rlp: precursor changes for trie, p2p

rlp/decode.go

@@ -173,6 +173,8 @@ var (
 func makeDecoder(typ reflect.Type, tags tags) (dec decoder, err error) {
 	kind := typ.Kind()
 	switch {
+	case typ == rawValueType:
+		return decodeRawValue, nil
 	case typ.Implements(decoderInterface):
 		return decodeDecoder, nil
 	case kind != reflect.Ptr && reflect.PtrTo(typ).Implements(decoderInterface):
@@ -203,6 +205,15 @@ func makeDecoder(typ reflect.Type, tags tags) (dec decoder, err error) {
 	}
 }
 
+func decodeRawValue(s *Stream, val reflect.Value) error {
+	r, err := s.Raw()
+	if err != nil {
+		return err
+	}
+	val.SetBytes(r)
+	return nil
+}
+
 func decodeUint(s *Stream, val reflect.Value) error {
 	typ := val.Type()
 	num, err := s.uint(typ.Bits())

rlp/decode_test.go

@@ -24,6 +24,7 @@ import (
 	"io"
 	"math/big"
 	"reflect"
+	"strings"
 	"testing"
 )
 
@@ -437,6 +438,11 @@ var decodeTests = []decodeTest{
 		error: "rlp: expected input string or byte for uint, decoding into (rlp.recstruct).Child.I",
 	},
 
+	// RawValue
+	{input: "01", ptr: new(RawValue), value: RawValue(unhex("01"))},
+	{input: "82FFFF", ptr: new(RawValue), value: RawValue(unhex("82FFFF"))},
+	{input: "C20102", ptr: new([]RawValue), value: []RawValue{unhex("01"), unhex("02")}},
+
 	// pointers
 	{input: "00", ptr: new(*[]byte), value: &[]byte{0}},
 	{input: "80", ptr: new(*uint), value: uintp(0)},
@@ -725,7 +731,7 @@ func encodeTestSlice(n uint) []byte {
 }
 
 func unhex(str string) []byte {
-	b, err := hex.DecodeString(str)
+	b, err := hex.DecodeString(strings.Replace(str, " ", "", -1))
 	if err != nil {
 		panic(fmt.Sprintf("invalid hex string: %q", str))
 	}

rlp/encode.go

@@ -45,12 +45,6 @@ type Encoder interface {
 	EncodeRLP(io.Writer) error
 }
 
-// ListSize returns the encoded size of an RLP list with the given
-// content size.
-func ListSize(contentSize uint64) uint64 {
-	return uint64(headsize(contentSize)) + contentSize
-}
-
 // Encode writes the RLP encoding of val to w. Note that Encode may
 // perform many small writes in some cases. Consider making w
 // buffered.
@@ -90,8 +84,8 @@ func Encode(w io.Writer, val interface{}) error {
 		return outer.encode(val)
 	}
 	eb := encbufPool.Get().(*encbuf)
-	eb.reset()
 	defer encbufPool.Put(eb)
+	eb.reset()
 	if err := eb.encode(val); err != nil {
 		return err
 	}
@@ -102,8 +96,8 @@ func Encode(w io.Writer, val interface{}) error {
 // Please see the documentation of Encode for the encoding rules.
 func EncodeToBytes(val interface{}) ([]byte, error) {
 	eb := encbufPool.Get().(*encbuf)
-	eb.reset()
 	defer encbufPool.Put(eb)
+	eb.reset()
 	if err := eb.encode(val); err != nil {
 		return nil, err
 	}
@@ -288,8 +282,13 @@ type encReader struct {
 func (r *encReader) Read(b []byte) (n int, err error) {
 	for {
 		if r.piece = r.next(); r.piece == nil {
-			encbufPool.Put(r.buf)
-			r.buf = nil
+			// Put the encode buffer back into the pool at EOF when it
+			// is first encountered. Subsequent calls still return EOF
+			// as the error but the buffer is no longer valid.
+			if r.buf != nil {
+				encbufPool.Put(r.buf)
+				r.buf = nil
+			}
 			return n, io.EOF
 		}
 		nn := copy(b[n:], r.piece)
@@ -349,6 +348,8 @@ var (
 func makeWriter(typ reflect.Type) (writer, error) {
 	kind := typ.Kind()
 	switch {
+	case typ == rawValueType:
+		return writeRawValue, nil
 	case typ.Implements(encoderInterface):
 		return writeEncoder, nil
 	case kind != reflect.Ptr && reflect.PtrTo(typ).Implements(encoderInterface):
@@ -384,6 +385,11 @@ func isByte(typ reflect.Type) bool {
 	return typ.Kind() == reflect.Uint8 && !typ.Implements(encoderInterface)
 }
 
+func writeRawValue(val reflect.Value, w *encbuf) error {
+	w.str = append(w.str, val.Bytes()...)
+	return nil
+}
+
 func writeUint(val reflect.Value, w *encbuf) error {
 	i := val.Uint()
 	if i == 0 {

rlp/encode_test.go

@@ -23,6 +23,7 @@ import (
 	"io"
 	"io/ioutil"
 	"math/big"
+	"sync"
 	"testing"
 )
 
@@ -203,6 +204,11 @@ var encTests = []encTest{
 		output: "F90200CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376CF84617364668471776572847A786376",
 	},
 
+	// RawValue
+	{val: RawValue(unhex("01")), output: "01"},
+	{val: RawValue(unhex("82FFFF")), output: "82FFFF"},
+	{val: []RawValue{unhex("01"), unhex("02")}, output: "C20102"},
+
 	// structs
 	{val: simplestruct{}, output: "C28080"},
 	{val: simplestruct{A: 3, B: "foo"}, output: "C50383666F6F"},
@@ -306,3 +312,25 @@ func TestEncodeToReaderPiecewise(t *testing.T) {
 		return output, nil
 	})
 }
+
+// This is a regression test verifying that encReader
+// returns its encbuf to the pool only once.
+func TestEncodeToReaderReturnToPool(t *testing.T) {
+	buf := make([]byte, 50)
+	wg := new(sync.WaitGroup)
+	for i := 0; i < 5; i++ {
+		wg.Add(1)
+		go func() {
+			for i := 0; i < 1000; i++ {
+				_, r, _ := EncodeToReader("foo")
+				ioutil.ReadAll(r)
+				r.Read(buf)
+				r.Read(buf)
+				r.Read(buf)
+				r.Read(buf)
+			}
+			wg.Done()
+		}()
+	}
+	wg.Wait()
+}

rlp/raw.go

@@ -0,0 +1,156 @@
+// Copyright 2015 The go-ethereum Authors
+// This file is part of the go-ethereum library.
+//
+// The go-ethereum library is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Lesser General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// The go-ethereum library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Lesser General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public License
+// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
+
+package rlp
+
+import (
+	"io"
+	"reflect"
+)
+
+// RawValue represents an encoded RLP value and can be used to delay
+// RLP decoding or precompute an encoding. Note that the decoder does
+// not verify whether the content of RawValues is valid RLP.
+type RawValue []byte
+
+var rawValueType = reflect.TypeOf(RawValue{})
+
+// ListSize returns the encoded size of an RLP list with the given
+// content size.
+func ListSize(contentSize uint64) uint64 {
+	return uint64(headsize(contentSize)) + contentSize
+}
+
+// Split returns the content of first RLP value and any
+// bytes after the value as subslices of b.
+func Split(b []byte) (k Kind, content, rest []byte, err error) {
+	k, ts, cs, err := readKind(b)
+	if err != nil {
+		return 0, nil, b, err
+	}
+	return k, b[ts : ts+cs], b[ts+cs:], nil
+}
+
+// SplitString splits b into the content of an RLP string
+// and any remaining bytes after the string.
+func SplitString(b []byte) (content, rest []byte, err error) {
+	k, content, rest, err := Split(b)
+	if err != nil {
+		return nil, b, err
+	}
+	if k == List {
+		return nil, b, ErrExpectedString
+	}
+	return content, rest, nil
+}
+
+// SplitList splits b into the content of a list and any remaining
+// bytes after the list.
+func SplitList(b []byte) (content, rest []byte, err error) {
+	k, content, rest, err := Split(b)
+	if err != nil {
+		return nil, b, err
+	}
+	if k != List {
+		return nil, b, ErrExpectedList
+	}
+	return content, rest, nil
+}
+
+// CountValues counts the number of encoded values in b.
+func CountValues(b []byte) (int, error) {
+	i := 0
+	for ; len(b) > 0; i++ {
+		_, tagsize, size, err := readKind(b)
+		if err != nil {
+			return 0, err
+		}
+		b = b[tagsize+size:]
+	}
+	return i, nil
+}
+
+func readKind(buf []byte) (k Kind, tagsize, contentsize uint64, err error) {
+	if len(buf) == 0 {
+		return 0, 0, 0, io.ErrUnexpectedEOF
+	}
+	b := buf[0]
+	switch {
+	case b < 0x80:
+		k = Byte
+		tagsize = 0
+		contentsize = 1
+	case b < 0xB8:
+		k = String
+		tagsize = 1
+		contentsize = uint64(b - 0x80)
+		// Reject strings that should've been single bytes.
+		if contentsize == 1 && buf[1] < 128 {
+			return 0, 0, 0, ErrCanonSize
+		}
+	case b < 0xC0:
+		k = String
+		tagsize = uint64(b-0xB7) + 1
+		contentsize, err = readSize(buf[1:], b-0xB7)
+	case b < 0xF8:
+		k = List
+		tagsize = 1
+		contentsize = uint64(b - 0xC0)
+	default:
+		k = List
+		tagsize = uint64(b-0xF7) + 1
+		contentsize, err = readSize(buf[1:], b-0xF7)
+	}
+	if err != nil {
+		return 0, 0, 0, err
+	}
+	// Reject values larger than the input slice.
+	if contentsize > uint64(len(buf))-tagsize {
+		return 0, 0, 0, ErrValueTooLarge
+	}
+	return k, tagsize, contentsize, err
+}
+
+func readSize(b []byte, slen byte) (uint64, error) {
+	if int(slen) > len(b) {
+		return 0, io.ErrUnexpectedEOF
+	}
+	var s uint64
+	switch slen {
+	case 1:
+		s = uint64(b[0])
+	case 2:
+		s = uint64(b[0])<<8 | uint64(b[1])
+	case 3:
+		s = uint64(b[0])<<16 | uint64(b[1])<<8 | uint64(b[2])
+	case 4:
+		s = uint64(b[0])<<24 | uint64(b[1])<<16 | uint64(b[2])<<8 | uint64(b[3])
+	case 5:
+		s = uint64(b[0])<<32 | uint64(b[1])<<24 | uint64(b[2])<<16 | uint64(b[3])<<8 | uint64(b[4])
+	case 6:
+		s = uint64(b[0])<<40 | uint64(b[1])<<32 | uint64(b[2])<<24 | uint64(b[3])<<16 | uint64(b[4])<<8 | uint64(b[5])
+	case 7:
+		s = uint64(b[0])<<48 | uint64(b[1])<<40 | uint64(b[2])<<32 | uint64(b[3])<<24 | uint64(b[4])<<16 | uint64(b[5])<<8 | uint64(b[6])
+	case 8:
+		s = uint64(b[0])<<56 | uint64(b[1])<<48 | uint64(b[2])<<40 | uint64(b[3])<<32 | uint64(b[4])<<24 | uint64(b[5])<<16 | uint64(b[6])<<8 | uint64(b[7])
+	}
+	// Reject sizes < 56 (shouldn't have separate size) and sizes with
+	// leading zero bytes.
+	if s < 56 || b[0] == 0 {
+		return 0, ErrCanonSize
+	}
+	return s, nil
+}