DX
+	MOVQ src_len+8(FP), R14
+	MOVQ i+24(FP), R15
+	MOVQ j+32(FP), SI
+	ADDQ DX, R14
+	ADDQ DX, R15
+	ADDQ DX, SI
+	MOVQ R14, R13
+	SUBQ $8, R13
+
+cmp8:
+	// As long as we are 8 or more bytes before the end of src, we can load and
+	// compare 8 bytes at a time. If those 8 bytes are equal, repeat.
+	CMPQ SI, R13
+	JA   cmp1
+	MOVQ (R15), AX
+	MOVQ (SI), BX
+	CMPQ AX, BX
+	JNE  bsf
+	ADDQ $8, R15
+	ADDQ $8, SI
+	JMP  cmp8
+
+bsf:
+	// If those 8 bytes were not equal, XOR the two 8 byte values, and return
+	// the index of the first byte that differs. The BSF instruction finds the
+	// least significant 1 bit, the amd64 architecture is little-endian, and
+	// the shift by 3 converts a bit index to a byte index.
+	XORQ AX, BX
+	BSFQ BX, BX
+	SHRQ $3, BX
+	ADDQ BX, SI
+
+	// Convert from &src[ret] to ret.
+	SUBQ DX, SI
+	MOVQ SI, ret+40(FP)
+	RET
+
+cmp1:
+	// In src's tail, compare 1 byte at a time.
+	CMPQ SI, R14
+	JAE  extendMatchEnd
+	MOVB (R15), AX
+	MOVB (SI), BX
+	CMPB AX, BX
+	JNE  extendMatchEnd
+	ADDQ $1, R15
+	ADDQ $1, SI
+	JMP  cmp1
+
+extendMatchEnd:
+	// Convert from &src[ret] to ret.
+	SUBQ DX, SI
+	MOVQ SI, ret+40(FP)
+	RET
+
+// ----------------------------------------------------------------------------
+
+// func encodeBlock(dst, src []byte) (d int)
+//
+// All local variables fit into registers, other than "var table". The register
+// allocation:
+//	- AX	.	.
+//	- BX	.	.
+//	- CX	56	shift (note that amd64 shifts by non-immediates must use CX).
+//	- DX	64	&src[0], tableSize
+//	- SI	72	&src[s]
+//	- DI	80	&dst[d]
+//	- R9	88	sLimit
+//	- R10	.	&src[nextEmit]
+//	- R11	96	prevHash, currHash, nextHash, offset
+//	- R12	104	&src[base], skip
+//	- R13	.	&src[nextS], &src[len(src) - 8]
+//	- R14	.	len(src), bytesBetweenHashLookups, &src[len(src)], x
+//	- R15	112	candidate
+//
+// The second column (56, 64, etc) is the stack offset to spill the registers
+// when calling other functions. We could pack this slightly tighter, but it's
+// simpler to have a dedicated spill map independent of the function called.
+//
+// "var table [maxTableSize]uint16" takes up 32768 bytes of stack space. An
+// extra 56 bytes, to call other functions, and an extra 64 bytes, to spill
+// local variables (registers) during calls gives 32768 + 56 + 64 = 32888.
+TEXT ·encodeBlock(SB), 0, $32888-56
+	MOVQ dst_base+0(FP), DI
+	MOVQ src_base+24(FP), SI
+	MOVQ src_len+32(FP), R14
+
+	// shift, tableSize := uint32(32-8), 1<<8
+	MOVQ $24, CX
+	MOVQ $256, DX
+
+calcShift:
+	// for ; tableSize < maxTableSize && tableSize < len(src); tableSize *= 2 {
+	//	shift--
+	// }
+	CMPQ DX, $16384
+	JGE  varTable
+	CMPQ DX, R14
+	JGE  varTable
+	SUBQ $1, CX
+	SHLQ $1, DX
+	JMP  calcShift
+
+varTable:
+	// var table [maxTableSize]uint16
+	//
+	// In the asm code, unlike the Go code, we can zero-initialize only the
+	// first tableSize elements. Each uint16 element is 2 bytes and each MOVOU
+	// writes 16 bytes, so we can do only tableSize/8 writes instead of the
+	// 2048 writes that would zero-initialize all of table's 32768 bytes.
+	SHRQ $3, DX
+	LEAQ table-32768(SP), BX
+	PXOR X0, X0
+
+memclr:
+	MOVOU X0, 0(BX)
+	ADDQ  $16, BX
+	SUBQ  $1, DX
+	JNZ   memclr
+
+	// !!! DX = &src[0]
+	MOVQ SI, DX
+
+	// sLimit := len(src) - inputMargin
+	MOVQ R14, R9
+	SUBQ $15, R9
+
+	// !!! Pre-emptively spill CX, DX and R9 to the stack. Their values don't
+	// change for the rest of the function.
+	MOVQ CX, 56(SP)
+	MOVQ DX, 64(SP)
+	MOVQ R9, 88(SP)
+
+	// nextEmit := 0
+	MOVQ DX, R10
+
+	// s := 1
+	ADDQ $1, SI
+
+	// nextHash := hash(load32(src, s), shift)
+	MOVL  0(SI), R11
+	IMULL $0x1e35a7bd, R11
+	SHRL  CX, R11
+
+outer:
+	// for { etc }
+
+	// skip := 32
+	MOVQ $32, R12
+
+	// nextS := s
+	MOVQ SI, R13
+
+	// candidate := 0
+	MOVQ $0, R15
+
+inner0:
+	// for { etc }
+
+	// s := nextS
+	MOVQ R13, SI
+
+	// bytesBetweenHashLookups := skip >> 5
+	MOVQ R12, R14
+	SHRQ $5, R14
+
+	// nextS = s + bytesBetweenHashLookups
+	ADDQ R14, R13
+
+	// skip += bytesBetweenHashLookups
+	ADDQ R14, R12
+
+	// if nextS > sLimit { goto emitRemainder }
+	MOVQ R13, AX
+	SUBQ DX, AX
+	CMPQ AX, R9
+	JA   emitRemainder
+
+	// candidate = int(table[nextHash])
+	// XXX: MOVWQZX table-32768(SP)(R11*2), R15
+	// XXX: 4e 0f b7 7c 5c 78       movzwq 0x78(%rsp,%r11,2),%r15
+	BYTE $0x4e
+	BYTE $0x0f
+	BYTE $0xb7
+	BYTE $0x7c
+	BYTE $0x5c
+	BYTE $0x78
+
+	// table[nextHash] = uint16(s)
+	MOVQ SI, AX
+	SUBQ DX, AX
+
+	// XXX: MOVW AX, table-32768(SP)(R11*2)
+	// XXX: 66 42 89 44 5c 78       mov    %ax,0x78(%rsp,%r11,2)
+	BYTE $0x66
+	BYTE $0x42
+	BYTE $0x89
+	BYTE $0x44
+	BYTE $0x5c
+	BYTE $0x78
+
+	// nextHash = hash(load32(src, nextS), shift)
+	MOVL  0(R13), R11
+	IMULL $0x1e35a7bd, R11
+	SHRL  CX, R11
+
+	// if load32(src, s) != load32(src, candidate) { continue } break
+	MOVL 0(SI), AX
+	MOVL (DX)(R15*1), BX
+	CMPL AX, BX
+	JNE  inner0
+
+fourByteMatch:
+	// As per the encode_other.go code:
+	//
+	// A 4-byte match has been found. We'll later see etc.
+
+	// !!! Jump to a fast path for short (<= 16 byte) literals. See the comment
+	// on inputMargin in encode.go.
+	MOVQ SI, AX
+	SUBQ R10, AX
+	CMPQ AX, $16
+	JLE  emitLiteralFastPath
+
+	// ----------------------------------------
+	// Begin inline of the emitLiteral call.
+	//
+	// d += emitLiteral(dst[d:], src[nextEmit:s])
+
+	MOVL AX, BX
+	SUBL $1, BX
+
+	CMPL BX, $60
+	JLT  inlineEmitLiteralOneByte
+	CMPL BX, $256
+	JLT  inlineEmitLiteralTwoBytes
+
+inlineEmitLiteralThreeBytes:
+	MOVB $0xf4, 0(DI)
+	MOVW BX, 1(DI)
+	ADDQ $3, DI
+	JMP  inlineEmitLiteralMemmove
+
+inlineEmitLiteralTwoBytes:
+	MOVB $0xf0, 0(DI)
+	MOVB BX, 1(DI)
+	ADDQ $2, DI
+	JMP  inlineEmitLiteralMemmove
+
+inlineEmitLiteralOneByte:
+	SHLB $2, BX
+	MOVB BX, 0(DI)
+	ADDQ $1, DI
+
+inlineEmitLiteralMemmove:
+	// Spill local variables (registers) onto the stack; call; unspill.
+	//
+	// copy(dst[i:], lit)
+	//
+	// This means calling runtime·memmove(&dst[i], &lit[0], len(lit)), so we push
+	// DI, R10 and AX as arguments.
+	MOVQ DI, 0(SP)
+	MOVQ R10, 8(SP)
+	MOVQ AX, 16(SP)
+	ADDQ AX, DI              // Finish the "d +=" part of "d += emitLiteral(etc)".
+	MOVQ SI, 72(SP)
+	MOVQ DI, 80(SP)
+	MOVQ R15, 112(SP)
+	CALL runtime·memmove(SB)
+	MOVQ 56(SP), CX
+	MOVQ 64(SP), DX
+	MOVQ 72(SP), SI
+	MOVQ 80(SP), DI
+	MOVQ 88(SP), R9
+	MOVQ 112(SP), R15
+	JMP  inner1
+
+inlineEmitLiteralEnd:
+	// End inline of the emitLiteral call.
+	// ----------------------------------------
+
+emitLiteralFastPath:
+	// !!! Emit the 1-byte encoding "uint8(len(lit)-1)<<2".
+	MOVB AX, BX
+	SUBB $1, BX
+	SHLB $2, BX
+	MOVB BX, (DI)
+	ADDQ $1, DI
+
+	// !!! Implement the copy from lit to dst as a 16-byte load and store.
+	// (Encode's documentation says that dst and src must not overlap.)
+	//
+	// This always copies 16 bytes, instead of only len(lit) bytes, but that's
+	// OK. Subsequent iterations will fix up the overrun.
+	//
+	// Note that on amd64, it is legal and cheap to issue unaligned 8-byte or
+	// 16-byte loads and stores. This technique probably wouldn't be as
+	// effective on architectures that are fussier about alignment.
+	MOVOU 0(R10), X0
+	MOVOU X0, 0(DI)
+	ADDQ  AX, DI
+
+inner1:
+	// for { etc }
+
+	// base := s
+	MOVQ SI, R12
+
+	// !!! offset := base - candidate
+	MOVQ R12, R11
+	SUBQ R15, R11
+	SUBQ DX, R11
+
+	// ----------------------------------------
+	// Begin inline of the extendMatch call.
+	//
+	// s = extendMatch(src, candidate+4, s+4)
+
+	// !!! R14 = &src[len(src)]
+	MOVQ src_len+32(FP), R14
+	ADDQ DX, R14
+
+	// !!! R13 = &src[len(src) - 8]
+	MOVQ R14, R13
+	SUBQ $8, R13
+
+	// !!! R15 = &src[candidate + 4]
+	ADDQ $4, R15
+	ADDQ DX, R15
+
+	// !!! s += 4
+	ADDQ $4, SI
+
+inlineExtendMatchCmp8:
+	// As long as we are 8 or more bytes before the end of src, we can load and
+	// compare 8 bytes at a time. If those 8 bytes are equal, repeat.
+	CMPQ SI, R13
+	JA   inlineExtendMatchCmp1
+	MOVQ (R15), AX
+	MOVQ (SI), BX
+	CMPQ AX, BX
+	JNE  inlineExtendMatchBSF
+	ADDQ $8, R15
+	ADDQ $8, SI
+	JMP  inlineExtendMatchCmp8
+
+inlineExtendMatchBSF:
+	// If those 8 bytes were not equal, XOR the two 8 byte values, and return
+	// the index of the first byte that differs. The BSF instruction finds the
+	// least significant 1 bit, the amd64 architecture is little-endian, and
+	// the shift by 3 converts a bit index to a byte index.
+	XORQ AX, BX
+	BSFQ BX, BX
+	SHRQ $3, BX
+	ADDQ BX, SI
+	JMP  inlineExtendMatchEnd
+
+inlineExtendMatchCmp1:
+	// In src's tail, compare 1 byte at a time.
+	CMPQ SI, R14
+	JAE  inlineExtendMatchEnd
+	MOVB (R15), AX
+	MOVB (SI), BX
+	CMPB AX, BX
+	JNE  inlineExtendMatchEnd
+	ADDQ $1, R15
+	ADDQ $1, SI
+	JMP  inlineExtendMatchCmp1
+
+inlineExtendMatchEnd:
+	// End inline of the extendMatch call.
+	// ----------------------------------------
+
+	// ----------------------------------------
+	// Begin inline of the emitCopy call.
+	//
+	// d += emitCopy(dst[d:], base-candidate, s-base)
+
+	// !!! length := s - base
+	MOVQ SI, AX
+	SUBQ R12, AX
+
+inlineEmitCopyLoop0:
+	// for length >= 68 { etc }
+	CMPL AX, $68
+	JLT  inlineEmitCopyStep1
+
+	// Emit a length 64 copy, encoded as 3 bytes.
+	MOVB $0xfe, 0(DI)
+	MOVW R11, 1(DI)
+	ADDQ $3, DI
+	SUBL $64, AX
+	JMP  inlineEmitCopyLoop0
+
+inlineEmitCopyStep1:
+	// if length > 64 { etc }
+	CMPL AX, $64
+	JLE  inlineEmitCopyStep2
+
+	// Emit a length 60 copy, encoded as 3 bytes.
+	MOVB $0xee, 0(DI)
+	MOVW R11, 1(DI)
+	ADDQ $3, DI
+	SUBL $60, AX
+
+inlineEmitCopyStep2:
+	// if length >= 12 || offset >= 2048 { goto inlineEmitCopyStep3 }
+	CMPL AX, $12
+	JGE  inlineEmitCopyStep3
+	CMPL R11, $2048
+	JGE  inlineEmitCopyStep3
+
+	// Emit the remaining copy, encoded as 2 bytes.
+	MOVB R11, 1(DI)
+	SHRL $8, R11
+	SHLB $5, R11
+	SUBB $4, AX
+	SHLB $2, AX
+	ORB  AX, R11
+	ORB  $1, R11
+	MOVB R11, 0(DI)
+	ADDQ $2, DI
+	JMP  inlineEmitCopyEnd
+
+inlineEmitCopyStep3:
+	// Emit the remaining copy, encoded as 3 bytes.
+	SUBL $1, AX
+	SHLB $2, AX
+	ORB  $2, AX
+	MOVB AX, 0(DI)
+	MOVW R11, 1(DI)
+	ADDQ $3, DI
+
+inlineEmitCopyEnd:
+	// End inline of the emitCopy call.
+	// ----------------------------------------
+
+	// nextEmit = s
+	MOVQ SI, R10
+
+	// if s >= sLimit { goto emitRemainder }
+	MOVQ SI, AX
+	SUBQ DX, AX
+	CMPQ AX, R9
+	JAE  emitRemainder
+
+	// As per the encode_other.go code:
+	//
+	// We could immediately etc.
+
+	// x := load64(src, s-1)
+	MOVQ -1(SI), R14
+
+	// prevHash := hash(uint32(x>>0), shift)
+	MOVL  R14, R11
+	IMULL $0x1e35a7bd, R11
+	SHRL  CX, R11
+
+	// table[prevHash] = uint16(s-1)
+	MOVQ SI, AX
+	SUBQ DX, AX
+	SUBQ $1, AX
+
+	// XXX: MOVW AX, table-32768(SP)(R11*2)
+	// XXX: 66 42 89 44 5c 78       mov    %ax,0x78(%rsp,%r11,2)
+	BYTE $0x66
+	BYTE $0x42
+	BYTE $0x89
+	BYTE $0x44
+	BYTE $0x5c
+	BYTE $0x78
+
+	// currHash := hash(uint32(x>>8), shift)
+	SHRQ  $8, R14
+	MOVL  R14, R11
+	IMULL $0x1e35a7bd, R11
+	SHRL  CX, R11
+
+	// candidate = int(table[currHash])
+	// XXX: MOVWQZX table-32768(SP)(R11*2), R15
+	// XXX: 4e 0f b7 7c 5c 78       movzwq 0x78(%rsp,%r11,2),%r15
+	BYTE $0x4e
+	BYTE $0x0f
+	BYTE $0xb7
+	BYTE $0x7c
+	BYTE $0x5c
+	BYTE $0x78
+
+	// table[currHash] = uint16(s)
+	ADDQ $1, AX
+
+	// XXX: MOVW AX, table-32768(SP)(R11*2)
+	// XXX: 66 42 89 44 5c 78       mov    %ax,0x78(%rsp,%r11,2)
+	BYTE $0x66
+	BYTE $0x42
+	BYTE $0x89
+	BYTE $0x44
+	BYTE $0x5c
+	BYTE $0x78
+
+	// if uint32(x>>8) == load32(src, candidate) { continue }
+	MOVL (DX)(R15*1), BX
+	CMPL R14, BX
+	JEQ  inner1
+
+	// nextHash = hash(uint32(x>>16), shift)
+	SHRQ  $8, R14
+	MOVL  R14, R11
+	IMULL $0x1e35a7bd, R11
+	SHRL  CX, R11
+
+	// s++
+	ADDQ $1, SI
+
+	// break out of the inner1 for loop, i.e. continue the outer loop.
+	JMP outer
+
+emitRemainder:
+	// if nextEmit < len(src) { etc }
+	MOVQ src_len+32(FP), AX
+	ADDQ DX, AX
+	CMPQ R10, AX
+	JEQ  encodeBlockEnd
+
+	// d += emitLiteral(dst[d:], src[nextEmit:])
+	//
+	// Push args.
+	MOVQ DI, 0(SP)
+	MOVQ $0, 8(SP)   // Unnecessary, as the callee ignores it, but conservative.
+	MOVQ $0, 16(SP)  // Unnecessary, as the callee ignores it, but conservative.
+	MOVQ R10, 24(SP)
+	SUBQ R10, AX
+	MOVQ AX, 32(SP)
+	MOVQ AX, 40(SP)  // Unnecessary, as the callee ignores it, but conservative.
+
+	// Spill local variables (registers) onto the stack; call; unspill.
+	MOVQ DI, 80(SP)
+	CALL ·emitLiteral(SB)
+	MOVQ 80(SP), DI
+
+	// Finish the "d +=" part of "d += emitLiteral(etc)".
+	ADDQ 48(SP), DI
+
+encodeBlockEnd:
+	MOVQ dst_base+0(FP), AX
+	SUBQ AX, DI
+	MOVQ DI, d+48(FP)
+	RET

_vendor/vendor/github.com/golang/snappy/encode_other.go

@@ -0,0 +1,238 @@
+// Copyright 2016 The Snappy-Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// +build !amd64 appengine !gc noasm
+
+package snappy
+
+func load32(b []byte, i int) uint32 {
+	b = b[i : i+4 : len(b)] // Help the compiler eliminate bounds checks on the next line.
+	return uint32(b[0]) | uint32(b[1])<<8 | uint32(b[2])<<16 | uint32(b[3])<<24
+}
+
+func load64(b []byte, i int) uint64 {
+	b = b[i : i+8 : len(b)] // Help the compiler eliminate bounds checks on the next line.
+	return uint64(b[0]) | uint64(b[1])<<8 | uint64(b[2])<<16 | uint64(b[3])<<24 |
+		uint64(b[4])<<32 | uint64(b[5])<<40 | uint64(b[6])<<48 | uint64(b[7])<<56
+}
+
+// emitLiteral writes a literal chunk and returns the number of bytes written.
+//
+// It assumes that:
+//	dst is long enough to hold the encoded bytes
+//	1 <= len(lit) && len(lit) <= 65536
+func emitLiteral(dst, lit []byte) int {
+	i, n := 0, uint(len(lit)-1)
+	switch {
+	case n < 60:
+		dst[0] = uint8(n)<<2 | tagLiteral
+		i = 1
+	case n < 1<<8:
+		dst[0] = 60<<2 | tagLiteral
+		dst[1] = uint8(n)
+		i = 2
+	default:
+		dst[0] = 61<<2 | tagLiteral
+		dst[1] = uint8(n)
+		dst[2] = uint8(n >> 8)
+		i = 3
+	}
+	return i + copy(dst[i:], lit)
+}
+
+// emitCopy writes a copy chunk and returns the number of bytes written.
+//
+// It assumes that:
+//	dst is long enough to hold the encoded bytes
+//	1 <= offset && offset <= 65535
+//	4 <= length && length <= 65535
+func emitCopy(dst []byte, offset, length int) int {
+	i := 0
+	// The maximum length for a single tagCopy1 or tagCopy2 op is 64 bytes. The
+	// threshold for this loop is a little higher (at 68 = 64 + 4), and the
+	// length emitted down below is is a little lower (at 60 = 64 - 4), because
+	// it's shorter to encode a length 67 copy as a length 60 tagCopy2 followed
+	// by a length 7 tagCopy1 (which encodes as 3+2 bytes) than to encode it as
+	// a length 64 tagCopy2 followed by a length 3 tagCopy2 (which encodes as
+	// 3+3 bytes). The magic 4 in the 64±4 is because the minimum length for a
+	// tagCopy1 op is 4 bytes, which is why a length 3 copy has to be an
+	// encodes-as-3-bytes tagCopy2 instead of an encodes-as-2-bytes tagCopy1.
+	for length >= 68 {
+		// Emit a length 64 copy, encoded as 3 bytes.
+		dst[i+0] = 63<<2 | tagCopy2
+		dst[i+1] = uint8(offset)
+		dst[i+2] = uint8(offset >> 8)
+		i += 3
+		length -= 64
+	}
+	if length > 64 {
+		// Emit a length 60 copy, encoded as 3 bytes.
+		dst[i+0] = 59<<2 | tagCopy2
+		dst[i+1] = uint8(offset)
+		dst[i+2] = uint8(offset >> 8)
+		i += 3
+		length -= 60
+	}
+	if length >= 12 || offset >= 2048 {
+		// Emit the remaining copy, encoded as 3 bytes.
+		dst[i+0] = uint8(length-1)<<2 | tagCopy2
+		dst[i+1] = uint8(offset)
+		dst[i+2] = uint8(offset >> 8)
+		return i + 3
+	}
+	// Emit the remaining copy, encoded as 2 bytes.
+	dst[i+0] = uint8(offset>>8)<<5 | uint8(length-4)<<2 | tagCopy1
+	dst[i+1] = uint8(offset)
+	return i + 2
+}
+
+// extendMatch returns the largest k such that k <= len(src) and that
+// src[i:i+k-j] and src[j:k] have the same contents.
+//
+// It assumes that:
+//	0 <= i && i < j && j <= len(src)
+func extendMatch(src []byte, i, j int) int {
+	for ; j < len(src) && src[i] == src[j]; i, j = i+1, j+1 {
+	}
+	return j
+}
+
+func hash(u, shift uint32) uint32 {
+	return (u * 0x1e35a7bd) >> shift
+}
+
+// encodeBlock encodes a non-empty src to a guaranteed-large-enough dst. It
+// assumes that the varint-encoded length of the decompressed bytes has already
+// been written.
+//
+// It also assumes that:
+//	len(dst) >= MaxEncodedLen(len(src)) &&
+// 	minNonLiteralBlockSize <= len(src) && len(src) <= maxBlockSize
+func encodeBlock(dst, src []byte) (d int) {
+	// Initialize the hash table. Its size ranges from 1<<8 to 1<<14 inclusive.
+	// The table element type is uint16, as s < sLimit and sLimit < len(src)
+	// and len(src) <= maxBlockSize and maxBlockSize == 65536.
+	const (
+		maxTableSize = 1 << 14
+		// tableMask is redundant, but helps the compiler eliminate bounds
+		// checks.
+		tableMask = maxTableSize - 1
+	)
+	shift := uint32(32 - 8)
+	for tableSize := 1 << 8; tableSize < maxTableSize && tableSize < len(src); tableSize *= 2 {
+		shift--
+	}
+	// In Go, all array elements are zero-initialized, so there is no advantage
+	// to a smaller tableSize per se. However, it matches the C++ algorithm,
+	// and in the asm versions of this code, we can get away with zeroing only
+	// the first tableSize elements.
+	var table [maxTableSize]uint16
+
+	// sLimit is when to stop looking for offset/length copies. The inputMargin
+	// lets us use a fast path for emitLiteral in the main loop, while we are
+	// looking for copies.
+	sLimit := len(src) - inputMargin
+
+	// nextEmit is where in src the next emitLiteral should start from.
+	nextEmit := 0
+
+	// The encoded form must start with a literal, as there are no previous
+	// bytes to copy, so we start looking for hash matches at s == 1.
+	s := 1
+	nextHash := hash(load32(src, s), shift)
+
+	for {
+		// Copied from the C++ snappy implementation:
+		//
+		// Heuristic match skipping: If 32 bytes are scanned with no matches
+		// found, start looking only at every other byte. If 32 more bytes are
+		// scanned (or skipped), look at every third byte, etc.. When a match
+		// is found, immediately go back to looking at every byte. This is a
+		// small loss (~5% performance, ~0.1% density) for compressible data
+		// due to more bookkeeping, but for non-compressible data (such as
+		// JPEG) it's a huge win since the compressor quickly "realizes" the
+		// data is incompressible and doesn't bother looking for matches
+		// everywhere.
+		//
+		// The "skip" variable keeps track of how many bytes there are since
+		// the last match; dividing it by 32 (ie. right-shifting by five) gives
+		// the number of bytes to move ahead for each iteration.
+		skip := 32
+
+		nextS := s
+		candidate := 0
+		for {
+			s = nextS
+			bytesBetweenHashLookups := skip >> 5
+			nextS = s + bytesBetweenHashLookups
+			skip += bytesBetweenHashLookups
+			if nextS > sLimit {
+				goto emitRemainder
+			}
+			candidate = int(table[nextHash&tableMask])
+			table[nextHash&tableMask] = uint16(s)
+			nextHash = hash(load32(src, nextS), shift)
+			if load32(src, s) == load32(src, candidate) {
+				break
+			}
+		}
+
+		// A 4-byte match has been found. We'll later see if more than 4 bytes
+		// match. But, prior to the match, src[nextEmit:s] are unmatched. Emit
+		// them as literal bytes.
+		d += emitLiteral(dst[d:], src[nextEmit:s])
+
+		// Call emitCopy, and then see if another emitCopy could be our next
+		// move. Repeat until we find no match for the input immediately after
+		// what was consumed by the last emitCopy call.
+		//
+		// If we exit this loop normally then we need to call emitLiteral next,
+		// though we don't yet know how big the literal will be. We handle that
+		// by proceeding to the next iteration of the main loop. We also can
+		// exit this loop via goto if we get close to exhausting the input.
+		for {
+			// Invariant: we have a 4-byte match at s, and no need to emit any
+			// literal bytes prior to s.
+			base := s
+
+			// Extend the 4-byte match as long as possible.
+			//
+			// This is an inlined version of:
+			//	s = extendMatch(src, candidate+4, s+4)
+			s += 4
+			for i := candidate + 4; s < len(src) && src[i] == src[s]; i, s = i+1, s+1 {
+			}
+
+			d += emitCopy(dst[d:], base-candidate, s-base)
+			nextEmit = s
+			if s >= sLimit {
+				goto emitRemainder
+			}
+
+			// We could immediately start working at s now, but to improve
+			// compression we first update the hash table at s-1 and at s. If
+			// another emitCopy is not our next move, also calculate nextHash
+			// at s+1. At least on GOARCH=amd64, these three hash calculations
+			// are faster as one load64 call (with some shifts) instead of
+			// three load32 calls.
+			x := load64(src, s-1)
+			prevHash := hash(uint32(x>>0), shift)
+			table[prevHash&tableMask] = uint16(s - 1)
+			currHash := hash(uint32(x>>8), shift)
+			candidate = int(table[currHash&tableMask])
+			table[currHash&tableMask] = uint16(s)
+			if uint32(x>>8) != load32(src, candidate) {
+				nextHash = hash(uint32(x>>16), shift)
+				s++
+				break
+			}
+		}
+	}
+
+emitRemainder:
+	if nextEmit < len(src) {
+		d += emitLiteral(dst[d:], src[nextEmit:])
+	}
+	return d
+}

_vendor/vendor/github.com/golang/snappy/snappy.go

@@ -6,7 +6,7 @@
 // It aims for very high speeds and reasonable compression.
 //
 // The C++ snappy implementation is at https://github.com/google/snappy
-package snappy
+package snappy // import "github.com/golang/snappy"
 
 import (
 	"hash/crc32"
@@ -32,7 +32,10 @@ Lempel-Ziv compression algorithms. In particular:
   - For l == 2, the offset ranges in [0, 1<<16) and the length in [1, 65).
     The length is 1 + m. The offset is the little-endian unsigned integer
     denoted by the next 2 bytes.
-  - For l == 3, this tag is a legacy format that is no longer supported.
+  - For l == 3, this tag is a legacy format that is no longer issued by most
+    encoders. Nonetheless, the offset ranges in [0, 1<<32) and the length in
+    [1, 65). The length is 1 + m. The offset is the little-endian unsigned
+    integer denoted by the next 4 bytes.
 */
 const (
 	tagLiteral = 0x00
@@ -46,9 +49,25 @@ const (
 	chunkHeaderSize = 4
 	magicChunk      = "\xff\x06\x00\x00" + magicBody
 	magicBody       = "sNaPpY"
+
+	// maxBlockSize is the maximum size of the input to encodeBlock. It is not
+	// part of the wire format per se, but some parts of the encoder assume
+	// that an offset fits into a uint16.
+	//
+	// Also, for the framing format (Writer type instead of Encode function),
 	// https://github.com/google/snappy/blob/master/framing_format.txt says
-	// that "the uncompressed data in a chunk must be no longer than 65536 bytes".
-	maxUncompressedChunkLen = 65536
+	// that "the uncompressed data in a chunk must be no longer than 65536
+	// bytes".
+	maxBlockSize = 65536
+
+	// maxEncodedLenOfMaxBlockSize equals MaxEncodedLen(maxBlockSize), but is
+	// hard coded to be a const instead of a variable, so that obufLen can also
+	// be a const. Their equivalence is confirmed by
+	// TestMaxEncodedLenOfMaxBlockSize.
+	maxEncodedLenOfMaxBlockSize = 76490
+
+	obufHeaderLen = len(magicChunk) + checksumSize + chunkHeaderSize
+	obufLen       = obufHeaderLen + maxEncodedLenOfMaxBlockSize
 )
 
 const (

_vendor/vendor/github.com/peterh/liner/bsdinput.go

@@ -37,3 +37,5 @@ type termios struct {
 	Ispeed int32
 	Ospeed int32
 }
+
+const cursorColumn = false

_vendor/vendor/github.com/peterh/liner/common.go

@@ -7,7 +7,6 @@ package liner
 
 import (
 	"bufio"
-	"bytes"
 	"container/ring"
 	"errors"
 	"fmt"
@@ -32,6 +31,8 @@ type commonState struct {
 	multiLineMode     bool
 	cursorRows        int
 	maxRows           int
+	shouldRestart     ShouldRestart
+	needRefresh       bool
 }
 
 // TabStyle is used to select how tab completions are displayed.
@@ -58,7 +59,12 @@ var ErrPromptAborted = errors.New("prompt aborted")
 // platform is normally supported, but stdout has been redirected
 var ErrNotTerminalOutput = errors.New("standard output is not a terminal")
 
-// Max elements to save on the killring
+// ErrInvalidPrompt is returned from Prompt or PasswordPrompt if the
+// prompt contains any unprintable runes (including substrings that could
+// be colour codes on some platforms).
+var ErrInvalidPrompt = errors.New("invalid prompt")
+
+// KillRingMax is the max number of elements to save on the killring.
 const KillRingMax = 60
 
 // HistoryLimit is the maximum number of entries saved in the scrollback history.
@@ -133,6 +139,13 @@ func (s *State) AppendHistory(item string) {
 	}
 }
 
+// ClearHistory clears the scroollback history.
+func (s *State) ClearHistory() {
+	s.historyMutex.Lock()
+	defer s.historyMutex.Unlock()
+	s.history = nil
+}
+
 // Returns the history lines starting with prefix
 func (s *State) getHistoryByPrefix(prefix string) (ph []string) {
 	for _, h := range s.history {
@@ -215,6 +228,16 @@ func (s *State) SetMultiLineMode(mlmode bool) {
 	s.multiLineMode = mlmode
 }
 
+// ShouldRestart is passed the error generated by readNext and returns true if
+// the the read should be restarted or false if the error should be returned.
+type ShouldRestart func(err error) bool
+
+// SetShouldRestart sets the restart function that Liner will call to determine
+// whether to retry the call to, or return the error returned by, readNext.
+func (s *State) SetShouldRestart(f ShouldRestart) {
+	s.shouldRestart = f
+}
+
 func (s *State) promptUnsupported(p string) (string, error) {
 	if !s.inputRedirected || !s.terminalSupported {
 		fmt.Print(p)
@@ -223,5 +246,5 @@ func (s *State) promptUnsupported(p string) (string, error) {
 	if err != nil {
 		return "", err
 	}
-	return string(bytes.TrimSpace(linebuf)), nil
+	return string(linebuf), nil
 }

_vendor/vendor/github.com/peterh/liner/input.go

@@ -31,11 +31,6 @@ type State struct {
 
 // NewLiner initializes a new *State, and sets the terminal into raw mode. To
 // restore the terminal to its previous state, call State.Close().
-//
-// Note if you are still using Go 1.0: NewLiner handles SIGWINCH, so it will
-// leak a channel every time you call it. Therefore, it is recommened that you
-// upgrade to a newer release of Go, or ensure that NewLiner is only called
-// once.
 func NewLiner() *State {
 	var s State
 	s.r = bufio.NewReader(os.Stdin)
@@ -67,8 +62,7 @@ func NewLiner() *State {
 	}
 
 	if !s.outputRedirected {
-		s.getColumns()
-		s.outputRedirected = s.columns <= 0
+		s.outputRedirected = !s.getColumns()
 	}
 
 	return &s
@@ -88,8 +82,12 @@ func (s *State) startPrompt() {
 	s.restartPrompt()
 }
 
+func (s *State) inputWaiting() bool {
+	return len(s.next) > 0
+}
+
 func (s *State) restartPrompt() {
-	next := make(chan nexter)
+	next := make(chan nexter, 200)
 	go func() {
 		for {
 			var n nexter
@@ -127,8 +125,6 @@ func (s *State) nextPending(timeout <-chan time.Time) (rune, error) {
 		s.pending = s.pending[1:]
 		return rv, errTimedOut
 	}
-	// not reached
-	return 0, nil
 }
 
 func (s *State) readNext() (interface{}, error) {
@@ -350,7 +346,7 @@ func (s *State) readNext() (interface{}, error) {
 
 // Close returns the terminal to its previous mode
 func (s *State) Close() error {
-	stopSignal(s.winch)
+	signal.Stop(s.winch)
 	if !s.inputRedirected {
 		s.origMode.ApplyMode()
 	}

_vendor/vendor/github.com/peterh/liner/input_darwin.go

@@ -37,3 +37,7 @@ type termios struct {
 	Ispeed uintptr
 	Ospeed uintptr
 }
+
+// Terminal.app needs a column for the cursor when the input line is at the
+// bottom of the window.
+const cursorColumn = true

_vendor/vendor/github.com/peterh/liner/input_linux.go

@@ -24,3 +24,5 @@ const (
 type termios struct {
 	syscall.Termios
 }
+
+const cursorColumn = false

_vendor/vendor/github.com/peterh/liner/input_windows.go

@@ -10,13 +10,14 @@ import (
 var (
 	kernel32 = syscall.NewLazyDLL("kernel32.dll")
 
-	procGetStdHandle               = kernel32.NewProc("GetStdHandle")
-	procReadConsoleInput           = kernel32.NewProc("ReadConsoleInputW")
-	procGetConsoleMode             = kernel32.NewProc("GetConsoleMode")
-	procSetConsoleMode             = kernel32.NewProc("SetConsoleMode")
-	procSetConsoleCursorPosition   = kernel32.NewProc("SetConsoleCursorPosition")
-	procGetConsoleScreenBufferInfo = kernel32.NewProc("GetConsoleScreenBufferInfo")
-	procFillConsoleOutputCharacter = kernel32.NewProc("FillConsoleOutputCharacterW")
+	procGetStdHandle                  = kernel32.NewProc("GetStdHandle")
+	procReadConsoleInput              = kernel32.NewProc("ReadConsoleInputW")
+	procGetNumberOfConsoleInputEvents = kernel32.NewProc("GetNumberOfConsoleInputEvents")
+	procGetConsoleMode                = kernel32.NewProc("GetConsoleMode")
+	procSetConsoleMode                = kernel32.NewProc("SetConsoleMode")
+	procSetConsoleCursorPosition      = kernel32.NewProc("SetConsoleCursorPosition")
+	procGetConsoleScreenBufferInfo    = kernel32.NewProc("GetConsoleScreenBufferInfo")
+	procFillConsoleOutputCharacter    = kernel32.NewProc("FillConsoleOutputCharacterW")
 )
 
 // These names are from the Win32 api, so they use underscores (contrary to
@@ -147,6 +148,21 @@ const (
 	modKeys = shiftPressed | leftAltPressed | rightAltPressed | leftCtrlPressed | rightCtrlPressed
 )
 
+// inputWaiting only returns true if the next call to readNext will return immediately.
+func (s *State) inputWaiting() bool {
+	var num uint32
+	ok, _, _ := procGetNumberOfConsoleInputEvents.Call(uintptr(s.handle), uintptr(unsafe.Pointer(&num)))
+	if ok == 0 {
+		// call failed, so we cannot guarantee a non-blocking readNext
+		return false
+	}
+
+	// during a "paste" input events are always an odd number, and
+	// the last one results in a blocking readNext, so return false
+	// when num is 1 or 0.
+	return num > 1
+}
+
 func (s *State) readNext() (interface{}, error) {
 	if s.repeat > 0 {
 		s.repeat--
@@ -168,6 +184,9 @@ func (s *State) readNext() (interface{}, error) {
 		if input.eventType == window_buffer_size_event {
 			xy := (*coord)(unsafe.Pointer(&input.blob[0]))
 			s.columns = int(xy.x)
+			if s.columns > 1 {
+				s.columns--
+			}
 			return winch, nil
 		}
 		if input.eventType != key_event {
@@ -260,7 +279,6 @@ func (s *State) readNext() (interface{}, error) {
 		}
 		return s.key, nil
 	}
-	return unknown, nil
 }
 
 // Close returns the terminal to its previous mode

_vendor/vendor/github.com/peterh/liner/line.go

@@ -90,11 +90,11 @@ const (
 )
 
 func (s *State) refresh(prompt []rune, buf []rune, pos int) error {
+	s.needRefresh = false
 	if s.multiLineMode {
 		return s.refreshMultiLine(prompt, buf, pos)
-	} else {
-		return s.refreshSingleLine(prompt, buf, pos)
 	}
+	return s.refreshSingleLine(prompt, buf, pos)
 }
 
 func (s *State) refreshSingleLine(prompt []rune, buf []rune, pos int) error {
@@ -387,8 +387,6 @@ func (s *State) tabComplete(p []rune, line []rune, pos int) ([]rune, int, interf
 		}
 		return []rune(head + pick + tail), hl + utf8.RuneCountInString(pick), next, nil
 	}
-	// Not reached
-	return line, pos, rune(esc), nil
 }
 
 // reverse intelligent search, implements a bash-like history search.
@@ -556,14 +554,26 @@ func (s *State) yank(p []rune, text []rune, pos int) ([]rune, int, interface{},
 			}
 		}
 	}
-
-	return line, pos, esc, nil
 }
 
 // Prompt displays p and returns a line of user input, not including a trailing
 // newline character. An io.EOF error is returned if the user signals end-of-file
 // by pressing Ctrl-D. Prompt allows line editing if the terminal supports it.
 func (s *State) Prompt(prompt string) (string, error) {
+	return s.PromptWithSuggestion(prompt, "", 0)
+}
+
+// PromptWithSuggestion displays prompt and an editable text with cursor at
+// given position. The cursor will be set to the end of the line if given position
+// is negative or greater than length of text. Returns a line of user input, not
+// including a trailing newline character. An io.EOF error is returned if the user
+// signals end-of-file by pressing Ctrl-D.
+func (s *State) PromptWithSuggestion(prompt string, text string, pos int) (string, error) {
+	for _, r := range prompt {
+		if unicode.Is(unicode.C, r) {
+			return "", ErrInvalidPrompt
+		}
+	}
 	if s.inputRedirected || !s.terminalSupported {
 		return s.promptUnsupported(prompt)
 	}
@@ -574,24 +584,37 @@ func (s *State) Prompt(prompt string) (string, error) {
 	s.historyMutex.RLock()
 	defer s.historyMutex.RUnlock()
 
-	s.startPrompt()
-	defer s.stopPrompt()
-	s.getColumns()
-
 	fmt.Print(prompt)
 	p := []rune(prompt)
-	var line []rune
-	pos := 0
+	var line = []rune(text)
 	historyEnd := ""
-	prefixHistory := s.getHistoryByPrefix(string(line))
-	historyPos := len(prefixHistory)
+	var historyPrefix []string
+	historyPos := 0
+	historyStale := true
 	historyAction := false // used to mark history related actions
 	killAction := 0        // used to mark kill related actions
+
+	defer s.stopPrompt()
+
+	if pos < 0 || len(text) < pos {
+		pos = len(text)
+	}
+	if len(line) > 0 {
+		s.refresh(p, line, pos)
+	}
+
+restart:
+	s.startPrompt()
+	s.getColumns()
+
 mainLoop:
 	for {
 		next, err := s.readNext()
 	haveNext:
 		if err != nil {
+			if s.shouldRestart != nil && s.shouldRestart(err) {
+				goto restart
+			}
 			return "", err
 		}
 
@@ -600,6 +623,9 @@ mainLoop:
 		case rune:
 			switch v {
 			case cr, lf:
+				if s.needRefresh {
+					s.refresh(p, line, pos)
+				}
 				if s.multiLineMode {
 					s.resetMultiLine(p, line, pos)
 				}
@@ -607,21 +633,21 @@ mainLoop:
 				break mainLoop
 			case ctrlA: // Start of line
 				pos = 0
-				s.refresh(p, line, pos)
+				s.needRefresh = true
 			case ctrlE: // End of line
 				pos = len(line)
-				s.refresh(p, line, pos)
+				s.needRefresh = true
 			case ctrlB: // left
 				if pos > 0 {
 					pos -= len(getSuffixGlyphs(line[:pos], 1))
-					s.refresh(p, line, pos)
+					s.needRefresh = true
 				} else {
 					fmt.Print(beep)
 				}
 			case ctrlF: // right
 				if pos < len(line) {
 					pos += len(getPrefixGlyphs(line[pos:], 1))
-					s.refresh(p, line, pos)
+					s.needRefresh = true
 				} else {
 					fmt.Print(beep)
 				}
@@ -640,7 +666,7 @@ mainLoop:
 				} else {
 					n := len(getPrefixGlyphs(line[pos:], 1))
 					line = append(line[:pos], line[pos+n:]...)
-					s.refresh(p, line, pos)
+					s.needRefresh = true
 				}
 			case ctrlK: // delete remainder of line
 				if pos >= len(line) {
@@ -654,32 +680,42 @@ mainLoop:
 
 					killAction = 2 // Mark that there was a kill action
 					line = line[:pos]
-					s.refresh(p, line, pos)
+					s.needRefresh = true
 				}
 			case ctrlP: // up
 				historyAction = true
+				if historyStale {
+					historyPrefix = s.getHistoryByPrefix(string(line))
+					historyPos = len(historyPrefix)
+					historyStale = false
+				}
 				if historyPos > 0 {
-					if historyPos == len(prefixHistory) {
+					if historyPos == len(historyPrefix) {
 						historyEnd = string(line)
 					}
 					historyPos--
-					line = []rune(prefixHistory[historyPos])
+					line = []rune(historyPrefix[historyPos])
 					pos = len(line)
-					s.refresh(p, line, pos)
+					s.needRefresh = true
 				} else {
 					fmt.Print(beep)
 				}
 			case ctrlN: // down
 				historyAction = true
-				if historyPos < len(prefixHistory) {
+				if historyStale {
+					historyPrefix = s.getHistoryByPrefix(string(line))
+					historyPos = len(historyPrefix)
+					historyStale = false
+				}
+				if historyPos < len(historyPrefix) {
 					historyPos++
-					if historyPos == len(prefixHistory) {
+					if historyPos == len(historyPrefix) {
 						line = []rune(historyEnd)
 					} else {
-						line = []rune(prefixHistory[historyPos])
+						line = []rune(historyPrefix[historyPos])
 					}
 					pos = len(line)
-					s.refresh(p, line, pos)
+					s.needRefresh = true
 				} else {
 					fmt.Print(beep)
 				}
@@ -697,11 +733,11 @@ mainLoop:
 					copy(line[pos-len(prev):], next)
 					copy(line[pos-len(prev)+len(next):], scratch)
 					pos += len(next)
-					s.refresh(p, line, pos)
+					s.needRefresh = true
 				}
 			case ctrlL: // clear screen
 				s.eraseScreen()
-				s.refresh(p, line, pos)
+				s.needRefresh = true
 			case ctrlC: // reset
 				fmt.Println("^C")
 				if s.multiLineMode {
@@ -721,7 +757,7 @@ mainLoop:
 					n := len(getSuffixGlyphs(line[:pos], 1))
 					line = append(line[:pos-n], line[pos:]...)
 					pos -= n
-					s.refresh(p, line, pos)
+					s.needRefresh = true
 				}
 			case ctrlU: // Erase line before cursor
 				if killAction > 0 {
@@ -733,7 +769,7 @@ mainLoop:
 				killAction = 2 // Mark that there was some killing
 				line = line[pos:]
 				pos = 0
-				s.refresh(p, line, pos)
+				s.needRefresh = true
 			case ctrlW: // Erase word
 				if pos == 0 {
 					fmt.Print(beep)
@@ -770,13 +806,13 @@ mainLoop:
 				}
 				killAction = 2 // Mark that there was some killing
 
-				s.refresh(p, line, pos)
+				s.needRefresh = true
 			case ctrlY: // Paste from Yank buffer
 				line, pos, next, err = s.yank(p, line, pos)
 				goto haveNext
 			case ctrlR: // Reverse Search
 				line, pos, next, err = s.reverseISearch(line, pos)
-				s.refresh(p, line, pos)
+				s.needRefresh = true
 				goto haveNext
 			case tab: // Tab completion
 				line, pos, next, err = s.tabComplete(p, line, pos)
@@ -791,14 +827,16 @@ mainLoop:
 			case 0, 28, 29, 30, 31:
 				fmt.Print(beep)
 			default:
-				if pos == len(line) && !s.multiLineMode && countGlyphs(p)+countGlyphs(line) < s.columns-1 {
+				if pos == len(line) && !s.multiLineMode &&
+					len(p)+len(line) < s.columns*4 && // Avoid countGlyphs on large lines
+					countGlyphs(p)+countGlyphs(line) < s.columns-1 {
 					line = append(line, v)
 					fmt.Printf("%c", v)
 					pos++
 				} else {
 					line = append(line[:pos], append([]rune{v}, line[pos:]...)...)
 					pos++
-					s.refresh(p, line, pos)
+					s.needRefresh = true
 				}
 			}
 		case action:
@@ -866,24 +904,34 @@ mainLoop:
 				}
 			case up:
 				historyAction = true
+				if historyStale {
+					historyPrefix = s.getHistoryByPrefix(string(line))
+					historyPos = len(historyPrefix)
+					historyStale = false
+				}
 				if historyPos > 0 {
-					if historyPos == len(prefixHistory) {
+					if historyPos == len(historyPrefix) {
 						historyEnd = string(line)
 					}
 					historyPos--
-					line = []rune(prefixHistory[historyPos])
+					line = []rune(historyPrefix[historyPos])
 					pos = len(line)
 				} else {
 					fmt.Print(beep)
 				}
 			case down:
 				historyAction = true
-				if historyPos < len(prefixHistory) {
+				if historyStale {
+					historyPrefix = s.getHistoryByPrefix(string(line))
+					historyPos = len(historyPrefix)
+					historyStale = false
+				}
+				if historyPos < len(historyPrefix) {
 					historyPos++
-					if historyPos == len(prefixHistory) {
+					if historyPos == len(historyPrefix) {
 						line = []rune(historyEnd)
 					} else {
-						line = []rune(prefixHistory[historyPos])
+						line = []rune(historyPrefix[historyPos])
 					}
 					pos = len(line)
 				} else {
@@ -907,11 +955,13 @@ mainLoop:
 					s.cursorRows = 1
 				}
 			}
+			s.needRefresh = true
+		}
+		if s.needRefresh && !s.inputWaiting() {
 			s.refresh(p, line, pos)
 		}
 		if !historyAction {
-			prefixHistory = s.getHistoryByPrefix(string(line))
-			historyPos = len(prefixHistory)
+			historyStale = true
 		}
 		if killAction > 0 {
 			killAction--
@@ -923,6 +973,11 @@ mainLoop:
 // PasswordPrompt displays p, and then waits for user input. The input typed by
 // the user is not displayed in the terminal.
 func (s *State) PasswordPrompt(prompt string) (string, error) {
+	for _, r := range prompt {
+		if unicode.Is(unicode.C, r) {
+			return "", ErrInvalidPrompt
+		}
+	}
 	if !s.terminalSupported {
 		return "", errors.New("liner: function not supported in this terminal")
 	}
@@ -933,8 +988,10 @@ func (s *State) PasswordPrompt(prompt string) (string, error) {
 		return "", ErrNotTerminalOutput
 	}
 
-	s.startPrompt()
 	defer s.stopPrompt()
+
+restart:
+	s.startPrompt()
 	s.getColumns()
 
 	fmt.Print(prompt)
@@ -946,6 +1003,9 @@ mainLoop:
 	for {
 		next, err := s.readNext()
 		if err != nil {
+			if s.shouldRestart != nil && s.shouldRestart(err) {
+				goto restart
+			}
 			return "", err
 		}
 
@@ -953,6 +1013,9 @@ mainLoop:
 		case rune:
 			switch v {
 			case cr, lf:
+				if s.needRefresh {
+					s.refresh(p, line, pos)
+				}
 				if s.multiLineMode {
 					s.resetMultiLine(p, line, pos)
 				}

_vendor/vendor/github.com/peterh/liner/output.go

@@ -48,14 +48,18 @@ type winSize struct {
 	xpixel, ypixel uint16
 }
 
-func (s *State) getColumns() {
+func (s *State) getColumns() bool {
 	var ws winSize
 	ok, _, _ := syscall.Syscall(syscall.SYS_IOCTL, uintptr(syscall.Stdout),
 		syscall.TIOCGWINSZ, uintptr(unsafe.Pointer(&ws)))
-	if ok < 0 {
-		s.columns = 80
+	if int(ok) < 0 {
+		return false
 	}
 	s.columns = int(ws.col)
+	if cursorColumn && s.columns > 1 {
+		s.columns--
+	}
+	return true
 }
 
 func (s *State) checkOutput() {

_vendor/vendor/github.com/peterh/liner/output_windows.go

@@ -69,4 +69,8 @@ func (s *State) getColumns() {
 	var sbi consoleScreenBufferInfo
 	procGetConsoleScreenBufferInfo.Call(uintptr(s.hOut), uintptr(unsafe.Pointer(&sbi)))
 	s.columns = int(sbi.dwSize.x)
+	if s.columns > 1 {
+		// Windows 10 needs a spare column for the cursor
+		s.columns--
+	}
 }

_vendor/vendor/github.com/peterh/liner/width.go

@@ -25,6 +25,12 @@ var doubleWidth = []*unicode.RangeTable{
 func countGlyphs(s []rune) int {
 	n := 0
 	for _, r := range s {
+		// speed up the common case
+		if r < 127 {
+			n++
+			continue
+		}
+
 		switch {
 		case unicode.IsOneOf(zeroWidth, r):
 		case unicode.IsOneOf(doubleWidth, r):
@@ -39,6 +45,10 @@ func countGlyphs(s []rune) int {
 func countMultiLineGlyphs(s []rune, columns int, start int) int {
 	n := start
 	for _, r := range s {
+		if r < 127 {
+			n++
+			continue
+		}
 		switch {
 		case unicode.IsOneOf(zeroWidth, r):
 		case unicode.IsOneOf(doubleWidth, r):
@@ -58,6 +68,11 @@ func countMultiLineGlyphs(s []rune, columns int, start int) int {
 func getPrefixGlyphs(s []rune, num int) []rune {
 	p := 0
 	for n := 0; n < num && p < len(s); p++ {
+		// speed up the common case
+		if s[p] < 127 {
+			n++
+			continue
+		}
 		if !unicode.IsOneOf(zeroWidth, s[p]) {
 			n++
 		}
@@ -71,6 +86,11 @@ func getPrefixGlyphs(s []rune, num int) []rune {
 func getSuffixGlyphs(s []rune, num int) []rune {
 	p := len(s)
 	for n := 0; n < num && p > 0; p-- {
+		// speed up the common case
+		if s[p-1] < 127 {
+			n++
+			continue
+		}
 		if !unicode.IsOneOf(zeroWidth, s[p-1]) {
 			n++
 		}