diff options
author | upstream source tree <ports@midipix.org> | 2015-03-15 20:14:05 -0400 |
---|---|---|
committer | upstream source tree <ports@midipix.org> | 2015-03-15 20:14:05 -0400 |
commit | 554fd8c5195424bdbcabf5de30fdc183aba391bd (patch) | |
tree | 976dc5ab7fddf506dadce60ae936f43f58787092 /libgo/go/encoding/git85 | |
download | cbb-gcc-4.6.4-554fd8c5195424bdbcabf5de30fdc183aba391bd.tar.bz2 cbb-gcc-4.6.4-554fd8c5195424bdbcabf5de30fdc183aba391bd.tar.xz |
obtained gcc-4.6.4.tar.bz2 from upstream website;upstream
verified gcc-4.6.4.tar.bz2.sig;
imported gcc-4.6.4 source tree from verified upstream tarball.
downloading a git-generated archive based on the 'upstream' tag
should provide you with a source tree that is binary identical
to the one extracted from the above tarball.
if you have obtained the source via the command 'git clone',
however, do note that line-endings of files in your working
directory might differ from line-endings of the respective
files in the upstream repository.
Diffstat (limited to 'libgo/go/encoding/git85')
-rw-r--r-- | libgo/go/encoding/git85/git.go | 277 | ||||
-rw-r--r-- | libgo/go/encoding/git85/git_test.go | 194 |
2 files changed, 471 insertions, 0 deletions
diff --git a/libgo/go/encoding/git85/git.go b/libgo/go/encoding/git85/git.go new file mode 100644 index 000000000..09a45cd3c --- /dev/null +++ b/libgo/go/encoding/git85/git.go @@ -0,0 +1,277 @@ +// 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 git85 implements the radix 85 data encoding +// used in the Git version control system. +package git85 + +import ( + "bytes" + "io" + "os" + "strconv" +) + +type CorruptInputError int64 + +func (e CorruptInputError) String() string { + return "illegal git85 data at input byte " + strconv.Itoa64(int64(e)) +} + +const encode = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz!#$%&()*+-;<=>?@^_`{|}~" + +// The decodings are 1+ the actual value, so that the +// default zero value can be used to mean "not valid". +var decode = [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, +} + +// Encode encodes src into EncodedLen(len(src)) +// bytes of dst. As a convenience, it returns the number +// of bytes written to dst, but this value is always EncodedLen(len(src)). +// Encode 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 Encode(dst, src []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] = encode[v%85] + v /= 85 + } + ndst += 5 + } + dst[ndst] = '\n' + ndst++ + src = src[n:] + } + return ndst +} + +// EncodedLen returns the length of an encoding of n source bytes. +func EncodedLen(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'} + +// Decode decodes src into at most MaxDecodedLen(len(src)) +// bytes, returning the actual number of bytes written to dst. +// +// If Decode encounters invalid input, it returns a CorruptInputError. +// +func Decode(dst, src []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 := decode[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 MaxDecodedLen(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 NewEncoder(w io.Writer) io.WriteCloser { return &encoder{w: w} } + +type encoder struct { + w io.Writer + err os.Error + buf [52]byte + nbuf int + out [1024]byte + nout int +} + +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 < 52; i++ { + e.buf[e.nbuf] = p[i] + e.nbuf++ + } + n += i + p = p[i:] + if e.nbuf < 52 { + return + } + nout := Encode(e.out[0:], e.buf[0:]) + 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 := Encode(e.out[0:], p[0:nn]) + if _, e.err = e.w.Write(e.out[0:nout]); e.err != nil { + return n, e.err + } + } + n += nn + p = p[nn:] + } + + // Trailing fringe. + for i := 0; i < len(p); i++ { + e.buf[i] = p[i] + } + e.nbuf = len(p) + n += len(p) + return +} + +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.out[0:], e.buf[0:e.nbuf]) + e.nbuf = 0 + _, e.err = e.w.Write(e.out[0:nout]) + } + return e.err +} + +// NewDecoder returns a new Git base85 stream decoder. +func NewDecoder(r io.Reader) io.Reader { return &decoder{r: r} } + +type decoder struct { + r io.Reader + err os.Error + readErr os.Error + buf [1024]byte + nbuf int + out []byte + outbuf [1024]byte + off int64 +} + +func (d *decoder) 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 = copy(p, d.out) + d.out = d.out[n:] + 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:]) + d.nbuf += nn + + // Send complete lines to Decode. + nl := bytes.LastIndex(d.buf[0:d.nbuf], newline) + if nl < 0 { + continue + } + nn, d.err = Decode(d.outbuf[0:], d.buf[0:nl+1]) + if e, ok := d.err.(CorruptInputError); ok { + d.err = CorruptInputError(int64(e) + d.off) + } + d.out = d.outbuf[0:nn] + d.nbuf = copy(d.buf[0:], d.buf[nl+1:d.nbuf]) + d.off += int64(nl + 1) + } + panic("unreacahable") +} diff --git a/libgo/go/encoding/git85/git_test.go b/libgo/go/encoding/git85/git_test.go new file mode 100644 index 000000000..c76385c35 --- /dev/null +++ b/libgo/go/encoding/git85/git_test.go @@ -0,0 +1,194 @@ +// 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 git85 + +import ( + "bytes" + "io/ioutil" + "os" + "testing" +) + +type testpair struct { + decoded, encoded string +} + +func testEqual(t *testing.T, msg string, args ...interface{}) bool { + if args[len(args)-2] != args[len(args)-1] { + t.Errorf(msg, args...) + return false + } + return true +} + +func TestGitTable(t *testing.T) { + var saw [256]bool + for i, c := range encode { + if decode[c] != uint8(i+1) { + t.Errorf("decode['%c'] = %d, want %d", c, decode[c], i+1) + } + saw[c] = true + } + for i, b := range saw { + if !b && decode[i] != 0 { + t.Errorf("decode[%d] = %d, want 0", i, decode[i]) + } + } +} + +var gitPairs = []testpair{ + // Wikipedia example, adapted. + { + "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 TestEncode(t *testing.T) { + for _, p := range gitPairs { + buf := make([]byte, EncodedLen(len(p.decoded))) + n := Encode(buf, []byte(p.decoded)) + if n != len(buf) { + t.Errorf("EncodedLen does not agree with Encode") + } + buf = buf[0:n] + testEqual(t, "Encode(%q) = %q, want %q", p.decoded, string(buf), p.encoded) + } +} + +func TestEncoder(t *testing.T) { + for _, p := range gitPairs { + bb := &bytes.Buffer{} + encoder := NewEncoder(bb) + encoder.Write([]byte(p.decoded)) + encoder.Close() + testEqual(t, "Encode(%q) = %q, want %q", p.decoded, bb.String(), p.encoded) + } +} + +func TestEncoderBuffering(t *testing.T) { + input := []byte(gitBigtest.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, gitBigtest.decoded, bb.String(), gitBigtest.encoded) + } +} + +func TestDecode(t *testing.T) { + for _, p := range gitPairs { + dbuf := make([]byte, 4*len(p.encoded)) + ndst, err := Decode(dbuf, []byte(p.encoded)) + testEqual(t, "Decode(%q) = error %v, want %v", p.encoded, err, os.Error(nil)) + 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 gitPairs { + decoder := NewDecoder(bytes.NewBufferString(p.encoded)) + dbuf, err := ioutil.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(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 TestDecodeCorrupt(t *testing.T) { + type corrupt struct { + e string + p int + } + examples := []corrupt{ + {"v", 0}, + {"!z!!!!!!!!!", 0}, + } + + for _, e := range examples { + dbuf := make([]byte, 2*len(e.e)) + _, err := Decode(dbuf, []byte(e.e)) + 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 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 := 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 := ioutil.ReadAll(NewDecoder(encoded)) + if err != nil { + t.Fatalf("ioutil.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) + } +} |