@@ -0,0 +1,3570 @@
6+ * The following notes assume that you are familiar with the YAML specification
7+ * (http://yaml.org/spec/cvs/current.html). We mostly follow it, although in
8+ * some cases we are less restrictive that it requires.
10+ * The process of transforming a YAML stream into a sequence of events is
11+ * divided on two steps: Scanning and Parsing.
13+ * The Scanner transforms the input stream into a sequence of tokens, while the
14+ * parser transform the sequence of tokens produced by the Scanner into a
15+ * sequence of parsing events.
17+ * The Scanner is rather clever and complicated. The Parser, on the contrary,
18+ * is a straightforward implementation of a recursive-descendant parser (or,
19+ * LL(1) parser, as it is usually called).
21+ * Actually there are two issues of Scanning that might be called "clever", the
22+ * rest is quite straightforward. The issues are "block collection start" and
23+ * "simple keys". Both issues are explained below in details.
25+ * Here the Scanning step is explained and implemented. We start with the list
26+ * of all the tokens produced by the Scanner together with short descriptions.
30+ * STREAM-START(encoding) # The stream start.
31+ * STREAM-END # The stream end.
32+ * VERSION-DIRECTIVE(major,minor) # The '%YAML' directive.
33+ * TAG-DIRECTIVE(handle,prefix) # The '%TAG' directive.
34+ * DOCUMENT-START # '---'
35+ * DOCUMENT-END # '...'
36+ * BLOCK-SEQUENCE-START # Indentation increase denoting a block
37+ * BLOCK-MAPPING-START # sequence or a block mapping.
38+ * BLOCK-END # Indentation decrease.
39+ * FLOW-SEQUENCE-START # '['
40+ * FLOW-SEQUENCE-END # ']'
41+ * BLOCK-SEQUENCE-START # '{'
42+ * BLOCK-SEQUENCE-END # '}'
45+ * KEY # '?' or nothing (simple keys).
47+ * ALIAS(anchor) # '*anchor'
48+ * ANCHOR(anchor) # '&anchor'
49+ * TAG(handle,suffix) # '!handle!suffix'
50+ * SCALAR(value,style) # A scalar.
52+ * The following two tokens are "virtual" tokens denoting the beginning and the
55+ * STREAM-START(encoding)
58+ * We pass the information about the input stream encoding with the
59+ * STREAM-START token.
61+ * The next two tokens are responsible for tags:
63+ * VERSION-DIRECTIVE(major,minor)
64+ * TAG-DIRECTIVE(handle,prefix)
70+ * %TAG !yaml! tag:yaml.org,2002:
73+ * The correspoding sequence of tokens:
75+ * STREAM-START(utf-8)
76+ * VERSION-DIRECTIVE(1,1)
77+ * TAG-DIRECTIVE("!","!foo")
78+ * TAG-DIRECTIVE("!yaml","tag:yaml.org,2002:")
82+ * Note that the VERSION-DIRECTIVE and TAG-DIRECTIVE tokens occupy a whole
85+ * The document start and end indicators are represented by:
90+ * Note that if a YAML stream contains an implicit document (without '---'
91+ * and '...' indicators), no DOCUMENT-START and DOCUMENT-END tokens will be
94+ * In the following examples, we present whole documents together with the
97+ * 1. An implicit document:
103+ * STREAM-START(utf-8)
104+ * SCALAR("a scalar",single-quoted)
107+ * 2. An explicit document:
115+ * STREAM-START(utf-8)
117+ * SCALAR("a scalar",single-quoted)
121+ * 3. Several documents in a stream:
127+ * 'yet another scalar'
131+ * STREAM-START(utf-8)
132+ * SCALAR("a scalar",single-quoted)
134+ * SCALAR("another scalar",single-quoted)
136+ * SCALAR("yet another scalar",single-quoted)
139+ * We have already introduced the SCALAR token above. The following tokens are
140+ * used to describe aliases, anchors, tag, and scalars:
144+ * TAG(handle,suffix)
145+ * SCALAR(value,style)
147+ * The following series of examples illustrate the usage of these tokens:
149+ * 1. A recursive sequence:
155+ * STREAM-START(utf-8)
157+ * FLOW-SEQUENCE-START
162+ * 2. A tagged scalar:
164+ * !!float "3.14" # A good approximation.
168+ * STREAM-START(utf-8)
170+ * SCALAR("3.14",double-quoted)
173+ * 3. Various scalar styles:
175+ * --- # Implicit empty plain scalars do not produce tokens.
176+ * --- a plain scalar
177+ * --- 'a single-quoted scalar'
178+ * --- "a double-quoted scalar"
187+ * STREAM-START(utf-8)
190+ * SCALAR("a plain scalar",plain)
192+ * SCALAR("a single-quoted scalar",single-quoted)
194+ * SCALAR("a double-quoted scalar",double-quoted)
196+ * SCALAR("a literal scalar",literal)
198+ * SCALAR("a folded scalar",folded)
201+ * Now it's time to review collection-related tokens. We will start with
204+ * FLOW-SEQUENCE-START
206+ * FLOW-MAPPING-START
212+ * The tokens FLOW-SEQUENCE-START, FLOW-SEQUENCE-END, FLOW-MAPPING-START, and
213+ * FLOW-MAPPING-END represent the indicators '[', ']', '{', and '}'
214+ * correspondingly. FLOW-ENTRY represent the ',' indicator. Finally the
215+ * indicators '?' and ':', which are used for denoting mapping keys and values,
216+ * are represented by the KEY and VALUE tokens.
218+ * The following examples show flow collections:
220+ * 1. A flow sequence:
222+ * [item 1, item 2, item 3]
226+ * STREAM-START(utf-8)
227+ * FLOW-SEQUENCE-START
228+ * SCALAR("item 1",plain)
230+ * SCALAR("item 2",plain)
232+ * SCALAR("item 3",plain)
236+ * 2. A flow mapping:
239+ * a simple key: a value, # Note that the KEY token is produced.
240+ * ? a complex key: another value,
245+ * STREAM-START(utf-8)
246+ * FLOW-MAPPING-START
248+ * SCALAR("a simple key",plain)
250+ * SCALAR("a value",plain)
253+ * SCALAR("a complex key",plain)
255+ * SCALAR("another value",plain)
260+ * A simple key is a key which is not denoted by the '?' indicator. Note that
261+ * the Scanner still produce the KEY token whenever it encounters a simple key.
263+ * For scanning block collections, the following tokens are used (note that we
264+ * repeat KEY and VALUE here):
266+ * BLOCK-SEQUENCE-START
267+ * BLOCK-MAPPING-START
273+ * The tokens BLOCK-SEQUENCE-START and BLOCK-MAPPING-START denote indentation
274+ * increase that precedes a block collection (cf. the INDENT token in Python).
275+ * The token BLOCK-END denote indentation decrease that ends a block collection
276+ * (cf. the DEDENT token in Python). However YAML has some syntax pecularities
277+ * that makes detections of these tokens more complex.
279+ * The tokens BLOCK-ENTRY, KEY, and VALUE are used to represent the indicators
280+ * '-', '?', and ':' correspondingly.
282+ * The following examples show how the tokens BLOCK-SEQUENCE-START,
283+ * BLOCK-MAPPING-START, and BLOCK-END are emitted by the Scanner:
285+ * 1. Block sequences:
298+ * STREAM-START(utf-8)
299+ * BLOCK-SEQUENCE-START
301+ * SCALAR("item 1",plain)
303+ * SCALAR("item 2",plain)
305+ * BLOCK-SEQUENCE-START
307+ * SCALAR("item 3.1",plain)
309+ * SCALAR("item 3.2",plain)
312+ * BLOCK-MAPPING-START
314+ * SCALAR("key 1",plain)
316+ * SCALAR("value 1",plain)
318+ * SCALAR("key 2",plain)
320+ * SCALAR("value 2",plain)
325+ * 2. Block mappings:
327+ * a simple key: a value # The KEY token is produced here.
339+ * STREAM-START(utf-8)
340+ * BLOCK-MAPPING-START
342+ * SCALAR("a simple key",plain)
344+ * SCALAR("a value",plain)
346+ * SCALAR("a complex key",plain)
348+ * SCALAR("another value",plain)
350+ * SCALAR("a mapping",plain)
351+ * BLOCK-MAPPING-START
353+ * SCALAR("key 1",plain)
355+ * SCALAR("value 1",plain)
357+ * SCALAR("key 2",plain)
359+ * SCALAR("value 2",plain)
362+ * SCALAR("a sequence",plain)
364+ * BLOCK-SEQUENCE-START
366+ * SCALAR("item 1",plain)
368+ * SCALAR("item 2",plain)
373+ * YAML does not always require to start a new block collection from a new
374+ * line. If the current line contains only '-', '?', and ':' indicators, a new
375+ * block collection may start at the current line. The following examples
376+ * illustrate this case:
378+ * 1. Collections in a sequence:
389+ * STREAM-START(utf-8)
390+ * BLOCK-SEQUENCE-START
392+ * BLOCK-SEQUENCE-START
394+ * SCALAR("item 1",plain)
396+ * SCALAR("item 2",plain)
399+ * BLOCK-MAPPING-START
401+ * SCALAR("key 1",plain)
403+ * SCALAR("value 1",plain)
405+ * SCALAR("key 2",plain)
407+ * SCALAR("value 2",plain)
410+ * BLOCK-MAPPING-START
412+ * SCALAR("complex key")
414+ * SCALAR("complex value")
419+ * 2. Collections in a mapping:
430+ * STREAM-START(utf-8)
431+ * BLOCK-MAPPING-START
433+ * SCALAR("a sequence",plain)
435+ * BLOCK-SEQUENCE-START
437+ * SCALAR("item 1",plain)
439+ * SCALAR("item 2",plain)
442+ * SCALAR("a mapping",plain)
444+ * BLOCK-MAPPING-START
446+ * SCALAR("key 1",plain)
448+ * SCALAR("value 1",plain)
450+ * SCALAR("key 2",plain)
452+ * SCALAR("value 2",plain)
457+ * YAML also permits non-indented sequences if they are included into a block
458+ * mapping. In this case, the token BLOCK-SEQUENCE-START is not produced:
461+ * - item 1 # BLOCK-SEQUENCE-START is NOT produced here.
466+ * STREAM-START(utf-8)
467+ * BLOCK-MAPPING-START
469+ * SCALAR("key",plain)
472+ * SCALAR("item 1",plain)
474+ * SCALAR("item 2",plain)
478+#include "yaml_private.h"
481+ * Ensure that the buffer contains the required number of characters.
482+ * Return 1 on success, 0 on failure (reader error or memory error).
485+#define CACHE(parser,length) \
486+ (parser->unread >= (length) \
488+ : yaml_parser_update_buffer(parser, (length)))
491+ * Advance the buffer pointer.
494+#define SKIP(parser) \
495+ (parser->mark.index ++, \
496+ parser->mark.column ++, \
497+ parser->unread --, \
498+ parser->buffer.pointer += WIDTH(parser->buffer))
500+#define SKIP_LINE(parser) \
501+ (IS_CRLF(parser->buffer) ? \
502+ (parser->mark.index += 2, \
503+ parser->mark.column = 0, \
504+ parser->mark.line ++, \
505+ parser->unread -= 2, \
506+ parser->buffer.pointer += 2) : \
507+ IS_BREAK(parser->buffer) ? \
508+ (parser->mark.index ++, \
509+ parser->mark.column = 0, \
510+ parser->mark.line ++, \
511+ parser->unread --, \
512+ parser->buffer.pointer += WIDTH(parser->buffer)) : 0)
515+ * Copy a character to a string buffer and advance pointers.
518+#define READ(parser,string) \
519+ (STRING_EXTEND(parser,string) ? \
520+ (COPY(string,parser->buffer), \
521+ parser->mark.index ++, \
522+ parser->mark.column ++, \
523+ parser->unread --, \
527+ * Copy a line break character to a string buffer and advance pointers.
530+#define READ_LINE(parser,string) \
531+ (STRING_EXTEND(parser,string) ? \
532+ (((CHECK_AT(parser->buffer,'\r',0) \
533+ && CHECK_AT(parser->buffer,'\n',1)) ? /* CR LF -> LF */ \
534+ (*((string).pointer++) = (yaml_char_t) '\n', \
535+ parser->buffer.pointer += 2, \
536+ parser->mark.index += 2, \
537+ parser->mark.column = 0, \
538+ parser->mark.line ++, \
539+ parser->unread -= 2) : \
540+ (CHECK_AT(parser->buffer,'\r',0) \
541+ || CHECK_AT(parser->buffer,'\n',0)) ? /* CR|LF -> LF */ \
542+ (*((string).pointer++) = (yaml_char_t) '\n', \
543+ parser->buffer.pointer ++, \
544+ parser->mark.index ++, \
545+ parser->mark.column = 0, \
546+ parser->mark.line ++, \
547+ parser->unread --) : \
548+ (CHECK_AT(parser->buffer,'\xC2',0) \
549+ && CHECK_AT(parser->buffer,'\x85',1)) ? /* NEL -> LF */ \
550+ (*((string).pointer++) = (yaml_char_t) '\n', \
551+ parser->buffer.pointer += 2, \
552+ parser->mark.index ++, \
553+ parser->mark.column = 0, \
554+ parser->mark.line ++, \
555+ parser->unread --) : \
556+ (CHECK_AT(parser->buffer,'\xE2',0) && \
557+ CHECK_AT(parser->buffer,'\x80',1) && \
558+ (CHECK_AT(parser->buffer,'\xA8',2) || \
559+ CHECK_AT(parser->buffer,'\xA9',2))) ? /* LS|PS -> LS|PS */ \
560+ (*((string).pointer++) = *(parser->buffer.pointer++), \
561+ *((string).pointer++) = *(parser->buffer.pointer++), \
562+ *((string).pointer++) = *(parser->buffer.pointer++), \
563+ parser->mark.index ++, \
564+ parser->mark.column = 0, \
565+ parser->mark.line ++, \
566+ parser->unread --) : 0), \
570+ * Public API declarations.
574+yaml_parser_scan(yaml_parser_t *parser, yaml_token_t *token);
581+yaml_parser_set_scanner_error(yaml_parser_t *parser, const char *context,
582+ yaml_mark_t context_mark, const char *problem);
585+ * High-level token API.
589+yaml_parser_fetch_more_tokens(yaml_parser_t *parser);
592+yaml_parser_fetch_next_token(yaml_parser_t *parser);
595+ * Potential simple keys.
599+yaml_parser_stale_simple_keys(yaml_parser_t *parser);
602+yaml_parser_save_simple_key(yaml_parser_t *parser);
605+yaml_parser_remove_simple_key(yaml_parser_t *parser);
608+yaml_parser_increase_flow_level(yaml_parser_t *parser);
611+yaml_parser_decrease_flow_level(yaml_parser_t *parser);
614+ * Indentation treatment.
618+yaml_parser_roll_indent(yaml_parser_t *parser, int column,
619+ int number, yaml_token_type_t type, yaml_mark_t mark);
622+yaml_parser_unroll_indent(yaml_parser_t *parser, int column);
629+yaml_parser_fetch_stream_start(yaml_parser_t *parser);
632+yaml_parser_fetch_stream_end(yaml_parser_t *parser);
635+yaml_parser_fetch_directive(yaml_parser_t *parser);
638+yaml_parser_fetch_document_indicator(yaml_parser_t *parser,
639+ yaml_token_type_t type);
642+yaml_parser_fetch_flow_collection_start(yaml_parser_t *parser,
643+ yaml_token_type_t type);
646+yaml_parser_fetch_flow_collection_end(yaml_parser_t *parser,
647+ yaml_token_type_t type);
650+yaml_parser_fetch_flow_entry(yaml_parser_t *parser);
653+yaml_parser_fetch_block_entry(yaml_parser_t *parser);
656+yaml_parser_fetch_key(yaml_parser_t *parser);
659+yaml_parser_fetch_value(yaml_parser_t *parser);
662+yaml_parser_fetch_anchor(yaml_parser_t *parser, yaml_token_type_t type);
665+yaml_parser_fetch_tag(yaml_parser_t *parser);
668+yaml_parser_fetch_block_scalar(yaml_parser_t *parser, int literal);
671+yaml_parser_fetch_flow_scalar(yaml_parser_t *parser, int single);
674+yaml_parser_fetch_plain_scalar(yaml_parser_t *parser);
681+yaml_parser_scan_to_next_token(yaml_parser_t *parser);
684+yaml_parser_scan_directive(yaml_parser_t *parser, yaml_token_t *token);
687+yaml_parser_scan_directive_name(yaml_parser_t *parser,
688+ yaml_mark_t start_mark, yaml_char_t **name);
691+yaml_parser_scan_version_directive_value(yaml_parser_t *parser,
692+ yaml_mark_t start_mark, int *major, int *minor);
695+yaml_parser_scan_version_directive_number(yaml_parser_t *parser,
696+ yaml_mark_t start_mark, int *number);
699+yaml_parser_scan_tag_directive_value(yaml_parser_t *parser,
700+ yaml_mark_t mark, yaml_char_t **handle, yaml_char_t **prefix);
703+yaml_parser_scan_anchor(yaml_parser_t *parser, yaml_token_t *token,
704+ yaml_token_type_t type);
707+yaml_parser_scan_tag(yaml_parser_t *parser, yaml_token_t *token);
710+yaml_parser_scan_tag_handle(yaml_parser_t *parser, int directive,
711+ yaml_mark_t start_mark, yaml_char_t **handle);
714+yaml_parser_scan_tag_uri(yaml_parser_t *parser, int directive,
715+ yaml_char_t *head, yaml_mark_t start_mark, yaml_char_t **uri);
718+yaml_parser_scan_uri_escapes(yaml_parser_t *parser, int directive,
719+ yaml_mark_t start_mark, yaml_string_t *string);
722+yaml_parser_scan_block_scalar(yaml_parser_t *parser, yaml_token_t *token,
726+yaml_parser_scan_block_scalar_breaks(yaml_parser_t *parser,
727+ int *indent, yaml_string_t *breaks,
728+ yaml_mark_t start_mark, yaml_mark_t *end_mark);
731+yaml_parser_scan_flow_scalar(yaml_parser_t *parser, yaml_token_t *token,
735+yaml_parser_scan_plain_scalar(yaml_parser_t *parser, yaml_token_t *token);
738+ * Get the next token.
742+yaml_parser_scan(yaml_parser_t *parser, yaml_token_t *token)
744+ assert(parser); /* Non-NULL parser object is expected. */
745+ assert(token); /* Non-NULL token object is expected. */
747+ /* Erase the token object. */
749+ memset(token, 0, sizeof(yaml_token_t));
751+ /* No tokens after STREAM-END or error. */
753+ if (parser->stream_end_produced || parser->error) {
757+ /* Ensure that the tokens queue contains enough tokens. */
759+ if (!parser->token_available) {
760+ if (!yaml_parser_fetch_more_tokens(parser))
764+ /* Fetch the next token from the queue. */
766+ *token = DEQUEUE(parser, parser->tokens);
767+ parser->token_available = 0;
768+ parser->tokens_parsed ++;
770+ if (token->type == YAML_STREAM_END_TOKEN) {
771+ parser->stream_end_produced = 1;
778+ * Set the scanner error and return 0.
782+yaml_parser_set_scanner_error(yaml_parser_t *parser, const char *context,
783+ yaml_mark_t context_mark, const char *problem)
785+ parser->error = YAML_SCANNER_ERROR;
786+ parser->context = context;
787+ parser->context_mark = context_mark;
788+ parser->problem = problem;
789+ parser->problem_mark = parser->mark;
795+ * Ensure that the tokens queue contains at least one token which can be
796+ * returned to the Parser.
800+yaml_parser_fetch_more_tokens(yaml_parser_t *parser)
802+ int need_more_tokens;
804+ /* While we need more tokens to fetch, do it. */
809+ * Check if we really need to fetch more tokens.
812+ need_more_tokens = 0;
814+ if (parser->tokens.head == parser->tokens.tail)
816+ /* Queue is empty. */
818+ need_more_tokens = 1;
822+ yaml_simple_key_t *simple_key;
824+ /* Check if any potential simple key may occupy the head position. */
826+ if (!yaml_parser_stale_simple_keys(parser))
829+ for (simple_key = parser->simple_keys.start;
830+ simple_key != parser->simple_keys.top; simple_key++) {
831+ if (simple_key->possible
832+ && simple_key->token_number == parser->tokens_parsed) {
833+ need_more_tokens = 1;
839+ /* We are finished. */
841+ if (!need_more_tokens)
844+ /* Fetch the next token. */
846+ if (!yaml_parser_fetch_next_token(parser))
850+ parser->token_available = 1;
856+ * The dispatcher for token fetchers.
860+yaml_parser_fetch_next_token(yaml_parser_t *parser)
862+ /* Ensure that the buffer is initialized. */
864+ if (!CACHE(parser, 1))
867+ /* Check if we just started scanning. Fetch STREAM-START then. */
869+ if (!parser->stream_start_produced)
870+ return yaml_parser_fetch_stream_start(parser);
872+ /* Eat whitespaces and comments until we reach the next token. */
874+ if (!yaml_parser_scan_to_next_token(parser))
877+ /* Remove obsolete potential simple keys. */
879+ if (!yaml_parser_stale_simple_keys(parser))
882+ /* Check the indentation level against the current column. */
884+ if (!yaml_parser_unroll_indent(parser, parser->mark.column))
888+ * Ensure that the buffer contains at least 4 characters. 4 is the length
889+ * of the longest indicators ('--- ' and '... ').
892+ if (!CACHE(parser, 4))
895+ /* Is it the end of the stream? */
897+ if (IS_Z(parser->buffer))
898+ return yaml_parser_fetch_stream_end(parser);
900+ /* Is it a directive? */
902+ if (parser->mark.column == 0 && CHECK(parser->buffer, '%'))
903+ return yaml_parser_fetch_directive(parser);
905+ /* Is it the document start indicator? */
907+ if (parser->mark.column == 0
908+ && CHECK_AT(parser->buffer, '-', 0)
909+ && CHECK_AT(parser->buffer, '-', 1)
910+ && CHECK_AT(parser->buffer, '-', 2)
911+ && IS_BLANKZ_AT(parser->buffer, 3))
912+ return yaml_parser_fetch_document_indicator(parser,
913+ YAML_DOCUMENT_START_TOKEN);
915+ /* Is it the document end indicator? */
917+ if (parser->mark.column == 0
918+ && CHECK_AT(parser->buffer, '.', 0)
919+ && CHECK_AT(parser->buffer, '.', 1)
920+ && CHECK_AT(parser->buffer, '.', 2)
921+ && IS_BLANKZ_AT(parser->buffer, 3))
922+ return yaml_parser_fetch_document_indicator(parser,
923+ YAML_DOCUMENT_END_TOKEN);
925+ /* Is it the flow sequence start indicator? */
927+ if (CHECK(parser->buffer, '['))
928+ return yaml_parser_fetch_flow_collection_start(parser,
929+ YAML_FLOW_SEQUENCE_START_TOKEN);
931+ /* Is it the flow mapping start indicator? */
933+ if (CHECK(parser->buffer, '{'))
934+ return yaml_parser_fetch_flow_collection_start(parser,
935+ YAML_FLOW_MAPPING_START_TOKEN);
937+ /* Is it the flow sequence end indicator? */
939+ if (CHECK(parser->buffer, ']'))
940+ return yaml_parser_fetch_flow_collection_end(parser,
941+ YAML_FLOW_SEQUENCE_END_TOKEN);
943+ /* Is it the flow mapping end indicator? */
945+ if (CHECK(parser->buffer, '}'))
946+ return yaml_parser_fetch_flow_collection_end(parser,
947+ YAML_FLOW_MAPPING_END_TOKEN);
949+ /* Is it the flow entry indicator? */
951+ if (CHECK(parser->buffer, ','))
952+ return yaml_parser_fetch_flow_entry(parser);
954+ /* Is it the block entry indicator? */
956+ if (CHECK(parser->buffer, '-') && IS_BLANKZ_AT(parser->buffer, 1))
957+ return yaml_parser_fetch_block_entry(parser);
959+ /* Is it the key indicator? */
961+ if (CHECK(parser->buffer, '?')
962+ && (parser->flow_level || IS_BLANKZ_AT(parser->buffer, 1)))
963+ return yaml_parser_fetch_key(parser);
965+ /* Is it the value indicator? */
967+ if (CHECK(parser->buffer, ':')
968+ && (parser->flow_level || IS_BLANKZ_AT(parser->buffer, 1)))
969+ return yaml_parser_fetch_value(parser);
971+ /* Is it an alias? */
973+ if (CHECK(parser->buffer, '*'))
974+ return yaml_parser_fetch_anchor(parser, YAML_ALIAS_TOKEN);
976+ /* Is it an anchor? */
978+ if (CHECK(parser->buffer, '&'))
979+ return yaml_parser_fetch_anchor(parser, YAML_ANCHOR_TOKEN);
983+ if (CHECK(parser->buffer, '!'))
984+ return yaml_parser_fetch_tag(parser);
986+ /* Is it a literal scalar? */
988+ if (CHECK(parser->buffer, '|') && !parser->flow_level)
989+ return yaml_parser_fetch_block_scalar(parser, 1);
991+ /* Is it a folded scalar? */
993+ if (CHECK(parser->buffer, '>') && !parser->flow_level)
994+ return yaml_parser_fetch_block_scalar(parser, 0);
996+ /* Is it a single-quoted scalar? */
998+ if (CHECK(parser->buffer, '\''))
999+ return yaml_parser_fetch_flow_scalar(parser, 1);
1001+ /* Is it a double-quoted scalar? */
1003+ if (CHECK(parser->buffer, '"'))
1004+ return yaml_parser_fetch_flow_scalar(parser, 0);
1007+ * Is it a plain scalar?
1009+ * A plain scalar may start with any non-blank characters except
1011+ * '-', '?', ':', ',', '[', ']', '{', '}',
1012+ * '#', '&', '*', '!', '|', '>', '\'', '\"',
1015+ * In the block context (and, for the '-' indicator, in the flow context
1016+ * too), it may also start with the characters
1020+ * if it is followed by a non-space character.
1022+ * The last rule is more restrictive than the specification requires.
1025+ if (!(IS_BLANKZ(parser->buffer) || CHECK(parser->buffer, '-')
1026+ || CHECK(parser->buffer, '?') || CHECK(parser->buffer, ':')
1027+ || CHECK(parser->buffer, ',') || CHECK(parser->buffer, '[')
1028+ || CHECK(parser->buffer, ']') || CHECK(parser->buffer, '{')
1029+ || CHECK(parser->buffer, '}') || CHECK(parser->buffer, '#')
1030+ || CHECK(parser->buffer, '&') || CHECK(parser->buffer, '*')
1031+ || CHECK(parser->buffer, '!') || CHECK(parser->buffer, '|')
1032+ || CHECK(parser->buffer, '>') || CHECK(parser->buffer, '\'')
1033+ || CHECK(parser->buffer, '"') || CHECK(parser->buffer, '%')
1034+ || CHECK(parser->buffer, '@') || CHECK(parser->buffer, '`')) ||
1035+ (CHECK(parser->buffer, '-') && !IS_BLANK_AT(parser->buffer, 1)) ||
1036+ (!parser->flow_level &&
1037+ (CHECK(parser->buffer, '?') || CHECK(parser->buffer, ':'))
1038+ && !IS_BLANKZ_AT(parser->buffer, 1)))
1039+ return yaml_parser_fetch_plain_scalar(parser);
1042+ * If we don't determine the token type so far, it is an error.
1045+ return yaml_parser_set_scanner_error(parser,
1046+ "while scanning for the next token", parser->mark,
1047+ "found character that cannot start any token");
1051+ * Check the list of potential simple keys and remove the positions that
1052+ * cannot contain simple keys anymore.
1056+yaml_parser_stale_simple_keys(yaml_parser_t *parser)
1058+ yaml_simple_key_t *simple_key;
1060+ /* Check for a potential simple key for each flow level. */
1062+ for (simple_key = parser->simple_keys.start;
1063+ simple_key != parser->simple_keys.top; simple_key ++)
1066+ * The specification requires that a simple key
1068+ * - is limited to a single line,
1069+ * - is shorter than 1024 characters.
1072+ if (simple_key->possible
1073+ && (simple_key->mark.line < parser->mark.line
1074+ || simple_key->mark.index+1024 < parser->mark.index)) {
1076+ /* Check if the potential simple key to be removed is required. */
1078+ if (simple_key->required) {
1079+ return yaml_parser_set_scanner_error(parser,
1080+ "while scanning a simple key", simple_key->mark,
1081+ "could not find expected ':'");
1084+ simple_key->possible = 0;
1092+ * Check if a simple key may start at the current position and add it if
1097+yaml_parser_save_simple_key(yaml_parser_t *parser)
1100+ * A simple key is required at the current position if the scanner is in
1101+ * the block context and the current column coincides with the indentation
1105+ int required = (!parser->flow_level
1106+ && parser->indent == (int)parser->mark.column);
1109+ * A simple key is required only when it is the first token in the current
1110+ * line. Therefore it is always allowed. But we add a check anyway.
1113+ assert(parser->simple_key_allowed || !required); /* Impossible. */
1116+ * If the current position may start a simple key, save it.
1119+ if (parser->simple_key_allowed)
1121+ yaml_simple_key_t simple_key;
1122+ simple_key.possible = 1;
1123+ simple_key.required = required;
1124+ simple_key.token_number =
1125+ parser->tokens_parsed + (parser->tokens.tail - parser->tokens.head);
1126+ simple_key.mark = parser->mark;
1128+ if (!yaml_parser_remove_simple_key(parser)) return 0;
1130+ *(parser->simple_keys.top-1) = simple_key;
1137+ * Remove a potential simple key at the current flow level.
1141+yaml_parser_remove_simple_key(yaml_parser_t *parser)
1143+ yaml_simple_key_t *simple_key = parser->simple_keys.top-1;
1145+ if (simple_key->possible)
1147+ /* If the key is required, it is an error. */
1149+ if (simple_key->required) {
1150+ return yaml_parser_set_scanner_error(parser,
1151+ "while scanning a simple key", simple_key->mark,
1152+ "could not find expected ':'");
1156+ /* Remove the key from the stack. */
1158+ simple_key->possible = 0;
1164+ * Increase the flow level and resize the simple key list if needed.
1168+yaml_parser_increase_flow_level(yaml_parser_t *parser)
1170+ yaml_simple_key_t empty_simple_key = { 0, 0, 0, { 0, 0, 0 } };
1172+ /* Reset the simple key on the next level. */
1174+ if (!PUSH(parser, parser->simple_keys, empty_simple_key))
1177+ /* Increase the flow level. */
1179+ parser->flow_level++;
1185+ * Decrease the flow level.
1189+yaml_parser_decrease_flow_level(yaml_parser_t *parser)
1191+ yaml_simple_key_t dummy_key; /* Used to eliminate a compiler warning. */
1193+ if (parser->flow_level) {
1194+ parser->flow_level --;
1195+ dummy_key = POP(parser, parser->simple_keys);
1202+ * Push the current indentation level to the stack and set the new level
1203+ * the current column is greater than the indentation level. In this case,
1204+ * append or insert the specified token into the token queue.
1209+yaml_parser_roll_indent(yaml_parser_t *parser, int column,
1210+ int number, yaml_token_type_t type, yaml_mark_t mark)
1212+ yaml_token_t token;
1214+ /* In the flow context, do nothing. */
1216+ if (parser->flow_level)
1219+ if (parser->indent < column)
1222+ * Push the current indentation level to the stack and set the new
1223+ * indentation level.
1226+ if (!PUSH(parser, parser->indents, parser->indent))
1229+ parser->indent = column;
1231+ /* Create a token and insert it into the queue. */
1233+ TOKEN_INIT(token, type, mark, mark);
1235+ if (number == -1) {
1236+ if (!ENQUEUE(parser, parser->tokens, token))
1240+ if (!QUEUE_INSERT(parser,
1241+ parser->tokens, number - parser->tokens_parsed, token))
1250+ * Pop indentation levels from the indents stack until the current level
1251+ * becomes less or equal to the column. For each intendation level, append
1252+ * the BLOCK-END token.
1257+yaml_parser_unroll_indent(yaml_parser_t *parser, int column)
1259+ yaml_token_t token;
1261+ /* In the flow context, do nothing. */
1263+ if (parser->flow_level)
1266+ /* Loop through the intendation levels in the stack. */
1268+ while (parser->indent > column)
1270+ /* Create a token and append it to the queue. */
1272+ TOKEN_INIT(token, YAML_BLOCK_END_TOKEN, parser->mark, parser->mark);
1274+ if (!ENQUEUE(parser, parser->tokens, token))
1277+ /* Pop the indentation level. */
1279+ parser->indent = POP(parser, parser->indents);
1286+ * Initialize the scanner and produce the STREAM-START token.
1290+yaml_parser_fetch_stream_start(yaml_parser_t *parser)
1292+ yaml_simple_key_t simple_key = { 0, 0, 0, { 0, 0, 0 } };
1293+ yaml_token_t token;
1295+ /* Set the initial indentation. */
1297+ parser->indent = -1;
1299+ /* Initialize the simple key stack. */
1301+ if (!PUSH(parser, parser->simple_keys, simple_key))
1304+ /* A simple key is allowed at the beginning of the stream. */
1306+ parser->simple_key_allowed = 1;
1308+ /* We have started. */
1310+ parser->stream_start_produced = 1;
1312+ /* Create the STREAM-START token and append it to the queue. */
1314+ STREAM_START_TOKEN_INIT(token, parser->encoding,
1315+ parser->mark, parser->mark);
1317+ if (!ENQUEUE(parser, parser->tokens, token))
1324+ * Produce the STREAM-END token and shut down the scanner.
1328+yaml_parser_fetch_stream_end(yaml_parser_t *parser)
1330+ yaml_token_t token;
1332+ /* Force new line. */
1334+ if (parser->mark.column != 0) {
1335+ parser->mark.column = 0;
1336+ parser->mark.line ++;
1339+ /* Reset the indentation level. */
1341+ if (!yaml_parser_unroll_indent(parser, -1))
1344+ /* Reset simple keys. */
1346+ if (!yaml_parser_remove_simple_key(parser))
1349+ parser->simple_key_allowed = 0;
1351+ /* Create the STREAM-END token and append it to the queue. */
1353+ STREAM_END_TOKEN_INIT(token, parser->mark, parser->mark);
1355+ if (!ENQUEUE(parser, parser->tokens, token))
1362+ * Produce a VERSION-DIRECTIVE or TAG-DIRECTIVE token.
1366+yaml_parser_fetch_directive(yaml_parser_t *parser)
1368+ yaml_token_t token;
1370+ /* Reset the indentation level. */
1372+ if (!yaml_parser_unroll_indent(parser, -1))
1375+ /* Reset simple keys. */
1377+ if (!yaml_parser_remove_simple_key(parser))
1380+ parser->simple_key_allowed = 0;
1382+ /* Create the YAML-DIRECTIVE or TAG-DIRECTIVE token. */
1384+ if (!yaml_parser_scan_directive(parser, &token))
1387+ /* Append the token to the queue. */
1389+ if (!ENQUEUE(parser, parser->tokens, token)) {
1390+ yaml_token_delete(&token);
1398+ * Produce the DOCUMENT-START or DOCUMENT-END token.
1402+yaml_parser_fetch_document_indicator(yaml_parser_t *parser,
1403+ yaml_token_type_t type)
1405+ yaml_mark_t start_mark, end_mark;
1406+ yaml_token_t token;
1408+ /* Reset the indentation level. */
1410+ if (!yaml_parser_unroll_indent(parser, -1))
1413+ /* Reset simple keys. */
1415+ if (!yaml_parser_remove_simple_key(parser))
1418+ parser->simple_key_allowed = 0;
1420+ /* Consume the token. */
1422+ start_mark = parser->mark;
1428+ end_mark = parser->mark;
1430+ /* Create the DOCUMENT-START or DOCUMENT-END token. */
1432+ TOKEN_INIT(token, type, start_mark, end_mark);
1434+ /* Append the token to the queue. */
1436+ if (!ENQUEUE(parser, parser->tokens, token))
1443+ * Produce the FLOW-SEQUENCE-START or FLOW-MAPPING-START token.
1447+yaml_parser_fetch_flow_collection_start(yaml_parser_t *parser,
1448+ yaml_token_type_t type)
1450+ yaml_mark_t start_mark, end_mark;
1451+ yaml_token_t token;
1453+ /* The indicators '[' and '{' may start a simple key. */
1455+ if (!yaml_parser_save_simple_key(parser))
1458+ /* Increase the flow level. */
1460+ if (!yaml_parser_increase_flow_level(parser))
1463+ /* A simple key may follow the indicators '[' and '{'. */
1465+ parser->simple_key_allowed = 1;
1467+ /* Consume the token. */
1469+ start_mark = parser->mark;
1471+ end_mark = parser->mark;
1473+ /* Create the FLOW-SEQUENCE-START of FLOW-MAPPING-START token. */
1475+ TOKEN_INIT(token, type, start_mark, end_mark);
1477+ /* Append the token to the queue. */
1479+ if (!ENQUEUE(parser, parser->tokens, token))
1486+ * Produce the FLOW-SEQUENCE-END or FLOW-MAPPING-END token.
1490+yaml_parser_fetch_flow_collection_end(yaml_parser_t *parser,
1491+ yaml_token_type_t type)
1493+ yaml_mark_t start_mark, end_mark;
1494+ yaml_token_t token;
1496+ /* Reset any potential simple key on the current flow level. */
1498+ if (!yaml_parser_remove_simple_key(parser))
1501+ /* Decrease the flow level. */
1503+ if (!yaml_parser_decrease_flow_level(parser))
1506+ /* No simple keys after the indicators ']' and '}'. */
1508+ parser->simple_key_allowed = 0;
1510+ /* Consume the token. */
1512+ start_mark = parser->mark;
1514+ end_mark = parser->mark;
1516+ /* Create the FLOW-SEQUENCE-END of FLOW-MAPPING-END token. */
1518+ TOKEN_INIT(token, type, start_mark, end_mark);
1520+ /* Append the token to the queue. */
1522+ if (!ENQUEUE(parser, parser->tokens, token))
1529+ * Produce the FLOW-ENTRY token.
1533+yaml_parser_fetch_flow_entry(yaml_parser_t *parser)
1535+ yaml_mark_t start_mark, end_mark;
1536+ yaml_token_t token;
1538+ /* Reset any potential simple keys on the current flow level. */
1540+ if (!yaml_parser_remove_simple_key(parser))
1543+ /* Simple keys are allowed after ','. */
1545+ parser->simple_key_allowed = 1;
1547+ /* Consume the token. */
1549+ start_mark = parser->mark;
1551+ end_mark = parser->mark;
1553+ /* Create the FLOW-ENTRY token and append it to the queue. */
1555+ TOKEN_INIT(token, YAML_FLOW_ENTRY_TOKEN, start_mark, end_mark);
1557+ if (!ENQUEUE(parser, parser->tokens, token))
1564+ * Produce the BLOCK-ENTRY token.
1568+yaml_parser_fetch_block_entry(yaml_parser_t *parser)
1570+ yaml_mark_t start_mark, end_mark;
1571+ yaml_token_t token;
1573+ /* Check if the scanner is in the block context. */
1575+ if (!parser->flow_level)
1577+ /* Check if we are allowed to start a new entry. */
1579+ if (!parser->simple_key_allowed) {
1580+ return yaml_parser_set_scanner_error(parser, NULL, parser->mark,
1581+ "block sequence entries are not allowed in this context");
1584+ /* Add the BLOCK-SEQUENCE-START token if needed. */
1586+ if (!yaml_parser_roll_indent(parser, parser->mark.column, -1,
1587+ YAML_BLOCK_SEQUENCE_START_TOKEN, parser->mark))
1593+ * It is an error for the '-' indicator to occur in the flow context,
1594+ * but we let the Parser detect and report about it because the Parser
1595+ * is able to point to the context.
1599+ /* Reset any potential simple keys on the current flow level. */
1601+ if (!yaml_parser_remove_simple_key(parser))
1604+ /* Simple keys are allowed after '-'. */
1606+ parser->simple_key_allowed = 1;
1608+ /* Consume the token. */
1610+ start_mark = parser->mark;
1612+ end_mark = parser->mark;
1614+ /* Create the BLOCK-ENTRY token and append it to the queue. */
1616+ TOKEN_INIT(token, YAML_BLOCK_ENTRY_TOKEN, start_mark, end_mark);
1618+ if (!ENQUEUE(parser, parser->tokens, token))
1625+ * Produce the KEY token.
1629+yaml_parser_fetch_key(yaml_parser_t *parser)
1631+ yaml_mark_t start_mark, end_mark;
1632+ yaml_token_t token;
1634+ /* In the block context, additional checks are required. */
1636+ if (!parser->flow_level)
1638+ /* Check if we are allowed to start a new key (not nessesary simple). */
1640+ if (!parser->simple_key_allowed) {
1641+ return yaml_parser_set_scanner_error(parser, NULL, parser->mark,
1642+ "mapping keys are not allowed in this context");
1645+ /* Add the BLOCK-MAPPING-START token if needed. */
1647+ if (!yaml_parser_roll_indent(parser, parser->mark.column, -1,
1648+ YAML_BLOCK_MAPPING_START_TOKEN, parser->mark))
1652+ /* Reset any potential simple keys on the current flow level. */
1654+ if (!yaml_parser_remove_simple_key(parser))
1657+ /* Simple keys are allowed after '?' in the block context. */
1659+ parser->simple_key_allowed = (!parser->flow_level);
1661+ /* Consume the token. */
1663+ start_mark = parser->mark;
1665+ end_mark = parser->mark;
1667+ /* Create the KEY token and append it to the queue. */
1669+ TOKEN_INIT(token, YAML_KEY_TOKEN, start_mark, end_mark);
1671+ if (!ENQUEUE(parser, parser->tokens, token))
1678+ * Produce the VALUE token.
1682+yaml_parser_fetch_value(yaml_parser_t *parser)
1684+ yaml_mark_t start_mark, end_mark;
1685+ yaml_token_t token;
1686+ yaml_simple_key_t *simple_key = parser->simple_keys.top-1;
1688+ /* Have we found a simple key? */
1690+ if (simple_key->possible)
1693+ /* Create the KEY token and insert it into the queue. */
1695+ TOKEN_INIT(token, YAML_KEY_TOKEN, simple_key->mark, simple_key->mark);
1697+ if (!QUEUE_INSERT(parser, parser->tokens,
1698+ simple_key->token_number - parser->tokens_parsed, token))
1701+ /* In the block context, we may need to add the BLOCK-MAPPING-START token. */
1703+ if (!yaml_parser_roll_indent(parser, simple_key->mark.column,
1704+ simple_key->token_number,
1705+ YAML_BLOCK_MAPPING_START_TOKEN, simple_key->mark))
1708+ /* Remove the simple key. */
1710+ simple_key->possible = 0;
1712+ /* A simple key cannot follow another simple key. */
1714+ parser->simple_key_allowed = 0;
1718+ /* The ':' indicator follows a complex key. */
1720+ /* In the block context, extra checks are required. */
1722+ if (!parser->flow_level)
1724+ /* Check if we are allowed to start a complex value. */
1726+ if (!parser->simple_key_allowed) {
1727+ return yaml_parser_set_scanner_error(parser, NULL, parser->mark,
1728+ "mapping values are not allowed in this context");
1731+ /* Add the BLOCK-MAPPING-START token if needed. */
1733+ if (!yaml_parser_roll_indent(parser, parser->mark.column, -1,
1734+ YAML_BLOCK_MAPPING_START_TOKEN, parser->mark))
1738+ /* Simple keys after ':' are allowed in the block context. */
1740+ parser->simple_key_allowed = (!parser->flow_level);
1743+ /* Consume the token. */
1745+ start_mark = parser->mark;
1747+ end_mark = parser->mark;
1749+ /* Create the VALUE token and append it to the queue. */
1751+ TOKEN_INIT(token, YAML_VALUE_TOKEN, start_mark, end_mark);
1753+ if (!ENQUEUE(parser, parser->tokens, token))
1760+ * Produce the ALIAS or ANCHOR token.
1764+yaml_parser_fetch_anchor(yaml_parser_t *parser, yaml_token_type_t type)
1766+ yaml_token_t token;
1768+ /* An anchor or an alias could be a simple key. */
1770+ if (!yaml_parser_save_simple_key(parser))
1773+ /* A simple key cannot follow an anchor or an alias. */
1775+ parser->simple_key_allowed = 0;
1777+ /* Create the ALIAS or ANCHOR token and append it to the queue. */
1779+ if (!yaml_parser_scan_anchor(parser, &token, type))
1782+ if (!ENQUEUE(parser, parser->tokens, token)) {
1783+ yaml_token_delete(&token);
1790+ * Produce the TAG token.
1794+yaml_parser_fetch_tag(yaml_parser_t *parser)
1796+ yaml_token_t token;
1798+ /* A tag could be a simple key. */
1800+ if (!yaml_parser_save_simple_key(parser))
1803+ /* A simple key cannot follow a tag. */
1805+ parser->simple_key_allowed = 0;
1807+ /* Create the TAG token and append it to the queue. */
1809+ if (!yaml_parser_scan_tag(parser, &token))
1812+ if (!ENQUEUE(parser, parser->tokens, token)) {
1813+ yaml_token_delete(&token);
1821+ * Produce the SCALAR(...,literal) or SCALAR(...,folded) tokens.
1825+yaml_parser_fetch_block_scalar(yaml_parser_t *parser, int literal)
1827+ yaml_token_t token;
1829+ /* Remove any potential simple keys. */
1831+ if (!yaml_parser_remove_simple_key(parser))
1834+ /* A simple key may follow a block scalar. */
1836+ parser->simple_key_allowed = 1;
1838+ /* Create the SCALAR token and append it to the queue. */
1840+ if (!yaml_parser_scan_block_scalar(parser, &token, literal))
1843+ if (!ENQUEUE(parser, parser->tokens, token)) {
1844+ yaml_token_delete(&token);
1852+ * Produce the SCALAR(...,single-quoted) or SCALAR(...,double-quoted) tokens.
1856+yaml_parser_fetch_flow_scalar(yaml_parser_t *parser, int single)
1858+ yaml_token_t token;
1860+ /* A plain scalar could be a simple key. */
1862+ if (!yaml_parser_save_simple_key(parser))
1865+ /* A simple key cannot follow a flow scalar. */
1867+ parser->simple_key_allowed = 0;
1869+ /* Create the SCALAR token and append it to the queue. */
1871+ if (!yaml_parser_scan_flow_scalar(parser, &token, single))
1874+ if (!ENQUEUE(parser, parser->tokens, token)) {
1875+ yaml_token_delete(&token);
1883+ * Produce the SCALAR(...,plain) token.
1887+yaml_parser_fetch_plain_scalar(yaml_parser_t *parser)
1889+ yaml_token_t token;
1891+ /* A plain scalar could be a simple key. */
1893+ if (!yaml_parser_save_simple_key(parser))
1896+ /* A simple key cannot follow a flow scalar. */
1898+ parser->simple_key_allowed = 0;
1900+ /* Create the SCALAR token and append it to the queue. */
1902+ if (!yaml_parser_scan_plain_scalar(parser, &token))
1905+ if (!ENQUEUE(parser, parser->tokens, token)) {
1906+ yaml_token_delete(&token);
1914+ * Eat whitespaces and comments until the next token is found.
1918+yaml_parser_scan_to_next_token(yaml_parser_t *parser)
1920+ /* Until the next token is not found. */
1924+ /* Allow the BOM mark to start a line. */
1926+ if (!CACHE(parser, 1)) return 0;
1928+ if (parser->mark.column == 0 && IS_BOM(parser->buffer))
1934+ * Tabs are allowed:
1936+ * - in the flow context;
1937+ * - in the block context, but not at the beginning of the line or
1938+ * after '-', '?', or ':' (complex value).
1941+ if (!CACHE(parser, 1)) return 0;
1943+ while (CHECK(parser->buffer,' ') ||
1944+ ((parser->flow_level || !parser->simple_key_allowed) &&
1945+ CHECK(parser->buffer, '\t'))) {
1947+ if (!CACHE(parser, 1)) return 0;
1950+ /* Eat a comment until a line break. */
1952+ if (CHECK(parser->buffer, '#')) {
1953+ while (!IS_BREAKZ(parser->buffer)) {
1955+ if (!CACHE(parser, 1)) return 0;
1959+ /* If it is a line break, eat it. */
1961+ if (IS_BREAK(parser->buffer))
1963+ if (!CACHE(parser, 2)) return 0;
1966+ /* In the block context, a new line may start a simple key. */
1968+ if (!parser->flow_level) {
1969+ parser->simple_key_allowed = 1;
1974+ /* We have found a token. */
1984+ * Scan a YAML-DIRECTIVE or TAG-DIRECTIVE token.
1987+ * %YAML 1.1 # a comment \n
1988+ * ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1989+ * %TAG !yaml! tag:yaml.org,2002: \n
1990+ * ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1994+yaml_parser_scan_directive(yaml_parser_t *parser, yaml_token_t *token)
1996+ yaml_mark_t start_mark, end_mark;
1997+ yaml_char_t *name = NULL;
1999+ yaml_char_t *handle = NULL, *prefix = NULL;
2003+ start_mark = parser->mark;
2007+ /* Scan the directive name. */
2009+ if (!yaml_parser_scan_directive_name(parser, start_mark, &name))
2012+ /* Is it a YAML directive? */
2014+ if (strcmp((char *)name, "YAML") == 0)
2016+ /* Scan the VERSION directive value. */
2018+ if (!yaml_parser_scan_version_directive_value(parser, start_mark,
2022+ end_mark = parser->mark;
2024+ /* Create a VERSION-DIRECTIVE token. */
2026+ VERSION_DIRECTIVE_TOKEN_INIT(*token, major, minor,
2027+ start_mark, end_mark);
2030+ /* Is it a TAG directive? */
2032+ else if (strcmp((char *)name, "TAG") == 0)
2034+ /* Scan the TAG directive value. */
2036+ if (!yaml_parser_scan_tag_directive_value(parser, start_mark,
2040+ end_mark = parser->mark;
2042+ /* Create a TAG-DIRECTIVE token. */
2044+ TAG_DIRECTIVE_TOKEN_INIT(*token, handle, prefix,
2045+ start_mark, end_mark);
2048+ /* Unknown directive. */
2052+ yaml_parser_set_scanner_error(parser, "while scanning a directive",
2053+ start_mark, "found uknown directive name");
2057+ /* Eat the rest of the line including any comments. */
2059+ if (!CACHE(parser, 1)) goto error;
2061+ while (IS_BLANK(parser->buffer)) {
2063+ if (!CACHE(parser, 1)) goto error;
2066+ if (CHECK(parser->buffer, '#')) {
2067+ while (!IS_BREAKZ(parser->buffer)) {
2069+ if (!CACHE(parser, 1)) goto error;
2073+ /* Check if we are at the end of the line. */
2075+ if (!IS_BREAKZ(parser->buffer)) {
2076+ yaml_parser_set_scanner_error(parser, "while scanning a directive",
2077+ start_mark, "did not find expected comment or line break");
2081+ /* Eat a line break. */
2083+ if (IS_BREAK(parser->buffer)) {
2084+ if (!CACHE(parser, 2)) goto error;
2100+ * Scan the directive name.
2103+ * %YAML 1.1 # a comment \n
2105+ * %TAG !yaml! tag:yaml.org,2002: \n
2110+yaml_parser_scan_directive_name(yaml_parser_t *parser,
2111+ yaml_mark_t start_mark, yaml_char_t **name)
2113+ yaml_string_t string = NULL_STRING;
2115+ if (!STRING_INIT(parser, string, INITIAL_STRING_SIZE)) goto error;
2117+ /* Consume the directive name. */
2119+ if (!CACHE(parser, 1)) goto error;
2121+ while (IS_ALPHA(parser->buffer))
2123+ if (!READ(parser, string)) goto error;
2124+ if (!CACHE(parser, 1)) goto error;
2127+ /* Check if the name is empty. */
2129+ if (string.start == string.pointer) {
2130+ yaml_parser_set_scanner_error(parser, "while scanning a directive",
2131+ start_mark, "could not find expected directive name");
2135+ /* Check for an blank character after the name. */
2137+ if (!IS_BLANKZ(parser->buffer)) {
2138+ yaml_parser_set_scanner_error(parser, "while scanning a directive",
2139+ start_mark, "found unexpected non-alphabetical character");
2143+ *name = string.start;
2148+ STRING_DEL(parser, string);
2153+ * Scan the value of VERSION-DIRECTIVE.
2156+ * %YAML 1.1 # a comment \n
2161+yaml_parser_scan_version_directive_value(yaml_parser_t *parser,
2162+ yaml_mark_t start_mark, int *major, int *minor)
2164+ /* Eat whitespaces. */
2166+ if (!CACHE(parser, 1)) return 0;
2168+ while (IS_BLANK(parser->buffer)) {
2170+ if (!CACHE(parser, 1)) return 0;
2173+ /* Consume the major version number. */
2175+ if (!yaml_parser_scan_version_directive_number(parser, start_mark, major))
2180+ if (!CHECK(parser->buffer, '.')) {
2181+ return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive",
2182+ start_mark, "did not find expected digit or '.' character");
2187+ /* Consume the minor version number. */
2189+ if (!yaml_parser_scan_version_directive_number(parser, start_mark, minor))
2195+#define MAX_NUMBER_LENGTH 9
2198+ * Scan the version number of VERSION-DIRECTIVE.
2201+ * %YAML 1.1 # a comment \n
2203+ * %YAML 1.1 # a comment \n
2208+yaml_parser_scan_version_directive_number(yaml_parser_t *parser,
2209+ yaml_mark_t start_mark, int *number)
2214+ /* Repeat while the next character is digit. */
2216+ if (!CACHE(parser, 1)) return 0;
2218+ while (IS_DIGIT(parser->buffer))
2220+ /* Check if the number is too long. */
2222+ if (++length > MAX_NUMBER_LENGTH) {
2223+ return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive",
2224+ start_mark, "found extremely long version number");
2227+ value = value*10 + AS_DIGIT(parser->buffer);
2231+ if (!CACHE(parser, 1)) return 0;
2234+ /* Check if the number was present. */
2237+ return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive",
2238+ start_mark, "did not find expected version number");
2247+ * Scan the value of a TAG-DIRECTIVE token.
2250+ * %TAG !yaml! tag:yaml.org,2002: \n
2251+ * ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2255+yaml_parser_scan_tag_directive_value(yaml_parser_t *parser,
2256+ yaml_mark_t start_mark, yaml_char_t **handle, yaml_char_t **prefix)
2258+ yaml_char_t *handle_value = NULL;
2259+ yaml_char_t *prefix_value = NULL;
2261+ /* Eat whitespaces. */
2263+ if (!CACHE(parser, 1)) goto error;
2265+ while (IS_BLANK(parser->buffer)) {
2267+ if (!CACHE(parser, 1)) goto error;
2270+ /* Scan a handle. */
2272+ if (!yaml_parser_scan_tag_handle(parser, 1, start_mark, &handle_value))
2275+ /* Expect a whitespace. */
2277+ if (!CACHE(parser, 1)) goto error;
2279+ if (!IS_BLANK(parser->buffer)) {
2280+ yaml_parser_set_scanner_error(parser, "while scanning a %TAG directive",
2281+ start_mark, "did not find expected whitespace");
2285+ /* Eat whitespaces. */
2287+ while (IS_BLANK(parser->buffer)) {
2289+ if (!CACHE(parser, 1)) goto error;
2292+ /* Scan a prefix. */
2294+ if (!yaml_parser_scan_tag_uri(parser, 1, NULL, start_mark, &prefix_value))
2297+ /* Expect a whitespace or line break. */
2299+ if (!CACHE(parser, 1)) goto error;
2301+ if (!IS_BLANKZ(parser->buffer)) {
2302+ yaml_parser_set_scanner_error(parser, "while scanning a %TAG directive",
2303+ start_mark, "did not find expected whitespace or line break");
2307+ *handle = handle_value;
2308+ *prefix = prefix_value;
2313+ yaml_free(handle_value);
2314+ yaml_free(prefix_value);
2319+yaml_parser_scan_anchor(yaml_parser_t *parser, yaml_token_t *token,
2320+ yaml_token_type_t type)
2323+ yaml_mark_t start_mark, end_mark;
2324+ yaml_string_t string = NULL_STRING;
2326+ if (!STRING_INIT(parser, string, INITIAL_STRING_SIZE)) goto error;
2328+ /* Eat the indicator character. */
2330+ start_mark = parser->mark;
2334+ /* Consume the value. */
2336+ if (!CACHE(parser, 1)) goto error;
2338+ while (IS_ALPHA(parser->buffer)) {
2339+ if (!READ(parser, string)) goto error;
2340+ if (!CACHE(parser, 1)) goto error;
2344+ end_mark = parser->mark;
2347+ * Check if length of the anchor is greater than 0 and it is followed by
2348+ * a whitespace character or one of the indicators:
2350+ * '?', ':', ',', ']', '}', '%', '@', '`'.
2353+ if (!length || !(IS_BLANKZ(parser->buffer) || CHECK(parser->buffer, '?')
2354+ || CHECK(parser->buffer, ':') || CHECK(parser->buffer, ',')
2355+ || CHECK(parser->buffer, ']') || CHECK(parser->buffer, '}')
2356+ || CHECK(parser->buffer, '%') || CHECK(parser->buffer, '@')
2357+ || CHECK(parser->buffer, '`'))) {
2358+ yaml_parser_set_scanner_error(parser, type == YAML_ANCHOR_TOKEN ?
2359+ "while scanning an anchor" : "while scanning an alias", start_mark,
2360+ "did not find expected alphabetic or numeric character");
2364+ /* Create a token. */
2366+ if (type == YAML_ANCHOR_TOKEN) {
2367+ ANCHOR_TOKEN_INIT(*token, string.start, start_mark, end_mark);
2370+ ALIAS_TOKEN_INIT(*token, string.start, start_mark, end_mark);
2376+ STRING_DEL(parser, string);
2381+ * Scan a TAG token.
2385+yaml_parser_scan_tag(yaml_parser_t *parser, yaml_token_t *token)
2387+ yaml_char_t *handle = NULL;
2388+ yaml_char_t *suffix = NULL;
2389+ yaml_mark_t start_mark, end_mark;
2391+ start_mark = parser->mark;
2393+ /* Check if the tag is in the canonical form. */
2395+ if (!CACHE(parser, 2)) goto error;
2397+ if (CHECK_AT(parser->buffer, '<', 1))
2399+ /* Set the handle to '' */
2401+ handle = yaml_malloc(1);
2402+ if (!handle) goto error;
2410+ /* Consume the tag value. */
2412+ if (!yaml_parser_scan_tag_uri(parser, 0, NULL, start_mark, &suffix))
2415+ /* Check for '>' and eat it. */
2417+ if (!CHECK(parser->buffer, '>')) {
2418+ yaml_parser_set_scanner_error(parser, "while scanning a tag",
2419+ start_mark, "did not find the expected '>'");
2427+ /* The tag has either the '!suffix' or the '!handle!suffix' form. */
2429+ /* First, try to scan a handle. */
2431+ if (!yaml_parser_scan_tag_handle(parser, 0, start_mark, &handle))
2434+ /* Check if it is, indeed, handle. */
2436+ if (handle[0] == '!' && handle[1] != '\0' && handle[strlen((char *)handle)-1] == '!')
2438+ /* Scan the suffix now. */
2440+ if (!yaml_parser_scan_tag_uri(parser, 0, NULL, start_mark, &suffix))
2445+ /* It wasn't a handle after all. Scan the rest of the tag. */
2447+ if (!yaml_parser_scan_tag_uri(parser, 0, handle, start_mark, &suffix))
2450+ /* Set the handle to '!'. */
2453+ handle = yaml_malloc(2);
2454+ if (!handle) goto error;
2459+ * A special case: the '!' tag. Set the handle to '' and the
2463+ if (suffix[0] == '\0') {
2464+ yaml_char_t *tmp = handle;
2471+ /* Check the character which ends the tag. */
2473+ if (!CACHE(parser, 1)) goto error;
2475+ if (!IS_BLANKZ(parser->buffer)) {
2476+ yaml_parser_set_scanner_error(parser, "while scanning a tag",
2477+ start_mark, "did not find expected whitespace or line break");
2481+ end_mark = parser->mark;
2483+ /* Create a token. */
2485+ TAG_TOKEN_INIT(*token, handle, suffix, start_mark, end_mark);
2496+ * Scan a tag handle.
2500+yaml_parser_scan_tag_handle(yaml_parser_t *parser, int directive,
2501+ yaml_mark_t start_mark, yaml_char_t **handle)
2503+ yaml_string_t string = NULL_STRING;
2505+ if (!STRING_INIT(parser, string, INITIAL_STRING_SIZE)) goto error;
2507+ /* Check the initial '!' character. */
2509+ if (!CACHE(parser, 1)) goto error;
2511+ if (!CHECK(parser->buffer, '!')) {
2512+ yaml_parser_set_scanner_error(parser, directive ?
2513+ "while scanning a tag directive" : "while scanning a tag",
2514+ start_mark, "did not find expected '!'");
2518+ /* Copy the '!' character. */
2520+ if (!READ(parser, string)) goto error;
2522+ /* Copy all subsequent alphabetical and numerical characters. */
2524+ if (!CACHE(parser, 1)) goto error;
2526+ while (IS_ALPHA(parser->buffer))
2528+ if (!READ(parser, string)) goto error;
2529+ if (!CACHE(parser, 1)) goto error;
2532+ /* Check if the trailing character is '!' and copy it. */
2534+ if (CHECK(parser->buffer, '!'))
2536+ if (!READ(parser, string)) goto error;
2541+ * It's either the '!' tag or not really a tag handle. If it's a %TAG
2542+ * directive, it's an error. If it's a tag token, it must be a part of
2546+ if (directive && !(string.start[0] == '!' && string.start[1] == '\0')) {
2547+ yaml_parser_set_scanner_error(parser, "while parsing a tag directive",
2548+ start_mark, "did not find expected '!'");
2553+ *handle = string.start;
2558+ STRING_DEL(parser, string);
2567+yaml_parser_scan_tag_uri(yaml_parser_t *parser, int directive,
2568+ yaml_char_t *head, yaml_mark_t start_mark, yaml_char_t **uri)
2570+ size_t length = head ? strlen((char *)head) : 0;
2571+ yaml_string_t string = NULL_STRING;
2573+ if (!STRING_INIT(parser, string, INITIAL_STRING_SIZE)) goto error;
2575+ /* Resize the string to include the head. */
2577+ while (string.end - string.start <= (int)length) {
2578+ if (!yaml_string_extend(&string.start, &string.pointer, &string.end)) {
2579+ parser->error = YAML_MEMORY_ERROR;
2585+ * Copy the head if needed.
2587+ * Note that we don't copy the leading '!' character.
2591+ memcpy(string.start, head+1, length-1);
2592+ string.pointer += length-1;
2595+ /* Scan the tag. */
2597+ if (!CACHE(parser, 1)) goto error;
2600+ * The set of characters that may appear in URI is as follows:
2602+ * '0'-'9', 'A'-'Z', 'a'-'z', '_', '-', ';', '/', '?', ':', '@', '&',
2603+ * '=', '+', '$', ',', '.', '!', '~', '*', '\'', '(', ')', '[', ']',
2607+ while (IS_ALPHA(parser->buffer) || CHECK(parser->buffer, ';')
2608+ || CHECK(parser->buffer, '/') || CHECK(parser->buffer, '?')
2609+ || CHECK(parser->buffer, ':') || CHECK(parser->buffer, '@')
2610+ || CHECK(parser->buffer, '&') || CHECK(parser->buffer, '=')
2611+ || CHECK(parser->buffer, '+') || CHECK(parser->buffer, '$')
2612+ || CHECK(parser->buffer, ',') || CHECK(parser->buffer, '.')
2613+ || CHECK(parser->buffer, '!') || CHECK(parser->buffer, '~')
2614+ || CHECK(parser->buffer, '*') || CHECK(parser->buffer, '\'')
2615+ || CHECK(parser->buffer, '(') || CHECK(parser->buffer, ')')
2616+ || CHECK(parser->buffer, '[') || CHECK(parser->buffer, ']')
2617+ || CHECK(parser->buffer, '%'))
2619+ /* Check if it is a URI-escape sequence. */
2621+ if (CHECK(parser->buffer, '%')) {
2622+ if (!yaml_parser_scan_uri_escapes(parser,
2623+ directive, start_mark, &string)) goto error;
2626+ if (!READ(parser, string)) goto error;
2630+ if (!CACHE(parser, 1)) goto error;
2633+ /* Check if the tag is non-empty. */
2636+ if (!STRING_EXTEND(parser, string))
2639+ yaml_parser_set_scanner_error(parser, directive ?
2640+ "while parsing a %TAG directive" : "while parsing a tag",
2641+ start_mark, "did not find expected tag URI");
2645+ *uri = string.start;
2650+ STRING_DEL(parser, string);
2655+ * Decode an URI-escape sequence corresponding to a single UTF-8 character.
2659+yaml_parser_scan_uri_escapes(yaml_parser_t *parser, int directive,
2660+ yaml_mark_t start_mark, yaml_string_t *string)
2664+ /* Decode the required number of characters. */
2668+ unsigned char octet = 0;
2670+ /* Check for a URI-escaped octet. */
2672+ if (!CACHE(parser, 3)) return 0;
2674+ if (!(CHECK(parser->buffer, '%')
2675+ && IS_HEX_AT(parser->buffer, 1)
2676+ && IS_HEX_AT(parser->buffer, 2))) {
2677+ return yaml_parser_set_scanner_error(parser, directive ?
2678+ "while parsing a %TAG directive" : "while parsing a tag",
2679+ start_mark, "did not find URI escaped octet");
2682+ /* Get the octet. */
2684+ octet = (AS_HEX_AT(parser->buffer, 1) << 4) + AS_HEX_AT(parser->buffer, 2);
2686+ /* If it is the leading octet, determine the length of the UTF-8 sequence. */
2690+ width = (octet & 0x80) == 0x00 ? 1 :
2691+ (octet & 0xE0) == 0xC0 ? 2 :
2692+ (octet & 0xF0) == 0xE0 ? 3 :
2693+ (octet & 0xF8) == 0xF0 ? 4 : 0;
2695+ return yaml_parser_set_scanner_error(parser, directive ?
2696+ "while parsing a %TAG directive" : "while parsing a tag",
2697+ start_mark, "found an incorrect leading UTF-8 octet");
2702+ /* Check if the trailing octet is correct. */
2704+ if ((octet & 0xC0) != 0x80) {
2705+ return yaml_parser_set_scanner_error(parser, directive ?
2706+ "while parsing a %TAG directive" : "while parsing a tag",
2707+ start_mark, "found an incorrect trailing UTF-8 octet");
2711+ /* Copy the octet and move the pointers. */
2713+ *(string->pointer++) = octet;
2724+ * Scan a block scalar.
2728+yaml_parser_scan_block_scalar(yaml_parser_t *parser, yaml_token_t *token,
2731+ yaml_mark_t start_mark;
2732+ yaml_mark_t end_mark;
2733+ yaml_string_t string = NULL_STRING;
2734+ yaml_string_t leading_break = NULL_STRING;
2735+ yaml_string_t trailing_breaks = NULL_STRING;
2739+ int leading_blank = 0;
2740+ int trailing_blank = 0;
2742+ if (!STRING_INIT(parser, string, INITIAL_STRING_SIZE)) goto error;
2743+ if (!STRING_INIT(parser, leading_break, INITIAL_STRING_SIZE)) goto error;
2744+ if (!STRING_INIT(parser, trailing_breaks, INITIAL_STRING_SIZE)) goto error;
2746+ /* Eat the indicator '|' or '>'. */
2748+ start_mark = parser->mark;
2752+ /* Scan the additional block scalar indicators. */
2754+ if (!CACHE(parser, 1)) goto error;
2756+ /* Check for a chomping indicator. */
2758+ if (CHECK(parser->buffer, '+') || CHECK(parser->buffer, '-'))
2760+ /* Set the chomping method and eat the indicator. */
2762+ chomping = CHECK(parser->buffer, '+') ? +1 : -1;
2766+ /* Check for an indentation indicator. */
2768+ if (!CACHE(parser, 1)) goto error;
2770+ if (IS_DIGIT(parser->buffer))
2772+ /* Check that the intendation is greater than 0. */
2774+ if (CHECK(parser->buffer, '0')) {
2775+ yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
2776+ start_mark, "found an intendation indicator equal to 0");
2780+ /* Get the intendation level and eat the indicator. */
2782+ increment = AS_DIGIT(parser->buffer);
2788+ /* Do the same as above, but in the opposite order. */
2790+ else if (IS_DIGIT(parser->buffer))
2792+ if (CHECK(parser->buffer, '0')) {
2793+ yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
2794+ start_mark, "found an intendation indicator equal to 0");
2798+ increment = AS_DIGIT(parser->buffer);
2802+ if (!CACHE(parser, 1)) goto error;
2804+ if (CHECK(parser->buffer, '+') || CHECK(parser->buffer, '-')) {
2805+ chomping = CHECK(parser->buffer, '+') ? +1 : -1;
2811+ /* Eat whitespaces and comments to the end of the line. */
2813+ if (!CACHE(parser, 1)) goto error;
2815+ while (IS_BLANK(parser->buffer)) {
2817+ if (!CACHE(parser, 1)) goto error;
2820+ if (CHECK(parser->buffer, '#')) {
2821+ while (!IS_BREAKZ(parser->buffer)) {
2823+ if (!CACHE(parser, 1)) goto error;
2827+ /* Check if we are at the end of the line. */
2829+ if (!IS_BREAKZ(parser->buffer)) {
2830+ yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
2831+ start_mark, "did not find expected comment or line break");
2835+ /* Eat a line break. */
2837+ if (IS_BREAK(parser->buffer)) {
2838+ if (!CACHE(parser, 2)) goto error;
2842+ end_mark = parser->mark;
2844+ /* Set the intendation level if it was specified. */
2847+ indent = parser->indent >= 0 ? parser->indent+increment : increment;
2850+ /* Scan the leading line breaks and determine the indentation level if needed. */
2852+ if (!yaml_parser_scan_block_scalar_breaks(parser, &indent, &trailing_breaks,
2853+ start_mark, &end_mark)) goto error;
2855+ /* Scan the block scalar content. */
2857+ if (!CACHE(parser, 1)) goto error;
2859+ while ((int)parser->mark.column == indent && !IS_Z(parser->buffer))
2862+ * We are at the beginning of a non-empty line.
2865+ /* Is it a trailing whitespace? */
2867+ trailing_blank = IS_BLANK(parser->buffer);
2869+ /* Check if we need to fold the leading line break. */
2871+ if (!literal && (*leading_break.start == '\n')
2872+ && !leading_blank && !trailing_blank)
2874+ /* Do we need to join the lines by space? */
2876+ if (*trailing_breaks.start == '\0') {
2877+ if (!STRING_EXTEND(parser, string)) goto error;
2878+ *(string.pointer ++) = ' ';
2881+ CLEAR(parser, leading_break);
2884+ if (!JOIN(parser, string, leading_break)) goto error;
2885+ CLEAR(parser, leading_break);
2888+ /* Append the remaining line breaks. */
2890+ if (!JOIN(parser, string, trailing_breaks)) goto error;
2891+ CLEAR(parser, trailing_breaks);
2893+ /* Is it a leading whitespace? */
2895+ leading_blank = IS_BLANK(parser->buffer);
2897+ /* Consume the current line. */
2899+ while (!IS_BREAKZ(parser->buffer)) {
2900+ if (!READ(parser, string)) goto error;
2901+ if (!CACHE(parser, 1)) goto error;
2904+ /* Consume the line break. */
2906+ if (!CACHE(parser, 2)) goto error;
2908+ if (!READ_LINE(parser, leading_break)) goto error;
2910+ /* Eat the following intendation spaces and line breaks. */
2912+ if (!yaml_parser_scan_block_scalar_breaks(parser,
2913+ &indent, &trailing_breaks, start_mark, &end_mark)) goto error;
2916+ /* Chomp the tail. */
2918+ if (chomping != -1) {
2919+ if (!JOIN(parser, string, leading_break)) goto error;
2921+ if (chomping == 1) {
2922+ if (!JOIN(parser, string, trailing_breaks)) goto error;
2925+ /* Create a token. */
2927+ SCALAR_TOKEN_INIT(*token, string.start, string.pointer-string.start,
2928+ literal ? YAML_LITERAL_SCALAR_STYLE : YAML_FOLDED_SCALAR_STYLE,
2929+ start_mark, end_mark);
2931+ STRING_DEL(parser, leading_break);
2932+ STRING_DEL(parser, trailing_breaks);
2937+ STRING_DEL(parser, string);
2938+ STRING_DEL(parser, leading_break);
2939+ STRING_DEL(parser, trailing_breaks);
2945+ * Scan intendation spaces and line breaks for a block scalar. Determine the
2946+ * intendation level if needed.
2950+yaml_parser_scan_block_scalar_breaks(yaml_parser_t *parser,
2951+ int *indent, yaml_string_t *breaks,
2952+ yaml_mark_t start_mark, yaml_mark_t *end_mark)
2954+ int max_indent = 0;
2956+ *end_mark = parser->mark;
2958+ /* Eat the intendation spaces and line breaks. */
2962+ /* Eat the intendation spaces. */
2964+ if (!CACHE(parser, 1)) return 0;
2966+ while ((!*indent || (int)parser->mark.column < *indent)
2967+ && IS_SPACE(parser->buffer)) {
2969+ if (!CACHE(parser, 1)) return 0;
2972+ if ((int)parser->mark.column > max_indent)
2973+ max_indent = (int)parser->mark.column;
2975+ /* Check for a tab character messing the intendation. */
2977+ if ((!*indent || (int)parser->mark.column < *indent)
2978+ && IS_TAB(parser->buffer)) {
2979+ return yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
2980+ start_mark, "found a tab character where an intendation space is expected");
2983+ /* Have we found a non-empty line? */
2985+ if (!IS_BREAK(parser->buffer)) break;
2987+ /* Consume the line break. */
2989+ if (!CACHE(parser, 2)) return 0;
2990+ if (!READ_LINE(parser, *breaks)) return 0;
2991+ *end_mark = parser->mark;
2994+ /* Determine the indentation level if needed. */
2997+ *indent = max_indent;
2998+ if (*indent < parser->indent + 1)
2999+ *indent = parser->indent + 1;
3008+ * Scan a quoted scalar.
3012+yaml_parser_scan_flow_scalar(yaml_parser_t *parser, yaml_token_t *token,
3015+ yaml_mark_t start_mark;
3016+ yaml_mark_t end_mark;
3017+ yaml_string_t string = NULL_STRING;
3018+ yaml_string_t leading_break = NULL_STRING;
3019+ yaml_string_t trailing_breaks = NULL_STRING;
3020+ yaml_string_t whitespaces = NULL_STRING;
3021+ int leading_blanks;
3023+ if (!STRING_INIT(parser, string, INITIAL_STRING_SIZE)) goto error;
3024+ if (!STRING_INIT(parser, leading_break, INITIAL_STRING_SIZE)) goto error;
3025+ if (!STRING_INIT(parser, trailing_breaks, INITIAL_STRING_SIZE)) goto error;
3026+ if (!STRING_INIT(parser, whitespaces, INITIAL_STRING_SIZE)) goto error;
3028+ /* Eat the left quote. */
3030+ start_mark = parser->mark;
3034+ /* Consume the content of the quoted scalar. */
3038+ /* Check that there are no document indicators at the beginning of the line. */
3040+ if (!CACHE(parser, 4)) goto error;
3042+ if (parser->mark.column == 0 &&
3043+ ((CHECK_AT(parser->buffer, '-', 0) &&
3044+ CHECK_AT(parser->buffer, '-', 1) &&
3045+ CHECK_AT(parser->buffer, '-', 2)) ||
3046+ (CHECK_AT(parser->buffer, '.', 0) &&
3047+ CHECK_AT(parser->buffer, '.', 1) &&
3048+ CHECK_AT(parser->buffer, '.', 2))) &&
3049+ IS_BLANKZ_AT(parser->buffer, 3))
3051+ yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar",
3052+ start_mark, "found unexpected document indicator");
3056+ /* Check for EOF. */
3058+ if (IS_Z(parser->buffer)) {
3059+ yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar",
3060+ start_mark, "found unexpected end of stream");
3064+ /* Consume non-blank characters. */
3066+ if (!CACHE(parser, 2)) goto error;
3068+ leading_blanks = 0;
3070+ while (!IS_BLANKZ(parser->buffer))
3072+ /* Check for an escaped single quote. */
3074+ if (single && CHECK_AT(parser->buffer, '\'', 0)
3075+ && CHECK_AT(parser->buffer, '\'', 1))
3077+ if (!STRING_EXTEND(parser, string)) goto error;
3078+ *(string.pointer++) = '\'';
3083+ /* Check for the right quote. */
3085+ else if (CHECK(parser->buffer, single ? '\'' : '"'))
3090+ /* Check for an escaped line break. */
3092+ else if (!single && CHECK(parser->buffer, '\\')
3093+ && IS_BREAK_AT(parser->buffer, 1))
3095+ if (!CACHE(parser, 3)) goto error;
3098+ leading_blanks = 1;
3102+ /* Check for an escape sequence. */
3104+ else if (!single && CHECK(parser->buffer, '\\'))
3106+ size_t code_length = 0;
3108+ if (!STRING_EXTEND(parser, string)) goto error;
3110+ /* Check the escape character. */
3112+ switch (parser->buffer.pointer[1])
3115+ *(string.pointer++) = '\0';
3119+ *(string.pointer++) = '\x07';
3123+ *(string.pointer++) = '\x08';
3128+ *(string.pointer++) = '\x09';
3132+ *(string.pointer++) = '\x0A';
3136+ *(string.pointer++) = '\x0B';
3140+ *(string.pointer++) = '\x0C';
3144+ *(string.pointer++) = '\x0D';
3148+ *(string.pointer++) = '\x1B';
3152+ *(string.pointer++) = '\x20';
3156+ *(string.pointer++) = '"';
3160+ *(string.pointer++) = '\'';
3164+ *(string.pointer++) = '\\';
3167+ case 'N': /* NEL (#x85) */
3168+ *(string.pointer++) = '\xC2';
3169+ *(string.pointer++) = '\x85';
3172+ case '_': /* #xA0 */
3173+ *(string.pointer++) = '\xC2';
3174+ *(string.pointer++) = '\xA0';
3177+ case 'L': /* LS (#x2028) */
3178+ *(string.pointer++) = '\xE2';
3179+ *(string.pointer++) = '\x80';
3180+ *(string.pointer++) = '\xA8';
3183+ case 'P': /* PS (#x2029) */
3184+ *(string.pointer++) = '\xE2';
3185+ *(string.pointer++) = '\x80';
3186+ *(string.pointer++) = '\xA9';
3202+ yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar",
3203+ start_mark, "found unknown escape character");
3210+ /* Consume an arbitrary escape code. */
3214+ unsigned int value = 0;
3217+ /* Scan the character value. */
3219+ if (!CACHE(parser, code_length)) goto error;
3221+ for (k = 0; k < code_length; k ++) {
3222+ if (!IS_HEX_AT(parser->buffer, k)) {
3223+ yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar",
3224+ start_mark, "did not find expected hexdecimal number");
3227+ value = (value << 4) + AS_HEX_AT(parser->buffer, k);
3230+ /* Check the value and write the character. */
3232+ if ((value >= 0xD800 && value <= 0xDFFF) || value > 0x10FFFF) {
3233+ yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar",
3234+ start_mark, "found invalid Unicode character escape code");
3238+ if (value <= 0x7F) {
3239+ *(string.pointer++) = value;
3241+ else if (value <= 0x7FF) {
3242+ *(string.pointer++) = 0xC0 + (value >> 6);
3243+ *(string.pointer++) = 0x80 + (value & 0x3F);
3245+ else if (value <= 0xFFFF) {
3246+ *(string.pointer++) = 0xE0 + (value >> 12);
3247+ *(string.pointer++) = 0x80 + ((value >> 6) & 0x3F);
3248+ *(string.pointer++) = 0x80 + (value & 0x3F);
3251+ *(string.pointer++) = 0xF0 + (value >> 18);
3252+ *(string.pointer++) = 0x80 + ((value >> 12) & 0x3F);
3253+ *(string.pointer++) = 0x80 + ((value >> 6) & 0x3F);
3254+ *(string.pointer++) = 0x80 + (value & 0x3F);
3257+ /* Advance the pointer. */
3259+ for (k = 0; k < code_length; k ++) {
3267+ /* It is a non-escaped non-blank character. */
3269+ if (!READ(parser, string)) goto error;
3272+ if (!CACHE(parser, 2)) goto error;
3275+ /* Check if we are at the end of the scalar. */
3277+ if (CHECK(parser->buffer, single ? '\'' : '"'))
3280+ /* Consume blank characters. */
3282+ if (!CACHE(parser, 1)) goto error;
3284+ while (IS_BLANK(parser->buffer) || IS_BREAK(parser->buffer))
3286+ if (IS_BLANK(parser->buffer))
3288+ /* Consume a space or a tab character. */
3290+ if (!leading_blanks) {
3291+ if (!READ(parser, whitespaces)) goto error;
3299+ if (!CACHE(parser, 2)) goto error;
3301+ /* Check if it is a first line break. */
3303+ if (!leading_blanks)
3305+ CLEAR(parser, whitespaces);
3306+ if (!READ_LINE(parser, leading_break)) goto error;
3307+ leading_blanks = 1;
3311+ if (!READ_LINE(parser, trailing_breaks)) goto error;
3314+ if (!CACHE(parser, 1)) goto error;
3317+ /* Join the whitespaces or fold line breaks. */
3319+ if (leading_blanks)
3321+ /* Do we need to fold line breaks? */
3323+ if (leading_break.start[0] == '\n') {
3324+ if (trailing_breaks.start[0] == '\0') {
3325+ if (!STRING_EXTEND(parser, string)) goto error;
3326+ *(string.pointer++) = ' ';
3329+ if (!JOIN(parser, string, trailing_breaks)) goto error;
3330+ CLEAR(parser, trailing_breaks);
3332+ CLEAR(parser, leading_break);
3335+ if (!JOIN(parser, string, leading_break)) goto error;
3336+ if (!JOIN(parser, string, trailing_breaks)) goto error;
3337+ CLEAR(parser, leading_break);
3338+ CLEAR(parser, trailing_breaks);
3343+ if (!JOIN(parser, string, whitespaces)) goto error;
3344+ CLEAR(parser, whitespaces);
3348+ /* Eat the right quote. */
3352+ end_mark = parser->mark;
3354+ /* Create a token. */
3356+ SCALAR_TOKEN_INIT(*token, string.start, string.pointer-string.start,
3357+ single ? YAML_SINGLE_QUOTED_SCALAR_STYLE : YAML_DOUBLE_QUOTED_SCALAR_STYLE,
3358+ start_mark, end_mark);
3360+ STRING_DEL(parser, leading_break);
3361+ STRING_DEL(parser, trailing_breaks);
3362+ STRING_DEL(parser, whitespaces);
3367+ STRING_DEL(parser, string);
3368+ STRING_DEL(parser, leading_break);
3369+ STRING_DEL(parser, trailing_breaks);
3370+ STRING_DEL(parser, whitespaces);
3376+ * Scan a plain scalar.
3380+yaml_parser_scan_plain_scalar(yaml_parser_t *parser, yaml_token_t *token)
3382+ yaml_mark_t start_mark;
3383+ yaml_mark_t end_mark;
3384+ yaml_string_t string = NULL_STRING;
3385+ yaml_string_t leading_break = NULL_STRING;
3386+ yaml_string_t trailing_breaks = NULL_STRING;
3387+ yaml_string_t whitespaces = NULL_STRING;
3388+ int leading_blanks = 0;
3389+ int indent = parser->indent+1;
3391+ if (!STRING_INIT(parser, string, INITIAL_STRING_SIZE)) goto error;
3392+ if (!STRING_INIT(parser, leading_break, INITIAL_STRING_SIZE)) goto error;
3393+ if (!STRING_INIT(parser, trailing_breaks, INITIAL_STRING_SIZE)) goto error;
3394+ if (!STRING_INIT(parser, whitespaces, INITIAL_STRING_SIZE)) goto error;
3396+ start_mark = end_mark = parser->mark;
3398+ /* Consume the content of the plain scalar. */
3402+ /* Check for a document indicator. */
3404+ if (!CACHE(parser, 4)) goto error;
3406+ if (parser->mark.column == 0 &&
3407+ ((CHECK_AT(parser->buffer, '-', 0) &&
3408+ CHECK_AT(parser->buffer, '-', 1) &&
3409+ CHECK_AT(parser->buffer, '-', 2)) ||
3410+ (CHECK_AT(parser->buffer, '.', 0) &&
3411+ CHECK_AT(parser->buffer, '.', 1) &&
3412+ CHECK_AT(parser->buffer, '.', 2))) &&
3413+ IS_BLANKZ_AT(parser->buffer, 3)) break;
3415+ /* Check for a comment. */
3417+ if (CHECK(parser->buffer, '#'))
3420+ /* Consume non-blank characters. */
3422+ while (!IS_BLANKZ(parser->buffer))
3424+ /* Check for 'x:x' in the flow context. TODO: Fix the test "spec-08-13". */
3426+ if (parser->flow_level
3427+ && CHECK(parser->buffer, ':')
3428+ && !IS_BLANKZ_AT(parser->buffer, 1)) {
3429+ yaml_parser_set_scanner_error(parser, "while scanning a plain scalar",
3430+ start_mark, "found unexpected ':'");
3434+ /* Check for indicators that may end a plain scalar. */
3436+ if ((CHECK(parser->buffer, ':') && IS_BLANKZ_AT(parser->buffer, 1))
3437+ || (parser->flow_level &&
3438+ (CHECK(parser->buffer, ',') || CHECK(parser->buffer, ':')
3439+ || CHECK(parser->buffer, '?') || CHECK(parser->buffer, '[')
3440+ || CHECK(parser->buffer, ']') || CHECK(parser->buffer, '{')
3441+ || CHECK(parser->buffer, '}'))))
3444+ /* Check if we need to join whitespaces and breaks. */
3446+ if (leading_blanks || whitespaces.start != whitespaces.pointer)
3448+ if (leading_blanks)
3450+ /* Do we need to fold line breaks? */
3452+ if (leading_break.start[0] == '\n') {
3453+ if (trailing_breaks.start[0] == '\0') {
3454+ if (!STRING_EXTEND(parser, string)) goto error;
3455+ *(string.pointer++) = ' ';
3458+ if (!JOIN(parser, string, trailing_breaks)) goto error;
3459+ CLEAR(parser, trailing_breaks);
3461+ CLEAR(parser, leading_break);
3464+ if (!JOIN(parser, string, leading_break)) goto error;
3465+ if (!JOIN(parser, string, trailing_breaks)) goto error;
3466+ CLEAR(parser, leading_break);
3467+ CLEAR(parser, trailing_breaks);
3470+ leading_blanks = 0;
3474+ if (!JOIN(parser, string, whitespaces)) goto error;
3475+ CLEAR(parser, whitespaces);
3479+ /* Copy the character. */
3481+ if (!READ(parser, string)) goto error;
3483+ end_mark = parser->mark;
3485+ if (!CACHE(parser, 2)) goto error;
3488+ /* Is it the end? */
3490+ if (!(IS_BLANK(parser->buffer) || IS_BREAK(parser->buffer)))
3493+ /* Consume blank characters. */
3495+ if (!CACHE(parser, 1)) goto error;
3497+ while (IS_BLANK(parser->buffer) || IS_BREAK(parser->buffer))
3499+ if (IS_BLANK(parser->buffer))
3501+ /* Check for tab character that abuse intendation. */
3503+ if (leading_blanks && (int)parser->mark.column < indent
3504+ && IS_TAB(parser->buffer)) {
3505+ yaml_parser_set_scanner_error(parser, "while scanning a plain scalar",
3506+ start_mark, "found a tab character that violate intendation");
3510+ /* Consume a space or a tab character. */
3512+ if (!leading_blanks) {
3513+ if (!READ(parser, whitespaces)) goto error;
3521+ if (!CACHE(parser, 2)) goto error;
3523+ /* Check if it is a first line break. */
3525+ if (!leading_blanks)
3527+ CLEAR(parser, whitespaces);
3528+ if (!READ_LINE(parser, leading_break)) goto error;
3529+ leading_blanks = 1;
3533+ if (!READ_LINE(parser, trailing_breaks)) goto error;
3536+ if (!CACHE(parser, 1)) goto error;
3539+ /* Check intendation level. */
3541+ if (!parser->flow_level && (int)parser->mark.column < indent)
3545+ /* Create a token. */
3547+ SCALAR_TOKEN_INIT(*token, string.start, string.pointer-string.start,
3548+ YAML_PLAIN_SCALAR_STYLE, start_mark, end_mark);
3550+ /* Note that we change the 'simple_key_allowed' flag. */
3552+ if (leading_blanks) {
3553+ parser->simple_key_allowed = 1;
3556+ STRING_DEL(parser, leading_break);
3557+ STRING_DEL(parser, trailing_breaks);
3558+ STRING_DEL(parser, whitespaces);
3563+ STRING_DEL(parser, string);
3564+ STRING_DEL(parser, leading_break);
3565+ STRING_DEL(parser, trailing_breaks);
3566+ STRING_DEL(parser, whitespaces);