diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-16 19:23:18 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-16 19:23:18 +0000 |
commit | 43a123c1ae6613b3efeed291fa552ecd909d3acf (patch) | |
tree | fd92518b7024bc74031f78a1cf9e454b65e73665 /src/encoding/base32 | |
parent | Initial commit. (diff) | |
download | golang-1.20-upstream.tar.xz golang-1.20-upstream.zip |
Adding upstream version 1.20.14.upstream/1.20.14upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'src/encoding/base32')
-rw-r--r-- | src/encoding/base32/base32.go | 549 | ||||
-rw-r--r-- | src/encoding/base32/base32_test.go | 820 | ||||
-rw-r--r-- | src/encoding/base32/example_test.go | 68 |
3 files changed, 1437 insertions, 0 deletions
diff --git a/src/encoding/base32/base32.go b/src/encoding/base32/base32.go new file mode 100644 index 0000000..41d343a --- /dev/null +++ b/src/encoding/base32/base32.go @@ -0,0 +1,549 @@ +// 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" + "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 [32]byte + decodeMap [256]byte + padChar rune +} + +const ( + StdPadding rune = '=' // Standard padding character + NoPadding rune = -1 // No padding + decodeMapInitialize = "" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" + + "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" +) + +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 { + if len(encoder) != 32 { + panic("encoding alphabet is not 32-bytes long") + } + + e := new(Encoding) + e.padChar = StdPadding + copy(e.encode[:], encoder) + copy(e.decodeMap[:], decodeMapInitialize) + + 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) + +// WithPadding creates a new encoding identical to enc except +// with a specified padding character, or NoPadding to disable padding. +// The padding character must not be '\r' or '\n', must not +// be contained in the encoding's alphabet and must be a rune equal or +// below '\xff'. +func (enc Encoding) WithPadding(padding rune) *Encoding { + if padding == '\r' || padding == '\n' || padding > 0xff { + panic("invalid padding") + } + + for i := 0; i < len(enc.encode); i++ { + if rune(enc.encode[i]) == padding { + panic("padding contained in alphabet") + } + } + + enc.padChar = padding + return &enc +} + +/* + * 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) { + for len(src) > 0 { + var b [8]byte + + // Unpack 8x 5-bit source blocks into a 5 byte + // destination quantum + switch len(src) { + default: + b[7] = src[4] & 0x1F + b[6] = src[4] >> 5 + fallthrough + case 4: + b[6] |= (src[3] << 3) & 0x1F + b[5] = (src[3] >> 2) & 0x1F + b[4] = src[3] >> 7 + fallthrough + case 3: + b[4] |= (src[2] << 1) & 0x1F + b[3] = (src[2] >> 4) & 0x1F + fallthrough + case 2: + b[3] |= (src[1] << 4) & 0x1F + b[2] = (src[1] >> 1) & 0x1F + b[1] = (src[1] >> 6) & 0x1F + fallthrough + case 1: + b[1] |= (src[0] << 2) & 0x1F + b[0] = src[0] >> 3 + } + + // Encode 5-bit blocks using the base32 alphabet + size := len(dst) + if size >= 8 { + // Common case, unrolled for extra performance + dst[0] = enc.encode[b[0]&31] + dst[1] = enc.encode[b[1]&31] + dst[2] = enc.encode[b[2]&31] + dst[3] = enc.encode[b[3]&31] + dst[4] = enc.encode[b[4]&31] + dst[5] = enc.encode[b[5]&31] + dst[6] = enc.encode[b[6]&31] + dst[7] = enc.encode[b[7]&31] + } else { + for i := 0; i < size; i++ { + dst[i] = enc.encode[b[i]&31] + } + } + + // Pad the final quantum + if len(src) < 5 { + if enc.padChar == NoPadding { + break + } + + dst[7] = byte(enc.padChar) + if len(src) < 4 { + dst[6] = byte(enc.padChar) + dst[5] = byte(enc.padChar) + if len(src) < 3 { + dst[4] = byte(enc.padChar) + if len(src) < 2 { + dst[3] = byte(enc.padChar) + dst[2] = byte(enc.padChar) + } + } + } + + break + } + + src = src[5:] + dst = dst[8:] + } +} + +// EncodeToString returns the base32 encoding of src. +func (enc *Encoding) EncodeToString(src []byte) string { + buf := make([]byte, enc.EncodedLen(len(src))) + enc.Encode(buf, src) + return string(buf) +} + +type encoder struct { + err 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 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 + } + 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. + copy(e.buf[:], p) + 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() 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]) + encodedLen := e.enc.EncodedLen(e.nbuf) + e.nbuf = 0 + _, e.err = e.w.Write(e.out[0:encodedLen]) + } + 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 { + if enc.padChar == NoPadding { + return (n*8 + 4) / 5 + } + return (n + 4) / 5 * 8 +} + +/* + * Decoder + */ + +type CorruptInputError int64 + +func (e CorruptInputError) Error() string { + return "illegal base32 data at input byte " + strconv.FormatInt(int64(e), 10) +} + +// 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. This method assumes that src has been +// stripped of all supported whitespace ('\r' and '\n'). +func (enc *Encoding) decode(dst, src []byte) (n int, end bool, err error) { + // Lift the nil check outside of the loop. + _ = enc.decodeMap + + dsti := 0 + olen := len(src) + + for len(src) > 0 && !end { + // Decode quantum using the base32 alphabet + var dbuf [8]byte + dlen := 8 + + for j := 0; j < 8; { + + if len(src) == 0 { + if enc.padChar != NoPadding { + // We have reached the end and are missing padding + return n, false, CorruptInputError(olen - len(src) - j) + } + // We have reached the end and are not expecting any padding + dlen, end = j, true + break + } + in := src[0] + src = src[1:] + if in == byte(enc.padChar) && j >= 2 && len(src) < 8 { + // We've reached the end and there's padding + if len(src)+j < 8-1 { + // not enough padding + return n, false, CorruptInputError(olen) + } + for k := 0; k < 8-1-j; k++ { + if len(src) > k && src[k] != byte(enc.padChar) { + // incorrect padding + return n, false, CorruptInputError(olen - len(src) + k - 1) + } + } + dlen, end = j, true + // 7, 5 and 2 are not valid padding lengths, and so 1, 3 and 6 are not + // valid dlen values. See RFC 4648 Section 6 "Base 32 Encoding" listing + // the five valid padding lengths, and Section 9 "Illustrations and + // Examples" for an illustration for how the 1st, 3rd and 6th base32 + // src bytes do not yield enough information to decode a dst byte. + if dlen == 1 || dlen == 3 || dlen == 6 { + return n, false, CorruptInputError(olen - len(src) - 1) + } + break + } + dbuf[j] = enc.decodeMap[in] + if dbuf[j] == 0xFF { + return n, false, CorruptInputError(olen - len(src) - 1) + } + j++ + } + + // Pack 8x 5-bit source blocks into 5 byte destination + // quantum + switch dlen { + case 8: + dst[dsti+4] = dbuf[6]<<5 | dbuf[7] + n++ + fallthrough + case 7: + dst[dsti+3] = dbuf[4]<<7 | dbuf[5]<<2 | dbuf[6]>>3 + n++ + fallthrough + case 5: + dst[dsti+2] = dbuf[3]<<4 | dbuf[4]>>1 + n++ + fallthrough + case 4: + dst[dsti+1] = dbuf[1]<<6 | dbuf[2]<<1 | dbuf[3]>>4 + n++ + fallthrough + case 2: + dst[dsti+0] = dbuf[0]<<3 | dbuf[1]>>2 + n++ + } + dsti += 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. +// New line characters (\r and \n) are ignored. +func (enc *Encoding) Decode(dst, src []byte) (n int, err error) { + buf := make([]byte, len(src)) + l := stripNewlines(buf, src) + n, _, err = enc.decode(dst, buf[:l]) + return +} + +// DecodeString returns the bytes represented by the base32 string s. +func (enc *Encoding) DecodeString(s string) ([]byte, error) { + buf := []byte(s) + l := stripNewlines(buf, buf) + n, _, err := enc.decode(buf, buf[:l]) + return buf[:n], err +} + +type decoder struct { + err 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 readEncodedData(r io.Reader, buf []byte, min int, expectsPadding bool) (n int, err error) { + for n < min && err == nil { + var nn int + nn, err = r.Read(buf[n:]) + n += nn + } + // data was read, less than min bytes could be read + if n < min && n > 0 && err == io.EOF { + err = io.ErrUnexpectedEOF + } + // no data was read, the buffer already contains some data + // when padding is disabled this is not an error, as the message can be of + // any length + if expectsPadding && min < 8 && n == 0 && err == io.EOF { + err = io.ErrUnexpectedEOF + } + return +} + +func (d *decoder) Read(p []byte) (n int, err error) { + // Use leftover decoded output from last read. + if len(d.out) > 0 { + n = copy(p, d.out) + d.out = d.out[n:] + if len(d.out) == 0 { + return n, d.err + } + return n, nil + } + + if d.err != nil { + return 0, d.err + } + + // Read a chunk. + nn := len(p) / 5 * 8 + if nn < 8 { + nn = 8 + } + if nn > len(d.buf) { + nn = len(d.buf) + } + + // Minimum amount of bytes that needs to be read each cycle + var min int + var expectsPadding bool + if d.enc.padChar == NoPadding { + min = 1 + expectsPadding = false + } else { + min = 8 - d.nbuf + expectsPadding = true + } + + nn, d.err = readEncodedData(d.r, d.buf[d.nbuf:nn], min, expectsPadding) + d.nbuf += nn + if d.nbuf < min { + return 0, d.err + } + if nn > 0 && d.end { + return 0, CorruptInputError(0) + } + + // Decode chunk into p, or d.out and then p if p is too small. + var nr int + if d.enc.padChar == NoPadding { + nr = d.nbuf + } else { + nr = d.nbuf / 8 * 8 + } + nw := d.enc.DecodedLen(d.nbuf) + + if nw > len(p) { + nw, d.end, 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, 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 err != nil && (d.err == nil || d.err == io.EOF) { + d.err = err + } + + if len(d.out) > 0 { + // We cannot return all the decoded bytes to the caller in this + // invocation of Read, so we return a nil error to ensure that Read + // will be called again. The error stored in d.err, if any, will be + // returned with the last set of decoded bytes. + return n, nil + } + + return n, d.err +} + +type newlineFilteringReader struct { + wrapped io.Reader +} + +// stripNewlines removes newline characters and returns the number +// of non-newline characters copied to dst. +func stripNewlines(dst, src []byte) int { + offset := 0 + for _, b := range src { + if b == '\r' || b == '\n' { + continue + } + dst[offset] = b + offset++ + } + return offset +} + +func (r *newlineFilteringReader) Read(p []byte) (int, error) { + n, err := r.wrapped.Read(p) + for n > 0 { + s := p[0:n] + offset := stripNewlines(s, s) + if err != nil || offset > 0 { + return offset, err + } + // Previous buffer entirely whitespace, read again + n, err = r.wrapped.Read(p) + } + return n, err +} + +// NewDecoder constructs a new base32 stream decoder. +func NewDecoder(enc *Encoding, r io.Reader) io.Reader { + return &decoder{enc: enc, r: &newlineFilteringReader{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 { + if enc.padChar == NoPadding { + return n * 5 / 8 + } + + return n / 8 * 5 +} diff --git a/src/encoding/base32/base32_test.go b/src/encoding/base32/base32_test.go new file mode 100644 index 0000000..8118531 --- /dev/null +++ b/src/encoding/base32/base32_test.go @@ -0,0 +1,820 @@ +// 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" + "errors" + "io" + "strings" + "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 ...any) bool { + t.Helper() + 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 { + got := StdEncoding.EncodeToString([]byte(p.decoded)) + testEqual(t, "Encode(%q) = %q, want %q", p.decoded, got, p.encoded) + } +} + +func TestEncoder(t *testing.T) { + for _, p := range pairs { + bb := &strings.Builder{} + 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 := &strings.Builder{} + 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, 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, 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, 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) + + dbuf, err = StdEncoding.DecodeString(p.encoded) + testEqual(t, "DecodeString(%q) = error %v, want %v", p.encoded, err, error(nil)) + testEqual(t, "DecodeString(%q) = %q, want %q", p.encoded, string(dbuf), p.decoded) + } +} + +func TestDecoder(t *testing.T) { + for _, p := range pairs { + decoder := NewDecoder(StdEncoding, strings.NewReader(p.encoded)) + dbuf := make([]byte, StdEncoding.DecodedLen(len(p.encoded))) + count, err := decoder.Read(dbuf) + if err != nil && err != io.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 != io.EOF { + _, err = decoder.Read(dbuf) + } + testEqual(t, "Read from %q = %v, want %v", p.encoded, err, io.EOF) + } +} + +type badReader struct { + data []byte + errs []error + called int + limit int +} + +// Populates p with data, returns a count of the bytes written and an +// error. The error returned is taken from badReader.errs, with each +// invocation of Read returning the next error in this slice, or io.EOF, +// if all errors from the slice have already been returned. The +// number of bytes returned is determined by the size of the input buffer +// the test passes to decoder.Read and will be a multiple of 8, unless +// badReader.limit is non zero. +func (b *badReader) Read(p []byte) (int, error) { + lim := len(p) + if b.limit != 0 && b.limit < lim { + lim = b.limit + } + if len(b.data) < lim { + lim = len(b.data) + } + for i := range p[:lim] { + p[i] = b.data[i] + } + b.data = b.data[lim:] + err := io.EOF + if b.called < len(b.errs) { + err = b.errs[b.called] + } + b.called++ + return lim, err +} + +// TestIssue20044 tests that decoder.Read behaves correctly when the caller +// supplied reader returns an error. +func TestIssue20044(t *testing.T) { + badErr := errors.New("bad reader error") + testCases := []struct { + r badReader + res string + err error + dbuflen int + }{ + // Check valid input data accompanied by an error is processed and the error is propagated. + {r: badReader{data: []byte("MY======"), errs: []error{badErr}}, + res: "f", err: badErr}, + // Check a read error accompanied by input data consisting of newlines only is propagated. + {r: badReader{data: []byte("\n\n\n\n\n\n\n\n"), errs: []error{badErr, nil}}, + res: "", err: badErr}, + // Reader will be called twice. The first time it will return 8 newline characters. The + // second time valid base32 encoded data and an error. The data should be decoded + // correctly and the error should be propagated. + {r: badReader{data: []byte("\n\n\n\n\n\n\n\nMY======"), errs: []error{nil, badErr}}, + res: "f", err: badErr, dbuflen: 8}, + // Reader returns invalid input data (too short) and an error. Verify the reader + // error is returned. + {r: badReader{data: []byte("MY====="), errs: []error{badErr}}, + res: "", err: badErr}, + // Reader returns invalid input data (too short) but no error. Verify io.ErrUnexpectedEOF + // is returned. + {r: badReader{data: []byte("MY====="), errs: []error{nil}}, + res: "", err: io.ErrUnexpectedEOF}, + // Reader returns invalid input data and an error. Verify the reader and not the + // decoder error is returned. + {r: badReader{data: []byte("Ma======"), errs: []error{badErr}}, + res: "", err: badErr}, + // Reader returns valid data and io.EOF. Check data is decoded and io.EOF is propagated. + {r: badReader{data: []byte("MZXW6YTB"), errs: []error{io.EOF}}, + res: "fooba", err: io.EOF}, + // Check errors are properly reported when decoder.Read is called multiple times. + // decoder.Read will be called 8 times, badReader.Read will be called twice, returning + // valid data both times but an error on the second call. + {r: badReader{data: []byte("NRSWC43VOJSS4==="), errs: []error{nil, badErr}}, + res: "leasure.", err: badErr, dbuflen: 1}, + // Check io.EOF is properly reported when decoder.Read is called multiple times. + // decoder.Read will be called 8 times, badReader.Read will be called twice, returning + // valid data both times but io.EOF on the second call. + {r: badReader{data: []byte("NRSWC43VOJSS4==="), errs: []error{nil, io.EOF}}, + res: "leasure.", err: io.EOF, dbuflen: 1}, + // The following two test cases check that errors are propagated correctly when more than + // 8 bytes are read at a time. + {r: badReader{data: []byte("NRSWC43VOJSS4==="), errs: []error{io.EOF}}, + res: "leasure.", err: io.EOF, dbuflen: 11}, + {r: badReader{data: []byte("NRSWC43VOJSS4==="), errs: []error{badErr}}, + res: "leasure.", err: badErr, dbuflen: 11}, + // Check that errors are correctly propagated when the reader returns valid bytes in + // groups that are not divisible by 8. The first read will return 11 bytes and no + // error. The second will return 7 and an error. The data should be decoded correctly + // and the error should be propagated. + {r: badReader{data: []byte("NRSWC43VOJSS4==="), errs: []error{nil, badErr}, limit: 11}, + res: "leasure.", err: badErr}, + } + + for _, tc := range testCases { + input := tc.r.data + decoder := NewDecoder(StdEncoding, &tc.r) + var dbuflen int + if tc.dbuflen > 0 { + dbuflen = tc.dbuflen + } else { + dbuflen = StdEncoding.DecodedLen(len(input)) + } + dbuf := make([]byte, dbuflen) + var err error + var res []byte + for err == nil { + var n int + n, err = decoder.Read(dbuf) + if n > 0 { + res = append(res, dbuf[:n]...) + } + } + + testEqual(t, "Decoding of %q = %q, want %q", string(input), string(res), tc.res) + testEqual(t, "Decoding of %q err = %v, expected %v", string(input), err, tc.err) + } +} + +// TestDecoderError verifies decode errors are propagated when there are no read +// errors. +func TestDecoderError(t *testing.T) { + for _, readErr := range []error{io.EOF, nil} { + input := "MZXW6YTb" + dbuf := make([]byte, StdEncoding.DecodedLen(len(input))) + br := badReader{data: []byte(input), errs: []error{readErr}} + decoder := NewDecoder(StdEncoding, &br) + n, err := decoder.Read(dbuf) + testEqual(t, "Read after EOF, n = %d, expected %d", n, 0) + if _, ok := err.(CorruptInputError); !ok { + t.Errorf("Corrupt input error expected. Found %T", err) + } + } +} + +// TestReaderEOF ensures decoder.Read behaves correctly when input data is +// exhausted. +func TestReaderEOF(t *testing.T) { + for _, readErr := range []error{io.EOF, nil} { + input := "MZXW6YTB" + br := badReader{data: []byte(input), errs: []error{nil, readErr}} + decoder := NewDecoder(StdEncoding, &br) + dbuf := make([]byte, StdEncoding.DecodedLen(len(input))) + n, err := decoder.Read(dbuf) + testEqual(t, "Decoding of %q err = %v, expected %v", input, err, error(nil)) + n, err = decoder.Read(dbuf) + testEqual(t, "Read after EOF, n = %d, expected %d", n, 0) + testEqual(t, "Read after EOF, err = %v, expected %v", err, io.EOF) + n, err = decoder.Read(dbuf) + testEqual(t, "Read after EOF, n = %d, expected %d", n, 0) + testEqual(t, "Read after EOF, err = %v, expected %v", err, io.EOF) + } +} + +func TestDecoderBuffering(t *testing.T) { + for bs := 1; bs <= 12; bs++ { + decoder := NewDecoder(StdEncoding, strings.NewReader(bigtest.encoded)) + buf := make([]byte, len(bigtest.decoded)+12) + var total int + var n int + var err error + for total = 0; total < len(bigtest.decoded) && err == nil; { + n, err = decoder.Read(buf[total : total+bs]) + total += n + } + if err != nil && err != io.EOF { + t.Errorf("Read from %q at pos %d = %d, unexpected error %v", bigtest.encoded, total, n, err) + } + testEqual(t, "Decoding/%d of %q = %q, want %q", bs, bigtest.encoded, string(buf[0:total]), bigtest.decoded) + } +} + +func TestDecodeCorrupt(t *testing.T) { + testCases := []struct { + input string + offset int // -1 means no corruption. + }{ + {"", -1}, + {"!!!!", 0}, + {"x===", 0}, + {"AA=A====", 2}, + {"AAA=AAAA", 3}, + {"MMMMMMMMM", 8}, + {"MMMMMM", 0}, + {"A=", 1}, + {"AA=", 3}, + {"AA==", 4}, + {"AA===", 5}, + {"AAAA=", 5}, + {"AAAA==", 6}, + {"AAAAA=", 6}, + {"AAAAA==", 7}, + {"A=======", 1}, + {"AA======", -1}, + {"AAA=====", 3}, + {"AAAA====", -1}, + {"AAAAA===", -1}, + {"AAAAAA==", 6}, + {"AAAAAAA=", -1}, + {"AAAAAAAA", -1}, + } + for _, tc := range testCases { + dbuf := make([]byte, StdEncoding.DecodedLen(len(tc.input))) + _, err := StdEncoding.Decode(dbuf, []byte(tc.input)) + if tc.offset == -1 { + if err != nil { + t.Error("Decoder wrongly detected corruption in", tc.input) + } + continue + } + switch err := err.(type) { + case CorruptInputError: + testEqual(t, "Corruption in %q at offset %v, want %v", tc.input, int(err), tc.offset) + default: + t.Error("Decoder failed to detect corruption in", tc) + } + } +} + +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 := io.ReadAll(NewDecoder(StdEncoding, 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) + } +} + +func testStringEncoding(t *testing.T, expected string, examples []string) { + for _, e := range examples { + buf, err := StdEncoding.DecodeString(e) + if err != nil { + t.Errorf("Decode(%q) failed: %v", e, err) + continue + } + if s := string(buf); s != expected { + t.Errorf("Decode(%q) = %q, want %q", e, s, expected) + } + } +} + +func TestNewLineCharacters(t *testing.T) { + // Each of these should decode to the string "sure", without errors. + examples := []string{ + "ON2XEZI=", + "ON2XEZI=\r", + "ON2XEZI=\n", + "ON2XEZI=\r\n", + "ON2XEZ\r\nI=", + "ON2X\rEZ\nI=", + "ON2X\nEZ\rI=", + "ON2XEZ\nI=", + "ON2XEZI\n=", + } + testStringEncoding(t, "sure", examples) + + // Each of these should decode to the string "foobar", without errors. + examples = []string{ + "MZXW6YTBOI======", + "MZXW6YTBOI=\r\n=====", + } + testStringEncoding(t, "foobar", examples) +} + +func TestDecoderIssue4779(t *testing.T) { + encoded := `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` + encodedShort := strings.ReplaceAll(encoded, "\n", "") + + dec := NewDecoder(StdEncoding, strings.NewReader(encoded)) + res1, err := io.ReadAll(dec) + if err != nil { + t.Errorf("ReadAll failed: %v", err) + } + + dec = NewDecoder(StdEncoding, strings.NewReader(encodedShort)) + var res2 []byte + res2, err = io.ReadAll(dec) + if err != nil { + t.Errorf("ReadAll failed: %v", err) + } + + if !bytes.Equal(res1, res2) { + t.Error("Decoded results not equal") + } +} + +func BenchmarkEncode(b *testing.B) { + data := make([]byte, 8192) + buf := make([]byte, StdEncoding.EncodedLen(len(data))) + b.SetBytes(int64(len(data))) + for i := 0; i < b.N; i++ { + StdEncoding.Encode(buf, data) + } +} + +func BenchmarkEncodeToString(b *testing.B) { + data := make([]byte, 8192) + b.SetBytes(int64(len(data))) + for i := 0; i < b.N; i++ { + StdEncoding.EncodeToString(data) + } +} + +func BenchmarkDecode(b *testing.B) { + data := make([]byte, StdEncoding.EncodedLen(8192)) + StdEncoding.Encode(data, make([]byte, 8192)) + buf := make([]byte, 8192) + b.SetBytes(int64(len(data))) + for i := 0; i < b.N; i++ { + StdEncoding.Decode(buf, data) + } +} +func BenchmarkDecodeString(b *testing.B) { + data := StdEncoding.EncodeToString(make([]byte, 8192)) + b.SetBytes(int64(len(data))) + for i := 0; i < b.N; i++ { + StdEncoding.DecodeString(data) + } +} + +func TestWithCustomPadding(t *testing.T) { + for _, testcase := range pairs { + defaultPadding := StdEncoding.EncodeToString([]byte(testcase.decoded)) + customPadding := StdEncoding.WithPadding('@').EncodeToString([]byte(testcase.decoded)) + expected := strings.ReplaceAll(defaultPadding, "=", "@") + + if expected != customPadding { + t.Errorf("Expected custom %s, got %s", expected, customPadding) + } + if testcase.encoded != defaultPadding { + t.Errorf("Expected %s, got %s", testcase.encoded, defaultPadding) + } + } +} + +func TestWithoutPadding(t *testing.T) { + for _, testcase := range pairs { + defaultPadding := StdEncoding.EncodeToString([]byte(testcase.decoded)) + customPadding := StdEncoding.WithPadding(NoPadding).EncodeToString([]byte(testcase.decoded)) + expected := strings.TrimRight(defaultPadding, "=") + + if expected != customPadding { + t.Errorf("Expected custom %s, got %s", expected, customPadding) + } + if testcase.encoded != defaultPadding { + t.Errorf("Expected %s, got %s", testcase.encoded, defaultPadding) + } + } +} + +func TestDecodeWithPadding(t *testing.T) { + encodings := []*Encoding{ + StdEncoding, + StdEncoding.WithPadding('-'), + StdEncoding.WithPadding(NoPadding), + } + + for i, enc := range encodings { + for _, pair := range pairs { + + input := pair.decoded + encoded := enc.EncodeToString([]byte(input)) + + decoded, err := enc.DecodeString(encoded) + if err != nil { + t.Errorf("DecodeString Error for encoding %d (%q): %v", i, input, err) + } + + if input != string(decoded) { + t.Errorf("Unexpected result for encoding %d: got %q; want %q", i, decoded, input) + } + } + } +} + +func TestDecodeWithWrongPadding(t *testing.T) { + encoded := StdEncoding.EncodeToString([]byte("foobar")) + + _, err := StdEncoding.WithPadding('-').DecodeString(encoded) + if err == nil { + t.Error("expected error") + } + + _, err = StdEncoding.WithPadding(NoPadding).DecodeString(encoded) + if err == nil { + t.Error("expected error") + } +} + +func TestBufferedDecodingSameError(t *testing.T) { + testcases := []struct { + prefix string + chunkCombinations [][]string + expected error + }{ + // NBSWY3DPO5XXE3DE == helloworld + // Test with "ZZ" as extra input + {"helloworld", [][]string{ + {"NBSW", "Y3DP", "O5XX", "E3DE", "ZZ"}, + {"NBSWY3DPO5XXE3DE", "ZZ"}, + {"NBSWY3DPO5XXE3DEZZ"}, + {"NBS", "WY3", "DPO", "5XX", "E3D", "EZZ"}, + {"NBSWY3DPO5XXE3", "DEZZ"}, + }, io.ErrUnexpectedEOF}, + + // Test with "ZZY" as extra input + {"helloworld", [][]string{ + {"NBSW", "Y3DP", "O5XX", "E3DE", "ZZY"}, + {"NBSWY3DPO5XXE3DE", "ZZY"}, + {"NBSWY3DPO5XXE3DEZZY"}, + {"NBS", "WY3", "DPO", "5XX", "E3D", "EZZY"}, + {"NBSWY3DPO5XXE3", "DEZZY"}, + }, io.ErrUnexpectedEOF}, + + // Normal case, this is valid input + {"helloworld", [][]string{ + {"NBSW", "Y3DP", "O5XX", "E3DE"}, + {"NBSWY3DPO5XXE3DE"}, + {"NBS", "WY3", "DPO", "5XX", "E3D", "E"}, + {"NBSWY3DPO5XXE3", "DE"}, + }, nil}, + + // MZXW6YTB = fooba + {"fooba", [][]string{ + {"MZXW6YTBZZ"}, + {"MZXW6YTBZ", "Z"}, + {"MZXW6YTB", "ZZ"}, + {"MZXW6YT", "BZZ"}, + {"MZXW6Y", "TBZZ"}, + {"MZXW6Y", "TB", "ZZ"}, + {"MZXW6", "YTBZZ"}, + {"MZXW6", "YTB", "ZZ"}, + {"MZXW6", "YT", "BZZ"}, + }, io.ErrUnexpectedEOF}, + + // Normal case, this is valid input + {"fooba", [][]string{ + {"MZXW6YTB"}, + {"MZXW6YT", "B"}, + {"MZXW6Y", "TB"}, + {"MZXW6", "YTB"}, + {"MZXW6", "YT", "B"}, + {"MZXW", "6YTB"}, + {"MZXW", "6Y", "TB"}, + }, nil}, + } + + for _, testcase := range testcases { + for _, chunks := range testcase.chunkCombinations { + pr, pw := io.Pipe() + + // Write the encoded chunks into the pipe + go func() { + for _, chunk := range chunks { + pw.Write([]byte(chunk)) + } + pw.Close() + }() + + decoder := NewDecoder(StdEncoding, pr) + _, err := io.ReadAll(decoder) + + if err != testcase.expected { + t.Errorf("Expected %v, got %v; case %s %+v", testcase.expected, err, testcase.prefix, chunks) + } + } + } +} + +func TestBufferedDecodingPadding(t *testing.T) { + testcases := []struct { + chunks []string + expectedError string + }{ + {[]string{ + "I4======", + "==", + }, "unexpected EOF"}, + + {[]string{ + "I4======N4======", + }, "illegal base32 data at input byte 2"}, + + {[]string{ + "I4======", + "N4======", + }, "illegal base32 data at input byte 0"}, + + {[]string{ + "I4======", + "========", + }, "illegal base32 data at input byte 0"}, + + {[]string{ + "I4I4I4I4", + "I4======", + "I4======", + }, "illegal base32 data at input byte 0"}, + } + + for _, testcase := range testcases { + testcase := testcase + pr, pw := io.Pipe() + go func() { + for _, chunk := range testcase.chunks { + _, _ = pw.Write([]byte(chunk)) + } + _ = pw.Close() + }() + + decoder := NewDecoder(StdEncoding, pr) + _, err := io.ReadAll(decoder) + + if err == nil && len(testcase.expectedError) != 0 { + t.Errorf("case %q: got nil error, want %v", testcase.chunks, testcase.expectedError) + } else if err.Error() != testcase.expectedError { + t.Errorf("case %q: got %v, want %v", testcase.chunks, err, testcase.expectedError) + } + } +} + +func TestEncodedDecodedLen(t *testing.T) { + type test struct { + in int + wantEnc int + wantDec int + } + data := bytes.Repeat([]byte("x"), 100) + for _, test := range []struct { + name string + enc *Encoding + cases []test + }{ + {"StdEncoding", StdEncoding, []test{ + {0, 0, 0}, + {1, 8, 5}, + {5, 8, 5}, + {6, 16, 10}, + {10, 16, 10}, + }}, + {"NoPadding", StdEncoding.WithPadding(NoPadding), []test{ + {0, 0, 0}, + {1, 2, 1}, + {2, 4, 2}, + {5, 8, 5}, + {6, 10, 6}, + {7, 12, 7}, + {10, 16, 10}, + {11, 18, 11}, + }}, + } { + t.Run(test.name, func(t *testing.T) { + for _, tc := range test.cases { + encLen := test.enc.EncodedLen(tc.in) + decLen := test.enc.DecodedLen(encLen) + enc := test.enc.EncodeToString(data[:tc.in]) + if len(enc) != encLen { + t.Fatalf("EncodedLen(%d) = %d but encoded to %q (%d)", tc.in, encLen, enc, len(enc)) + } + if encLen != tc.wantEnc { + t.Fatalf("EncodedLen(%d) = %d; want %d", tc.in, encLen, tc.wantEnc) + } + if decLen != tc.wantDec { + t.Fatalf("DecodedLen(%d) = %d; want %d", encLen, decLen, tc.wantDec) + } + } + }) + } +} + +func TestWithoutPaddingClose(t *testing.T) { + encodings := []*Encoding{ + StdEncoding, + StdEncoding.WithPadding(NoPadding), + } + + for _, encoding := range encodings { + for _, testpair := range pairs { + + var buf strings.Builder + encoder := NewEncoder(encoding, &buf) + encoder.Write([]byte(testpair.decoded)) + encoder.Close() + + expected := testpair.encoded + if encoding.padChar == NoPadding { + expected = strings.ReplaceAll(expected, "=", "") + } + + res := buf.String() + + if res != expected { + t.Errorf("Expected %s got %s; padChar=%d", expected, res, encoding.padChar) + } + } + } +} + +func TestDecodeReadAll(t *testing.T) { + encodings := []*Encoding{ + StdEncoding, + StdEncoding.WithPadding(NoPadding), + } + + for _, pair := range pairs { + for encIndex, encoding := range encodings { + encoded := pair.encoded + if encoding.padChar == NoPadding { + encoded = strings.ReplaceAll(encoded, "=", "") + } + + decReader, err := io.ReadAll(NewDecoder(encoding, strings.NewReader(encoded))) + if err != nil { + t.Errorf("NewDecoder error: %v", err) + } + + if pair.decoded != string(decReader) { + t.Errorf("Expected %s got %s; Encoding %d", pair.decoded, decReader, encIndex) + } + } + } +} + +func TestDecodeSmallBuffer(t *testing.T) { + encodings := []*Encoding{ + StdEncoding, + StdEncoding.WithPadding(NoPadding), + } + + for bufferSize := 1; bufferSize < 200; bufferSize++ { + for _, pair := range pairs { + for encIndex, encoding := range encodings { + encoded := pair.encoded + if encoding.padChar == NoPadding { + encoded = strings.ReplaceAll(encoded, "=", "") + } + + decoder := NewDecoder(encoding, strings.NewReader(encoded)) + + var allRead []byte + + for { + buf := make([]byte, bufferSize) + n, err := decoder.Read(buf) + allRead = append(allRead, buf[0:n]...) + if err == io.EOF { + break + } + if err != nil { + t.Error(err) + } + } + + if pair.decoded != string(allRead) { + t.Errorf("Expected %s got %s; Encoding %d; bufferSize %d", pair.decoded, allRead, encIndex, bufferSize) + } + } + } + } +} diff --git a/src/encoding/base32/example_test.go b/src/encoding/base32/example_test.go new file mode 100644 index 0000000..251624f --- /dev/null +++ b/src/encoding/base32/example_test.go @@ -0,0 +1,68 @@ +// Copyright 2012 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. + +// Keep in sync with ../base64/example_test.go. + +package base32_test + +import ( + "encoding/base32" + "fmt" + "os" +) + +func ExampleEncoding_EncodeToString() { + data := []byte("any + old & data") + str := base32.StdEncoding.EncodeToString(data) + fmt.Println(str) + // Output: + // MFXHSIBLEBXWYZBAEYQGIYLUME====== +} + +func ExampleEncoding_Encode() { + data := []byte("Hello, world!") + dst := make([]byte, base32.StdEncoding.EncodedLen(len(data))) + base32.StdEncoding.Encode(dst, data) + fmt.Println(string(dst)) + // Output: + // JBSWY3DPFQQHO33SNRSCC=== +} + +func ExampleEncoding_DecodeString() { + str := "ONXW2ZJAMRQXIYJAO5UXI2BAAAQGC3TEEDX3XPY=" + data, err := base32.StdEncoding.DecodeString(str) + if err != nil { + fmt.Println("error:", err) + return + } + fmt.Printf("%q\n", data) + // Output: + // "some data with \x00 and \ufeff" +} + +func ExampleEncoding_Decode() { + str := "JBSWY3DPFQQHO33SNRSCC===" + dst := make([]byte, base32.StdEncoding.DecodedLen(len(str))) + n, err := base32.StdEncoding.Decode(dst, []byte(str)) + if err != nil { + fmt.Println("decode error:", err) + return + } + dst = dst[:n] + fmt.Printf("%q\n", dst) + // Output: + // "Hello, world!" +} + +func ExampleNewEncoder() { + input := []byte("foo\x00bar") + encoder := base32.NewEncoder(base32.StdEncoding, os.Stdout) + encoder.Write(input) + // Must close the encoder when finished to flush any partial blocks. + // If you comment out the following line, the last partial block "r" + // won't be encoded. + encoder.Close() + // Output: + // MZXW6ADCMFZA==== +} |