-
Notifications
You must be signed in to change notification settings - Fork 60
/
diagnose.go
724 lines (615 loc) · 17.4 KB
/
diagnose.go
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
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
// Copyright (c) Faye Amacker. All rights reserved.
// Licensed under the MIT License. See LICENSE in the project root for license information.
package cbor
import (
"bytes"
"encoding/base32"
"encoding/base64"
"encoding/hex"
"errors"
"fmt"
"io"
"math"
"math/big"
"strconv"
"unicode/utf16"
"unicode/utf8"
"github.com/x448/float16"
)
// DiagMode is the main interface for CBOR diagnostic notation.
type DiagMode interface {
// Diagnose returns extended diagnostic notation (EDN) of CBOR data items using this DiagMode.
Diagnose([]byte) (string, error)
// DiagnoseFirst returns extended diagnostic notation (EDN) of the first CBOR data item using the DiagMode. Any remaining bytes are returned in rest.
DiagnoseFirst([]byte) (string, []byte, error)
// DiagOptions returns user specified options used to create this DiagMode.
DiagOptions() DiagOptions
}
// ByteStringEncoding specifies the base encoding that byte strings are notated.
type ByteStringEncoding uint8
const (
// ByteStringBase16Encoding encodes byte strings in base16, without padding.
ByteStringBase16Encoding ByteStringEncoding = iota
// ByteStringBase32Encoding encodes byte strings in base32, without padding.
ByteStringBase32Encoding
// ByteStringBase32HexEncoding encodes byte strings in base32hex, without padding.
ByteStringBase32HexEncoding
// ByteStringBase64Encoding encodes byte strings in base64url, without padding.
ByteStringBase64Encoding
maxByteStringEncoding
)
func (bse ByteStringEncoding) valid() error {
if bse >= maxByteStringEncoding {
return errors.New("cbor: invalid ByteStringEncoding " + strconv.Itoa(int(bse)))
}
return nil
}
// DiagOptions specifies Diag options.
type DiagOptions struct {
// ByteStringEncoding specifies the base encoding that byte strings are notated.
// Default is ByteStringBase16Encoding.
ByteStringEncoding ByteStringEncoding
// ByteStringHexWhitespace specifies notating with whitespace in byte string
// when ByteStringEncoding is ByteStringBase16Encoding.
ByteStringHexWhitespace bool
// ByteStringText specifies notating with text in byte string
// if it is a valid UTF-8 text.
ByteStringText bool
// ByteStringEmbeddedCBOR specifies notating embedded CBOR in byte string
// if it is a valid CBOR bytes.
ByteStringEmbeddedCBOR bool
// CBORSequence specifies notating CBOR sequences.
// otherwise, it returns an error if there are more bytes after the first CBOR.
CBORSequence bool
// FloatPrecisionIndicator specifies appending a suffix to indicate float precision.
// Refer to https://www.rfc-editor.org/rfc/rfc8949.html#name-encoding-indicators.
FloatPrecisionIndicator bool
// MaxNestedLevels specifies the max nested levels allowed for any combination of CBOR array, maps, and tags.
// Default is 32 levels and it can be set to [4, 65535]. Note that higher maximum levels of nesting can
// require larger amounts of stack to deserialize. Don't increase this higher than you require.
MaxNestedLevels int
// MaxArrayElements specifies the max number of elements for CBOR arrays.
// Default is 128*1024=131072 and it can be set to [16, 2147483647]
MaxArrayElements int
// MaxMapPairs specifies the max number of key-value pairs for CBOR maps.
// Default is 128*1024=131072 and it can be set to [16, 2147483647]
MaxMapPairs int
}
// DiagMode returns a DiagMode with immutable options.
func (opts DiagOptions) DiagMode() (DiagMode, error) {
return opts.diagMode()
}
func (opts DiagOptions) diagMode() (*diagMode, error) {
if err := opts.ByteStringEncoding.valid(); err != nil {
return nil, err
}
decMode, err := DecOptions{
MaxNestedLevels: opts.MaxNestedLevels,
MaxArrayElements: opts.MaxArrayElements,
MaxMapPairs: opts.MaxMapPairs,
}.decMode()
if err != nil {
return nil, err
}
return &diagMode{
byteStringEncoding: opts.ByteStringEncoding,
byteStringHexWhitespace: opts.ByteStringHexWhitespace,
byteStringText: opts.ByteStringText,
byteStringEmbeddedCBOR: opts.ByteStringEmbeddedCBOR,
cborSequence: opts.CBORSequence,
floatPrecisionIndicator: opts.FloatPrecisionIndicator,
decMode: decMode,
}, nil
}
type diagMode struct {
byteStringEncoding ByteStringEncoding
byteStringHexWhitespace bool
byteStringText bool
byteStringEmbeddedCBOR bool
cborSequence bool
floatPrecisionIndicator bool
decMode *decMode
}
// DiagOptions returns user specified options used to create this DiagMode.
func (dm *diagMode) DiagOptions() DiagOptions {
return DiagOptions{
ByteStringEncoding: dm.byteStringEncoding,
ByteStringHexWhitespace: dm.byteStringHexWhitespace,
ByteStringText: dm.byteStringText,
ByteStringEmbeddedCBOR: dm.byteStringEmbeddedCBOR,
CBORSequence: dm.cborSequence,
FloatPrecisionIndicator: dm.floatPrecisionIndicator,
MaxNestedLevels: dm.decMode.maxNestedLevels,
MaxArrayElements: dm.decMode.maxArrayElements,
MaxMapPairs: dm.decMode.maxMapPairs,
}
}
// Diagnose returns extended diagnostic notation (EDN) of CBOR data items using the DiagMode.
func (dm *diagMode) Diagnose(data []byte) (string, error) {
return newDiagnose(data, dm.decMode, dm).diag(dm.cborSequence)
}
// DiagnoseFirst returns extended diagnostic notation (EDN) of the first CBOR data item using the DiagMode. Any remaining bytes are returned in rest.
func (dm *diagMode) DiagnoseFirst(data []byte) (diagNotation string, rest []byte, err error) {
return newDiagnose(data, dm.decMode, dm).diagFirst()
}
var defaultDiagMode, _ = DiagOptions{}.diagMode()
// Diagnose returns extended diagnostic notation (EDN) of CBOR data items
// using the default diagnostic mode.
//
// Refer to https://www.rfc-editor.org/rfc/rfc8949.html#name-diagnostic-notation.
func Diagnose(data []byte) (string, error) {
return defaultDiagMode.Diagnose(data)
}
// Diagnose returns extended diagnostic notation (EDN) of the first CBOR data item using the DiagMode. Any remaining bytes are returned in rest.
func DiagnoseFirst(data []byte) (diagNotation string, rest []byte, err error) {
return defaultDiagMode.DiagnoseFirst(data)
}
type diagnose struct {
dm *diagMode
d *decoder
w *bytes.Buffer
}
func newDiagnose(data []byte, decm *decMode, diagm *diagMode) *diagnose {
return &diagnose{
dm: diagm,
d: &decoder{data: data, dm: decm},
w: &bytes.Buffer{},
}
}
func (di *diagnose) diag(cborSequence bool) (string, error) {
// CBOR Sequence
firstItem := true
for {
switch err := di.wellformed(cborSequence); err {
case nil:
if !firstItem {
di.w.WriteString(", ")
}
firstItem = false
if itemErr := di.item(); itemErr != nil {
return di.w.String(), itemErr
}
case io.EOF:
if firstItem {
return di.w.String(), err
}
return di.w.String(), nil
default:
return di.w.String(), err
}
}
}
func (di *diagnose) diagFirst() (diagNotation string, rest []byte, err error) {
err = di.wellformed(true)
if err == nil {
err = di.item()
}
if err == nil {
// Return EDN and the rest of the data slice (which might be len 0)
return di.w.String(), di.d.data[di.d.off:], nil
}
return di.w.String(), nil, err
}
func (di *diagnose) wellformed(allowExtraData bool) error {
off := di.d.off
err := di.d.wellformed(allowExtraData, false)
di.d.off = off
return err
}
func (di *diagnose) item() error { //nolint:gocyclo
initialByte := di.d.data[di.d.off]
switch initialByte {
case cborByteStringWithIndefiniteLengthHead,
cborTextStringWithIndefiniteLengthHead: // indefinite-length byte/text string
di.d.off++
if isBreakFlag(di.d.data[di.d.off]) {
di.d.off++
switch initialByte {
case cborByteStringWithIndefiniteLengthHead:
// indefinite-length bytes with no chunks.
di.w.WriteString(`''_`)
return nil
case cborTextStringWithIndefiniteLengthHead:
// indefinite-length text with no chunks.
di.w.WriteString(`""_`)
return nil
}
}
di.w.WriteString("(_ ")
i := 0
for !di.d.foundBreak() {
if i > 0 {
di.w.WriteString(", ")
}
i++
// wellformedIndefiniteString() already checked that the next item is a byte/text string.
if err := di.item(); err != nil {
return err
}
}
di.w.WriteByte(')')
return nil
case cborArrayWithIndefiniteLengthHead: // indefinite-length array
di.d.off++
di.w.WriteString("[_ ")
i := 0
for !di.d.foundBreak() {
if i > 0 {
di.w.WriteString(", ")
}
i++
if err := di.item(); err != nil {
return err
}
}
di.w.WriteByte(']')
return nil
case cborMapWithIndefiniteLengthHead: // indefinite-length map
di.d.off++
di.w.WriteString("{_ ")
i := 0
for !di.d.foundBreak() {
if i > 0 {
di.w.WriteString(", ")
}
i++
// key
if err := di.item(); err != nil {
return err
}
di.w.WriteString(": ")
// value
if err := di.item(); err != nil {
return err
}
}
di.w.WriteByte('}')
return nil
}
t := di.d.nextCBORType()
switch t {
case cborTypePositiveInt:
_, _, val := di.d.getHead()
di.w.WriteString(strconv.FormatUint(val, 10))
return nil
case cborTypeNegativeInt:
_, _, val := di.d.getHead()
if val > math.MaxInt64 {
// CBOR negative integer overflows int64, use big.Int to store value.
bi := new(big.Int)
bi.SetUint64(val)
bi.Add(bi, big.NewInt(1))
bi.Neg(bi)
di.w.WriteString(bi.String())
return nil
}
nValue := int64(-1) ^ int64(val)
di.w.WriteString(strconv.FormatInt(nValue, 10))
return nil
case cborTypeByteString:
b, _ := di.d.parseByteString()
return di.encodeByteString(b)
case cborTypeTextString:
b, err := di.d.parseTextString()
if err != nil {
return err
}
return di.encodeTextString(string(b), '"')
case cborTypeArray:
_, _, val := di.d.getHead()
count := int(val)
di.w.WriteByte('[')
for i := 0; i < count; i++ {
if i > 0 {
di.w.WriteString(", ")
}
if err := di.item(); err != nil {
return err
}
}
di.w.WriteByte(']')
return nil
case cborTypeMap:
_, _, val := di.d.getHead()
count := int(val)
di.w.WriteByte('{')
for i := 0; i < count; i++ {
if i > 0 {
di.w.WriteString(", ")
}
// key
if err := di.item(); err != nil {
return err
}
di.w.WriteString(": ")
// value
if err := di.item(); err != nil {
return err
}
}
di.w.WriteByte('}')
return nil
case cborTypeTag:
_, _, tagNum := di.d.getHead()
switch tagNum {
case tagNumUnsignedBignum:
if nt := di.d.nextCBORType(); nt != cborTypeByteString {
return newInadmissibleTagContentTypeError(
tagNumUnsignedBignum,
"byte string",
nt.String())
}
b, _ := di.d.parseByteString()
bi := new(big.Int).SetBytes(b)
di.w.WriteString(bi.String())
return nil
case tagNumNegativeBignum:
if nt := di.d.nextCBORType(); nt != cborTypeByteString {
return newInadmissibleTagContentTypeError(
tagNumNegativeBignum,
"byte string",
nt.String(),
)
}
b, _ := di.d.parseByteString()
bi := new(big.Int).SetBytes(b)
bi.Add(bi, big.NewInt(1))
bi.Neg(bi)
di.w.WriteString(bi.String())
return nil
default:
di.w.WriteString(strconv.FormatUint(tagNum, 10))
di.w.WriteByte('(')
if err := di.item(); err != nil {
return err
}
di.w.WriteByte(')')
return nil
}
case cborTypePrimitives:
_, ai, val := di.d.getHead()
switch ai {
case additionalInformationAsFalse:
di.w.WriteString("false")
return nil
case additionalInformationAsTrue:
di.w.WriteString("true")
return nil
case additionalInformationAsNull:
di.w.WriteString("null")
return nil
case additionalInformationAsUndefined:
di.w.WriteString("undefined")
return nil
case additionalInformationAsFloat16,
additionalInformationAsFloat32,
additionalInformationAsFloat64:
return di.encodeFloat(ai, val)
default:
di.w.WriteString("simple(")
di.w.WriteString(strconv.FormatUint(val, 10))
di.w.WriteByte(')')
return nil
}
}
return nil
}
// writeU16 format a rune as "\uxxxx"
func (di *diagnose) writeU16(val rune) {
di.w.WriteString("\\u")
var in [2]byte
in[0] = byte(val >> 8)
in[1] = byte(val)
sz := hex.EncodedLen(len(in))
di.w.Grow(sz)
dst := di.w.Bytes()[di.w.Len() : di.w.Len()+sz]
hex.Encode(dst, in[:])
di.w.Write(dst)
}
var rawBase32Encoding = base32.StdEncoding.WithPadding(base32.NoPadding)
var rawBase32HexEncoding = base32.HexEncoding.WithPadding(base32.NoPadding)
func (di *diagnose) encodeByteString(val []byte) error {
if len(val) > 0 {
if di.dm.byteStringText && utf8.Valid(val) {
return di.encodeTextString(string(val), '\'')
}
if di.dm.byteStringEmbeddedCBOR {
di2 := newDiagnose(val, di.dm.decMode, di.dm)
// should always notating embedded CBOR sequence.
if str, err := di2.diag(true); err == nil {
di.w.WriteString("<<")
di.w.WriteString(str)
di.w.WriteString(">>")
return nil
}
}
}
switch di.dm.byteStringEncoding {
case ByteStringBase16Encoding:
di.w.WriteString("h'")
if di.dm.byteStringHexWhitespace {
sz := hex.EncodedLen(len(val))
if len(val) > 0 {
sz += len(val) - 1
}
di.w.Grow(sz)
dst := di.w.Bytes()[di.w.Len():]
for i := range val {
if i > 0 {
dst = append(dst, ' ')
}
hex.Encode(dst[len(dst):len(dst)+2], val[i:i+1])
dst = dst[:len(dst)+2]
}
di.w.Write(dst)
} else {
sz := hex.EncodedLen(len(val))
di.w.Grow(sz)
dst := di.w.Bytes()[di.w.Len() : di.w.Len()+sz]
hex.Encode(dst, val)
di.w.Write(dst)
}
di.w.WriteByte('\'')
return nil
case ByteStringBase32Encoding:
di.w.WriteString("b32'")
sz := rawBase32Encoding.EncodedLen(len(val))
di.w.Grow(sz)
dst := di.w.Bytes()[di.w.Len() : di.w.Len()+sz]
rawBase32Encoding.Encode(dst, val)
di.w.Write(dst)
di.w.WriteByte('\'')
return nil
case ByteStringBase32HexEncoding:
di.w.WriteString("h32'")
sz := rawBase32HexEncoding.EncodedLen(len(val))
di.w.Grow(sz)
dst := di.w.Bytes()[di.w.Len() : di.w.Len()+sz]
rawBase32HexEncoding.Encode(dst, val)
di.w.Write(dst)
di.w.WriteByte('\'')
return nil
case ByteStringBase64Encoding:
di.w.WriteString("b64'")
sz := base64.RawURLEncoding.EncodedLen(len(val))
di.w.Grow(sz)
dst := di.w.Bytes()[di.w.Len() : di.w.Len()+sz]
base64.RawURLEncoding.Encode(dst, val)
di.w.Write(dst)
di.w.WriteByte('\'')
return nil
default:
// It should not be possible for users to construct a *diagMode with an invalid byte
// string encoding.
panic(fmt.Sprintf("diagmode has invalid ByteStringEncoding %v", di.dm.byteStringEncoding))
}
}
const utf16SurrSelf = rune(0x10000)
// quote should be either `'` or `"`
func (di *diagnose) encodeTextString(val string, quote byte) error {
di.w.WriteByte(quote)
for i := 0; i < len(val); {
if b := val[i]; b < utf8.RuneSelf {
switch {
case b == '\t', b == '\n', b == '\r', b == '\\', b == quote:
di.w.WriteByte('\\')
switch b {
case '\t':
b = 't'
case '\n':
b = 'n'
case '\r':
b = 'r'
}
di.w.WriteByte(b)
case b >= ' ' && b <= '~':
di.w.WriteByte(b)
default:
di.writeU16(rune(b))
}
i++
continue
}
c, size := utf8.DecodeRuneInString(val[i:])
switch {
case c == utf8.RuneError:
return &SemanticError{"cbor: invalid UTF-8 string"}
case c < utf16SurrSelf:
di.writeU16(c)
default:
c1, c2 := utf16.EncodeRune(c)
di.writeU16(c1)
di.writeU16(c2)
}
i += size
}
di.w.WriteByte(quote)
return nil
}
func (di *diagnose) encodeFloat(ai byte, val uint64) error {
f64 := float64(0)
switch ai {
case additionalInformationAsFloat16:
f16 := float16.Frombits(uint16(val))
switch {
case f16.IsNaN():
di.w.WriteString("NaN")
return nil
case f16.IsInf(1):
di.w.WriteString("Infinity")
return nil
case f16.IsInf(-1):
di.w.WriteString("-Infinity")
return nil
default:
f64 = float64(f16.Float32())
}
case additionalInformationAsFloat32:
f32 := math.Float32frombits(uint32(val))
switch {
case f32 != f32:
di.w.WriteString("NaN")
return nil
case f32 > math.MaxFloat32:
di.w.WriteString("Infinity")
return nil
case f32 < -math.MaxFloat32:
di.w.WriteString("-Infinity")
return nil
default:
f64 = float64(f32)
}
case additionalInformationAsFloat64:
f64 = math.Float64frombits(val)
switch {
case f64 != f64:
di.w.WriteString("NaN")
return nil
case f64 > math.MaxFloat64:
di.w.WriteString("Infinity")
return nil
case f64 < -math.MaxFloat64:
di.w.WriteString("-Infinity")
return nil
}
}
// Use ES6 number to string conversion which should match most JSON generators.
// Inspired by https://github.com/golang/go/blob/4df10fba1687a6d4f51d7238a403f8f2298f6a16/src/encoding/json/encode.go#L585
const bitSize = 64
b := make([]byte, 0, 32)
if abs := math.Abs(f64); abs != 0 && (abs < 1e-6 || abs >= 1e21) {
b = strconv.AppendFloat(b, f64, 'e', -1, bitSize)
// clean up e-09 to e-9
n := len(b)
if n >= 4 && string(b[n-4:n-1]) == "e-0" {
b = append(b[:n-2], b[n-1])
}
} else {
b = strconv.AppendFloat(b, f64, 'f', -1, bitSize)
}
// add decimal point and trailing zero if needed
if bytes.IndexByte(b, '.') < 0 {
if i := bytes.IndexByte(b, 'e'); i < 0 {
b = append(b, '.', '0')
} else {
b = append(b[:i+2], b[i:]...)
b[i] = '.'
b[i+1] = '0'
}
}
di.w.WriteString(string(b))
if di.dm.floatPrecisionIndicator {
switch ai {
case additionalInformationAsFloat16:
di.w.WriteString("_1")
return nil
case additionalInformationAsFloat32:
di.w.WriteString("_2")
return nil
case additionalInformationAsFloat64:
di.w.WriteString("_3")
return nil
}
}
return nil
}