blob: ca0070108f4ebe6a09a222075267e0ffca996e72 [file] [log] [blame]
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001//
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
Jonathan Hart4b110f62020-03-13 17:36:19 -070023package yaml
24
25import (
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).
507func 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.
513func skip(parser *yaml_parser_t) {
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000514 if !is_blank(parser.buffer, parser.buffer_pos) {
515 parser.newlines = 0
516 }
Jonathan Hart4b110f62020-03-13 17:36:19 -0700517 parser.mark.index++
518 parser.mark.column++
519 parser.unread--
520 parser.buffer_pos += width(parser.buffer[parser.buffer_pos])
521}
522
523func 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
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000530 parser.newlines++
Jonathan Hart4b110f62020-03-13 17:36:19 -0700531 } 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])
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000537 parser.newlines++
Jonathan Hart4b110f62020-03-13 17:36:19 -0700538 }
539}
540
541// Copy a character to a string buffer and advance pointers.
542func read(parser *yaml_parser_t, s []byte) []byte {
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000543 if !is_blank(parser.buffer, parser.buffer_pos) {
544 parser.newlines = 0
545 }
Jonathan Hart4b110f62020-03-13 17:36:19 -0700546 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.
568func 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--
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000597 parser.newlines++
Jonathan Hart4b110f62020-03-13 17:36:19 -0700598 return s
599}
600
601// Get the next token.
602func 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.
631func 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
640func 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
648func 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.
657func yaml_parser_fetch_more_tokens(parser *yaml_parser_t) bool {
658 // While we need more tokens to fetch, do it.
659 for {
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000660 // [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 {
Jonathan Hart4b110f62020-03-13 17:36:19 -0700671 return false
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000672 } else if !valid {
673 break
Jonathan Hart4b110f62020-03-13 17:36:19 -0700674 }
Jonathan Hart4b110f62020-03-13 17:36:19 -0700675 }
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.
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000687func yaml_parser_fetch_next_token(parser *yaml_parser_t) (ok bool) {
Jonathan Hart4b110f62020-03-13 17:36:19 -0700688 // 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
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000698 scan_mark := parser.mark
699
Jonathan Hart4b110f62020-03-13 17:36:19 -0700700 // Eat whitespaces and comments until we reach the next token.
701 if !yaml_parser_scan_to_next_token(parser) {
702 return false
703 }
704
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000705 // [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.
Jonathan Hart4b110f62020-03-13 17:36:19 -0700708
709 // Check the indentation level against the current column.
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000710 if !yaml_parser_unroll_indent(parser, parser.mark.column, scan_mark) {
Jonathan Hart4b110f62020-03-13 17:36:19 -0700711 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
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000743 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
Jonathan Hart4b110f62020-03-13 17:36:19 -0700763 // 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.
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000856 // [Go] TODO Make this logic more reasonable.
Jonathan Hart4b110f62020-03-13 17:36:19 -0700857 //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
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000883func 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
Jonathan Hart4b110f62020-03-13 17:36:19 -0700886 }
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000887
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
Jonathan Hart4b110f62020-03-13 17:36:19 -0700907}
908
909// Check if a simple key may start at the current position and add it if
910// needed.
911func 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),
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000926 mark: parser.mark,
Jonathan Hart4b110f62020-03-13 17:36:19 -0700927 }
Jonathan Hart4b110f62020-03-13 17:36:19 -0700928
929 if !yaml_parser_remove_simple_key(parser) {
930 return false
931 }
932 parser.simple_keys[len(parser.simple_keys)-1] = simple_key
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000933 parser.simple_keys_by_tok[simple_key.token_number] = len(parser.simple_keys) - 1
Jonathan Hart4b110f62020-03-13 17:36:19 -0700934 }
935 return true
936}
937
938// Remove a potential simple key at the current flow level.
939func 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 }
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000948 // 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)
Jonathan Hart4b110f62020-03-13 17:36:19 -0700951 }
Jonathan Hart4b110f62020-03-13 17:36:19 -0700952 return true
953}
954
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000955// max_flow_level limits the flow_level
956const max_flow_level = 10000
957
Jonathan Hart4b110f62020-03-13 17:36:19 -0700958// Increase the flow level and resize the simple key list if needed.
959func yaml_parser_increase_flow_level(parser *yaml_parser_t) bool {
960 // Reset the simple key on the next level.
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000961 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 })
Jonathan Hart4b110f62020-03-13 17:36:19 -0700967
968 // Increase the flow level.
969 parser.flow_level++
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000970 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 }
Jonathan Hart4b110f62020-03-13 17:36:19 -0700975 return true
976}
977
978// Decrease the flow level.
979func yaml_parser_decrease_flow_level(parser *yaml_parser_t) bool {
980 if parser.flow_level > 0 {
981 parser.flow_level--
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000982 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]
Jonathan Hart4b110f62020-03-13 17:36:19 -0700985 }
986 return true
987}
988
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +0000989// max_indents limits the indents stack size
990const max_indents = 10000
991
Jonathan Hart4b110f62020-03-13 17:36:19 -0700992// 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.
995func 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
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001006 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 }
Jonathan Hart4b110f62020-03-13 17:36:19 -07001011
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.
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001029func yaml_parser_unroll_indent(parser *yaml_parser_t, column int, scan_mark yaml_mark_t) bool {
Jonathan Hart4b110f62020-03-13 17:36:19 -07001030 // In the flow context, do nothing.
1031 if parser.flow_level > 0 {
1032 return true
1033 }
1034
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001035 block_mark := scan_mark
1036 block_mark.index--
1037
Jonathan Hart4b110f62020-03-13 17:36:19 -07001038 // Loop through the indentation levels in the stack.
1039 for parser.indent > column {
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001040
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
Jonathan Hart4b110f62020-03-13 17:36:19 -07001067 // Create a token and append it to the queue.
1068 token := yaml_token_t{
1069 typ: yaml_BLOCK_END_TOKEN,
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001070 start_mark: block_mark,
1071 end_mark: block_mark,
Jonathan Hart4b110f62020-03-13 17:36:19 -07001072 }
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.
1083func 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
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001091 parser.simple_keys_by_tok = make(map[int]int)
1092
Jonathan Hart4b110f62020-03-13 17:36:19 -07001093 // 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.
1111func 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.
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001120 if !yaml_parser_unroll_indent(parser, -1, parser.mark) {
Jonathan Hart4b110f62020-03-13 17:36:19 -07001121 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.
1142func yaml_parser_fetch_directive(parser *yaml_parser_t) bool {
1143 // Reset the indentation level.
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001144 if !yaml_parser_unroll_indent(parser, -1, parser.mark) {
Jonathan Hart4b110f62020-03-13 17:36:19 -07001145 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.
1166func yaml_parser_fetch_document_indicator(parser *yaml_parser_t, typ yaml_token_type_t) bool {
1167 // Reset the indentation level.
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001168 if !yaml_parser_unroll_indent(parser, -1, parser.mark) {
Jonathan Hart4b110f62020-03-13 17:36:19 -07001169 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.
1200func yaml_parser_fetch_flow_collection_start(parser *yaml_parser_t, typ yaml_token_type_t) bool {
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001201
Jonathan Hart4b110f62020-03-13 17:36:19 -07001202 // 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.
1232func 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.
1264func 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.
1289func 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.
1331func 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.
1370func 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?
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001375 if valid, ok := yaml_simple_key_is_valid(parser, simple_key); !ok {
1376 return false
1377
1378 } else if valid {
1379
Jonathan Hart4b110f62020-03-13 17:36:19 -07001380 // 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
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001397 delete(parser.simple_keys_by_tok, simple_key.token_number)
Jonathan Hart4b110f62020-03-13 17:36:19 -07001398
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.
1440func 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.
1459func 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.
1478func 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.
1497func 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.
1516func 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.
1535func yaml_parser_scan_to_next_token(parser *yaml_parser_t) bool {
1536
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001537 scan_mark := parser.mark
1538
Jonathan Hart4b110f62020-03-13 17:36:19 -07001539 // 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
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001565 // 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
Jonathan Hart4b110f62020-03-13 17:36:19 -07001588 // Eat a comment until a line break.
1589 if parser.buffer[parser.buffer_pos] == '#' {
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001590 if !yaml_parser_scan_comments(parser, scan_mark) {
1591 return false
Jonathan Hart4b110f62020-03-13 17:36:19 -07001592 }
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//
1622func 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] == '#' {
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00001689 // [Go] Discard this inline comment for the time being.
1690 //if !yaml_parser_scan_line_comment(parser, start_mark) {
1691 // return false
1692 //}
Jonathan Hart4b110f62020-03-13 17:36:19 -07001693 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//
1727func 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// ^^^^^^
1763func 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
1795const 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// ^
1804func 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//
1840func 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
1898func 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
1956func 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.
2041func 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.
2086func 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 // '=', '+', '$', ',', '.', '!', '~', '*', '\'', '(', ')', '[', ']',
2107 // '%'.
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00002108 // [Go] TODO Convert this into more reasonable logic.
Jonathan Hart4b110f62020-03-13 17:36:19 -07002109 for is_alpha(parser.buffer, parser.buffer_pos) || parser.buffer[parser.buffer_pos] == ';' ||
2110 parser.buffer[parser.buffer_pos] == '/' || parser.buffer[parser.buffer_pos] == '?' ||
2111 parser.buffer[parser.buffer_pos] == ':' || parser.buffer[parser.buffer_pos] == '@' ||
2112 parser.buffer[parser.buffer_pos] == '&' || parser.buffer[parser.buffer_pos] == '=' ||
2113 parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '$' ||
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.
2144func 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.
2190func 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] == '#' {
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00002263 if !yaml_parser_scan_line_comment(parser, start_mark) {
2264 return false
2265 }
Jonathan Hart4b110f62020-03-13 17:36:19 -07002266 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.
2381func 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.
2435func 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.
2691func 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}
David K. Bainbridgee05cf0c2021-08-19 03:16:50 +00002833
2834func 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
2885func 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}