mirror of
https://github.com/golang/go
synced 2024-11-12 07:40:23 -07:00
Add base85, based on base64.
Also add git version of base85, which is what I really wanted but didn't recognize as different until it was too late. R=austin DELTA=980 (972 added, 4 deleted, 4 changed) OCL=35580 CL=35921
This commit is contained in:
Russ Cox
parent
60ff8cc86f
commit
fcc4be8c65
@ -1,6 +1,7 @@
|
||||
archive/tar.install: bytes.install io.install os.install strconv.install strings.install
|
||||
asn1.install: fmt.install os.install reflect.install strconv.install strings.install time.install
|
||||
base64.install: bytes.install io.install os.install strconv.install
|
||||
base85.install: bytes.install io.install os.install strconv.install
|
||||
big.install:
|
||||
bignum.install: fmt.install
|
||||
bufio.install: io.install os.install strconv.install utf8.install
|
||||
|
||||
@ -15,6 +15,7 @@ DIRS=\
|
||||
archive/tar\
|
||||
asn1\
|
||||
base64\
|
||||
base85\
|
||||
big\
|
||||
bignum\
|
||||
bufio\
|
||||
|
||||
@ -16,7 +16,7 @@ import (
|
||||
* Encodings
|
||||
*/
|
||||
|
||||
// Encoding is a radix 64 encoding/decoding scheme, defined by a
|
||||
// An Encoding is a radix 64 encoding/decoding scheme, defined by a
|
||||
// 64-character alphabet. The most common encoding is the "base64"
|
||||
// encoding defined in RFC 4648 and used in MIME (RFC 2045) and PEM
|
||||
// (RFC 1421). RFC 4648 also defines an alternate encoding, which is
|
||||
@ -56,7 +56,7 @@ var URLEncoding = NewEncoding(encodeURL)
|
||||
*/
|
||||
|
||||
// Encode encodes src using the encoding enc, writing
|
||||
// EncodedLen(len(input)) bytes to dst.
|
||||
// EncodedLen(len(src)) bytes to dst.
|
||||
//
|
||||
// The encoding pads the output to a multiple of 4 bytes,
|
||||
// so Encode is not appropriate for use on individual blocks
|
||||
@ -134,7 +134,6 @@ func (e *encoder) Write(p []byte) (n int, err os.Error) {
|
||||
return;
|
||||
}
|
||||
e.enc.Encode(&e.buf, &e.out);
|
||||
var _ int;
|
||||
if _, e.err = e.w.Write(e.out[0:4]); e.err != nil {
|
||||
return n, e.err;
|
||||
}
|
||||
@ -142,7 +141,7 @@ func (e *encoder) Write(p []byte) (n int, err os.Error) {
|
||||
}
|
||||
|
||||
// Large interior chunks.
|
||||
for len(p) > 3 {
|
||||
for len(p) >= 3 {
|
||||
nn := len(e.out)/4*3;
|
||||
if nn > len(p) {
|
||||
nn = len(p);
|
||||
@ -150,7 +149,6 @@ func (e *encoder) Write(p []byte) (n int, err os.Error) {
|
||||
nn -= nn%3;
|
||||
if nn > 0 {
|
||||
e.enc.Encode(p[0:nn], &e.out);
|
||||
var _ int;
|
||||
if _, e.err = e.w.Write(e.out[0 : nn/3*4]); e.err != nil {
|
||||
return n, e.err;
|
||||
}
|
||||
@ -175,7 +173,6 @@ func (e *encoder) Close() os.Error {
|
||||
if e.err == nil && e.nbuf > 0 {
|
||||
e.enc.Encode(e.buf[0 : e.nbuf], &e.out);
|
||||
e.nbuf = 0;
|
||||
var _ int;
|
||||
_, e.err = e.w.Write(e.out[0:4]);
|
||||
}
|
||||
return e.err;
|
||||
@ -206,7 +203,7 @@ func (e CorruptInputError) String() string {
|
||||
return "illegal base64 data at input byte" + strconv.Itoa64(int64(e));
|
||||
}
|
||||
|
||||
// decode is like Decode, but returns an additional 'end' value, which
|
||||
// decode is like Decode but returns an additional 'end' value, which
|
||||
// indicates if end-of-message padding was encountered and thus any
|
||||
// additional data is an error. decode also assumes len(src)%4==0,
|
||||
// since it is meant for internal use.
|
||||
@ -262,7 +259,6 @@ func (enc *Encoding) Decode(src, dst []byte) (n int, err os.Error) {
|
||||
return 0, CorruptInputError(len(src)/4*4);
|
||||
}
|
||||
|
||||
var _ bool;
|
||||
n, _, err = enc.decode(src, dst);
|
||||
return;
|
||||
}
|
||||
|
||||
12
src/pkg/base85/Makefile
Normal file
12
src/pkg/base85/Makefile
Normal file
@ -0,0 +1,12 @@
|
||||
# Copyright 2009 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.
|
||||
|
||||
include $(GOROOT)/src/Make.$(GOARCH)
|
||||
|
||||
TARG=base85
|
||||
GOFILES=\
|
||||
base85.go\
|
||||
git.go\
|
||||
|
||||
include $(GOROOT)/src/Make.pkg
|
||||
307
src/pkg/base85/base85.go
Normal file
307
src/pkg/base85/base85.go
Normal file
@ -0,0 +1,307 @@
|
||||
// Copyright 2009 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.
|
||||
|
||||
// Package base85 implements radix 85 encoding/decoding.
|
||||
package base85
|
||||
|
||||
import (
|
||||
"bytes";
|
||||
"io";
|
||||
"os";
|
||||
"strconv";
|
||||
)
|
||||
|
||||
/*
|
||||
* Encoder
|
||||
*/
|
||||
|
||||
// Encode encodes src into at most MaxEncodedLen(len(src))
|
||||
// bytes of dst, returning the actual number of bytes written.
|
||||
// Encode implements the ascii85 encoding as used in the btoa
|
||||
// tool and Adobe's PostScript and PDF document formats.
|
||||
//
|
||||
// The encoding handles 4-byte chunks, using a special encoding
|
||||
// for the last fragment, so Encode is not appropriate for use on
|
||||
// individual blocks of a large data stream. Use NewEncoder() instead.
|
||||
func Encode(src, dst []byte) int {
|
||||
if len(src) == 0 {
|
||||
return 0;
|
||||
}
|
||||
|
||||
n := 0;
|
||||
for len(src) > 0 {
|
||||
dst[0] = 0;
|
||||
dst[1] = 0;
|
||||
dst[2] = 0;
|
||||
dst[3] = 0;
|
||||
dst[4] = 0;
|
||||
|
||||
// Unpack 4 bytes into uint32 to repack into base 85 5-byte.
|
||||
var v uint32;
|
||||
switch len(src) {
|
||||
default:
|
||||
v |= uint32(src[3]);
|
||||
fallthrough;
|
||||
case 3:
|
||||
v |= uint32(src[2])<<8;
|
||||
fallthrough;
|
||||
case 2:
|
||||
v |= uint32(src[1])<<16;
|
||||
fallthrough;
|
||||
case 1:
|
||||
v |= uint32(src[0])<<24;
|
||||
}
|
||||
|
||||
// Special case: zero (!!!!!) shortens to z.
|
||||
if v == 0 && len(src) >= 4 {
|
||||
dst[0] = 'z';
|
||||
dst = dst[1:len(dst)];
|
||||
n++;
|
||||
continue;
|
||||
}
|
||||
|
||||
// Otherwise, 5 base 85 digits starting at !.
|
||||
for i := 4; i >= 0; i-- {
|
||||
dst[i] = '!' + byte(v%85);
|
||||
v /= 85;
|
||||
}
|
||||
|
||||
// If src was short, discard the low destination bytes.
|
||||
m := 5;
|
||||
if len(src) < 4 {
|
||||
m -= 4 - len(src);
|
||||
src = nil;
|
||||
} else {
|
||||
src = src[4:len(src)];
|
||||
}
|
||||
dst = dst[m:len(dst)];
|
||||
n += m;
|
||||
}
|
||||
return n;
|
||||
}
|
||||
|
||||
// MaxEncodedLen returns the maximum length of an encoding of n source bytes.
|
||||
func MaxEncodedLen(n int) int {
|
||||
return (n+3)/4*5;
|
||||
}
|
||||
|
||||
// NewEncoder returns a new ascii85 stream encoder. Data written to
|
||||
// the returned writer will be encoded and then written to w.
|
||||
// Ascii85 encodings operate in 32-bit blocks; when finished
|
||||
// writing, the caller must Close the returned encoder to flush any
|
||||
// trailing partial block.
|
||||
func NewEncoder(w io.Writer) io.WriteCloser {
|
||||
return &encoder{w: w};
|
||||
}
|
||||
|
||||
type encoder struct {
|
||||
err os.Error;
|
||||
w io.Writer;
|
||||
buf [4]byte; // buffered data waiting to be encoded
|
||||
nbuf int; // number of bytes in buf
|
||||
out [1024]byte; // output buffer
|
||||
}
|
||||
|
||||
func (e *encoder) Write(p []byte) (n int, err os.Error) {
|
||||
if e.err != nil {
|
||||
return 0, e.err;
|
||||
}
|
||||
|
||||
// Leading fringe.
|
||||
if e.nbuf > 0 {
|
||||
var i int;
|
||||
for i = 0; i < len(p) && e.nbuf < 4; i++ {
|
||||
e.buf[e.nbuf] = p[i];
|
||||
e.nbuf++;
|
||||
}
|
||||
n += i;
|
||||
p = p[i:len(p)];
|
||||
if e.nbuf < 4 {
|
||||
return;
|
||||
}
|
||||
nout := Encode(&e.buf, &e.out);
|
||||
var _ int;
|
||||
if _, e.err = e.w.Write(e.out[0:nout]); e.err != nil {
|
||||
return n, e.err;
|
||||
}
|
||||
e.nbuf = 0;
|
||||
}
|
||||
|
||||
// Large interior chunks.
|
||||
for len(p) >= 4 {
|
||||
nn := len(e.out)/5*4;
|
||||
if nn > len(p) {
|
||||
nn = len(p);
|
||||
}
|
||||
nn -= nn%4;
|
||||
if nn > 0 {
|
||||
nout := Encode(p[0:nn], &e.out);
|
||||
var _ int;
|
||||
if _, e.err = e.w.Write(e.out[0:nout]); e.err != nil {
|
||||
return n, e.err;
|
||||
}
|
||||
}
|
||||
n += nn;
|
||||
p = p[nn:len(p)];
|
||||
}
|
||||
|
||||
// Trailing fringe.
|
||||
for i := 0; i < len(p); i++ {
|
||||
e.buf[i] = p[i];
|
||||
}
|
||||
e.nbuf = len(p);
|
||||
n += len(p);
|
||||
return;
|
||||
}
|
||||
|
||||
// Close flushes any pending output from the encoder.
|
||||
// It is an error to call Write after calling Close.
|
||||
func (e *encoder) Close() os.Error {
|
||||
// If there's anything left in the buffer, flush it out
|
||||
if e.err == nil && e.nbuf > 0 {
|
||||
nout := Encode(e.buf[0:e.nbuf], &e.out);
|
||||
e.nbuf = 0;
|
||||
var _ int;
|
||||
_, e.err = e.w.Write(e.out[0:nout]);
|
||||
}
|
||||
return e.err;
|
||||
}
|
||||
|
||||
/*
|
||||
* Decoder
|
||||
*/
|
||||
|
||||
type CorruptInputError int64
|
||||
|
||||
func (e CorruptInputError) String() string {
|
||||
return "illegal ascii85 data at input byte" + strconv.Itoa64(int64(e));
|
||||
}
|
||||
|
||||
// Decode decodes src into dst, returning both the number
|
||||
// of bytes consumed from src and the number written to dst.
|
||||
// If src contains invalid ascii85 data, Decode will return the
|
||||
// number of bytes successfully written and a CorruptInputError.
|
||||
// Decode ignores space and control characters in src.
|
||||
//
|
||||
// If flush is true, Decode assumes that src represents the
|
||||
// end of the input stream and processes it completely rather
|
||||
// than wait for the completion of another 32-bit block.
|
||||
//
|
||||
// NewDecoder wraps an io.Reader interface around Decode.
|
||||
//
|
||||
func Decode(src, dst []byte, flush bool) (nsrc, ndst int, err os.Error) {
|
||||
var v uint32;
|
||||
var nb int;
|
||||
for i, b := range src {
|
||||
if len(dst) - ndst < 4 {
|
||||
return;
|
||||
}
|
||||
switch {
|
||||
case b <= ' ':
|
||||
continue;
|
||||
case b == 'z' && nb == 0:
|
||||
nb = 5;
|
||||
v = 0;
|
||||
case '!' <= b && b <= 'u':
|
||||
v = v*85 + uint32(b - '!');
|
||||
nb++;
|
||||
default:
|
||||
return 0, 0, CorruptInputError(i);
|
||||
}
|
||||
if nb == 5 {
|
||||
nsrc = i+1;
|
||||
dst[ndst] = byte(v>>24);
|
||||
dst[ndst+1] = byte(v>>16);
|
||||
dst[ndst+2] = byte(v>>8);
|
||||
dst[ndst+3] = byte(v);
|
||||
ndst += 4;
|
||||
nb = 0;
|
||||
v = 0;
|
||||
}
|
||||
}
|
||||
if flush {
|
||||
nsrc = len(src);
|
||||
if nb > 0 {
|
||||
// The number of output bytes in the last fragment
|
||||
// is the number of leftover input bytes - 1:
|
||||
// the extra byte provides enough bits to cover
|
||||
// the inefficiency of the encoding for the block.
|
||||
if nb == 1 {
|
||||
return 0, 0, CorruptInputError(len(src));
|
||||
}
|
||||
for i := nb; i < 5; i++ {
|
||||
// The short encoding truncated the output value.
|
||||
// We have to assume the worst case values (digit 84)
|
||||
// in order to ensure that the top bits are correct.
|
||||
v = v*85 + 84;
|
||||
}
|
||||
for i := 0; i < nb-1; i++ {
|
||||
dst[ndst] = byte(v>>24);
|
||||
v <<= 8;
|
||||
ndst++;
|
||||
}
|
||||
}
|
||||
}
|
||||
return;
|
||||
}
|
||||
|
||||
// NewDecoder constructs a new ascii85 stream decoder.
|
||||
func NewDecoder(r io.Reader) io.Reader {
|
||||
return &decoder{r: r};
|
||||
}
|
||||
|
||||
type decoder struct {
|
||||
err os.Error;
|
||||
readErr os.Error;
|
||||
r io.Reader;
|
||||
end bool; // saw end of message
|
||||
buf [1024]byte; // leftover input
|
||||
nbuf int;
|
||||
out []byte; // leftover decoded output
|
||||
outbuf [1024]byte;
|
||||
}
|
||||
|
||||
func (d *decoder) Read(p []byte) (n int, err os.Error) {
|
||||
if len(p) == 0 {
|
||||
return 0, nil;
|
||||
}
|
||||
if d.err != nil {
|
||||
return 0, d.err;
|
||||
}
|
||||
|
||||
for {
|
||||
// Copy leftover output from last decode.
|
||||
if len(d.out) > 0 {
|
||||
n = bytes.Copy(p, d.out);
|
||||
d.out = d.out[n:len(d.out)];
|
||||
return;
|
||||
}
|
||||
|
||||
// Decode leftover input from last read.
|
||||
var nn, nsrc, ndst int;
|
||||
if d.nbuf > 0 {
|
||||
nsrc, ndst, d.err = Decode(d.buf[0:d.nbuf], &d.outbuf, d.readErr != nil);
|
||||
if ndst > 0 {
|
||||
d.out = d.outbuf[0:ndst];
|
||||
d.nbuf = bytes.Copy(&d.buf, d.buf[nsrc:d.nbuf]);
|
||||
continue; // copy out and return
|
||||
}
|
||||
}
|
||||
|
||||
// Out of input, out of decoded output. Check errors.
|
||||
if d.err != nil {
|
||||
return 0, d.err;
|
||||
}
|
||||
if d.readErr != nil {
|
||||
d.err = d.readErr;
|
||||
return 0, d.err;
|
||||
}
|
||||
|
||||
// Read more data.
|
||||
nn, d.readErr = d.r.Read(d.buf[d.nbuf:len(d.buf)]);
|
||||
d.nbuf += nn;
|
||||
}
|
||||
panic("unreachable");
|
||||
}
|
||||
|
||||
193
src/pkg/base85/base85_test.go
Normal file
193
src/pkg/base85/base85_test.go
Normal file
@ -0,0 +1,193 @@
|
||||
// Copyright 2009 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.
|
||||
|
||||
package base85
|
||||
|
||||
import (
|
||||
"bytes";
|
||||
"io";
|
||||
"os";
|
||||
"reflect";
|
||||
"strings";
|
||||
"testing";
|
||||
)
|
||||
|
||||
type testpair struct {
|
||||
decoded, encoded string;
|
||||
}
|
||||
|
||||
var pairs = []testpair{
|
||||
// Wikipedia example
|
||||
testpair{
|
||||
"Man is distinguished, not only by his reason, but by this singular passion from "
|
||||
"other animals, which is a lust of the mind, that by a perseverance of delight in "
|
||||
"the continued and indefatigable generation of knowledge, exceeds the short "
|
||||
"vehemence of any carnal pleasure.",
|
||||
"9jqo^BlbD-BleB1DJ+*+F(f,q/0JhKF<GL>Cj@.4Gp$d7F!,L7@<6@)/0JDEF<G%<+EV:2F!,\n"
|
||||
"O<DJ+*.@<*K0@<6L(Df-\\0Ec5e;DffZ(EZee.Bl.9pF\"AGXBPCsi+DGm>@3BB/F*&OCAfu2/AKY\n"
|
||||
"i(DIb:@FD,*)+C]U=@3BN#EcYf8ATD3s@q?d$AftVqCh[NqF<G:8+EV:.+Cf>-FD5W8ARlolDIa\n"
|
||||
"l(DId<j@<?3r@:F%a+D58'ATD4$Bl@l3De:,-DJs`8ARoFb/0JMK@qB4^F!,R<AKZ&-DfTqBG%G\n"
|
||||
">uD.RTpAKYo'+CT/5+Cei#DII?(E,9)oF*2M7/c\n"
|
||||
}
|
||||
}
|
||||
|
||||
var bigtest = pairs[len(pairs)-1];
|
||||
|
||||
func testEqual(t *testing.T, msg string, args ...) bool {
|
||||
v := reflect.NewValue(args).(*reflect.StructValue);
|
||||
v1 := v.Field(v.NumField() - 2);
|
||||
v2 := v.Field(v.NumField() - 1);
|
||||
if v1.Interface() != v2.Interface() {
|
||||
t.Errorf(msg, args);
|
||||
return false;
|
||||
}
|
||||
return true;
|
||||
}
|
||||
|
||||
func strip85(s string) string {
|
||||
t := make([]byte, len(s));
|
||||
w := 0;
|
||||
for r := 0; r < len(s); r++ {
|
||||
c := s[r];
|
||||
if c > ' ' {
|
||||
t[w] = c;
|
||||
w++;
|
||||
}
|
||||
}
|
||||
return string(t[0:w]);
|
||||
}
|
||||
|
||||
func TestEncode(t *testing.T) {
|
||||
for _, p := range pairs {
|
||||
buf := make([]byte, MaxEncodedLen(len(p.decoded)));
|
||||
n := Encode(strings.Bytes(p.decoded), buf);
|
||||
buf = buf[0:n];
|
||||
testEqual(t, "Encode(%q) = %q, want %q", p.decoded, strip85(string(buf)), strip85(p.encoded));
|
||||
}
|
||||
}
|
||||
|
||||
func TestEncoder(t *testing.T) {
|
||||
for _, p := range pairs {
|
||||
bb := &bytes.Buffer{};
|
||||
encoder := NewEncoder(bb);
|
||||
encoder.Write(strings.Bytes(p.decoded));
|
||||
encoder.Close();
|
||||
testEqual(t, "Encode(%q) = %q, want %q", p.decoded, strip85(bb.String()), strip85(p.encoded));
|
||||
}
|
||||
}
|
||||
|
||||
func TestEncoderBuffering(t *testing.T) {
|
||||
input := strings.Bytes(bigtest.decoded);
|
||||
for bs := 1; bs <= 12; bs++ {
|
||||
bb := &bytes.Buffer{};
|
||||
encoder := NewEncoder(bb);
|
||||
for pos := 0; pos < len(input); pos += bs {
|
||||
end := pos+bs;
|
||||
if end > len(input) {
|
||||
end = len(input);
|
||||
}
|
||||
n, err := encoder.Write(input[pos:end]);
|
||||
testEqual(t, "Write(%q) gave error %v, want %v", input[pos:end], err, os.Error(nil));
|
||||
testEqual(t, "Write(%q) gave length %v, want %v", input[pos:end], n, end-pos);
|
||||
}
|
||||
err := encoder.Close();
|
||||
testEqual(t, "Close gave error %v, want %v", err, os.Error(nil));
|
||||
testEqual(t, "Encoding/%d of %q = %q, want %q", bs, bigtest.decoded, strip85(bb.String()), strip85(bigtest.encoded));
|
||||
}
|
||||
}
|
||||
|
||||
func TestDecode(t *testing.T) {
|
||||
for _, p := range pairs {
|
||||
dbuf := make([]byte, 4*len(p.encoded));
|
||||
nsrc, ndst, err := Decode(strings.Bytes(p.encoded), dbuf, true);
|
||||
testEqual(t, "Decode(%q) = error %v, want %v", p.encoded, err, os.Error(nil));
|
||||
testEqual(t, "Decode(%q) = nsrc %v, want %v", p.encoded, nsrc, len(p.encoded));
|
||||
testEqual(t, "Decode(%q) = ndst %v, want %v", p.encoded, ndst, len(p.decoded));
|
||||
testEqual(t, "Decode(%q) = %q, want %q", p.encoded, string(dbuf[0:ndst]), p.decoded);
|
||||
}
|
||||
}
|
||||
|
||||
func TestDecoder(t *testing.T) {
|
||||
for _, p := range pairs {
|
||||
decoder := NewDecoder(bytes.NewBufferString(p.encoded));
|
||||
dbuf, err := io.ReadAll(decoder);
|
||||
if err != nil {
|
||||
t.Fatal("Read failed", err);
|
||||
}
|
||||
testEqual(t, "Read from %q = length %v, want %v", p.encoded, len(dbuf), len(p.decoded));
|
||||
testEqual(t, "Decoding of %q = %q, want %q", p.encoded, string(dbuf), p.decoded);
|
||||
if err != nil {
|
||||
testEqual(t, "Read from %q = %v, want %v", p.encoded, err, os.EOF);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
func TestDecoderBuffering(t *testing.T) {
|
||||
for bs := 1; bs <= 12; bs++ {
|
||||
decoder := NewDecoder(bytes.NewBufferString(bigtest.encoded));
|
||||
buf := make([]byte, len(bigtest.decoded)+12);
|
||||
var total int;
|
||||
for total = 0; total < len(bigtest.decoded); {
|
||||
n, err := decoder.Read(buf[total : total+bs]);
|
||||
testEqual(t, "Read from %q at pos %d = %d, %v, want _, %v", bigtest.encoded, total, n, err, os.Error(nil));
|
||||
total += n;
|
||||
}
|
||||
testEqual(t, "Decoding/%d of %q = %q, want %q", bs, bigtest.encoded, string(buf[0:total]), bigtest.decoded);
|
||||
}
|
||||
}
|
||||
|
||||
func TestDecodeCorrupt(t *testing.T) {
|
||||
type corrupt struct {
|
||||
e string;
|
||||
p int;
|
||||
}
|
||||
examples := []corrupt{
|
||||
corrupt{"v", 0},
|
||||
corrupt{"!z!!!!!!!!!", 1},
|
||||
};
|
||||
|
||||
for _, e := range examples {
|
||||
dbuf := make([]byte, 4*len(e.e));
|
||||
_, _, err := Decode(strings.Bytes(e.e), dbuf, true);
|
||||
switch err := err.(type) {
|
||||
case CorruptInputError:
|
||||
testEqual(t, "Corruption in %q at offset %v, want %v", e.e, int(err), e.p);
|
||||
default:
|
||||
t.Error("Decoder failed to detect corruption in", e);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
func TestBig(t *testing.T) {
|
||||
n := 3*1000 + 1;
|
||||
raw := make([]byte, n);
|
||||
const alpha = "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ";
|
||||
for i := 0; i < n; i++ {
|
||||
raw[i] = alpha[i%len(alpha)];
|
||||
}
|
||||
encoded := new(bytes.Buffer);
|
||||
w := NewEncoder(encoded);
|
||||
nn, err := w.Write(raw);
|
||||
if nn != n || err != nil {
|
||||
t.Fatalf("Encoder.Write(raw) = %d, %v want %d, nil", nn, err, n);
|
||||
}
|
||||
err = w.Close();
|
||||
if err != nil {
|
||||
t.Fatalf("Encoder.Close() = %v want nil", err);
|
||||
}
|
||||
decoded, err := io.ReadAll(NewDecoder(encoded));
|
||||
if err != nil {
|
||||
t.Fatalf("io.ReadAll(NewDecoder(...)): %v", err);
|
||||
}
|
||||
|
||||
if !bytes.Equal(raw, decoded) {
|
||||
var i int;
|
||||
for i = 0; i < len(decoded) && i < len(raw); i++ {
|
||||
if decoded[i] != raw[i] {
|
||||
break;
|
||||
}
|
||||
}
|
||||
t.Errorf("Decode(Encode(%d-byte string)) failed at offset %d", n, i);
|
||||
}
|
||||
}
|
||||
275
src/pkg/base85/git.go
Normal file
275
src/pkg/base85/git.go
Normal file
@ -0,0 +1,275 @@
|
||||
// Copyright 2009 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.
|
||||
|
||||
package base85
|
||||
|
||||
import (
|
||||
"bytes";
|
||||
"io";
|
||||
"os";
|
||||
)
|
||||
|
||||
const gitEncode = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz!#$%&()*+-;<=>?@^_`{|}~"
|
||||
|
||||
// The decodings are 1+ the actual value, so that the
|
||||
// default zero value can be used to mean "not valid".
|
||||
var gitDecode = [256]uint8{
|
||||
'0': 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
|
||||
'A': 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23,
|
||||
24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36,
|
||||
'a': 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49,
|
||||
50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62,
|
||||
'!': 63,
|
||||
'#': 64, 65, 66, 67,
|
||||
'(': 68, 69, 70, 71,
|
||||
'-': 72,
|
||||
';': 73,
|
||||
'<': 74, 75, 76, 77,
|
||||
'@': 78,
|
||||
'^': 79, 80, 81,
|
||||
'{': 82, 83, 84, 85
|
||||
}
|
||||
|
||||
// GitEncode encodes src into GitEncodedLen(len(src))
|
||||
// bytes of dst. As a convenience, it returns the number
|
||||
// of bytes written to dst, but this value is always GitEncodedLen(len(src)).
|
||||
// GitEncode implements the radix 85 encoding used in the
|
||||
// Git version control tool.
|
||||
//
|
||||
// The encoding splits src into chunks of at most 52 bytes
|
||||
// and encodes each chunk on its own line.
|
||||
func GitEncode(src, dst []byte) int {
|
||||
ndst := 0;
|
||||
for len(src) > 0 {
|
||||
n := len(src);
|
||||
if n > 52 {
|
||||
n = 52;
|
||||
}
|
||||
if n <= 27 {
|
||||
dst[ndst] = byte('A' + n - 1);
|
||||
} else {
|
||||
dst[ndst] = byte('a' + n - 26 - 1);
|
||||
}
|
||||
ndst++;
|
||||
for i := 0; i < n; i += 4 {
|
||||
var v uint32;
|
||||
for j := 0; j < 4 && i+j < n; j++ {
|
||||
v |= uint32(src[i+j]) << uint(24 - j*8);
|
||||
}
|
||||
for j := 4; j >= 0; j-- {
|
||||
dst[ndst+j] = gitEncode[v%85];
|
||||
v /= 85;
|
||||
}
|
||||
ndst += 5;
|
||||
}
|
||||
dst[ndst] = '\n';
|
||||
ndst++;
|
||||
src = src[n:len(src)];
|
||||
}
|
||||
return ndst;
|
||||
}
|
||||
|
||||
// GitEncodedLen returns the length of an encoding of n source bytes.
|
||||
func GitEncodedLen(n int) int {
|
||||
if n == 0 {
|
||||
return 0;
|
||||
}
|
||||
// 5 bytes per 4 bytes of input, rounded up.
|
||||
// 2 extra bytes for each line of 52 src bytes, rounded up.
|
||||
return (n+3)/4*5 + (n+51)/52*2;
|
||||
}
|
||||
|
||||
var newline = []byte{'\n'}
|
||||
|
||||
// GitDecode decodes src into at most MaxGitDecodedLen(len(src))
|
||||
// bytes, returning the actual number of bytes written to dst.
|
||||
//
|
||||
// If GitDecode encounters invalid input, it returns a CorruptInputError.
|
||||
//
|
||||
func GitDecode(src, dst []byte) (n int, err os.Error) {
|
||||
ndst := 0;
|
||||
nsrc := 0;
|
||||
for nsrc < len(src) {
|
||||
var l int;
|
||||
switch ch := int(src[nsrc]); {
|
||||
case 'A' <= ch && ch <= 'Z':
|
||||
l = ch - 'A' + 1;
|
||||
case 'a' <= ch && ch <= 'z':
|
||||
l = ch - 'a' + 26 + 1;
|
||||
default:
|
||||
return ndst, CorruptInputError(nsrc);
|
||||
}
|
||||
if nsrc+1+l > len(src) {
|
||||
return ndst, CorruptInputError(nsrc);
|
||||
}
|
||||
el := (l+3)/4*5; // encoded len
|
||||
if nsrc+1+el+1 > len(src) || src[nsrc+1+el] != '\n' {
|
||||
return ndst, CorruptInputError(nsrc);
|
||||
}
|
||||
line := src[nsrc+1:nsrc+1+el];
|
||||
for i := 0; i < el; i += 5 {
|
||||
var v uint32;
|
||||
for j := 0; j < 5; j++ {
|
||||
ch := gitDecode[line[i+j]];
|
||||
if ch == 0 {
|
||||
return ndst, CorruptInputError(nsrc+1+i+j);
|
||||
}
|
||||
v = v*85 + uint32(ch-1);
|
||||
}
|
||||
for j := 0; j < 4; j++ {
|
||||
dst[ndst] = byte(v>>24);
|
||||
v <<= 8;
|
||||
ndst++;
|
||||
}
|
||||
}
|
||||
// Last fragment may have run too far (but there was room in dst).
|
||||
// Back up.
|
||||
if l%4 != 0 {
|
||||
ndst -= 4 - l%4;
|
||||
}
|
||||
nsrc += 1+el+1;
|
||||
}
|
||||
return ndst, nil;
|
||||
}
|
||||
|
||||
func MaxGitDecodedLen(n int) int {
|
||||
return n/5*4;
|
||||
}
|
||||
|
||||
// NewEncoder returns a new Git base85 stream encoder. Data written to
|
||||
// the returned writer will be encoded and then written to w.
|
||||
// The Git encoding operates on 52-byte blocks; when finished
|
||||
// writing, the caller must Close the returned encoder to flush any
|
||||
// partially written blocks.
|
||||
func NewGitEncoder(w io.Writer) io.WriteCloser {
|
||||
return &gitEncoder{w: w};
|
||||
}
|
||||
|
||||
type gitEncoder struct {
|
||||
w io.Writer;
|
||||
err os.Error;
|
||||
buf [52]byte;
|
||||
nbuf int;
|
||||
out [1024]byte;
|
||||
nout int;
|
||||
}
|
||||
|
||||
func (e *gitEncoder) Write(p []byte) (n int, err os.Error) {
|
||||
if e.err != nil {
|
||||
return 0, e.err;
|
||||
}
|
||||
|
||||
// Leading fringe.
|
||||
if e.nbuf > 0 {
|
||||
var i int;
|
||||
for i = 0; i < len(p) && e.nbuf < 52; i++ {
|
||||
e.buf[e.nbuf] = p[i];
|
||||
e.nbuf++;
|
||||
}
|
||||
n += i;
|
||||
p = p[i:len(p)];
|
||||
if e.nbuf < 52 {
|
||||
return;
|
||||
}
|
||||
nout := GitEncode(&e.buf, &e.out);
|
||||
if _, e.err = e.w.Write(e.out[0:nout]); e.err != nil {
|
||||
return n, e.err;
|
||||
}
|
||||
e.nbuf = 0;
|
||||
}
|
||||
|
||||
// Large interior chunks.
|
||||
for len(p) >= 52 {
|
||||
nn := len(e.out)/(1+52/4*5+1) * 52;
|
||||
if nn > len(p) {
|
||||
nn = len(p)/52 * 52;
|
||||
}
|
||||
if nn > 0 {
|
||||
nout := GitEncode(p[0:nn], &e.out);
|
||||
if _, e.err = e.w.Write(e.out[0:nout]); e.err != nil {
|
||||
return n, e.err;
|
||||
}
|
||||
}
|
||||
n += nn;
|
||||
p = p[nn:len(p)];
|
||||
}
|
||||
|
||||
// Trailing fringe.
|
||||
for i := 0; i < len(p); i++ {
|
||||
e.buf[i] = p[i];
|
||||
}
|
||||
e.nbuf = len(p);
|
||||
n += len(p);
|
||||
return;
|
||||
}
|
||||
|
||||
func (e *gitEncoder) Close() os.Error {
|
||||
// If there's anything left in the buffer, flush it out
|
||||
if e.err == nil && e.nbuf > 0 {
|
||||
nout := GitEncode(e.buf[0:e.nbuf], &e.out);
|
||||
e.nbuf = 0;
|
||||
_, e.err = e.w.Write(e.out[0:nout]);
|
||||
}
|
||||
return e.err;
|
||||
}
|
||||
|
||||
// NewGitDecoder returns a new Git base85 stream decoder.
|
||||
func NewGitDecoder(r io.Reader) io.Reader {
|
||||
return &gitDecoder{r: r};
|
||||
}
|
||||
|
||||
type gitDecoder struct {
|
||||
r io.Reader;
|
||||
err os.Error;
|
||||
readErr os.Error;
|
||||
buf [1024]byte;
|
||||
nbuf int;
|
||||
out []byte;
|
||||
outbuf [1024]byte;
|
||||
off int64;
|
||||
}
|
||||
|
||||
func (d *gitDecoder) Read(p []byte) (n int, err os.Error) {
|
||||
if len(p) == 0 {
|
||||
return 0, nil;
|
||||
}
|
||||
|
||||
for {
|
||||
// Copy leftover output from last decode.
|
||||
if len(d.out) > 0 {
|
||||
n = bytes.Copy(p, d.out);
|
||||
d.out = d.out[n:len(d.out)];
|
||||
return;
|
||||
}
|
||||
|
||||
// Out of decoded output. Check errors.
|
||||
if d.err != nil {
|
||||
return 0, d.err;
|
||||
}
|
||||
if d.readErr != nil {
|
||||
d.err = d.readErr;
|
||||
return 0, d.err;
|
||||
}
|
||||
|
||||
// Read and decode more input.
|
||||
var nn int;
|
||||
nn, d.readErr = d.r.Read(d.buf[d.nbuf:len(d.buf)]);
|
||||
d.nbuf += nn;
|
||||
|
||||
// Send complete lines to GitDecode.
|
||||
nl := bytes.LastIndex(d.buf[0:d.nbuf], newline);
|
||||
if nl < 0 {
|
||||
continue;
|
||||
}
|
||||
nn, d.err = GitDecode(d.buf[0:nl+1], &d.outbuf);
|
||||
if e, ok := d.err.(CorruptInputError); ok {
|
||||
d.err = CorruptInputError(int64(e)+d.off);
|
||||
}
|
||||
d.out = d.outbuf[0:nn];
|
||||
d.nbuf = bytes.Copy(&d.buf, d.buf[nl+1:d.nbuf]);
|
||||
d.off += int64(nl+1);
|
||||
}
|
||||
panic("unreacahable");
|
||||
}
|
||||
|
||||
183
src/pkg/base85/git_test.go
Normal file
183
src/pkg/base85/git_test.go
Normal file
@ -0,0 +1,183 @@
|
||||
// Copyright 2009 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.
|
||||
|
||||
package base85
|
||||
|
||||
import (
|
||||
"bytes";
|
||||
"io";
|
||||
"os";
|
||||
"strings";
|
||||
"testing";
|
||||
)
|
||||
|
||||
func TestGitTable(t *testing.T) {
|
||||
var saw [256]bool;
|
||||
for i, c := range gitEncode {
|
||||
if gitDecode[c] != uint8(i+1) {
|
||||
t.Errorf("gitDecode['%c'] = %d, want %d", c, gitDecode[c], i+1);
|
||||
}
|
||||
saw[c] = true;
|
||||
}
|
||||
for i, b := range saw {
|
||||
if !b && gitDecode[i] != 0 {
|
||||
t.Errorf("gitDecode[%d] = %d, want 0", i, gitDecode[i]);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
var gitPairs = []testpair{
|
||||
// Wikipedia example, adapted.
|
||||
testpair{
|
||||
"Man is distinguished, not only by his reason, but by this singular passion from "
|
||||
"other animals, which is a lust of the mind, that by a perseverance of delight in "
|
||||
"the continued and indefatigable generation of knowledge, exceeds the short "
|
||||
"vehemence of any carnal pleasure.",
|
||||
|
||||
"zO<`^zX>%ZCX>)XGZfA9Ab7*B`EFf-gbRchTY<VDJc_3(Mb0BhMVRLV8EFfZabRc4R\n"
|
||||
"zAarPHb0BkRZfA9DVR9gFVRLh7Z*CxFa&K)QZ**v7av))DX>DO_b1WctXlY|;AZc?T\n"
|
||||
"zVIXXEb95kYW*~HEWgu;7Ze%PVbZB98AYyqSVIXj2a&u*NWpZI|V`U(3W*}r`Y-wj`\n"
|
||||
"zbRcPNAarPDAY*TCbZKsNWn>^>Ze$>7Ze(R<VRUI{VPb4$AZKN6WpZJ3X>V>IZ)PBC\n"
|
||||
"zZf|#NWn^b%EFfigV`XJzb0BnRWgv5CZ*p`Xc4cT~ZDnp_Wgu^6AYpEKAY);2ZeeU7\n"
|
||||
"IaBO8^b9HiME&u=k\n"
|
||||
}
|
||||
}
|
||||
|
||||
var gitBigtest = gitPairs[len(gitPairs)-1];
|
||||
|
||||
func TestGitEncode(t *testing.T) {
|
||||
for _, p := range gitPairs {
|
||||
buf := make([]byte, GitEncodedLen(len(p.decoded)));
|
||||
n := GitEncode(strings.Bytes(p.decoded), buf);
|
||||
if n != len(buf) {
|
||||
t.Errorf("GitEncodedLen does not agree with GitEncode");
|
||||
}
|
||||
buf = buf[0:n];
|
||||
testEqual(t, "GitEncode(%q) = %q, want %q", p.decoded, string(buf), p.encoded);
|
||||
}
|
||||
}
|
||||
|
||||
func TestGitEncoder(t *testing.T) {
|
||||
for _, p := range gitPairs {
|
||||
bb := &bytes.Buffer{};
|
||||
encoder := NewGitEncoder(bb);
|
||||
encoder.Write(strings.Bytes(p.decoded));
|
||||
encoder.Close();
|
||||
testEqual(t, "GitEncode(%q) = %q, want %q", p.decoded, bb.String(), p.encoded);
|
||||
}
|
||||
}
|
||||
|
||||
func TestGitEncoderBuffering(t *testing.T) {
|
||||
input := strings.Bytes(gitBigtest.decoded);
|
||||
for bs := 1; bs <= 12; bs++ {
|
||||
bb := &bytes.Buffer{};
|
||||
encoder := NewGitEncoder(bb);
|
||||
for pos := 0; pos < len(input); pos += bs {
|
||||
end := pos+bs;
|
||||
if end > len(input) {
|
||||
end = len(input);
|
||||
}
|
||||
n, err := encoder.Write(input[pos:end]);
|
||||
testEqual(t, "Write(%q) gave error %v, want %v", input[pos:end], err, os.Error(nil));
|
||||
testEqual(t, "Write(%q) gave length %v, want %v", input[pos:end], n, end-pos);
|
||||
}
|
||||
err := encoder.Close();
|
||||
testEqual(t, "Close gave error %v, want %v", err, os.Error(nil));
|
||||
testEqual(t, "Encoding/%d of %q = %q, want %q", bs, gitBigtest.decoded, bb.String(), gitBigtest.encoded);
|
||||
}
|
||||
}
|
||||
|
||||
func TestGitDecode(t *testing.T) {
|
||||
for _, p := range gitPairs {
|
||||
dbuf := make([]byte, 4*len(p.encoded));
|
||||
ndst, err := GitDecode(strings.Bytes(p.encoded), dbuf);
|
||||
testEqual(t, "GitDecode(%q) = error %v, want %v", p.encoded, err, os.Error(nil));
|
||||
testEqual(t, "GitDecode(%q) = ndst %v, want %v", p.encoded, ndst, len(p.decoded));
|
||||
testEqual(t, "GitDecode(%q) = %q, want %q", p.encoded, string(dbuf[0:ndst]), p.decoded);
|
||||
}
|
||||
}
|
||||
|
||||
func TestGitDecoder(t *testing.T) {
|
||||
for _, p := range gitPairs {
|
||||
decoder := NewGitDecoder(bytes.NewBufferString(p.encoded));
|
||||
dbuf, err := io.ReadAll(decoder);
|
||||
if err != nil {
|
||||
t.Fatal("Read failed", err);
|
||||
}
|
||||
testEqual(t, "Read from %q = length %v, want %v", p.encoded, len(dbuf), len(p.decoded));
|
||||
testEqual(t, "Decoding of %q = %q, want %q", p.encoded, string(dbuf), p.decoded);
|
||||
if err != nil {
|
||||
testEqual(t, "Read from %q = %v, want %v", p.encoded, err, os.EOF);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
func TestGitDecoderBuffering(t *testing.T) {
|
||||
for bs := 1; bs <= 12; bs++ {
|
||||
decoder := NewGitDecoder(bytes.NewBufferString(gitBigtest.encoded));
|
||||
buf := make([]byte, len(gitBigtest.decoded)+12);
|
||||
var total int;
|
||||
for total = 0; total < len(gitBigtest.decoded); {
|
||||
n, err := decoder.Read(buf[total : total+bs]);
|
||||
testEqual(t, "Read from %q at pos %d = %d, %v, want _, %v", gitBigtest.encoded, total, n, err, os.Error(nil));
|
||||
total += n;
|
||||
}
|
||||
testEqual(t, "Decoding/%d of %q = %q, want %q", bs, gitBigtest.encoded, string(buf[0:total]), gitBigtest.decoded);
|
||||
}
|
||||
}
|
||||
|
||||
func TestGitDecodeCorrupt(t *testing.T) {
|
||||
type corrupt struct {
|
||||
e string;
|
||||
p int;
|
||||
}
|
||||
examples := []corrupt{
|
||||
corrupt{"v", 0},
|
||||
corrupt{"!z!!!!!!!!!", 0},
|
||||
};
|
||||
|
||||
for _, e := range examples {
|
||||
dbuf := make([]byte, 2*len(e.e));
|
||||
_, err := GitDecode(strings.Bytes(e.e), dbuf);
|
||||
switch err := err.(type) {
|
||||
case CorruptInputError:
|
||||
testEqual(t, "Corruption in %q at offset %v, want %v", e.e, int(err), e.p);
|
||||
default:
|
||||
t.Error("GitDecoder failed to detect corruption in", e);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
func TestGitBig(t *testing.T) {
|
||||
n := 3*1000 + 1;
|
||||
raw := make([]byte, n);
|
||||
const alpha = "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ";
|
||||
for i := 0; i < n; i++ {
|
||||
raw[i] = alpha[i%len(alpha)];
|
||||
}
|
||||
encoded := new(bytes.Buffer);
|
||||
w := NewGitEncoder(encoded);
|
||||
nn, err := w.Write(raw);
|
||||
if nn != n || err != nil {
|
||||
t.Fatalf("GitEncoder.Write(raw) = %d, %v want %d, nil", nn, err, n);
|
||||
}
|
||||
err = w.Close();
|
||||
if err != nil {
|
||||
t.Fatalf("GitEncoder.Close() = %v want nil", err);
|
||||
}
|
||||
decoded, err := io.ReadAll(NewGitDecoder(encoded));
|
||||
if err != nil {
|
||||
t.Fatalf("io.ReadAll(NewGitDecoder(...)): %v", err);
|
||||
}
|
||||
|
||||
if !bytes.Equal(raw, decoded) {
|
||||
var i int;
|
||||
for i = 0; i < len(decoded) && i < len(raw); i++ {
|
||||
if decoded[i] != raw[i] {
|
||||
break;
|
||||
}
|
||||
}
|
||||
t.Errorf("GitDecode(GitEncode(%d-byte string)) failed at offset %d", n, i);
|
||||
}
|
||||
}
|
||||
Loading…
Reference in New Issue
Block a user