1
//
2
// Copyright (c) 2011-2019 Canonical Ltd
3
// Copyright (c) 2006-2010 Kirill Simonov
4
//
5
// Permission is hereby granted, free of charge, to any person obtaining a copy of
6
// this software and associated documentation files (the "Software"), to deal in
7
// the Software without restriction, including without limitation the rights to
8
// use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
9
// of the Software, and to permit persons to whom the Software is furnished to do
10
// so, subject to the following conditions:
11
//
12
// The above copyright notice and this permission notice shall be included in all
13
// copies or substantial portions of the Software.
14
//
15
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17
// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18
// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19
// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20
// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21
// SOFTWARE.
22

23
package yaml
24

25
import (
26
	"bytes"
27
	"fmt"
28
)
29

30
// Introduction
31
// ************
32
//
33
// The following notes assume that you are familiar with the YAML specification
34
// (http://yaml.org/spec/1.2/spec.html).  We mostly follow it, although in
35
// some cases we are less restrictive that it requires.
36
//
37
// The process of transforming a YAML stream into a sequence of events is
38
// divided on two steps: Scanning and Parsing.
39
//
40
// The Scanner transforms the input stream into a sequence of tokens, while the
41
// parser transform the sequence of tokens produced by the Scanner into a
42
// sequence of parsing events.
43
//
44
// The Scanner is rather clever and complicated. The Parser, on the contrary,
45
// is a straightforward implementation of a recursive-descendant parser (or,
46
// LL(1) parser, as it is usually called).
47
//
48
// Actually there are two issues of Scanning that might be called "clever", the
49
// rest is quite straightforward.  The issues are "block collection start" and
50
// "simple keys".  Both issues are explained below in details.
51
//
52
// Here the Scanning step is explained and implemented.  We start with the list
53
// of all the tokens produced by the Scanner together with short descriptions.
54
//
55
// Now, tokens:
56
//
57
//      STREAM-START(encoding)          # The stream start.
58
//      STREAM-END                      # The stream end.
59
//      VERSION-DIRECTIVE(major,minor)  # The '%YAML' directive.
60
//      TAG-DIRECTIVE(handle,prefix)    # The '%TAG' directive.
61
//      DOCUMENT-START                  # '---'
62
//      DOCUMENT-END                    # '...'
63
//      BLOCK-SEQUENCE-START            # Indentation increase denoting a block
64
//      BLOCK-MAPPING-START             # sequence or a block mapping.
65
//      BLOCK-END                       # Indentation decrease.
66
//      FLOW-SEQUENCE-START             # '['
67
//      FLOW-SEQUENCE-END               # ']'
68
//      BLOCK-SEQUENCE-START            # '{'
69
//      BLOCK-SEQUENCE-END              # '}'
70
//      BLOCK-ENTRY                     # '-'
71
//      FLOW-ENTRY                      # ','
72
//      KEY                             # '?' or nothing (simple keys).
73
//      VALUE                           # ':'
74
//      ALIAS(anchor)                   # '*anchor'
75
//      ANCHOR(anchor)                  # '&anchor'
76
//      TAG(handle,suffix)              # '!handle!suffix'
77
//      SCALAR(value,style)             # A scalar.
78
//
79
// The following two tokens are "virtual" tokens denoting the beginning and the
80
// end of the stream:
81
//
82
//      STREAM-START(encoding)
83
//      STREAM-END
84
//
85
// We pass the information about the input stream encoding with the
86
// STREAM-START token.
87
//
88
// The next two tokens are responsible for tags:
89
//
90
//      VERSION-DIRECTIVE(major,minor)
91
//      TAG-DIRECTIVE(handle,prefix)
92
//
93
// Example:
94
//
95
//      %YAML   1.1
96
//      %TAG    !   !foo
97
//      %TAG    !yaml!  tag:yaml.org,2002:
98
//      ---
99
//
100
// The correspoding sequence of tokens:
101
//
102
//      STREAM-START(utf-8)
103
//      VERSION-DIRECTIVE(1,1)
104
//      TAG-DIRECTIVE("!","!foo")
105
//      TAG-DIRECTIVE("!yaml","tag:yaml.org,2002:")
106
//      DOCUMENT-START
107
//      STREAM-END
108
//
109
// Note that the VERSION-DIRECTIVE and TAG-DIRECTIVE tokens occupy a whole
110
// line.
111
//
112
// The document start and end indicators are represented by:
113
//
114
//      DOCUMENT-START
115
//      DOCUMENT-END
116
//
117
// Note that if a YAML stream contains an implicit document (without '---'
118
// and '...' indicators), no DOCUMENT-START and DOCUMENT-END tokens will be
119
// produced.
120
//
121
// In the following examples, we present whole documents together with the
122
// produced tokens.
123
//
124
//      1. An implicit document:
125
//
126
//          'a scalar'
127
//
128
//      Tokens:
129
//
130
//          STREAM-START(utf-8)
131
//          SCALAR("a scalar",single-quoted)
132
//          STREAM-END
133
//
134
//      2. An explicit document:
135
//
136
//          ---
137
//          'a scalar'
138
//          ...
139
//
140
//      Tokens:
141
//
142
//          STREAM-START(utf-8)
143
//          DOCUMENT-START
144
//          SCALAR("a scalar",single-quoted)
145
//          DOCUMENT-END
146
//          STREAM-END
147
//
148
//      3. Several documents in a stream:
149
//
150
//          'a scalar'
151
//          ---
152
//          'another scalar'
153
//          ---
154
//          'yet another scalar'
155
//
156
//      Tokens:
157
//
158
//          STREAM-START(utf-8)
159
//          SCALAR("a scalar",single-quoted)
160
//          DOCUMENT-START
161
//          SCALAR("another scalar",single-quoted)
162
//          DOCUMENT-START
163
//          SCALAR("yet another scalar",single-quoted)
164
//          STREAM-END
165
//
166
// We have already introduced the SCALAR token above.  The following tokens are
167
// used to describe aliases, anchors, tag, and scalars:
168
//
169
//      ALIAS(anchor)
170
//      ANCHOR(anchor)
171
//      TAG(handle,suffix)
172
//      SCALAR(value,style)
173
//
174
// The following series of examples illustrate the usage of these tokens:
175
//
176
//      1. A recursive sequence:
177
//
178
//          &A [ *A ]
179
//
180
//      Tokens:
181
//
182
//          STREAM-START(utf-8)
183
//          ANCHOR("A")
184
//          FLOW-SEQUENCE-START
185
//          ALIAS("A")
186
//          FLOW-SEQUENCE-END
187
//          STREAM-END
188
//
189
//      2. A tagged scalar:
190
//
191
//          !!float "3.14"  # A good approximation.
192
//
193
//      Tokens:
194
//
195
//          STREAM-START(utf-8)
196
//          TAG("!!","float")
197
//          SCALAR("3.14",double-quoted)
198
//          STREAM-END
199
//
200
//      3. Various scalar styles:
201
//
202
//          --- # Implicit empty plain scalars do not produce tokens.
203
//          --- a plain scalar
204
//          --- 'a single-quoted scalar'
205
//          --- "a double-quoted scalar"
206
//          --- |-
207
//            a literal scalar
208
//          --- >-
209
//            a folded
210
//            scalar
211
//
212
//      Tokens:
213
//
214
//          STREAM-START(utf-8)
215
//          DOCUMENT-START
216
//          DOCUMENT-START
217
//          SCALAR("a plain scalar",plain)
218
//          DOCUMENT-START
219
//          SCALAR("a single-quoted scalar",single-quoted)
220
//          DOCUMENT-START
221
//          SCALAR("a double-quoted scalar",double-quoted)
222
//          DOCUMENT-START
223
//          SCALAR("a literal scalar",literal)
224
//          DOCUMENT-START
225
//          SCALAR("a folded scalar",folded)
226
//          STREAM-END
227
//
228
// Now it's time to review collection-related tokens. We will start with
229
// flow collections:
230
//
231
//      FLOW-SEQUENCE-START
232
//      FLOW-SEQUENCE-END
233
//      FLOW-MAPPING-START
234
//      FLOW-MAPPING-END
235
//      FLOW-ENTRY
236
//      KEY
237
//      VALUE
238
//
239
// The tokens FLOW-SEQUENCE-START, FLOW-SEQUENCE-END, FLOW-MAPPING-START, and
240
// FLOW-MAPPING-END represent the indicators '[', ']', '{', and '}'
241
// correspondingly.  FLOW-ENTRY represent the ',' indicator.  Finally the
242
// indicators '?' and ':', which are used for denoting mapping keys and values,
243
// are represented by the KEY and VALUE tokens.
244
//
245
// The following examples show flow collections:
246
//
247
//      1. A flow sequence:
248
//
249
//          [item 1, item 2, item 3]
250
//
251
//      Tokens:
252
//
253
//          STREAM-START(utf-8)
254
//          FLOW-SEQUENCE-START
255
//          SCALAR("item 1",plain)
256
//          FLOW-ENTRY
257
//          SCALAR("item 2",plain)
258
//          FLOW-ENTRY
259
//          SCALAR("item 3",plain)
260
//          FLOW-SEQUENCE-END
261
//          STREAM-END
262
//
263
//      2. A flow mapping:
264
//
265
//          {
266
//              a simple key: a value,  # Note that the KEY token is produced.
267
//              ? a complex key: another value,
268
//          }
269
//
270
//      Tokens:
271
//
272
//          STREAM-START(utf-8)
273
//          FLOW-MAPPING-START
274
//          KEY
275
//          SCALAR("a simple key",plain)
276
//          VALUE
277
//          SCALAR("a value",plain)
278
//          FLOW-ENTRY
279
//          KEY
280
//          SCALAR("a complex key",plain)
281
//          VALUE
282
//          SCALAR("another value",plain)
283
//          FLOW-ENTRY
284
//          FLOW-MAPPING-END
285
//          STREAM-END
286
//
287
// A simple key is a key which is not denoted by the '?' indicator.  Note that
288
// the Scanner still produce the KEY token whenever it encounters a simple key.
289
//
290
// For scanning block collections, the following tokens are used (note that we
291
// repeat KEY and VALUE here):
292
//
293
//      BLOCK-SEQUENCE-START
294
//      BLOCK-MAPPING-START
295
//      BLOCK-END
296
//      BLOCK-ENTRY
297
//      KEY
298
//      VALUE
299
//
300
// The tokens BLOCK-SEQUENCE-START and BLOCK-MAPPING-START denote indentation
301
// increase that precedes a block collection (cf. the INDENT token in Python).
302
// The token BLOCK-END denote indentation decrease that ends a block collection
303
// (cf. the DEDENT token in Python).  However YAML has some syntax pecularities
304
// that makes detections of these tokens more complex.
305
//
306
// The tokens BLOCK-ENTRY, KEY, and VALUE are used to represent the indicators
307
// '-', '?', and ':' correspondingly.
308
//
309
// The following examples show how the tokens BLOCK-SEQUENCE-START,
310
// BLOCK-MAPPING-START, and BLOCK-END are emitted by the Scanner:
311
//
312
//      1. Block sequences:
313
//
314
//          - item 1
315
//          - item 2
316
//          -
317
//            - item 3.1
318
//            - item 3.2
319
//          -
320
//            key 1: value 1
321
//            key 2: value 2
322
//
323
//      Tokens:
324
//
325
//          STREAM-START(utf-8)
326
//          BLOCK-SEQUENCE-START
327
//          BLOCK-ENTRY
328
//          SCALAR("item 1",plain)
329
//          BLOCK-ENTRY
330
//          SCALAR("item 2",plain)
331
//          BLOCK-ENTRY
332
//          BLOCK-SEQUENCE-START
333
//          BLOCK-ENTRY
334
//          SCALAR("item 3.1",plain)
335
//          BLOCK-ENTRY
336
//          SCALAR("item 3.2",plain)
337
//          BLOCK-END
338
//          BLOCK-ENTRY
339
//          BLOCK-MAPPING-START
340
//          KEY
341
//          SCALAR("key 1",plain)
342
//          VALUE
343
//          SCALAR("value 1",plain)
344
//          KEY
345
//          SCALAR("key 2",plain)
346
//          VALUE
347
//          SCALAR("value 2",plain)
348
//          BLOCK-END
349
//          BLOCK-END
350
//          STREAM-END
351
//
352
//      2. Block mappings:
353
//
354
//          a simple key: a value   # The KEY token is produced here.
355
//          ? a complex key
356
//          : another value
357
//          a mapping:
358
//            key 1: value 1
359
//            key 2: value 2
360
//          a sequence:
361
//            - item 1
362
//            - item 2
363
//
364
//      Tokens:
365
//
366
//          STREAM-START(utf-8)
367
//          BLOCK-MAPPING-START
368
//          KEY
369
//          SCALAR("a simple key",plain)
370
//          VALUE
371
//          SCALAR("a value",plain)
372
//          KEY
373
//          SCALAR("a complex key",plain)
374
//          VALUE
375
//          SCALAR("another value",plain)
376
//          KEY
377
//          SCALAR("a mapping",plain)
378
//          BLOCK-MAPPING-START
379
//          KEY
380
//          SCALAR("key 1",plain)
381
//          VALUE
382
//          SCALAR("value 1",plain)
383
//          KEY
384
//          SCALAR("key 2",plain)
385
//          VALUE
386
//          SCALAR("value 2",plain)
387
//          BLOCK-END
388
//          KEY
389
//          SCALAR("a sequence",plain)
390
//          VALUE
391
//          BLOCK-SEQUENCE-START
392
//          BLOCK-ENTRY
393
//          SCALAR("item 1",plain)
394
//          BLOCK-ENTRY
395
//          SCALAR("item 2",plain)
396
//          BLOCK-END
397
//          BLOCK-END
398
//          STREAM-END
399
//
400
// YAML does not always require to start a new block collection from a new
401
// line.  If the current line contains only '-', '?', and ':' indicators, a new
402
// block collection may start at the current line.  The following examples
403
// illustrate this case:
404
//
405
//      1. Collections in a sequence:
406
//
407
//          - - item 1
408
//            - item 2
409
//          - key 1: value 1
410
//            key 2: value 2
411
//          - ? complex key
412
//            : complex value
413
//
414
//      Tokens:
415
//
416
//          STREAM-START(utf-8)
417
//          BLOCK-SEQUENCE-START
418
//          BLOCK-ENTRY
419
//          BLOCK-SEQUENCE-START
420
//          BLOCK-ENTRY
421
//          SCALAR("item 1",plain)
422
//          BLOCK-ENTRY
423
//          SCALAR("item 2",plain)
424
//          BLOCK-END
425
//          BLOCK-ENTRY
426
//          BLOCK-MAPPING-START
427
//          KEY
428
//          SCALAR("key 1",plain)
429
//          VALUE
430
//          SCALAR("value 1",plain)
431
//          KEY
432
//          SCALAR("key 2",plain)
433
//          VALUE
434
//          SCALAR("value 2",plain)
435
//          BLOCK-END
436
//          BLOCK-ENTRY
437
//          BLOCK-MAPPING-START
438
//          KEY
439
//          SCALAR("complex key")
440
//          VALUE
441
//          SCALAR("complex value")
442
//          BLOCK-END
443
//          BLOCK-END
444
//          STREAM-END
445
//
446
//      2. Collections in a mapping:
447
//
448
//          ? a sequence
449
//          : - item 1
450
//            - item 2
451
//          ? a mapping
452
//          : key 1: value 1
453
//            key 2: value 2
454
//
455
//      Tokens:
456
//
457
//          STREAM-START(utf-8)
458
//          BLOCK-MAPPING-START
459
//          KEY
460
//          SCALAR("a sequence",plain)
461
//          VALUE
462
//          BLOCK-SEQUENCE-START
463
//          BLOCK-ENTRY
464
//          SCALAR("item 1",plain)
465
//          BLOCK-ENTRY
466
//          SCALAR("item 2",plain)
467
//          BLOCK-END
468
//          KEY
469
//          SCALAR("a mapping",plain)
470
//          VALUE
471
//          BLOCK-MAPPING-START
472
//          KEY
473
//          SCALAR("key 1",plain)
474
//          VALUE
475
//          SCALAR("value 1",plain)
476
//          KEY
477
//          SCALAR("key 2",plain)
478
//          VALUE
479
//          SCALAR("value 2",plain)
480
//          BLOCK-END
481
//          BLOCK-END
482
//          STREAM-END
483
//
484
// YAML also permits non-indented sequences if they are included into a block
485
// mapping.  In this case, the token BLOCK-SEQUENCE-START is not produced:
486
//
487
//      key:
488
//      - item 1    # BLOCK-SEQUENCE-START is NOT produced here.
489
//      - item 2
490
//
491
// Tokens:
492
//
493
//      STREAM-START(utf-8)
494
//      BLOCK-MAPPING-START
495
//      KEY
496
//      SCALAR("key",plain)
497
//      VALUE
498
//      BLOCK-ENTRY
499
//      SCALAR("item 1",plain)
500
//      BLOCK-ENTRY
501
//      SCALAR("item 2",plain)
502
//      BLOCK-END
503
//
504

505
// Ensure that the buffer contains the required number of characters.
506
// Return true on success, false on failure (reader error or memory error).
507
func cache(parser *yaml_parser_t, length int) bool {
508
	// [Go] This was inlined: !cache(A, B) -> unread < B && !update(A, B)
509
	return parser.unread >= length || yaml_parser_update_buffer(parser, length)
510
}
511

512
// Advance the buffer pointer.
513
func skip(parser *yaml_parser_t) {
514
	if !is_blank(parser.buffer, parser.buffer_pos) {
515
		parser.newlines = 0
516
	}
517
	parser.mark.index++
518
	parser.mark.column++
519
	parser.unread--
520
	parser.buffer_pos += width(parser.buffer[parser.buffer_pos])
521
}
522

523
func skip_line(parser *yaml_parser_t) {
524
	if is_crlf(parser.buffer, parser.buffer_pos) {
525
		parser.mark.index += 2
526
		parser.mark.column = 0
527
		parser.mark.line++
528
		parser.unread -= 2
529
		parser.buffer_pos += 2
530
		parser.newlines++
531
	} else if is_break(parser.buffer, parser.buffer_pos) {
532
		parser.mark.index++
533
		parser.mark.column = 0
534
		parser.mark.line++
535
		parser.unread--
536
		parser.buffer_pos += width(parser.buffer[parser.buffer_pos])
537
		parser.newlines++
538
	}
539
}
540

541
// Copy a character to a string buffer and advance pointers.
542
func read(parser *yaml_parser_t, s []byte) []byte {
543
	if !is_blank(parser.buffer, parser.buffer_pos) {
544
		parser.newlines = 0
545
	}
546
	w := width(parser.buffer[parser.buffer_pos])
547
	if w == 0 {
548
		panic("invalid character sequence")
549
	}
550
	if len(s) == 0 {
551
		s = make([]byte, 0, 32)
552
	}
553
	if w == 1 && len(s)+w <= cap(s) {
554
		s = s[:len(s)+1]
555
		s[len(s)-1] = parser.buffer[parser.buffer_pos]
556
		parser.buffer_pos++
557
	} else {
558
		s = append(s, parser.buffer[parser.buffer_pos:parser.buffer_pos+w]...)
559
		parser.buffer_pos += w
560
	}
561
	parser.mark.index++
562
	parser.mark.column++
563
	parser.unread--
564
	return s
565
}
566

567
// Copy a line break character to a string buffer and advance pointers.
568
func read_line(parser *yaml_parser_t, s []byte) []byte {
569
	buf := parser.buffer
570
	pos := parser.buffer_pos
571
	switch {
572
	case buf[pos] == '\r' && buf[pos+1] == '\n':
573
		// CR LF . LF
574
		s = append(s, '\n')
575
		parser.buffer_pos += 2
576
		parser.mark.index++
577
		parser.unread--
578
	case buf[pos] == '\r' || buf[pos] == '\n':
579
		// CR|LF . LF
580
		s = append(s, '\n')
581
		parser.buffer_pos += 1
582
	case buf[pos] == '\xC2' && buf[pos+1] == '\x85':
583
		// NEL . LF
584
		s = append(s, '\n')
585
		parser.buffer_pos += 2
586
	case buf[pos] == '\xE2' && buf[pos+1] == '\x80' && (buf[pos+2] == '\xA8' || buf[pos+2] == '\xA9'):
587
		// LS|PS . LS|PS
588
		s = append(s, buf[parser.buffer_pos:pos+3]...)
589
		parser.buffer_pos += 3
590
	default:
591
		return s
592
	}
593
	parser.mark.index++
594
	parser.mark.column = 0
595
	parser.mark.line++
596
	parser.unread--
597
	parser.newlines++
598
	return s
599
}
600

601
// Get the next token.
602
func yaml_parser_scan(parser *yaml_parser_t, token *yaml_token_t) bool {
603
	// Erase the token object.
604
	*token = yaml_token_t{} // [Go] Is this necessary?
605

606
	// No tokens after STREAM-END or error.
607
	if parser.stream_end_produced || parser.error != yaml_NO_ERROR {
608
		return true
609
	}
610

611
	// Ensure that the tokens queue contains enough tokens.
612
	if !parser.token_available {
613
		if !yaml_parser_fetch_more_tokens(parser) {
614
			return false
615
		}
616
	}
617

618
	// Fetch the next token from the queue.
619
	*token = parser.tokens[parser.tokens_head]
620
	parser.tokens_head++
621
	parser.tokens_parsed++
622
	parser.token_available = false
623

624
	if token.typ == yaml_STREAM_END_TOKEN {
625
		parser.stream_end_produced = true
626
	}
627
	return true
628
}
629

630
// Set the scanner error and return false.
631
func yaml_parser_set_scanner_error(parser *yaml_parser_t, context string, context_mark yaml_mark_t, problem string) bool {
632
	parser.error = yaml_SCANNER_ERROR
633
	parser.context = context
634
	parser.context_mark = context_mark
635
	parser.problem = problem
636
	parser.problem_mark = parser.mark
637
	return false
638
}
639

640
func yaml_parser_set_scanner_tag_error(parser *yaml_parser_t, directive bool, context_mark yaml_mark_t, problem string) bool {
641
	context := "while parsing a tag"
642
	if directive {
643
		context = "while parsing a %TAG directive"
644
	}
645
	return yaml_parser_set_scanner_error(parser, context, context_mark, problem)
646
}
647

648
func trace(args ...interface{}) func() {
649
	pargs := append([]interface{}{"+++"}, args...)
650
	fmt.Println(pargs...)
651
	pargs = append([]interface{}{"---"}, args...)
652
	return func() { fmt.Println(pargs...) }
653
}
654

655
// Ensure that the tokens queue contains at least one token which can be
656
// returned to the Parser.
657
func yaml_parser_fetch_more_tokens(parser *yaml_parser_t) bool {
658
	// While we need more tokens to fetch, do it.
659
	for {
660
		// [Go] The comment parsing logic requires a lookahead of two tokens
661
		// so that foot comments may be parsed in time of associating them
662
		// with the tokens that are parsed before them, and also for line
663
		// comments to be transformed into head comments in some edge cases.
664
		if parser.tokens_head < len(parser.tokens)-2 {
665
			// If a potential simple key is at the head position, we need to fetch
666
			// the next token to disambiguate it.
667
			head_tok_idx, ok := parser.simple_keys_by_tok[parser.tokens_parsed]
668
			if !ok {
669
				break
670
			} else if valid, ok := yaml_simple_key_is_valid(parser, &parser.simple_keys[head_tok_idx]); !ok {
671
				return false
672
			} else if !valid {
673
				break
674
			}
675
		}
676
		// Fetch the next token.
677
		if !yaml_parser_fetch_next_token(parser) {
678
			return false
679
		}
680
	}
681

682
	parser.token_available = true
683
	return true
684
}
685

686
// The dispatcher for token fetchers.
687
func yaml_parser_fetch_next_token(parser *yaml_parser_t) (ok bool) {
688
	// Ensure that the buffer is initialized.
689
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
690
		return false
691
	}
692

693
	// Check if we just started scanning.  Fetch STREAM-START then.
694
	if !parser.stream_start_produced {
695
		return yaml_parser_fetch_stream_start(parser)
696
	}
697

698
	scan_mark := parser.mark
699

700
	// Eat whitespaces and comments until we reach the next token.
701
	if !yaml_parser_scan_to_next_token(parser) {
702
		return false
703
	}
704

705
	// [Go] While unrolling indents, transform the head comments of prior
706
	// indentation levels observed after scan_start into foot comments at
707
	// the respective indexes.
708

709
	// Check the indentation level against the current column.
710
	if !yaml_parser_unroll_indent(parser, parser.mark.column, scan_mark) {
711
		return false
712
	}
713

714
	// Ensure that the buffer contains at least 4 characters.  4 is the length
715
	// of the longest indicators ('--- ' and '... ').
716
	if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) {
717
		return false
718
	}
719

720
	// Is it the end of the stream?
721
	if is_z(parser.buffer, parser.buffer_pos) {
722
		return yaml_parser_fetch_stream_end(parser)
723
	}
724

725
	// Is it a directive?
726
	if parser.mark.column == 0 && parser.buffer[parser.buffer_pos] == '%' {
727
		return yaml_parser_fetch_directive(parser)
728
	}
729

730
	buf := parser.buffer
731
	pos := parser.buffer_pos
732

733
	// Is it the document start indicator?
734
	if parser.mark.column == 0 && buf[pos] == '-' && buf[pos+1] == '-' && buf[pos+2] == '-' && is_blankz(buf, pos+3) {
735
		return yaml_parser_fetch_document_indicator(parser, yaml_DOCUMENT_START_TOKEN)
736
	}
737

738
	// Is it the document end indicator?
739
	if parser.mark.column == 0 && buf[pos] == '.' && buf[pos+1] == '.' && buf[pos+2] == '.' && is_blankz(buf, pos+3) {
740
		return yaml_parser_fetch_document_indicator(parser, yaml_DOCUMENT_END_TOKEN)
741
	}
742

743
	comment_mark := parser.mark
744
	if len(parser.tokens) > 0 && (parser.flow_level == 0 && buf[pos] == ':' || parser.flow_level > 0 && buf[pos] == ',') {
745
		// Associate any following comments with the prior token.
746
		comment_mark = parser.tokens[len(parser.tokens)-1].start_mark
747
	}
748
	defer func() {
749
		if !ok {
750
			return
751
		}
752
		if len(parser.tokens) > 0 && parser.tokens[len(parser.tokens)-1].typ == yaml_BLOCK_ENTRY_TOKEN {
753
			// Sequence indicators alone have no line comments. It becomes
754
			// a head comment for whatever follows.
755
			return
756
		}
757
		if !yaml_parser_scan_line_comment(parser, comment_mark) {
758
			ok = false
759
			return
760
		}
761
	}()
762

763
	// Is it the flow sequence start indicator?
764
	if buf[pos] == '[' {
765
		return yaml_parser_fetch_flow_collection_start(parser, yaml_FLOW_SEQUENCE_START_TOKEN)
766
	}
767

768
	// Is it the flow mapping start indicator?
769
	if parser.buffer[parser.buffer_pos] == '{' {
770
		return yaml_parser_fetch_flow_collection_start(parser, yaml_FLOW_MAPPING_START_TOKEN)
771
	}
772

773
	// Is it the flow sequence end indicator?
774
	if parser.buffer[parser.buffer_pos] == ']' {
775
		return yaml_parser_fetch_flow_collection_end(parser,
776
			yaml_FLOW_SEQUENCE_END_TOKEN)
777
	}
778

779
	// Is it the flow mapping end indicator?
780
	if parser.buffer[parser.buffer_pos] == '}' {
781
		return yaml_parser_fetch_flow_collection_end(parser,
782
			yaml_FLOW_MAPPING_END_TOKEN)
783
	}
784

785
	// Is it the flow entry indicator?
786
	if parser.buffer[parser.buffer_pos] == ',' {
787
		return yaml_parser_fetch_flow_entry(parser)
788
	}
789

790
	// Is it the block entry indicator?
791
	if parser.buffer[parser.buffer_pos] == '-' && is_blankz(parser.buffer, parser.buffer_pos+1) {
792
		return yaml_parser_fetch_block_entry(parser)
793
	}
794

795
	// Is it the key indicator?
796
	if parser.buffer[parser.buffer_pos] == '?' && (parser.flow_level > 0 || is_blankz(parser.buffer, parser.buffer_pos+1)) {
797
		return yaml_parser_fetch_key(parser)
798
	}
799

800
	// Is it the value indicator?
801
	if parser.buffer[parser.buffer_pos] == ':' && (parser.flow_level > 0 || is_blankz(parser.buffer, parser.buffer_pos+1)) {
802
		return yaml_parser_fetch_value(parser)
803
	}
804

805
	// Is it an alias?
806
	if parser.buffer[parser.buffer_pos] == '*' {
807
		return yaml_parser_fetch_anchor(parser, yaml_ALIAS_TOKEN)
808
	}
809

810
	// Is it an anchor?
811
	if parser.buffer[parser.buffer_pos] == '&' {
812
		return yaml_parser_fetch_anchor(parser, yaml_ANCHOR_TOKEN)
813
	}
814

815
	// Is it a tag?
816
	if parser.buffer[parser.buffer_pos] == '!' {
817
		return yaml_parser_fetch_tag(parser)
818
	}
819

820
	// Is it a literal scalar?
821
	if parser.buffer[parser.buffer_pos] == '|' && parser.flow_level == 0 {
822
		return yaml_parser_fetch_block_scalar(parser, true)
823
	}
824

825
	// Is it a folded scalar?
826
	if parser.buffer[parser.buffer_pos] == '>' && parser.flow_level == 0 {
827
		return yaml_parser_fetch_block_scalar(parser, false)
828
	}
829

830
	// Is it a single-quoted scalar?
831
	if parser.buffer[parser.buffer_pos] == '\'' {
832
		return yaml_parser_fetch_flow_scalar(parser, true)
833
	}
834

835
	// Is it a double-quoted scalar?
836
	if parser.buffer[parser.buffer_pos] == '"' {
837
		return yaml_parser_fetch_flow_scalar(parser, false)
838
	}
839

840
	// Is it a plain scalar?
841
	//
842
	// A plain scalar may start with any non-blank characters except
843
	//
844
	//      '-', '?', ':', ',', '[', ']', '{', '}',
845
	//      '#', '&', '*', '!', '|', '>', '\'', '\"',
846
	//      '%', '@', '`'.
847
	//
848
	// In the block context (and, for the '-' indicator, in the flow context
849
	// too), it may also start with the characters
850
	//
851
	//      '-', '?', ':'
852
	//
853
	// if it is followed by a non-space character.
854
	//
855
	// The last rule is more restrictive than the specification requires.
856
	// [Go] TODO Make this logic more reasonable.
857
	//switch parser.buffer[parser.buffer_pos] {
858
	//case '-', '?', ':', ',', '?', '-', ',', ':', ']', '[', '}', '{', '&', '#', '!', '*', '>', '|', '"', '\'', '@', '%', '-', '`':
859
	//}
860
	if !(is_blankz(parser.buffer, parser.buffer_pos) || parser.buffer[parser.buffer_pos] == '-' ||
861
		parser.buffer[parser.buffer_pos] == '?' || parser.buffer[parser.buffer_pos] == ':' ||
862
		parser.buffer[parser.buffer_pos] == ',' || parser.buffer[parser.buffer_pos] == '[' ||
863
		parser.buffer[parser.buffer_pos] == ']' || parser.buffer[parser.buffer_pos] == '{' ||
864
		parser.buffer[parser.buffer_pos] == '}' || parser.buffer[parser.buffer_pos] == '#' ||
865
		parser.buffer[parser.buffer_pos] == '&' || parser.buffer[parser.buffer_pos] == '*' ||
866
		parser.buffer[parser.buffer_pos] == '!' || parser.buffer[parser.buffer_pos] == '|' ||
867
		parser.buffer[parser.buffer_pos] == '>' || parser.buffer[parser.buffer_pos] == '\'' ||
868
		parser.buffer[parser.buffer_pos] == '"' || parser.buffer[parser.buffer_pos] == '%' ||
869
		parser.buffer[parser.buffer_pos] == '@' || parser.buffer[parser.buffer_pos] == '`') ||
870
		(parser.buffer[parser.buffer_pos] == '-' && !is_blank(parser.buffer, parser.buffer_pos+1)) ||
871
		(parser.flow_level == 0 &&
872
			(parser.buffer[parser.buffer_pos] == '?' || parser.buffer[parser.buffer_pos] == ':') &&
873
			!is_blankz(parser.buffer, parser.buffer_pos+1)) {
874
		return yaml_parser_fetch_plain_scalar(parser)
875
	}
876

877
	// If we don't determine the token type so far, it is an error.
878
	return yaml_parser_set_scanner_error(parser,
879
		"while scanning for the next token", parser.mark,
880
		"found character that cannot start any token")
881
}
882

883
func yaml_simple_key_is_valid(parser *yaml_parser_t, simple_key *yaml_simple_key_t) (valid, ok bool) {
884
	if !simple_key.possible {
885
		return false, true
886
	}
887

888
	// The 1.2 specification says:
889
	//
890
	//     "If the ? indicator is omitted, parsing needs to see past the
891
	//     implicit key to recognize it as such. To limit the amount of
892
	//     lookahead required, the “:” indicator must appear at most 1024
893
	//     Unicode characters beyond the start of the key. In addition, the key
894
	//     is restricted to a single line."
895
	//
896
	if simple_key.mark.line < parser.mark.line || simple_key.mark.index+1024 < parser.mark.index {
897
		// Check if the potential simple key to be removed is required.
898
		if simple_key.required {
899
			return false, yaml_parser_set_scanner_error(parser,
900
				"while scanning a simple key", simple_key.mark,
901
				"could not find expected ':'")
902
		}
903
		simple_key.possible = false
904
		return false, true
905
	}
906
	return true, true
907
}
908

909
// Check if a simple key may start at the current position and add it if
910
// needed.
911
func yaml_parser_save_simple_key(parser *yaml_parser_t) bool {
912
	// A simple key is required at the current position if the scanner is in
913
	// the block context and the current column coincides with the indentation
914
	// level.
915

916
	required := parser.flow_level == 0 && parser.indent == parser.mark.column
917

918
	//
919
	// If the current position may start a simple key, save it.
920
	//
921
	if parser.simple_key_allowed {
922
		simple_key := yaml_simple_key_t{
923
			possible:     true,
924
			required:     required,
925
			token_number: parser.tokens_parsed + (len(parser.tokens) - parser.tokens_head),
926
			mark:         parser.mark,
927
		}
928

929
		if !yaml_parser_remove_simple_key(parser) {
930
			return false
931
		}
932
		parser.simple_keys[len(parser.simple_keys)-1] = simple_key
933
		parser.simple_keys_by_tok[simple_key.token_number] = len(parser.simple_keys) - 1
934
	}
935
	return true
936
}
937

938
// Remove a potential simple key at the current flow level.
939
func yaml_parser_remove_simple_key(parser *yaml_parser_t) bool {
940
	i := len(parser.simple_keys) - 1
941
	if parser.simple_keys[i].possible {
942
		// If the key is required, it is an error.
943
		if parser.simple_keys[i].required {
944
			return yaml_parser_set_scanner_error(parser,
945
				"while scanning a simple key", parser.simple_keys[i].mark,
946
				"could not find expected ':'")
947
		}
948
		// Remove the key from the stack.
949
		parser.simple_keys[i].possible = false
950
		delete(parser.simple_keys_by_tok, parser.simple_keys[i].token_number)
951
	}
952
	return true
953
}
954

955
// max_flow_level limits the flow_level
956
const max_flow_level = 10000
957

958
// Increase the flow level and resize the simple key list if needed.
959
func yaml_parser_increase_flow_level(parser *yaml_parser_t) bool {
960
	// Reset the simple key on the next level.
961
	parser.simple_keys = append(parser.simple_keys, yaml_simple_key_t{
962
		possible:     false,
963
		required:     false,
964
		token_number: parser.tokens_parsed + (len(parser.tokens) - parser.tokens_head),
965
		mark:         parser.mark,
966
	})
967

968
	// Increase the flow level.
969
	parser.flow_level++
970
	if parser.flow_level > max_flow_level {
971
		return yaml_parser_set_scanner_error(parser,
972
			"while increasing flow level", parser.simple_keys[len(parser.simple_keys)-1].mark,
973
			fmt.Sprintf("exceeded max depth of %d", max_flow_level))
974
	}
975
	return true
976
}
977

978
// Decrease the flow level.
979
func yaml_parser_decrease_flow_level(parser *yaml_parser_t) bool {
980
	if parser.flow_level > 0 {
981
		parser.flow_level--
982
		last := len(parser.simple_keys) - 1
983
		delete(parser.simple_keys_by_tok, parser.simple_keys[last].token_number)
984
		parser.simple_keys = parser.simple_keys[:last]
985
	}
986
	return true
987
}
988

989
// max_indents limits the indents stack size
990
const max_indents = 10000
991

992
// Push the current indentation level to the stack and set the new level
993
// the current column is greater than the indentation level.  In this case,
994
// append or insert the specified token into the token queue.
995
func yaml_parser_roll_indent(parser *yaml_parser_t, column, number int, typ yaml_token_type_t, mark yaml_mark_t) bool {
996
	// In the flow context, do nothing.
997
	if parser.flow_level > 0 {
998
		return true
999
	}
1000

1001
	if parser.indent < column {
1002
		// Push the current indentation level to the stack and set the new
1003
		// indentation level.
1004
		parser.indents = append(parser.indents, parser.indent)
1005
		parser.indent = column
1006
		if len(parser.indents) > max_indents {
1007
			return yaml_parser_set_scanner_error(parser,
1008
				"while increasing indent level", parser.simple_keys[len(parser.simple_keys)-1].mark,
1009
				fmt.Sprintf("exceeded max depth of %d", max_indents))
1010
		}
1011

1012
		// Create a token and insert it into the queue.
1013
		token := yaml_token_t{
1014
			typ:        typ,
1015
			start_mark: mark,
1016
			end_mark:   mark,
1017
		}
1018
		if number > -1 {
1019
			number -= parser.tokens_parsed
1020
		}
1021
		yaml_insert_token(parser, number, &token)
1022
	}
1023
	return true
1024
}
1025

1026
// Pop indentation levels from the indents stack until the current level
1027
// becomes less or equal to the column.  For each indentation level, append
1028
// the BLOCK-END token.
1029
func yaml_parser_unroll_indent(parser *yaml_parser_t, column int, scan_mark yaml_mark_t) bool {
1030
	// In the flow context, do nothing.
1031
	if parser.flow_level > 0 {
1032
		return true
1033
	}
1034

1035
	block_mark := scan_mark
1036
	block_mark.index--
1037

1038
	// Loop through the indentation levels in the stack.
1039
	for parser.indent > column {
1040

1041
		// [Go] Reposition the end token before potential following
1042
		//      foot comments of parent blocks. For that, search
1043
		//      backwards for recent comments that were at the same
1044
		//      indent as the block that is ending now.
1045
		stop_index := block_mark.index
1046
		for i := len(parser.comments) - 1; i >= 0; i-- {
1047
			comment := &parser.comments[i]
1048

1049
			if comment.end_mark.index < stop_index {
1050
				// Don't go back beyond the start of the comment/whitespace scan, unless column < 0.
1051
				// If requested indent column is < 0, then the document is over and everything else
1052
				// is a foot anyway.
1053
				break
1054
			}
1055
			if comment.start_mark.column == parser.indent+1 {
1056
				// This is a good match. But maybe there's a former comment
1057
				// at that same indent level, so keep searching.
1058
				block_mark = comment.start_mark
1059
			}
1060

1061
			// While the end of the former comment matches with
1062
			// the start of the following one, we know there's
1063
			// nothing in between and scanning is still safe.
1064
			stop_index = comment.scan_mark.index
1065
		}
1066

1067
		// Create a token and append it to the queue.
1068
		token := yaml_token_t{
1069
			typ:        yaml_BLOCK_END_TOKEN,
1070
			start_mark: block_mark,
1071
			end_mark:   block_mark,
1072
		}
1073
		yaml_insert_token(parser, -1, &token)
1074

1075
		// Pop the indentation level.
1076
		parser.indent = parser.indents[len(parser.indents)-1]
1077
		parser.indents = parser.indents[:len(parser.indents)-1]
1078
	}
1079
	return true
1080
}
1081

1082
// Initialize the scanner and produce the STREAM-START token.
1083
func yaml_parser_fetch_stream_start(parser *yaml_parser_t) bool {
1084

1085
	// Set the initial indentation.
1086
	parser.indent = -1
1087

1088
	// Initialize the simple key stack.
1089
	parser.simple_keys = append(parser.simple_keys, yaml_simple_key_t{})
1090

1091
	parser.simple_keys_by_tok = make(map[int]int)
1092

1093
	// A simple key is allowed at the beginning of the stream.
1094
	parser.simple_key_allowed = true
1095

1096
	// We have started.
1097
	parser.stream_start_produced = true
1098

1099
	// Create the STREAM-START token and append it to the queue.
1100
	token := yaml_token_t{
1101
		typ:        yaml_STREAM_START_TOKEN,
1102
		start_mark: parser.mark,
1103
		end_mark:   parser.mark,
1104
		encoding:   parser.encoding,
1105
	}
1106
	yaml_insert_token(parser, -1, &token)
1107
	return true
1108
}
1109

1110
// Produce the STREAM-END token and shut down the scanner.
1111
func yaml_parser_fetch_stream_end(parser *yaml_parser_t) bool {
1112

1113
	// Force new line.
1114
	if parser.mark.column != 0 {
1115
		parser.mark.column = 0
1116
		parser.mark.line++
1117
	}
1118

1119
	// Reset the indentation level.
1120
	if !yaml_parser_unroll_indent(parser, -1, parser.mark) {
1121
		return false
1122
	}
1123

1124
	// Reset simple keys.
1125
	if !yaml_parser_remove_simple_key(parser) {
1126
		return false
1127
	}
1128

1129
	parser.simple_key_allowed = false
1130

1131
	// Create the STREAM-END token and append it to the queue.
1132
	token := yaml_token_t{
1133
		typ:        yaml_STREAM_END_TOKEN,
1134
		start_mark: parser.mark,
1135
		end_mark:   parser.mark,
1136
	}
1137
	yaml_insert_token(parser, -1, &token)
1138
	return true
1139
}
1140

1141
// Produce a VERSION-DIRECTIVE or TAG-DIRECTIVE token.
1142
func yaml_parser_fetch_directive(parser *yaml_parser_t) bool {
1143
	// Reset the indentation level.
1144
	if !yaml_parser_unroll_indent(parser, -1, parser.mark) {
1145
		return false
1146
	}
1147

1148
	// Reset simple keys.
1149
	if !yaml_parser_remove_simple_key(parser) {
1150
		return false
1151
	}
1152

1153
	parser.simple_key_allowed = false
1154

1155
	// Create the YAML-DIRECTIVE or TAG-DIRECTIVE token.
1156
	token := yaml_token_t{}
1157
	if !yaml_parser_scan_directive(parser, &token) {
1158
		return false
1159
	}
1160
	// Append the token to the queue.
1161
	yaml_insert_token(parser, -1, &token)
1162
	return true
1163
}
1164

1165
// Produce the DOCUMENT-START or DOCUMENT-END token.
1166
func yaml_parser_fetch_document_indicator(parser *yaml_parser_t, typ yaml_token_type_t) bool {
1167
	// Reset the indentation level.
1168
	if !yaml_parser_unroll_indent(parser, -1, parser.mark) {
1169
		return false
1170
	}
1171

1172
	// Reset simple keys.
1173
	if !yaml_parser_remove_simple_key(parser) {
1174
		return false
1175
	}
1176

1177
	parser.simple_key_allowed = false
1178

1179
	// Consume the token.
1180
	start_mark := parser.mark
1181

1182
	skip(parser)
1183
	skip(parser)
1184
	skip(parser)
1185

1186
	end_mark := parser.mark
1187

1188
	// Create the DOCUMENT-START or DOCUMENT-END token.
1189
	token := yaml_token_t{
1190
		typ:        typ,
1191
		start_mark: start_mark,
1192
		end_mark:   end_mark,
1193
	}
1194
	// Append the token to the queue.
1195
	yaml_insert_token(parser, -1, &token)
1196
	return true
1197
}
1198

1199
// Produce the FLOW-SEQUENCE-START or FLOW-MAPPING-START token.
1200
func yaml_parser_fetch_flow_collection_start(parser *yaml_parser_t, typ yaml_token_type_t) bool {
1201

1202
	// The indicators '[' and '{' may start a simple key.
1203
	if !yaml_parser_save_simple_key(parser) {
1204
		return false
1205
	}
1206

1207
	// Increase the flow level.
1208
	if !yaml_parser_increase_flow_level(parser) {
1209
		return false
1210
	}
1211

1212
	// A simple key may follow the indicators '[' and '{'.
1213
	parser.simple_key_allowed = true
1214

1215
	// Consume the token.
1216
	start_mark := parser.mark
1217
	skip(parser)
1218
	end_mark := parser.mark
1219

1220
	// Create the FLOW-SEQUENCE-START of FLOW-MAPPING-START token.
1221
	token := yaml_token_t{
1222
		typ:        typ,
1223
		start_mark: start_mark,
1224
		end_mark:   end_mark,
1225
	}
1226
	// Append the token to the queue.
1227
	yaml_insert_token(parser, -1, &token)
1228
	return true
1229
}
1230

1231
// Produce the FLOW-SEQUENCE-END or FLOW-MAPPING-END token.
1232
func yaml_parser_fetch_flow_collection_end(parser *yaml_parser_t, typ yaml_token_type_t) bool {
1233
	// Reset any potential simple key on the current flow level.
1234
	if !yaml_parser_remove_simple_key(parser) {
1235
		return false
1236
	}
1237

1238
	// Decrease the flow level.
1239
	if !yaml_parser_decrease_flow_level(parser) {
1240
		return false
1241
	}
1242

1243
	// No simple keys after the indicators ']' and '}'.
1244
	parser.simple_key_allowed = false
1245

1246
	// Consume the token.
1247

1248
	start_mark := parser.mark
1249
	skip(parser)
1250
	end_mark := parser.mark
1251

1252
	// Create the FLOW-SEQUENCE-END of FLOW-MAPPING-END token.
1253
	token := yaml_token_t{
1254
		typ:        typ,
1255
		start_mark: start_mark,
1256
		end_mark:   end_mark,
1257
	}
1258
	// Append the token to the queue.
1259
	yaml_insert_token(parser, -1, &token)
1260
	return true
1261
}
1262

1263
// Produce the FLOW-ENTRY token.
1264
func yaml_parser_fetch_flow_entry(parser *yaml_parser_t) bool {
1265
	// Reset any potential simple keys on the current flow level.
1266
	if !yaml_parser_remove_simple_key(parser) {
1267
		return false
1268
	}
1269

1270
	// Simple keys are allowed after ','.
1271
	parser.simple_key_allowed = true
1272

1273
	// Consume the token.
1274
	start_mark := parser.mark
1275
	skip(parser)
1276
	end_mark := parser.mark
1277

1278
	// Create the FLOW-ENTRY token and append it to the queue.
1279
	token := yaml_token_t{
1280
		typ:        yaml_FLOW_ENTRY_TOKEN,
1281
		start_mark: start_mark,
1282
		end_mark:   end_mark,
1283
	}
1284
	yaml_insert_token(parser, -1, &token)
1285
	return true
1286
}
1287

1288
// Produce the BLOCK-ENTRY token.
1289
func yaml_parser_fetch_block_entry(parser *yaml_parser_t) bool {
1290
	// Check if the scanner is in the block context.
1291
	if parser.flow_level == 0 {
1292
		// Check if we are allowed to start a new entry.
1293
		if !parser.simple_key_allowed {
1294
			return yaml_parser_set_scanner_error(parser, "", parser.mark,
1295
				"block sequence entries are not allowed in this context")
1296
		}
1297
		// Add the BLOCK-SEQUENCE-START token if needed.
1298
		if !yaml_parser_roll_indent(parser, parser.mark.column, -1, yaml_BLOCK_SEQUENCE_START_TOKEN, parser.mark) {
1299
			return false
1300
		}
1301
	} else {
1302
		// It is an error for the '-' indicator to occur in the flow context,
1303
		// but we let the Parser detect and report about it because the Parser
1304
		// is able to point to the context.
1305
	}
1306

1307
	// Reset any potential simple keys on the current flow level.
1308
	if !yaml_parser_remove_simple_key(parser) {
1309
		return false
1310
	}
1311

1312
	// Simple keys are allowed after '-'.
1313
	parser.simple_key_allowed = true
1314

1315
	// Consume the token.
1316
	start_mark := parser.mark
1317
	skip(parser)
1318
	end_mark := parser.mark
1319

1320
	// Create the BLOCK-ENTRY token and append it to the queue.
1321
	token := yaml_token_t{
1322
		typ:        yaml_BLOCK_ENTRY_TOKEN,
1323
		start_mark: start_mark,
1324
		end_mark:   end_mark,
1325
	}
1326
	yaml_insert_token(parser, -1, &token)
1327
	return true
1328
}
1329

1330
// Produce the KEY token.
1331
func yaml_parser_fetch_key(parser *yaml_parser_t) bool {
1332

1333
	// In the block context, additional checks are required.
1334
	if parser.flow_level == 0 {
1335
		// Check if we are allowed to start a new key (not nessesary simple).
1336
		if !parser.simple_key_allowed {
1337
			return yaml_parser_set_scanner_error(parser, "", parser.mark,
1338
				"mapping keys are not allowed in this context")
1339
		}
1340
		// Add the BLOCK-MAPPING-START token if needed.
1341
		if !yaml_parser_roll_indent(parser, parser.mark.column, -1, yaml_BLOCK_MAPPING_START_TOKEN, parser.mark) {
1342
			return false
1343
		}
1344
	}
1345

1346
	// Reset any potential simple keys on the current flow level.
1347
	if !yaml_parser_remove_simple_key(parser) {
1348
		return false
1349
	}
1350

1351
	// Simple keys are allowed after '?' in the block context.
1352
	parser.simple_key_allowed = parser.flow_level == 0
1353

1354
	// Consume the token.
1355
	start_mark := parser.mark
1356
	skip(parser)
1357
	end_mark := parser.mark
1358

1359
	// Create the KEY token and append it to the queue.
1360
	token := yaml_token_t{
1361
		typ:        yaml_KEY_TOKEN,
1362
		start_mark: start_mark,
1363
		end_mark:   end_mark,
1364
	}
1365
	yaml_insert_token(parser, -1, &token)
1366
	return true
1367
}
1368

1369
// Produce the VALUE token.
1370
func yaml_parser_fetch_value(parser *yaml_parser_t) bool {
1371

1372
	simple_key := &parser.simple_keys[len(parser.simple_keys)-1]
1373

1374
	// Have we found a simple key?
1375
	if valid, ok := yaml_simple_key_is_valid(parser, simple_key); !ok {
1376
		return false
1377

1378
	} else if valid {
1379

1380
		// Create the KEY token and insert it into the queue.
1381
		token := yaml_token_t{
1382
			typ:        yaml_KEY_TOKEN,
1383
			start_mark: simple_key.mark,
1384
			end_mark:   simple_key.mark,
1385
		}
1386
		yaml_insert_token(parser, simple_key.token_number-parser.tokens_parsed, &token)
1387

1388
		// In the block context, we may need to add the BLOCK-MAPPING-START token.
1389
		if !yaml_parser_roll_indent(parser, simple_key.mark.column,
1390
			simple_key.token_number,
1391
			yaml_BLOCK_MAPPING_START_TOKEN, simple_key.mark) {
1392
			return false
1393
		}
1394

1395
		// Remove the simple key.
1396
		simple_key.possible = false
1397
		delete(parser.simple_keys_by_tok, simple_key.token_number)
1398

1399
		// A simple key cannot follow another simple key.
1400
		parser.simple_key_allowed = false
1401

1402
	} else {
1403
		// The ':' indicator follows a complex key.
1404

1405
		// In the block context, extra checks are required.
1406
		if parser.flow_level == 0 {
1407

1408
			// Check if we are allowed to start a complex value.
1409
			if !parser.simple_key_allowed {
1410
				return yaml_parser_set_scanner_error(parser, "", parser.mark,
1411
					"mapping values are not allowed in this context")
1412
			}
1413

1414
			// Add the BLOCK-MAPPING-START token if needed.
1415
			if !yaml_parser_roll_indent(parser, parser.mark.column, -1, yaml_BLOCK_MAPPING_START_TOKEN, parser.mark) {
1416
				return false
1417
			}
1418
		}
1419

1420
		// Simple keys after ':' are allowed in the block context.
1421
		parser.simple_key_allowed = parser.flow_level == 0
1422
	}
1423

1424
	// Consume the token.
1425
	start_mark := parser.mark
1426
	skip(parser)
1427
	end_mark := parser.mark
1428

1429
	// Create the VALUE token and append it to the queue.
1430
	token := yaml_token_t{
1431
		typ:        yaml_VALUE_TOKEN,
1432
		start_mark: start_mark,
1433
		end_mark:   end_mark,
1434
	}
1435
	yaml_insert_token(parser, -1, &token)
1436
	return true
1437
}
1438

1439
// Produce the ALIAS or ANCHOR token.
1440
func yaml_parser_fetch_anchor(parser *yaml_parser_t, typ yaml_token_type_t) bool {
1441
	// An anchor or an alias could be a simple key.
1442
	if !yaml_parser_save_simple_key(parser) {
1443
		return false
1444
	}
1445

1446
	// A simple key cannot follow an anchor or an alias.
1447
	parser.simple_key_allowed = false
1448

1449
	// Create the ALIAS or ANCHOR token and append it to the queue.
1450
	var token yaml_token_t
1451
	if !yaml_parser_scan_anchor(parser, &token, typ) {
1452
		return false
1453
	}
1454
	yaml_insert_token(parser, -1, &token)
1455
	return true
1456
}
1457

1458
// Produce the TAG token.
1459
func yaml_parser_fetch_tag(parser *yaml_parser_t) bool {
1460
	// A tag could be a simple key.
1461
	if !yaml_parser_save_simple_key(parser) {
1462
		return false
1463
	}
1464

1465
	// A simple key cannot follow a tag.
1466
	parser.simple_key_allowed = false
1467

1468
	// Create the TAG token and append it to the queue.
1469
	var token yaml_token_t
1470
	if !yaml_parser_scan_tag(parser, &token) {
1471
		return false
1472
	}
1473
	yaml_insert_token(parser, -1, &token)
1474
	return true
1475
}
1476

1477
// Produce the SCALAR(...,literal) or SCALAR(...,folded) tokens.
1478
func yaml_parser_fetch_block_scalar(parser *yaml_parser_t, literal bool) bool {
1479
	// Remove any potential simple keys.
1480
	if !yaml_parser_remove_simple_key(parser) {
1481
		return false
1482
	}
1483

1484
	// A simple key may follow a block scalar.
1485
	parser.simple_key_allowed = true
1486

1487
	// Create the SCALAR token and append it to the queue.
1488
	var token yaml_token_t
1489
	if !yaml_parser_scan_block_scalar(parser, &token, literal) {
1490
		return false
1491
	}
1492
	yaml_insert_token(parser, -1, &token)
1493
	return true
1494
}
1495

1496
// Produce the SCALAR(...,single-quoted) or SCALAR(...,double-quoted) tokens.
1497
func yaml_parser_fetch_flow_scalar(parser *yaml_parser_t, single bool) bool {
1498
	// A plain scalar could be a simple key.
1499
	if !yaml_parser_save_simple_key(parser) {
1500
		return false
1501
	}
1502

1503
	// A simple key cannot follow a flow scalar.
1504
	parser.simple_key_allowed = false
1505

1506
	// Create the SCALAR token and append it to the queue.
1507
	var token yaml_token_t
1508
	if !yaml_parser_scan_flow_scalar(parser, &token, single) {
1509
		return false
1510
	}
1511
	yaml_insert_token(parser, -1, &token)
1512
	return true
1513
}
1514

1515
// Produce the SCALAR(...,plain) token.
1516
func yaml_parser_fetch_plain_scalar(parser *yaml_parser_t) bool {
1517
	// A plain scalar could be a simple key.
1518
	if !yaml_parser_save_simple_key(parser) {
1519
		return false
1520
	}
1521

1522
	// A simple key cannot follow a flow scalar.
1523
	parser.simple_key_allowed = false
1524

1525
	// Create the SCALAR token and append it to the queue.
1526
	var token yaml_token_t
1527
	if !yaml_parser_scan_plain_scalar(parser, &token) {
1528
		return false
1529
	}
1530
	yaml_insert_token(parser, -1, &token)
1531
	return true
1532
}
1533

1534
// Eat whitespaces and comments until the next token is found.
1535
func yaml_parser_scan_to_next_token(parser *yaml_parser_t) bool {
1536

1537
	scan_mark := parser.mark
1538

1539
	// Until the next token is not found.
1540
	for {
1541
		// Allow the BOM mark to start a line.
1542
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1543
			return false
1544
		}
1545
		if parser.mark.column == 0 && is_bom(parser.buffer, parser.buffer_pos) {
1546
			skip(parser)
1547
		}
1548

1549
		// Eat whitespaces.
1550
		// Tabs are allowed:
1551
		//  - in the flow context
1552
		//  - in the block context, but not at the beginning of the line or
1553
		//  after '-', '?', or ':' (complex value).
1554
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1555
			return false
1556
		}
1557

1558
		for parser.buffer[parser.buffer_pos] == ' ' || ((parser.flow_level > 0 || !parser.simple_key_allowed) && parser.buffer[parser.buffer_pos] == '\t') {
1559
			skip(parser)
1560
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1561
				return false
1562
			}
1563
		}
1564

1565
		// Check if we just had a line comment under a sequence entry that
1566
		// looks more like a header to the following content. Similar to this:
1567
		//
1568
		// - # The comment
1569
		//   - Some data
1570
		//
1571
		// If so, transform the line comment to a head comment and reposition.
1572
		if len(parser.comments) > 0 && len(parser.tokens) > 1 {
1573
			tokenA := parser.tokens[len(parser.tokens)-2]
1574
			tokenB := parser.tokens[len(parser.tokens)-1]
1575
			comment := &parser.comments[len(parser.comments)-1]
1576
			if tokenA.typ == yaml_BLOCK_SEQUENCE_START_TOKEN && tokenB.typ == yaml_BLOCK_ENTRY_TOKEN && len(comment.line) > 0 && !is_break(parser.buffer, parser.buffer_pos) {
1577
				// If it was in the prior line, reposition so it becomes a
1578
				// header of the follow up token. Otherwise, keep it in place
1579
				// so it becomes a header of the former.
1580
				comment.head = comment.line
1581
				comment.line = nil
1582
				if comment.start_mark.line == parser.mark.line-1 {
1583
					comment.token_mark = parser.mark
1584
				}
1585
			}
1586
		}
1587

1588
		// Eat a comment until a line break.
1589
		if parser.buffer[parser.buffer_pos] == '#' {
1590
			if !yaml_parser_scan_comments(parser, scan_mark) {
1591
				return false
1592
			}
1593
		}
1594

1595
		// If it is a line break, eat it.
1596
		if is_break(parser.buffer, parser.buffer_pos) {
1597
			if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
1598
				return false
1599
			}
1600
			skip_line(parser)
1601

1602
			// In the block context, a new line may start a simple key.
1603
			if parser.flow_level == 0 {
1604
				parser.simple_key_allowed = true
1605
			}
1606
		} else {
1607
			break // We have found a token.
1608
		}
1609
	}
1610

1611
	return true
1612
}
1613

1614
// Scan a YAML-DIRECTIVE or TAG-DIRECTIVE token.
1615
//
1616
// Scope:
1617
//      %YAML    1.1    # a comment \n
1618
//      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1619
//      %TAG    !yaml!  tag:yaml.org,2002:  \n
1620
//      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1621
//
1622
func yaml_parser_scan_directive(parser *yaml_parser_t, token *yaml_token_t) bool {
1623
	// Eat '%'.
1624
	start_mark := parser.mark
1625
	skip(parser)
1626

1627
	// Scan the directive name.
1628
	var name []byte
1629
	if !yaml_parser_scan_directive_name(parser, start_mark, &name) {
1630
		return false
1631
	}
1632

1633
	// Is it a YAML directive?
1634
	if bytes.Equal(name, []byte("YAML")) {
1635
		// Scan the VERSION directive value.
1636
		var major, minor int8
1637
		if !yaml_parser_scan_version_directive_value(parser, start_mark, &major, &minor) {
1638
			return false
1639
		}
1640
		end_mark := parser.mark
1641

1642
		// Create a VERSION-DIRECTIVE token.
1643
		*token = yaml_token_t{
1644
			typ:        yaml_VERSION_DIRECTIVE_TOKEN,
1645
			start_mark: start_mark,
1646
			end_mark:   end_mark,
1647
			major:      major,
1648
			minor:      minor,
1649
		}
1650

1651
		// Is it a TAG directive?
1652
	} else if bytes.Equal(name, []byte("TAG")) {
1653
		// Scan the TAG directive value.
1654
		var handle, prefix []byte
1655
		if !yaml_parser_scan_tag_directive_value(parser, start_mark, &handle, &prefix) {
1656
			return false
1657
		}
1658
		end_mark := parser.mark
1659

1660
		// Create a TAG-DIRECTIVE token.
1661
		*token = yaml_token_t{
1662
			typ:        yaml_TAG_DIRECTIVE_TOKEN,
1663
			start_mark: start_mark,
1664
			end_mark:   end_mark,
1665
			value:      handle,
1666
			prefix:     prefix,
1667
		}
1668

1669
		// Unknown directive.
1670
	} else {
1671
		yaml_parser_set_scanner_error(parser, "while scanning a directive",
1672
			start_mark, "found unknown directive name")
1673
		return false
1674
	}
1675

1676
	// Eat the rest of the line including any comments.
1677
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1678
		return false
1679
	}
1680

1681
	for is_blank(parser.buffer, parser.buffer_pos) {
1682
		skip(parser)
1683
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1684
			return false
1685
		}
1686
	}
1687

1688
	if parser.buffer[parser.buffer_pos] == '#' {
1689
		// [Go] Discard this inline comment for the time being.
1690
		//if !yaml_parser_scan_line_comment(parser, start_mark) {
1691
		//	return false
1692
		//}
1693
		for !is_breakz(parser.buffer, parser.buffer_pos) {
1694
			skip(parser)
1695
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1696
				return false
1697
			}
1698
		}
1699
	}
1700

1701
	// Check if we are at the end of the line.
1702
	if !is_breakz(parser.buffer, parser.buffer_pos) {
1703
		yaml_parser_set_scanner_error(parser, "while scanning a directive",
1704
			start_mark, "did not find expected comment or line break")
1705
		return false
1706
	}
1707

1708
	// Eat a line break.
1709
	if is_break(parser.buffer, parser.buffer_pos) {
1710
		if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
1711
			return false
1712
		}
1713
		skip_line(parser)
1714
	}
1715

1716
	return true
1717
}
1718

1719
// Scan the directive name.
1720
//
1721
// Scope:
1722
//      %YAML   1.1     # a comment \n
1723
//       ^^^^
1724
//      %TAG    !yaml!  tag:yaml.org,2002:  \n
1725
//       ^^^
1726
//
1727
func yaml_parser_scan_directive_name(parser *yaml_parser_t, start_mark yaml_mark_t, name *[]byte) bool {
1728
	// Consume the directive name.
1729
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1730
		return false
1731
	}
1732

1733
	var s []byte
1734
	for is_alpha(parser.buffer, parser.buffer_pos) {
1735
		s = read(parser, s)
1736
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1737
			return false
1738
		}
1739
	}
1740

1741
	// Check if the name is empty.
1742
	if len(s) == 0 {
1743
		yaml_parser_set_scanner_error(parser, "while scanning a directive",
1744
			start_mark, "could not find expected directive name")
1745
		return false
1746
	}
1747

1748
	// Check for an blank character after the name.
1749
	if !is_blankz(parser.buffer, parser.buffer_pos) {
1750
		yaml_parser_set_scanner_error(parser, "while scanning a directive",
1751
			start_mark, "found unexpected non-alphabetical character")
1752
		return false
1753
	}
1754
	*name = s
1755
	return true
1756
}
1757

1758
// Scan the value of VERSION-DIRECTIVE.
1759
//
1760
// Scope:
1761
//      %YAML   1.1     # a comment \n
1762
//           ^^^^^^
1763
func yaml_parser_scan_version_directive_value(parser *yaml_parser_t, start_mark yaml_mark_t, major, minor *int8) bool {
1764
	// Eat whitespaces.
1765
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1766
		return false
1767
	}
1768
	for is_blank(parser.buffer, parser.buffer_pos) {
1769
		skip(parser)
1770
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1771
			return false
1772
		}
1773
	}
1774

1775
	// Consume the major version number.
1776
	if !yaml_parser_scan_version_directive_number(parser, start_mark, major) {
1777
		return false
1778
	}
1779

1780
	// Eat '.'.
1781
	if parser.buffer[parser.buffer_pos] != '.' {
1782
		return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive",
1783
			start_mark, "did not find expected digit or '.' character")
1784
	}
1785

1786
	skip(parser)
1787

1788
	// Consume the minor version number.
1789
	if !yaml_parser_scan_version_directive_number(parser, start_mark, minor) {
1790
		return false
1791
	}
1792
	return true
1793
}
1794

1795
const max_number_length = 2
1796

1797
// Scan the version number of VERSION-DIRECTIVE.
1798
//
1799
// Scope:
1800
//      %YAML   1.1     # a comment \n
1801
//              ^
1802
//      %YAML   1.1     # a comment \n
1803
//                ^
1804
func yaml_parser_scan_version_directive_number(parser *yaml_parser_t, start_mark yaml_mark_t, number *int8) bool {
1805

1806
	// Repeat while the next character is digit.
1807
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1808
		return false
1809
	}
1810
	var value, length int8
1811
	for is_digit(parser.buffer, parser.buffer_pos) {
1812
		// Check if the number is too long.
1813
		length++
1814
		if length > max_number_length {
1815
			return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive",
1816
				start_mark, "found extremely long version number")
1817
		}
1818
		value = value*10 + int8(as_digit(parser.buffer, parser.buffer_pos))
1819
		skip(parser)
1820
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1821
			return false
1822
		}
1823
	}
1824

1825
	// Check if the number was present.
1826
	if length == 0 {
1827
		return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive",
1828
			start_mark, "did not find expected version number")
1829
	}
1830
	*number = value
1831
	return true
1832
}
1833

1834
// Scan the value of a TAG-DIRECTIVE token.
1835
//
1836
// Scope:
1837
//      %TAG    !yaml!  tag:yaml.org,2002:  \n
1838
//          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1839
//
1840
func yaml_parser_scan_tag_directive_value(parser *yaml_parser_t, start_mark yaml_mark_t, handle, prefix *[]byte) bool {
1841
	var handle_value, prefix_value []byte
1842

1843
	// Eat whitespaces.
1844
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1845
		return false
1846
	}
1847

1848
	for is_blank(parser.buffer, parser.buffer_pos) {
1849
		skip(parser)
1850
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1851
			return false
1852
		}
1853
	}
1854

1855
	// Scan a handle.
1856
	if !yaml_parser_scan_tag_handle(parser, true, start_mark, &handle_value) {
1857
		return false
1858
	}
1859

1860
	// Expect a whitespace.
1861
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1862
		return false
1863
	}
1864
	if !is_blank(parser.buffer, parser.buffer_pos) {
1865
		yaml_parser_set_scanner_error(parser, "while scanning a %TAG directive",
1866
			start_mark, "did not find expected whitespace")
1867
		return false
1868
	}
1869

1870
	// Eat whitespaces.
1871
	for is_blank(parser.buffer, parser.buffer_pos) {
1872
		skip(parser)
1873
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1874
			return false
1875
		}
1876
	}
1877

1878
	// Scan a prefix.
1879
	if !yaml_parser_scan_tag_uri(parser, true, nil, start_mark, &prefix_value) {
1880
		return false
1881
	}
1882

1883
	// Expect a whitespace or line break.
1884
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1885
		return false
1886
	}
1887
	if !is_blankz(parser.buffer, parser.buffer_pos) {
1888
		yaml_parser_set_scanner_error(parser, "while scanning a %TAG directive",
1889
			start_mark, "did not find expected whitespace or line break")
1890
		return false
1891
	}
1892

1893
	*handle = handle_value
1894
	*prefix = prefix_value
1895
	return true
1896
}
1897

1898
func yaml_parser_scan_anchor(parser *yaml_parser_t, token *yaml_token_t, typ yaml_token_type_t) bool {
1899
	var s []byte
1900

1901
	// Eat the indicator character.
1902
	start_mark := parser.mark
1903
	skip(parser)
1904

1905
	// Consume the value.
1906
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1907
		return false
1908
	}
1909

1910
	for is_alpha(parser.buffer, parser.buffer_pos) {
1911
		s = read(parser, s)
1912
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
1913
			return false
1914
		}
1915
	}
1916

1917
	end_mark := parser.mark
1918

1919
	/*
1920
	 * Check if length of the anchor is greater than 0 and it is followed by
1921
	 * a whitespace character or one of the indicators:
1922
	 *
1923
	 *      '?', ':', ',', ']', '}', '%', '@', '`'.
1924
	 */
1925

1926
	if len(s) == 0 ||
1927
		!(is_blankz(parser.buffer, parser.buffer_pos) || parser.buffer[parser.buffer_pos] == '?' ||
1928
			parser.buffer[parser.buffer_pos] == ':' || parser.buffer[parser.buffer_pos] == ',' ||
1929
			parser.buffer[parser.buffer_pos] == ']' || parser.buffer[parser.buffer_pos] == '}' ||
1930
			parser.buffer[parser.buffer_pos] == '%' || parser.buffer[parser.buffer_pos] == '@' ||
1931
			parser.buffer[parser.buffer_pos] == '`') {
1932
		context := "while scanning an alias"
1933
		if typ == yaml_ANCHOR_TOKEN {
1934
			context = "while scanning an anchor"
1935
		}
1936
		yaml_parser_set_scanner_error(parser, context, start_mark,
1937
			"did not find expected alphabetic or numeric character")
1938
		return false
1939
	}
1940

1941
	// Create a token.
1942
	*token = yaml_token_t{
1943
		typ:        typ,
1944
		start_mark: start_mark,
1945
		end_mark:   end_mark,
1946
		value:      s,
1947
	}
1948

1949
	return true
1950
}
1951

1952
/*
1953
 * Scan a TAG token.
1954
 */
1955

1956
func yaml_parser_scan_tag(parser *yaml_parser_t, token *yaml_token_t) bool {
1957
	var handle, suffix []byte
1958

1959
	start_mark := parser.mark
1960

1961
	// Check if the tag is in the canonical form.
1962
	if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
1963
		return false
1964
	}
1965

1966
	if parser.buffer[parser.buffer_pos+1] == '<' {
1967
		// Keep the handle as ''
1968

1969
		// Eat '!<'
1970
		skip(parser)
1971
		skip(parser)
1972

1973
		// Consume the tag value.
1974
		if !yaml_parser_scan_tag_uri(parser, false, nil, start_mark, &suffix) {
1975
			return false
1976
		}
1977

1978
		// Check for '>' and eat it.
1979
		if parser.buffer[parser.buffer_pos] != '>' {
1980
			yaml_parser_set_scanner_error(parser, "while scanning a tag",
1981
				start_mark, "did not find the expected '>'")
1982
			return false
1983
		}
1984

1985
		skip(parser)
1986
	} else {
1987
		// The tag has either the '!suffix' or the '!handle!suffix' form.
1988

1989
		// First, try to scan a handle.
1990
		if !yaml_parser_scan_tag_handle(parser, false, start_mark, &handle) {
1991
			return false
1992
		}
1993

1994
		// Check if it is, indeed, handle.
1995
		if handle[0] == '!' && len(handle) > 1 && handle[len(handle)-1] == '!' {
1996
			// Scan the suffix now.
1997
			if !yaml_parser_scan_tag_uri(parser, false, nil, start_mark, &suffix) {
1998
				return false
1999
			}
2000
		} else {
2001
			// It wasn't a handle after all.  Scan the rest of the tag.
2002
			if !yaml_parser_scan_tag_uri(parser, false, handle, start_mark, &suffix) {
2003
				return false
2004
			}
2005

2006
			// Set the handle to '!'.
2007
			handle = []byte{'!'}
2008

2009
			// A special case: the '!' tag.  Set the handle to '' and the
2010
			// suffix to '!'.
2011
			if len(suffix) == 0 {
2012
				handle, suffix = suffix, handle
2013
			}
2014
		}
2015
	}
2016

2017
	// Check the character which ends the tag.
2018
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2019
		return false
2020
	}
2021
	if !is_blankz(parser.buffer, parser.buffer_pos) {
2022
		yaml_parser_set_scanner_error(parser, "while scanning a tag",
2023
			start_mark, "did not find expected whitespace or line break")
2024
		return false
2025
	}
2026

2027
	end_mark := parser.mark
2028

2029
	// Create a token.
2030
	*token = yaml_token_t{
2031
		typ:        yaml_TAG_TOKEN,
2032
		start_mark: start_mark,
2033
		end_mark:   end_mark,
2034
		value:      handle,
2035
		suffix:     suffix,
2036
	}
2037
	return true
2038
}
2039

2040
// Scan a tag handle.
2041
func yaml_parser_scan_tag_handle(parser *yaml_parser_t, directive bool, start_mark yaml_mark_t, handle *[]byte) bool {
2042
	// Check the initial '!' character.
2043
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2044
		return false
2045
	}
2046
	if parser.buffer[parser.buffer_pos] != '!' {
2047
		yaml_parser_set_scanner_tag_error(parser, directive,
2048
			start_mark, "did not find expected '!'")
2049
		return false
2050
	}
2051

2052
	var s []byte
2053

2054
	// Copy the '!' character.
2055
	s = read(parser, s)
2056

2057
	// Copy all subsequent alphabetical and numerical characters.
2058
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2059
		return false
2060
	}
2061
	for is_alpha(parser.buffer, parser.buffer_pos) {
2062
		s = read(parser, s)
2063
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2064
			return false
2065
		}
2066
	}
2067

2068
	// Check if the trailing character is '!' and copy it.
2069
	if parser.buffer[parser.buffer_pos] == '!' {
2070
		s = read(parser, s)
2071
	} else {
2072
		// It's either the '!' tag or not really a tag handle.  If it's a %TAG
2073
		// directive, it's an error.  If it's a tag token, it must be a part of URI.
2074
		if directive && string(s) != "!" {
2075
			yaml_parser_set_scanner_tag_error(parser, directive,
2076
				start_mark, "did not find expected '!'")
2077
			return false
2078
		}
2079
	}
2080

2081
	*handle = s
2082
	return true
2083
}
2084

2085
// Scan a tag.
2086
func yaml_parser_scan_tag_uri(parser *yaml_parser_t, directive bool, head []byte, start_mark yaml_mark_t, uri *[]byte) bool {
2087
	//size_t length = head ? strlen((char *)head) : 0
2088
	var s []byte
2089
	hasTag := len(head) > 0
2090

2091
	// Copy the head if needed.
2092
	//
2093
	// Note that we don't copy the leading '!' character.
2094
	if len(head) > 1 {
2095
		s = append(s, head[1:]...)
2096
	}
2097

2098
	// Scan the tag.
2099
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2100
		return false
2101
	}
2102

2103
	// The set of characters that may appear in URI is as follows:
2104
	//
2105
	//      '0'-'9', 'A'-'Z', 'a'-'z', '_', '-', ';', '/', '?', ':', '@', '&',
2106
	//      '=', '+', '$', ',', '.', '!', '~', '*', '\'', '(', ')', '[', ']',
2107
	//      '%'.
2108
	// [Go] TODO Convert this into more reasonable logic.
2109
	for is_alpha(parser.buffer, parser.buffer_pos) || parser.buffer[parser.buffer_pos] == ';' ||
2110
		parser.buffer[parser.buffer_pos] == '/' || parser.buffer[parser.buffer_pos] == '?' ||
2111
		parser.buffer[parser.buffer_pos] == ':' || parser.buffer[parser.buffer_pos] == '@' ||
2112
		parser.buffer[parser.buffer_pos] == '&' || parser.buffer[parser.buffer_pos] == '=' ||
2113
		parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '$' ||
2114
		parser.buffer[parser.buffer_pos] == ',' || parser.buffer[parser.buffer_pos] == '.' ||
2115
		parser.buffer[parser.buffer_pos] == '!' || parser.buffer[parser.buffer_pos] == '~' ||
2116
		parser.buffer[parser.buffer_pos] == '*' || parser.buffer[parser.buffer_pos] == '\'' ||
2117
		parser.buffer[parser.buffer_pos] == '(' || parser.buffer[parser.buffer_pos] == ')' ||
2118
		parser.buffer[parser.buffer_pos] == '[' || parser.buffer[parser.buffer_pos] == ']' ||
2119
		parser.buffer[parser.buffer_pos] == '%' {
2120
		// Check if it is a URI-escape sequence.
2121
		if parser.buffer[parser.buffer_pos] == '%' {
2122
			if !yaml_parser_scan_uri_escapes(parser, directive, start_mark, &s) {
2123
				return false
2124
			}
2125
		} else {
2126
			s = read(parser, s)
2127
		}
2128
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2129
			return false
2130
		}
2131
		hasTag = true
2132
	}
2133

2134
	if !hasTag {
2135
		yaml_parser_set_scanner_tag_error(parser, directive,
2136
			start_mark, "did not find expected tag URI")
2137
		return false
2138
	}
2139
	*uri = s
2140
	return true
2141
}
2142

2143
// Decode an URI-escape sequence corresponding to a single UTF-8 character.
2144
func yaml_parser_scan_uri_escapes(parser *yaml_parser_t, directive bool, start_mark yaml_mark_t, s *[]byte) bool {
2145

2146
	// Decode the required number of characters.
2147
	w := 1024
2148
	for w > 0 {
2149
		// Check for a URI-escaped octet.
2150
		if parser.unread < 3 && !yaml_parser_update_buffer(parser, 3) {
2151
			return false
2152
		}
2153

2154
		if !(parser.buffer[parser.buffer_pos] == '%' &&
2155
			is_hex(parser.buffer, parser.buffer_pos+1) &&
2156
			is_hex(parser.buffer, parser.buffer_pos+2)) {
2157
			return yaml_parser_set_scanner_tag_error(parser, directive,
2158
				start_mark, "did not find URI escaped octet")
2159
		}
2160

2161
		// Get the octet.
2162
		octet := byte((as_hex(parser.buffer, parser.buffer_pos+1) << 4) + as_hex(parser.buffer, parser.buffer_pos+2))
2163

2164
		// If it is the leading octet, determine the length of the UTF-8 sequence.
2165
		if w == 1024 {
2166
			w = width(octet)
2167
			if w == 0 {
2168
				return yaml_parser_set_scanner_tag_error(parser, directive,
2169
					start_mark, "found an incorrect leading UTF-8 octet")
2170
			}
2171
		} else {
2172
			// Check if the trailing octet is correct.
2173
			if octet&0xC0 != 0x80 {
2174
				return yaml_parser_set_scanner_tag_error(parser, directive,
2175
					start_mark, "found an incorrect trailing UTF-8 octet")
2176
			}
2177
		}
2178

2179
		// Copy the octet and move the pointers.
2180
		*s = append(*s, octet)
2181
		skip(parser)
2182
		skip(parser)
2183
		skip(parser)
2184
		w--
2185
	}
2186
	return true
2187
}
2188

2189
// Scan a block scalar.
2190
func yaml_parser_scan_block_scalar(parser *yaml_parser_t, token *yaml_token_t, literal bool) bool {
2191
	// Eat the indicator '|' or '>'.
2192
	start_mark := parser.mark
2193
	skip(parser)
2194

2195
	// Scan the additional block scalar indicators.
2196
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2197
		return false
2198
	}
2199

2200
	// Check for a chomping indicator.
2201
	var chomping, increment int
2202
	if parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '-' {
2203
		// Set the chomping method and eat the indicator.
2204
		if parser.buffer[parser.buffer_pos] == '+' {
2205
			chomping = +1
2206
		} else {
2207
			chomping = -1
2208
		}
2209
		skip(parser)
2210

2211
		// Check for an indentation indicator.
2212
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2213
			return false
2214
		}
2215
		if is_digit(parser.buffer, parser.buffer_pos) {
2216
			// Check that the indentation is greater than 0.
2217
			if parser.buffer[parser.buffer_pos] == '0' {
2218
				yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
2219
					start_mark, "found an indentation indicator equal to 0")
2220
				return false
2221
			}
2222

2223
			// Get the indentation level and eat the indicator.
2224
			increment = as_digit(parser.buffer, parser.buffer_pos)
2225
			skip(parser)
2226
		}
2227

2228
	} else if is_digit(parser.buffer, parser.buffer_pos) {
2229
		// Do the same as above, but in the opposite order.
2230

2231
		if parser.buffer[parser.buffer_pos] == '0' {
2232
			yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
2233
				start_mark, "found an indentation indicator equal to 0")
2234
			return false
2235
		}
2236
		increment = as_digit(parser.buffer, parser.buffer_pos)
2237
		skip(parser)
2238

2239
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2240
			return false
2241
		}
2242
		if parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '-' {
2243
			if parser.buffer[parser.buffer_pos] == '+' {
2244
				chomping = +1
2245
			} else {
2246
				chomping = -1
2247
			}
2248
			skip(parser)
2249
		}
2250
	}
2251

2252
	// Eat whitespaces and comments to the end of the line.
2253
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2254
		return false
2255
	}
2256
	for is_blank(parser.buffer, parser.buffer_pos) {
2257
		skip(parser)
2258
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2259
			return false
2260
		}
2261
	}
2262
	if parser.buffer[parser.buffer_pos] == '#' {
2263
		if !yaml_parser_scan_line_comment(parser, start_mark) {
2264
			return false
2265
		}
2266
		for !is_breakz(parser.buffer, parser.buffer_pos) {
2267
			skip(parser)
2268
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2269
				return false
2270
			}
2271
		}
2272
	}
2273

2274
	// Check if we are at the end of the line.
2275
	if !is_breakz(parser.buffer, parser.buffer_pos) {
2276
		yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
2277
			start_mark, "did not find expected comment or line break")
2278
		return false
2279
	}
2280

2281
	// Eat a line break.
2282
	if is_break(parser.buffer, parser.buffer_pos) {
2283
		if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
2284
			return false
2285
		}
2286
		skip_line(parser)
2287
	}
2288

2289
	end_mark := parser.mark
2290

2291
	// Set the indentation level if it was specified.
2292
	var indent int
2293
	if increment > 0 {
2294
		if parser.indent >= 0 {
2295
			indent = parser.indent + increment
2296
		} else {
2297
			indent = increment
2298
		}
2299
	}
2300

2301
	// Scan the leading line breaks and determine the indentation level if needed.
2302
	var s, leading_break, trailing_breaks []byte
2303
	if !yaml_parser_scan_block_scalar_breaks(parser, &indent, &trailing_breaks, start_mark, &end_mark) {
2304
		return false
2305
	}
2306

2307
	// Scan the block scalar content.
2308
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2309
		return false
2310
	}
2311
	var leading_blank, trailing_blank bool
2312
	for parser.mark.column == indent && !is_z(parser.buffer, parser.buffer_pos) {
2313
		// We are at the beginning of a non-empty line.
2314

2315
		// Is it a trailing whitespace?
2316
		trailing_blank = is_blank(parser.buffer, parser.buffer_pos)
2317

2318
		// Check if we need to fold the leading line break.
2319
		if !literal && !leading_blank && !trailing_blank && len(leading_break) > 0 && leading_break[0] == '\n' {
2320
			// Do we need to join the lines by space?
2321
			if len(trailing_breaks) == 0 {
2322
				s = append(s, ' ')
2323
			}
2324
		} else {
2325
			s = append(s, leading_break...)
2326
		}
2327
		leading_break = leading_break[:0]
2328

2329
		// Append the remaining line breaks.
2330
		s = append(s, trailing_breaks...)
2331
		trailing_breaks = trailing_breaks[:0]
2332

2333
		// Is it a leading whitespace?
2334
		leading_blank = is_blank(parser.buffer, parser.buffer_pos)
2335

2336
		// Consume the current line.
2337
		for !is_breakz(parser.buffer, parser.buffer_pos) {
2338
			s = read(parser, s)
2339
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2340
				return false
2341
			}
2342
		}
2343

2344
		// Consume the line break.
2345
		if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
2346
			return false
2347
		}
2348

2349
		leading_break = read_line(parser, leading_break)
2350

2351
		// Eat the following indentation spaces and line breaks.
2352
		if !yaml_parser_scan_block_scalar_breaks(parser, &indent, &trailing_breaks, start_mark, &end_mark) {
2353
			return false
2354
		}
2355
	}
2356

2357
	// Chomp the tail.
2358
	if chomping != -1 {
2359
		s = append(s, leading_break...)
2360
	}
2361
	if chomping == 1 {
2362
		s = append(s, trailing_breaks...)
2363
	}
2364

2365
	// Create a token.
2366
	*token = yaml_token_t{
2367
		typ:        yaml_SCALAR_TOKEN,
2368
		start_mark: start_mark,
2369
		end_mark:   end_mark,
2370
		value:      s,
2371
		style:      yaml_LITERAL_SCALAR_STYLE,
2372
	}
2373
	if !literal {
2374
		token.style = yaml_FOLDED_SCALAR_STYLE
2375
	}
2376
	return true
2377
}
2378

2379
// Scan indentation spaces and line breaks for a block scalar.  Determine the
2380
// indentation level if needed.
2381
func yaml_parser_scan_block_scalar_breaks(parser *yaml_parser_t, indent *int, breaks *[]byte, start_mark yaml_mark_t, end_mark *yaml_mark_t) bool {
2382
	*end_mark = parser.mark
2383

2384
	// Eat the indentation spaces and line breaks.
2385
	max_indent := 0
2386
	for {
2387
		// Eat the indentation spaces.
2388
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2389
			return false
2390
		}
2391
		for (*indent == 0 || parser.mark.column < *indent) && is_space(parser.buffer, parser.buffer_pos) {
2392
			skip(parser)
2393
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2394
				return false
2395
			}
2396
		}
2397
		if parser.mark.column > max_indent {
2398
			max_indent = parser.mark.column
2399
		}
2400

2401
		// Check for a tab character messing the indentation.
2402
		if (*indent == 0 || parser.mark.column < *indent) && is_tab(parser.buffer, parser.buffer_pos) {
2403
			return yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
2404
				start_mark, "found a tab character where an indentation space is expected")
2405
		}
2406

2407
		// Have we found a non-empty line?
2408
		if !is_break(parser.buffer, parser.buffer_pos) {
2409
			break
2410
		}
2411

2412
		// Consume the line break.
2413
		if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
2414
			return false
2415
		}
2416
		// [Go] Should really be returning breaks instead.
2417
		*breaks = read_line(parser, *breaks)
2418
		*end_mark = parser.mark
2419
	}
2420

2421
	// Determine the indentation level if needed.
2422
	if *indent == 0 {
2423
		*indent = max_indent
2424
		if *indent < parser.indent+1 {
2425
			*indent = parser.indent + 1
2426
		}
2427
		if *indent < 1 {
2428
			*indent = 1
2429
		}
2430
	}
2431
	return true
2432
}
2433

2434
// Scan a quoted scalar.
2435
func yaml_parser_scan_flow_scalar(parser *yaml_parser_t, token *yaml_token_t, single bool) bool {
2436
	// Eat the left quote.
2437
	start_mark := parser.mark
2438
	skip(parser)
2439

2440
	// Consume the content of the quoted scalar.
2441
	var s, leading_break, trailing_breaks, whitespaces []byte
2442
	for {
2443
		// Check that there are no document indicators at the beginning of the line.
2444
		if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) {
2445
			return false
2446
		}
2447

2448
		if parser.mark.column == 0 &&
2449
			((parser.buffer[parser.buffer_pos+0] == '-' &&
2450
				parser.buffer[parser.buffer_pos+1] == '-' &&
2451
				parser.buffer[parser.buffer_pos+2] == '-') ||
2452
				(parser.buffer[parser.buffer_pos+0] == '.' &&
2453
					parser.buffer[parser.buffer_pos+1] == '.' &&
2454
					parser.buffer[parser.buffer_pos+2] == '.')) &&
2455
			is_blankz(parser.buffer, parser.buffer_pos+3) {
2456
			yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar",
2457
				start_mark, "found unexpected document indicator")
2458
			return false
2459
		}
2460

2461
		// Check for EOF.
2462
		if is_z(parser.buffer, parser.buffer_pos) {
2463
			yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar",
2464
				start_mark, "found unexpected end of stream")
2465
			return false
2466
		}
2467

2468
		// Consume non-blank characters.
2469
		leading_blanks := false
2470
		for !is_blankz(parser.buffer, parser.buffer_pos) {
2471
			if single && parser.buffer[parser.buffer_pos] == '\'' && parser.buffer[parser.buffer_pos+1] == '\'' {
2472
				// Is is an escaped single quote.
2473
				s = append(s, '\'')
2474
				skip(parser)
2475
				skip(parser)
2476

2477
			} else if single && parser.buffer[parser.buffer_pos] == '\'' {
2478
				// It is a right single quote.
2479
				break
2480
			} else if !single && parser.buffer[parser.buffer_pos] == '"' {
2481
				// It is a right double quote.
2482
				break
2483

2484
			} else if !single && parser.buffer[parser.buffer_pos] == '\\' && is_break(parser.buffer, parser.buffer_pos+1) {
2485
				// It is an escaped line break.
2486
				if parser.unread < 3 && !yaml_parser_update_buffer(parser, 3) {
2487
					return false
2488
				}
2489
				skip(parser)
2490
				skip_line(parser)
2491
				leading_blanks = true
2492
				break
2493

2494
			} else if !single && parser.buffer[parser.buffer_pos] == '\\' {
2495
				// It is an escape sequence.
2496
				code_length := 0
2497

2498
				// Check the escape character.
2499
				switch parser.buffer[parser.buffer_pos+1] {
2500
				case '0':
2501
					s = append(s, 0)
2502
				case 'a':
2503
					s = append(s, '\x07')
2504
				case 'b':
2505
					s = append(s, '\x08')
2506
				case 't', '\t':
2507
					s = append(s, '\x09')
2508
				case 'n':
2509
					s = append(s, '\x0A')
2510
				case 'v':
2511
					s = append(s, '\x0B')
2512
				case 'f':
2513
					s = append(s, '\x0C')
2514
				case 'r':
2515
					s = append(s, '\x0D')
2516
				case 'e':
2517
					s = append(s, '\x1B')
2518
				case ' ':
2519
					s = append(s, '\x20')
2520
				case '"':
2521
					s = append(s, '"')
2522
				case '\'':
2523
					s = append(s, '\'')
2524
				case '\\':
2525
					s = append(s, '\\')
2526
				case 'N': // NEL (#x85)
2527
					s = append(s, '\xC2')
2528
					s = append(s, '\x85')
2529
				case '_': // #xA0
2530
					s = append(s, '\xC2')
2531
					s = append(s, '\xA0')
2532
				case 'L': // LS (#x2028)
2533
					s = append(s, '\xE2')
2534
					s = append(s, '\x80')
2535
					s = append(s, '\xA8')
2536
				case 'P': // PS (#x2029)
2537
					s = append(s, '\xE2')
2538
					s = append(s, '\x80')
2539
					s = append(s, '\xA9')
2540
				case 'x':
2541
					code_length = 2
2542
				case 'u':
2543
					code_length = 4
2544
				case 'U':
2545
					code_length = 8
2546
				default:
2547
					yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar",
2548
						start_mark, "found unknown escape character")
2549
					return false
2550
				}
2551

2552
				skip(parser)
2553
				skip(parser)
2554

2555
				// Consume an arbitrary escape code.
2556
				if code_length > 0 {
2557
					var value int
2558

2559
					// Scan the character value.
2560
					if parser.unread < code_length && !yaml_parser_update_buffer(parser, code_length) {
2561
						return false
2562
					}
2563
					for k := 0; k < code_length; k++ {
2564
						if !is_hex(parser.buffer, parser.buffer_pos+k) {
2565
							yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar",
2566
								start_mark, "did not find expected hexdecimal number")
2567
							return false
2568
						}
2569
						value = (value << 4) + as_hex(parser.buffer, parser.buffer_pos+k)
2570
					}
2571

2572
					// Check the value and write the character.
2573
					if (value >= 0xD800 && value <= 0xDFFF) || value > 0x10FFFF {
2574
						yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar",
2575
							start_mark, "found invalid Unicode character escape code")
2576
						return false
2577
					}
2578
					if value <= 0x7F {
2579
						s = append(s, byte(value))
2580
					} else if value <= 0x7FF {
2581
						s = append(s, byte(0xC0+(value>>6)))
2582
						s = append(s, byte(0x80+(value&0x3F)))
2583
					} else if value <= 0xFFFF {
2584
						s = append(s, byte(0xE0+(value>>12)))
2585
						s = append(s, byte(0x80+((value>>6)&0x3F)))
2586
						s = append(s, byte(0x80+(value&0x3F)))
2587
					} else {
2588
						s = append(s, byte(0xF0+(value>>18)))
2589
						s = append(s, byte(0x80+((value>>12)&0x3F)))
2590
						s = append(s, byte(0x80+((value>>6)&0x3F)))
2591
						s = append(s, byte(0x80+(value&0x3F)))
2592
					}
2593

2594
					// Advance the pointer.
2595
					for k := 0; k < code_length; k++ {
2596
						skip(parser)
2597
					}
2598
				}
2599
			} else {
2600
				// It is a non-escaped non-blank character.
2601
				s = read(parser, s)
2602
			}
2603
			if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
2604
				return false
2605
			}
2606
		}
2607

2608
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2609
			return false
2610
		}
2611

2612
		// Check if we are at the end of the scalar.
2613
		if single {
2614
			if parser.buffer[parser.buffer_pos] == '\'' {
2615
				break
2616
			}
2617
		} else {
2618
			if parser.buffer[parser.buffer_pos] == '"' {
2619
				break
2620
			}
2621
		}
2622

2623
		// Consume blank characters.
2624
		for is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos) {
2625
			if is_blank(parser.buffer, parser.buffer_pos) {
2626
				// Consume a space or a tab character.
2627
				if !leading_blanks {
2628
					whitespaces = read(parser, whitespaces)
2629
				} else {
2630
					skip(parser)
2631
				}
2632
			} else {
2633
				if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
2634
					return false
2635
				}
2636

2637
				// Check if it is a first line break.
2638
				if !leading_blanks {
2639
					whitespaces = whitespaces[:0]
2640
					leading_break = read_line(parser, leading_break)
2641
					leading_blanks = true
2642
				} else {
2643
					trailing_breaks = read_line(parser, trailing_breaks)
2644
				}
2645
			}
2646
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2647
				return false
2648
			}
2649
		}
2650

2651
		// Join the whitespaces or fold line breaks.
2652
		if leading_blanks {
2653
			// Do we need to fold line breaks?
2654
			if len(leading_break) > 0 && leading_break[0] == '\n' {
2655
				if len(trailing_breaks) == 0 {
2656
					s = append(s, ' ')
2657
				} else {
2658
					s = append(s, trailing_breaks...)
2659
				}
2660
			} else {
2661
				s = append(s, leading_break...)
2662
				s = append(s, trailing_breaks...)
2663
			}
2664
			trailing_breaks = trailing_breaks[:0]
2665
			leading_break = leading_break[:0]
2666
		} else {
2667
			s = append(s, whitespaces...)
2668
			whitespaces = whitespaces[:0]
2669
		}
2670
	}
2671

2672
	// Eat the right quote.
2673
	skip(parser)
2674
	end_mark := parser.mark
2675

2676
	// Create a token.
2677
	*token = yaml_token_t{
2678
		typ:        yaml_SCALAR_TOKEN,
2679
		start_mark: start_mark,
2680
		end_mark:   end_mark,
2681
		value:      s,
2682
		style:      yaml_SINGLE_QUOTED_SCALAR_STYLE,
2683
	}
2684
	if !single {
2685
		token.style = yaml_DOUBLE_QUOTED_SCALAR_STYLE
2686
	}
2687
	return true
2688
}
2689

2690
// Scan a plain scalar.
2691
func yaml_parser_scan_plain_scalar(parser *yaml_parser_t, token *yaml_token_t) bool {
2692

2693
	var s, leading_break, trailing_breaks, whitespaces []byte
2694
	var leading_blanks bool
2695
	var indent = parser.indent + 1
2696

2697
	start_mark := parser.mark
2698
	end_mark := parser.mark
2699

2700
	// Consume the content of the plain scalar.
2701
	for {
2702
		// Check for a document indicator.
2703
		if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) {
2704
			return false
2705
		}
2706
		if parser.mark.column == 0 &&
2707
			((parser.buffer[parser.buffer_pos+0] == '-' &&
2708
				parser.buffer[parser.buffer_pos+1] == '-' &&
2709
				parser.buffer[parser.buffer_pos+2] == '-') ||
2710
				(parser.buffer[parser.buffer_pos+0] == '.' &&
2711
					parser.buffer[parser.buffer_pos+1] == '.' &&
2712
					parser.buffer[parser.buffer_pos+2] == '.')) &&
2713
			is_blankz(parser.buffer, parser.buffer_pos+3) {
2714
			break
2715
		}
2716

2717
		// Check for a comment.
2718
		if parser.buffer[parser.buffer_pos] == '#' {
2719
			break
2720
		}
2721

2722
		// Consume non-blank characters.
2723
		for !is_blankz(parser.buffer, parser.buffer_pos) {
2724

2725
			// Check for indicators that may end a plain scalar.
2726
			if (parser.buffer[parser.buffer_pos] == ':' && is_blankz(parser.buffer, parser.buffer_pos+1)) ||
2727
				(parser.flow_level > 0 &&
2728
					(parser.buffer[parser.buffer_pos] == ',' ||
2729
						parser.buffer[parser.buffer_pos] == '?' || parser.buffer[parser.buffer_pos] == '[' ||
2730
						parser.buffer[parser.buffer_pos] == ']' || parser.buffer[parser.buffer_pos] == '{' ||
2731
						parser.buffer[parser.buffer_pos] == '}')) {
2732
				break
2733
			}
2734

2735
			// Check if we need to join whitespaces and breaks.
2736
			if leading_blanks || len(whitespaces) > 0 {
2737
				if leading_blanks {
2738
					// Do we need to fold line breaks?
2739
					if leading_break[0] == '\n' {
2740
						if len(trailing_breaks) == 0 {
2741
							s = append(s, ' ')
2742
						} else {
2743
							s = append(s, trailing_breaks...)
2744
						}
2745
					} else {
2746
						s = append(s, leading_break...)
2747
						s = append(s, trailing_breaks...)
2748
					}
2749
					trailing_breaks = trailing_breaks[:0]
2750
					leading_break = leading_break[:0]
2751
					leading_blanks = false
2752
				} else {
2753
					s = append(s, whitespaces...)
2754
					whitespaces = whitespaces[:0]
2755
				}
2756
			}
2757

2758
			// Copy the character.
2759
			s = read(parser, s)
2760

2761
			end_mark = parser.mark
2762
			if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
2763
				return false
2764
			}
2765
		}
2766

2767
		// Is it the end?
2768
		if !(is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos)) {
2769
			break
2770
		}
2771

2772
		// Consume blank characters.
2773
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2774
			return false
2775
		}
2776

2777
		for is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos) {
2778
			if is_blank(parser.buffer, parser.buffer_pos) {
2779

2780
				// Check for tab characters that abuse indentation.
2781
				if leading_blanks && parser.mark.column < indent && is_tab(parser.buffer, parser.buffer_pos) {
2782
					yaml_parser_set_scanner_error(parser, "while scanning a plain scalar",
2783
						start_mark, "found a tab character that violates indentation")
2784
					return false
2785
				}
2786

2787
				// Consume a space or a tab character.
2788
				if !leading_blanks {
2789
					whitespaces = read(parser, whitespaces)
2790
				} else {
2791
					skip(parser)
2792
				}
2793
			} else {
2794
				if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
2795
					return false
2796
				}
2797

2798
				// Check if it is a first line break.
2799
				if !leading_blanks {
2800
					whitespaces = whitespaces[:0]
2801
					leading_break = read_line(parser, leading_break)
2802
					leading_blanks = true
2803
				} else {
2804
					trailing_breaks = read_line(parser, trailing_breaks)
2805
				}
2806
			}
2807
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2808
				return false
2809
			}
2810
		}
2811

2812
		// Check indentation level.
2813
		if parser.flow_level == 0 && parser.mark.column < indent {
2814
			break
2815
		}
2816
	}
2817

2818
	// Create a token.
2819
	*token = yaml_token_t{
2820
		typ:        yaml_SCALAR_TOKEN,
2821
		start_mark: start_mark,
2822
		end_mark:   end_mark,
2823
		value:      s,
2824
		style:      yaml_PLAIN_SCALAR_STYLE,
2825
	}
2826

2827
	// Note that we change the 'simple_key_allowed' flag.
2828
	if leading_blanks {
2829
		parser.simple_key_allowed = true
2830
	}
2831
	return true
2832
}
2833

2834
func yaml_parser_scan_line_comment(parser *yaml_parser_t, token_mark yaml_mark_t) bool {
2835
	if parser.newlines > 0 {
2836
		return true
2837
	}
2838

2839
	var start_mark yaml_mark_t
2840
	var text []byte
2841

2842
	for peek := 0; peek < 512; peek++ {
2843
		if parser.unread < peek+1 && !yaml_parser_update_buffer(parser, peek+1) {
2844
			break
2845
		}
2846
		if is_blank(parser.buffer, parser.buffer_pos+peek) {
2847
			continue
2848
		}
2849
		if parser.buffer[parser.buffer_pos+peek] == '#' {
2850
			seen := parser.mark.index+peek
2851
			for {
2852
				if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
2853
					return false
2854
				}
2855
				if is_breakz(parser.buffer, parser.buffer_pos) {
2856
					if parser.mark.index >= seen {
2857
						break
2858
					}
2859
					if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
2860
						return false
2861
					}
2862
					skip_line(parser)
2863
				} else if parser.mark.index >= seen {
2864
					if len(text) == 0 {
2865
						start_mark = parser.mark
2866
					}
2867
					text = read(parser, text)
2868
				} else {
2869
					skip(parser)
2870
				}
2871
			}
2872
		}
2873
		break
2874
	}
2875
	if len(text) > 0 {
2876
		parser.comments = append(parser.comments, yaml_comment_t{
2877
			token_mark: token_mark,
2878
			start_mark: start_mark,
2879
			line: text,
2880
		})
2881
	}
2882
	return true
2883
}
2884

2885
func yaml_parser_scan_comments(parser *yaml_parser_t, scan_mark yaml_mark_t) bool {
2886
	token := parser.tokens[len(parser.tokens)-1]
2887

2888
	if token.typ == yaml_FLOW_ENTRY_TOKEN && len(parser.tokens) > 1 {
2889
		token = parser.tokens[len(parser.tokens)-2]
2890
	}
2891

2892
	var token_mark = token.start_mark
2893
	var start_mark yaml_mark_t
2894
	var next_indent = parser.indent
2895
	if next_indent < 0 {
2896
		next_indent = 0
2897
	}
2898

2899
	var recent_empty = false
2900
	var first_empty = parser.newlines <= 1
2901

2902
	var line = parser.mark.line
2903
	var column = parser.mark.column
2904

2905
	var text []byte
2906

2907
	// The foot line is the place where a comment must start to
2908
	// still be considered as a foot of the prior content.
2909
	// If there's some content in the currently parsed line, then
2910
	// the foot is the line below it.
2911
	var foot_line = -1
2912
	if scan_mark.line > 0 {
2913
		foot_line = parser.mark.line-parser.newlines+1
2914
		if parser.newlines == 0 && parser.mark.column > 1 {
2915
			foot_line++
2916
		}
2917
	}
2918

2919
	var peek = 0
2920
	for ; peek < 512; peek++ {
2921
		if parser.unread < peek+1 && !yaml_parser_update_buffer(parser, peek+1) {
2922
			break
2923
		}
2924
		column++
2925
		if is_blank(parser.buffer, parser.buffer_pos+peek) {
2926
			continue
2927
		}
2928
		c := parser.buffer[parser.buffer_pos+peek]
2929
		var close_flow = parser.flow_level > 0 && (c == ']' || c == '}')
2930
		if close_flow || is_breakz(parser.buffer, parser.buffer_pos+peek) {
2931
			// Got line break or terminator.
2932
			if close_flow || !recent_empty {
2933
				if close_flow || first_empty && (start_mark.line == foot_line && token.typ != yaml_VALUE_TOKEN || start_mark.column-1 < next_indent) {
2934
					// This is the first empty line and there were no empty lines before,
2935
					// so this initial part of the comment is a foot of the prior token
2936
					// instead of being a head for the following one. Split it up.
2937
					// Alternatively, this might also be the last comment inside a flow
2938
					// scope, so it must be a footer.
2939
					if len(text) > 0 {
2940
						if start_mark.column-1 < next_indent {
2941
							// If dedented it's unrelated to the prior token.
2942
							token_mark = start_mark
2943
						}
2944
						parser.comments = append(parser.comments, yaml_comment_t{
2945
							scan_mark:  scan_mark,
2946
							token_mark: token_mark,
2947
							start_mark: start_mark,
2948
							end_mark:   yaml_mark_t{parser.mark.index + peek, line, column},
2949
							foot:       text,
2950
						})
2951
						scan_mark = yaml_mark_t{parser.mark.index + peek, line, column}
2952
						token_mark = scan_mark
2953
						text = nil
2954
					}
2955
				} else {
2956
					if len(text) > 0 && parser.buffer[parser.buffer_pos+peek] != 0 {
2957
						text = append(text, '\n')
2958
					}
2959
				}
2960
			}
2961
			if !is_break(parser.buffer, parser.buffer_pos+peek) {
2962
				break
2963
			}
2964
			first_empty = false
2965
			recent_empty = true
2966
			column = 0
2967
			line++
2968
			continue
2969
		}
2970

2971
		if len(text) > 0 && (close_flow || column-1 < next_indent && column != start_mark.column) {
2972
			// The comment at the different indentation is a foot of the
2973
			// preceding data rather than a head of the upcoming one.
2974
			parser.comments = append(parser.comments, yaml_comment_t{
2975
				scan_mark:  scan_mark,
2976
				token_mark: token_mark,
2977
				start_mark: start_mark,
2978
				end_mark:   yaml_mark_t{parser.mark.index + peek, line, column},
2979
				foot:       text,
2980
			})
2981
			scan_mark = yaml_mark_t{parser.mark.index + peek, line, column}
2982
			token_mark = scan_mark
2983
			text = nil
2984
		}
2985

2986
		if parser.buffer[parser.buffer_pos+peek] != '#' {
2987
			break
2988
		}
2989

2990
		if len(text) == 0 {
2991
			start_mark = yaml_mark_t{parser.mark.index + peek, line, column}
2992
		} else {
2993
			text = append(text, '\n')
2994
		}
2995

2996
		recent_empty = false
2997

2998
		// Consume until after the consumed comment line.
2999
		seen := parser.mark.index+peek
3000
		for {
3001
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
3002
				return false
3003
			}
3004
			if is_breakz(parser.buffer, parser.buffer_pos) {
3005
				if parser.mark.index >= seen {
3006
					break
3007
				}
3008
				if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
3009
					return false
3010
				}
3011
				skip_line(parser)
3012
			} else if parser.mark.index >= seen {
3013
				text = read(parser, text)
3014
			} else {
3015
				skip(parser)
3016
			}
3017
		}
3018

3019
		peek = 0
3020
		column = 0
3021
		line = parser.mark.line
3022
		next_indent = parser.indent
3023
		if next_indent < 0 {
3024
			next_indent = 0
3025
		}
3026
	}
3027

3028
	if len(text) > 0 {
3029
		parser.comments = append(parser.comments, yaml_comment_t{
3030
			scan_mark:  scan_mark,
3031
			token_mark: start_mark,
3032
			start_mark: start_mark,
3033
			end_mark:   yaml_mark_t{parser.mark.index + peek - 1, line, column},
3034
			head:       text,
3035
		})
3036
	}
3037
	return true
3038
}
3039

3040