summaryrefslogtreecommitdiffstats
path: root/src/encoding/base32/base32.go
blob: 41d343aaac0fcaad798a5738dbbe0fd3264cb9d2 (plain)
1
2
3
4
5
6
7
8
9
10
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
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
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
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
}