fix: add envs and reduce memory

external_jsonlib_test/go.mod

@@ -13,6 +13,7 @@ require (
 )
 
 require (
+	github.com/bytedance/gopkg v0.0.0-20240531030433-5df24c0168e2 // indirect
 	github.com/bytedance/sonic/loader v0.1.1 // indirect
 	github.com/cloudwego/base64x v0.1.4 // indirect
 	github.com/cloudwego/iasm v0.2.0 // indirect

external_jsonlib_test/go.sum

@@ -1,5 +1,7 @@
 github.com/buger/jsonparser v1.1.1 h1:2PnMjfWD7wBILjqQbt530v576A/cAbQvEW9gGIpYMUs=
 github.com/buger/jsonparser v1.1.1/go.mod h1:6RYKKt7H4d4+iWqouImQ9R2FZql3VbhNgx27UK13J/0=
+github.com/bytedance/gopkg v0.0.0-20240531030433-5df24c0168e2 h1:e+WTWDw35RetW+Zuhcy4YPFmNPRh7yPG7lAaWJeUQl0=
+github.com/bytedance/gopkg v0.0.0-20240531030433-5df24c0168e2/go.mod h1:FtQG3YbQG9L/91pbKSw787yBQPutC+457AvDW77fgUQ=
 github.com/bytedance/sonic/loader v0.1.1 h1:c+e5Pt1k/cy5wMveRDyk2X4B9hF4g7an8N3zCYjJFNM=
 github.com/bytedance/sonic/loader v0.1.1/go.mod h1:ncP89zfokxS5LZrJxl5z0UJcsk4M4yY2JpfqGeCtNLU=
 github.com/cloudwego/base64x v0.1.4 h1:jwCgWpFanWmN8xoIUHa2rtzmkd5J2plF/dnLS6Xd/0Y=
@@ -45,6 +47,13 @@ github.com/twitchyliquid64/golang-asm v0.15.1 h1:SU5vSMR7hnwNxj24w34ZyCi/FmDZTkS
 github.com/twitchyliquid64/golang-asm v0.15.1/go.mod h1:a1lVb/DtPvCB8fslRZhAngC2+aY1QWCk3Cedj/Gdt08=
 golang.org/x/arch v0.0.0-20210923205945-b76863e36670 h1:18EFjUmQOcUvxNYSkA6jO9VAiXCnxFY6NyDX0bHDmkU=
 golang.org/x/arch v0.0.0-20210923205945-b76863e36670/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
+golang.org/x/net v0.0.0-20221014081412-f15817d10f9b/go.mod h1:YDH+HFinaLZZlnHAfSS6ZXJJ9M9t4Dl22yv3iI2vPwk=
+golang.org/x/sync v0.0.0-20210220032951-036812b2e83c/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
+golang.org/x/sys v0.0.0-20210615035016-665e8c7367d1/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/sys v0.0.0-20220728004956-3c1f35247d10/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/term v0.0.0-20210927222741-03fcf44c2211/go.mod h1:jbD1KX2456YbFQfuXm/mYQcufACuNUgVhRMnK/tPxf8=
+golang.org/x/text v0.3.7/go.mod h1:u+2+/6zg+i71rQMx5EYifcz6MCKuco9NR6JIITiCfzQ=
+golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405 h1:yhCVgyC4o1eVCa2tZl7eS0r+SDo693bJlVdllGtEeKM=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=

fuzz/other_fuzz_test.go

@@ -51,7 +51,7 @@ func fuzzHtmlEscape(t *testing.T, data []byte){
 func fuzzStream(t *testing.T, data []byte) {
 	r := bytes.NewBuffer(data)
 	dc := decoder.NewStreamDecoder(r)
-	decoderEnableValidateString(dc)
+	dc.ValidateString()
 	r1 := bytes.NewBuffer(data)
 	dc1 := decoder.NewStreamDecoder(r1)
 

internal/decoder/api/decoder.go

@@ -17,14 +17,16 @@
 package api
 
 import (
-    `reflect`
-
-    `github.com/bytedance/sonic/internal/native`
-    `github.com/bytedance/sonic/internal/native/types`
-	`github.com/bytedance/sonic/internal/decoder/consts`
-	`github.com/bytedance/sonic/internal/decoder/errors`
-    `github.com/bytedance/sonic/internal/rt`
-    `github.com/bytedance/sonic/option`
+	"reflect"
+
+	"github.com/bytedance/sonic/internal/decoder/consts"
+	"github.com/bytedance/sonic/internal/decoder/errors"
+	"github.com/bytedance/sonic/internal/decoder/optdec"
+	"github.com/bytedance/sonic/internal/envs"
+	"github.com/bytedance/sonic/internal/native"
+	"github.com/bytedance/sonic/internal/native/types"
+	"github.com/bytedance/sonic/internal/rt"
+	"github.com/bytedance/sonic/option"
 )
 
 const (
@@ -64,11 +66,21 @@ type Decoder struct {
     i int
     f uint64
     s string
+	owned bool
 }
 
+var padding string = "x\"x\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00"
+
 // NewDecoder creates a new decoder instance.
 func NewDecoder(s string) *Decoder {
-    return &Decoder{s: s}
+	var owned = false
+	/* clone into buffer and padding it before unmarshal */
+	if envs.UseOptDec {
+		s = optdec.PaddingJson(s)
+		owned = true
+	}
+
+    return &Decoder{s: s, owned: owned}
 }
 
 // Pos returns the current decoding position.

internal/decoder/api/stream.go

@@ -61,25 +61,16 @@ func NewStreamDecoder(r io.Reader) *StreamDecoder {
 func (self *StreamDecoder) Decode(val interface{}) (err error) {
     // read more data into buf
     if self.More() {
-        // println(string(self.buf))
         var s = self.scanp
     try_skip:
         var e = len(self.buf)
-        // println("s:", s, "e:", e, "scanned:",self.scanned, "scanp:",self.scanp, self.buf)
         var src = rt.Mem2Str(self.buf[s:e])
-        // if len(src) > 5 {
-        //     println(src[:5], src[len(src)-5:])
-        // } else {
-        //     println(src)
-        // }
         // try skip
         var x = 0;
         if y := native.SkipOneFast(&src, &x); y < 0 {
             if self.readMore()  {
-                // println("more")
                 goto try_skip
             } else {
-                // println("no more")
                 err = SyntaxError{e, self.s, types.ParsingError(-s), ""}
                 self.setErr(err)
                 return
@@ -89,7 +80,6 @@ func (self *StreamDecoder) Decode(val interface{}) (err error) {
             e = x + s
         }
 
-        // println("decode: ", s, e)
         // must copy string here for safety
         self.Decoder.Reset(string(self.buf[s:e]))
         err = self.Decoder.Decode(val)
@@ -101,13 +91,11 @@ func (self *StreamDecoder) Decode(val interface{}) (err error) {
         self.scanp = e
         _, empty := self.scan()
         if empty {
-            // println("recycle")
             // no remain valid bytes, thus we just recycle buffer
             mem := self.buf
             self.buf = nil
             bufPool.Put(mem[:0])
         } else {
-            // println("keep")
             // remain undecoded bytes, move them onto head
             n := copy(self.buf, self.buf[self.scanp:])
             self.buf = self.buf[:n]
@@ -123,7 +111,6 @@ func (self *StreamDecoder) Decode(val interface{}) (err error) {
 // InputOffset returns the input stream byte offset of the current decoder position. 
 // The offset gives the location of the end of the most recently returned token and the beginning of the next token.
 func (self *StreamDecoder) InputOffset() int64 {
-    // println("input offset",self.scanned, self.scanp)
     return self.scanned + int64(self.scanp)
 }
 

internal/decoder/optdec/decoder.go

@@ -66,6 +66,7 @@ func Decode(s *string, i *int, f uint64, val interface{}) error {
 	if err != nil {
 		return err
 	}
+
 
 	/* parse into document */
 	ctx, err := NewContext(*s, *i, uint64(f), etp)

internal/decoder/optdec/native.go

@@ -12,6 +12,8 @@ import (
 )
 
 
+const _PADDING_SIZE = 64
+
 type ErrorCode int
 
 const (
@@ -85,10 +87,8 @@ func (self *nodeBuf) init(nodes []node) {
 // should consitent with native/parser.c
 type Parser struct {
 	Json    string
-	padded	[]byte
 	nodes 	[]node
 	dbuf 	[]byte
-	backup  []node
 
 	options uint64
 	// JSON cursor
@@ -116,105 +116,126 @@ type jsonStat struct {
 
 
 var (
-	defaultJsonPaddedCap uintptr =  1 << 20  // 1 Mb
 	defaultNodesCap      uintptr =  (1 << 20) / unsafe.Sizeof(node{})  // 1 Mb
 )
 
 var parsePool sync.Pool = sync.Pool {
 	New: func () interface{} {
 		return &Parser{
 			options: 0,
-			padded: make([]byte, 0, defaultJsonPaddedCap),
 			nodes: make([]node, defaultNodesCap, defaultNodesCap),
 			dbuf: make([]byte, types.MaxDigitNums, types.MaxDigitNums),
 		}
 	},
 }
 
-var padding string = "x\"x\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00"
+var _PADDING_STR string = "x\"x\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00"
+
+func PaddingJson(s string) string {
+	bytes := rt.Mallocgc(uintptr(len(s) + _PADDING_SIZE), rt.BytesType, false)
+	rt.Memmove(bytes, ((*rt.GoString)(unsafe.Pointer(&s))).Ptr, uintptr(len(s)))
+	rt.Memmove(unsafe.Pointer(uintptr(bytes) + uintptr(len(s))), ((*rt.GoString)(unsafe.Pointer(&_PADDING_STR))).Ptr, _PADDING_SIZE)
+	// not include padding buffer in string
+	owned := rt.GoString{Ptr: bytes, Len: len(s)}
+	return *(*string)(unsafe.Pointer(&owned))
+}
 
-func newParser(data string, pos int, opt uint64) *Parser {
+func newParser(owned string, pos int, opt uint64) *Parser {
 	p := parsePool.Get().(*Parser)
 
 	/* validate json if needed */
-	if (opt & (1 << _F_validate_string)) != 0  && !utf8.ValidateString(data){
-		dbuf := utf8.CorrectWith(nil, rt.Str2Mem(data[pos:]), "\ufffd")
-		dbuf = append(dbuf, padding...)
-		p.Json = rt.Mem2Str(dbuf[:len(dbuf) - len(padding)])
+	if (opt & (1 << _F_validate_string)) != 0  && !utf8.ValidateString(owned){
+		dbuf := utf8.CorrectWith(nil, rt.Str2Mem(owned[pos:]), "\ufffd")
+		dbuf = append(dbuf, _PADDING_STR...)
+		p.Json = rt.Mem2Str(dbuf[:len(dbuf) - len(_PADDING_STR)])
 		p.Utf8Inv = true
-		p.start = uintptr((*rt.GoString)(unsafe.Pointer(&p.Json)).Ptr)
 	} else {
-		p.Json = data
-		// TODO: prevent too large JSON
-		p.padded = append(p.padded, data[pos:]...)
-		p.padded = append(p.padded, padding...)
-		p.start = uintptr((*rt.GoSlice)(unsafe.Pointer(&p.padded)).Ptr)
+		p.Json = owned
 	}
 
+	p.start = uintptr((*rt.GoString)(unsafe.Pointer(&p.Json)).Ptr)
 	p.cur 	= p.start
 	p.end   = p.cur + uintptr(len(p.Json))
 	p.options = opt
+
+	if p.nodes == nil {
+		newn := rt.Mallocgc(uintptr(defaultNodesCap) * unsafe.Sizeof(node{}), rt.ByteType, false)
+		slice := rt.GoSlice {
+			Ptr: newn,
+			Len: int(defaultNodesCap),
+			Cap: int(defaultNodesCap),
+		}
+		p.nodes = *(*[]node)(unsafe.Pointer(&slice))
+	}
 	p.nbuf.init(p.nodes)
 	return p
 }
 
-
 func (p *Parser) Pos() int {
 	return int(p.cur - p.start)
 }
 
 func (p *Parser) JsonBytes() []byte {
-	if p.Utf8Inv {
-		return (rt.Str2Mem(p.Json))
-	} else {
-		return p.padded
+	return rt.Str2Mem(p.Json)
+}
+
+func (p *Parser) growNodeBuf() {
+	// calculate new capacity
+	offset := p.nbuf.ncur - p.nbuf.nstart
+	oldLen := int(offset / unsafe.Sizeof(node{}))
+	newCap := oldLen * 2
+	if newCap <= len(p.Json) / 32 {
+		newCap = len(p.Json) / 32
 	}
+
+	// grow node buffer
+	newn := rt.Mallocgc(uintptr(newCap) * unsafe.Sizeof(node{}), rt.ByteType, false)
+	rt.Memmove(newn, unsafe.Pointer(p.nbuf.nstart), offset)
+	slice := rt.GoSlice {
+		Ptr: newn,
+		Len: newCap,
+		Cap: newCap,
+	}
+	p.nodes = *(*[]node)(unsafe.Pointer(&slice))
+
+	// update cursor
+	p.nbuf.ncur = uintptr(newn) + offset
+	p.nbuf.nstart = uintptr(newn)
+	p.nbuf.nend = uintptr(newn) + uintptr(newCap) * unsafe.Sizeof(node{})
+
 }
 
 func (p *Parser) parse() ErrorCode {
-	var offset uintptr
+	var err ErrorCode
+
 	// when decode into struct, we should decode number as possible
 	old := p.options
 	if !p.isEface {
 		p.options &^= 1 << _F_use_number
 	}
 
-	// fast path with limited node buffer
-	err := ErrorCode(native.ParseWithPadding(unsafe.Pointer(p)))
-	if err != SONIC_VISIT_FAILED {
-		goto ret;
+	for {
+		// fast path with limited node buffer
+		err = ErrorCode(native.ParseWithPadding(unsafe.Pointer(p)))
+		if err != SONIC_VISIT_FAILED {
+			p.options = old
+			return err
+		}
+		p.growNodeBuf()
 	}
-
-	// fallback parse
-	// maybe node buf is not enough, continue
-	p.backup = p.nodes
-	p.nodes = make([]node, len(p.Json) / 2 + 2, len(p.Json) / 2 + 2)
-	copy(p.nodes, p.backup)
-	offset = (p.nbuf.ncur - p.nbuf.nstart) / unsafe.Sizeof(node{})
-	p.nbuf.nstart = uintptr(unsafe.Pointer(&p.nodes[0]))
-	p.nbuf.nend = p.nbuf.nstart + uintptr(cap(p.nodes)) * unsafe.Sizeof(node{})
-	p.nbuf.ncur = uintptr(unsafe.Pointer(&p.nodes[offset]))
-	err = ErrorCode(native.ParseWithPadding(unsafe.Pointer(p)))
-
-ret:
-	p.options = old
-	return err
 }
 
 func (p *Parser) reset() {
 	p.options = 0
-	p.padded = p.padded[:0]
-	// nodes is too large here, we will not reset it and use small backup nodes buffer
-	if p.backup != nil {
-		p.nodes = p.backup
-		p.backup = nil
-	}
 	p.start = 0
 	p.cur = 0
 	p.end = 0
 	p.Json = ""
 	p.nbuf = nodeBuf{}
 	p._nbk = _nospaceBlock{}
+	if len(p.nodes) > int(defaultNodesCap) {
+		p.nodes = nil
+	}
 	p.Utf8Inv = false
 	p.isEface = false
 }