diff --git a/src/pkg/Makefile b/src/pkg/Makefile index 3494e1c13a..6e3c7f05f7 100644 --- a/src/pkg/Makefile +++ b/src/pkg/Makefile @@ -58,6 +58,7 @@ DIRS=\ debug/proc\ ebnf\ encoding/ascii85\ + encoding/base32\ encoding/base64\ encoding/binary\ encoding/git85\ diff --git a/src/pkg/encoding/base32/Makefile b/src/pkg/encoding/base32/Makefile new file mode 100644 index 0000000000..c0e85b6448 --- /dev/null +++ b/src/pkg/encoding/base32/Makefile @@ -0,0 +1,11 @@ +# 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.inc + +TARG=encoding/base32 +GOFILES=\ + base32.go\ + +include $(GOROOT)/src/Make.pkg diff --git a/src/pkg/encoding/base32/base32.go b/src/pkg/encoding/base32/base32.go new file mode 100644 index 0000000000..acace30d6a --- /dev/null +++ b/src/pkg/encoding/base32/base32.go @@ -0,0 +1,368 @@ +// Copyright 2011 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 base32 implements base32 encoding as specified by RFC 4648. +package base32 + +import ( + "io" + "os" + "strconv" +) + +/* + * Encodings + */ + +// An Encoding is a radix 32 encoding/decoding scheme, defined by a +// 32-character alphabet. The most common is the "base32" encoding +// introduced for SASL GSSAPI and standardized in RFC 4648. +// The alternate "base32hex" encoding is used in DNSSEC. +type Encoding struct { + encode string + decodeMap [256]byte +} + +const encodeStd = "ABCDEFGHIJKLMNOPQRSTUVWXYZ234567" +const encodeHex = "0123456789ABCDEFGHIJKLMNOPQRSTUV" + +// NewEncoding returns a new Encoding defined by the given alphabet, +// which must be a 32-byte string. +func NewEncoding(encoder string) *Encoding { + e := new(Encoding) + e.encode = encoder + for i := 0; i < len(e.decodeMap); i++ { + e.decodeMap[i] = 0xFF + } + for i := 0; i < len(encoder); i++ { + e.decodeMap[encoder[i]] = byte(i) + } + return e +} + +// StdEncoding is the standard base32 encoding, as defined in +// RFC 4648. +var StdEncoding = NewEncoding(encodeStd) + +// HexEncoding is the ``Extended Hex Alphabet'' defined in RFC 4648. +// It is typically used in DNS. +var HexEncoding = NewEncoding(encodeHex) + +/* + * Encoder + */ + +// Encode encodes src using the encoding enc, writing +// EncodedLen(len(src)) bytes to dst. +// +// The encoding pads the output to a multiple of 8 bytes, +// so Encode is not appropriate for use on individual blocks +// of a large data stream. Use NewEncoder() instead. +func (enc *Encoding) Encode(dst, src []byte) { + if len(src) == 0 { + return + } + + for len(src) > 0 { + dst[0] = 0 + dst[1] = 0 + dst[2] = 0 + dst[3] = 0 + dst[4] = 0 + dst[5] = 0 + dst[6] = 0 + dst[7] = 0 + + // Unpack 8x 5-bit source blocks into a 5 byte + // destination quantum + switch len(src) { + default: + dst[7] |= src[4] & 0x1F + dst[6] |= src[4] >> 5 + fallthrough + case 4: + dst[6] |= (src[3] << 3) & 0x1F + dst[5] |= (src[3] >> 2) & 0x1F + dst[4] |= src[3] >> 7 + fallthrough + case 3: + dst[4] |= (src[2] << 1) & 0x1F + dst[3] |= (src[2] >> 4) & 0x1F + fallthrough + case 2: + dst[3] |= (src[1] << 4) & 0x1F + dst[2] |= (src[1] >> 1) & 0x1F + dst[1] |= (src[1] >> 6) & 0x1F + fallthrough + case 1: + dst[1] |= (src[0] << 2) & 0x1F + dst[0] |= src[0] >> 3 + } + + // Encode 5-bit blocks using the base32 alphabet + for j := 0; j < 8; j++ { + dst[j] = enc.encode[dst[j]] + } + + // Pad the final quantum + if len(src) < 5 { + dst[7] = '=' + if len(src) < 4 { + dst[6] = '=' + dst[5] = '=' + if len(src) < 3 { + dst[4] = '=' + if len(src) < 2 { + dst[3] = '=' + dst[2] = '=' + } + } + } + break + } + src = src[5:] + dst = dst[8:] + } +} + +type encoder struct { + err os.Error + enc *Encoding + w io.Writer + buf [5]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 < 5; i++ { + e.buf[e.nbuf] = p[i] + e.nbuf++ + } + n += i + p = p[i:] + if e.nbuf < 5 { + return + } + e.enc.Encode(e.out[0:], e.buf[0:]) + if _, e.err = e.w.Write(e.out[0:8]); e.err != nil { + return n, e.err + } + e.nbuf = 0 + } + + // Large interior chunks. + for len(p) >= 5 { + nn := len(e.out) / 8 * 5 + if nn > len(p) { + nn = len(p) + } + nn -= nn % 5 + if nn > 0 { + e.enc.Encode(e.out[0:], p[0:nn]) + if _, e.err = e.w.Write(e.out[0 : nn/5*8]); 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 +} + +// 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 { + e.enc.Encode(e.out[0:], e.buf[0:e.nbuf]) + e.nbuf = 0 + _, e.err = e.w.Write(e.out[0:8]) + } + return e.err +} + +// NewEncoder returns a new base32 stream encoder. Data written to +// the returned writer will be encoded using enc and then written to w. +// Base32 encodings operate in 5-byte blocks; when finished +// writing, the caller must Close the returned encoder to flush any +// partially written blocks. +func NewEncoder(enc *Encoding, w io.Writer) io.WriteCloser { + return &encoder{enc: enc, w: w} +} + +// EncodedLen returns the length in bytes of the base32 encoding +// of an input buffer of length n. +func (enc *Encoding) EncodedLen(n int) int { return (n + 4) / 5 * 8 } + +/* + * Decoder + */ + +type CorruptInputError int64 + +func (e CorruptInputError) String() string { + return "illegal base32 data at input byte " + strconv.Itoa64(int64(e)) +} + +// 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)%8==0, +// since it is meant for internal use. +func (enc *Encoding) decode(dst, src []byte) (n int, end bool, err os.Error) { + for i := 0; i < len(src)/8 && !end; i++ { + // Decode quantum using the base32 alphabet + var dbuf [8]byte + dlen := 8 + + // do the top bytes contain any data? + dbufloop: + for j := 0; j < 8; j++ { + in := src[i*8+j] + if in == '=' && j >= 2 && i == len(src)/8-1 { + // We've reached the end and there's + // padding, the rest should be padded + for k := j; k < 8; k++ { + if src[i*8+k] != '=' { + return n, false, CorruptInputError(i*8 + j) + } + } + dlen = j + end = true + break dbufloop + } + dbuf[j] = enc.decodeMap[in] + if dbuf[j] == 0xFF { + return n, false, CorruptInputError(i*8 + j) + } + } + + // Pack 8x 5-bit source blocks into 5 byte destination + // quantum + switch dlen { + case 7, 8: + dst[i*5+4] = dbuf[6]<<5 | dbuf[7] + fallthrough + case 6, 5: + dst[i*5+3] = dbuf[4]<<7 | dbuf[5]<<2 | dbuf[6]>>3 + fallthrough + case 4: + dst[i*5+2] = dbuf[3]<<4 | dbuf[4]>>1 + fallthrough + case 3: + dst[i*5+1] = dbuf[1]<<6 | dbuf[2]<<1 | dbuf[3]>>4 + fallthrough + case 2: + dst[i*5+0] = dbuf[0]<<3 | dbuf[1]>>2 + } + switch dlen { + case 2: + n += 1 + case 3, 4: + n += 2 + case 5: + n += 3 + case 6, 7: + n += 4 + case 8: + n += 5 + } + } + return n, end, nil +} + +// Decode decodes src using the encoding enc. It writes at most +// DecodedLen(len(src)) bytes to dst and returns the number of bytes +// written. If src contains invalid base32 data, it will return the +// number of bytes successfully written and CorruptInputError. +func (enc *Encoding) Decode(dst, src []byte) (n int, err os.Error) { + if len(src)%8 != 0 { + return 0, CorruptInputError(len(src) / 8 * 8) + } + + n, _, err = enc.decode(dst, src) + return +} + +type decoder struct { + err os.Error + enc *Encoding + r io.Reader + end bool // saw end of message + buf [1024]byte // leftover input + nbuf int + out []byte // leftover decoded output + outbuf [1024 / 8 * 5]byte +} + +func (d *decoder) Read(p []byte) (n int, err os.Error) { + if d.err != nil { + return 0, d.err + } + + // Use leftover decoded output from last read. + if len(d.out) > 0 { + n = copy(p, d.out) + d.out = d.out[n:] + return n, nil + } + + // Read a chunk. + nn := len(p) / 5 * 8 + if nn < 8 { + nn = 8 + } + if nn > len(d.buf) { + nn = len(d.buf) + } + nn, d.err = io.ReadAtLeast(d.r, d.buf[d.nbuf:nn], 8-d.nbuf) + d.nbuf += nn + if d.nbuf < 8 { + return 0, d.err + } + + // Decode chunk into p, or d.out and then p if p is too small. + nr := d.nbuf / 8 * 8 + nw := d.nbuf / 8 * 5 + if nw > len(p) { + nw, d.end, d.err = d.enc.decode(d.outbuf[0:], d.buf[0:nr]) + d.out = d.outbuf[0:nw] + n = copy(p, d.out) + d.out = d.out[n:] + } else { + n, d.end, d.err = d.enc.decode(p, d.buf[0:nr]) + } + d.nbuf -= nr + for i := 0; i < d.nbuf; i++ { + d.buf[i] = d.buf[i+nr] + } + + if d.err == nil { + d.err = err + } + return n, d.err +} + +// NewDecoder constructs a new base32 stream decoder. +func NewDecoder(enc *Encoding, r io.Reader) io.Reader { + return &decoder{enc: enc, r: r} +} + +// DecodedLen returns the maximum length in bytes of the decoded data +// corresponding to n bytes of base32-encoded data. +func (enc *Encoding) DecodedLen(n int) int { return n / 8 * 5 } diff --git a/src/pkg/encoding/base32/base32_test.go b/src/pkg/encoding/base32/base32_test.go new file mode 100644 index 0000000000..792e4dc635 --- /dev/null +++ b/src/pkg/encoding/base32/base32_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 base32 + +import ( + "bytes" + "io/ioutil" + "os" + "testing" +) + +type testpair struct { + decoded, encoded string +} + +var pairs = []testpair{ + // RFC 4648 examples + {"", ""}, + {"f", "MY======"}, + {"fo", "MZXQ===="}, + {"foo", "MZXW6==="}, + {"foob", "MZXW6YQ="}, + {"fooba", "MZXW6YTB"}, + {"foobar", "MZXW6YTBOI======"}, + + + // Wikipedia examples, converted to base32 + {"sure.", "ON2XEZJO"}, + {"sure", "ON2XEZI="}, + {"sur", "ON2XE==="}, + {"su", "ON2Q===="}, + {"leasure.", "NRSWC43VOJSS4==="}, + {"easure.", "MVQXG5LSMUXA===="}, + {"asure.", "MFZXK4TFFY======"}, + {"sure.", "ON2XEZJO"}, +} + +var bigtest = testpair{ + "Twas brillig, and the slithy toves", + "KR3WC4ZAMJZGS3DMNFTSYIDBNZSCA5DIMUQHG3DJORUHSIDUN53GK4Y=", +} + +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 TestEncode(t *testing.T) { + for _, p := range pairs { + buf := make([]byte, StdEncoding.EncodedLen(len(p.decoded))) + StdEncoding.Encode(buf, []byte(p.decoded)) + testEqual(t, "Encode(%q) = %q, want %q", p.decoded, string(buf), p.encoded) + } +} + +func TestEncoder(t *testing.T) { + for _, p := range pairs { + bb := &bytes.Buffer{} + encoder := NewEncoder(StdEncoding, 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(bigtest.decoded) + for bs := 1; bs <= 12; bs++ { + bb := &bytes.Buffer{} + encoder := NewEncoder(StdEncoding, 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, bb.String(), bigtest.encoded) + } +} + +func TestDecode(t *testing.T) { + for _, p := range pairs { + dbuf := make([]byte, StdEncoding.DecodedLen(len(p.encoded))) + count, end, err := StdEncoding.decode(dbuf, []byte(p.encoded)) + testEqual(t, "Decode(%q) = error %v, want %v", p.encoded, err, os.Error(nil)) + testEqual(t, "Decode(%q) = length %v, want %v", p.encoded, count, len(p.decoded)) + if len(p.encoded) > 0 { + testEqual(t, "Decode(%q) = end %v, want %v", p.encoded, end, (p.encoded[len(p.encoded)-1] == '=')) + } + testEqual(t, "Decode(%q) = %q, want %q", p.encoded, + string(dbuf[0:count]), + p.decoded) + } +} + +func TestDecoder(t *testing.T) { + for _, p := range pairs { + decoder := NewDecoder(StdEncoding, bytes.NewBufferString(p.encoded)) + dbuf := make([]byte, StdEncoding.DecodedLen(len(p.encoded))) + count, err := decoder.Read(dbuf) + if err != nil && err != os.EOF { + t.Fatal("Read failed", err) + } + testEqual(t, "Read from %q = length %v, want %v", p.encoded, count, len(p.decoded)) + testEqual(t, "Decoding of %q = %q, want %q", p.encoded, string(dbuf[0:count]), p.decoded) + if err != os.EOF { + count, err = decoder.Read(dbuf) + } + 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(StdEncoding, 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{ + {"!!!!", 0}, + {"x===", 0}, + {"AA=A====", 2}, + {"AAA=AAAA", 3}, + {"MMMMMMMMM", 8}, + {"MMMMMM", 0}, + } + + for _, e := range examples { + dbuf := make([]byte, StdEncoding.DecodedLen(len(e.e))) + _, err := StdEncoding.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 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(StdEncoding, 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(StdEncoding, 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) + } +}