commonmarker 0.23.9 → 1.0.0.pre

Sign up to get free protection for your applications and to get access to all the features.
Files changed (89) hide show
  1. checksums.yaml +4 -4
  2. data/README.md +70 -212
  3. data/commonmarker.gemspec +34 -31
  4. data/ext/commonmarker/Cargo.toml +12 -0
  5. data/ext/commonmarker/_util.rb +102 -0
  6. data/ext/commonmarker/extconf.rb +4 -5
  7. data/ext/commonmarker/src/comrak_options.rs +107 -0
  8. data/ext/commonmarker/src/lib.rs +27 -0
  9. data/lib/commonmarker/config.rb +57 -38
  10. data/lib/commonmarker/extension.rb +14 -0
  11. data/lib/commonmarker/renderer.rb +1 -127
  12. data/lib/commonmarker/version.rb +2 -2
  13. data/lib/commonmarker.rb +14 -29
  14. metadata +37 -181
  15. data/Rakefile +0 -109
  16. data/bin/commonmarker +0 -118
  17. data/ext/commonmarker/arena.c +0 -104
  18. data/ext/commonmarker/autolink.c +0 -508
  19. data/ext/commonmarker/autolink.h +0 -8
  20. data/ext/commonmarker/blocks.c +0 -1620
  21. data/ext/commonmarker/buffer.c +0 -278
  22. data/ext/commonmarker/buffer.h +0 -116
  23. data/ext/commonmarker/case_fold_switch.inc +0 -4327
  24. data/ext/commonmarker/chunk.h +0 -135
  25. data/ext/commonmarker/cmark-gfm-core-extensions.h +0 -54
  26. data/ext/commonmarker/cmark-gfm-extension_api.h +0 -737
  27. data/ext/commonmarker/cmark-gfm-extensions_export.h +0 -42
  28. data/ext/commonmarker/cmark-gfm.h +0 -833
  29. data/ext/commonmarker/cmark-gfm_export.h +0 -42
  30. data/ext/commonmarker/cmark-gfm_version.h +0 -7
  31. data/ext/commonmarker/cmark.c +0 -55
  32. data/ext/commonmarker/cmark_ctype.c +0 -44
  33. data/ext/commonmarker/cmark_ctype.h +0 -33
  34. data/ext/commonmarker/commonmark.c +0 -514
  35. data/ext/commonmarker/commonmarker.c +0 -1308
  36. data/ext/commonmarker/commonmarker.h +0 -16
  37. data/ext/commonmarker/config.h +0 -76
  38. data/ext/commonmarker/core-extensions.c +0 -27
  39. data/ext/commonmarker/entities.inc +0 -2138
  40. data/ext/commonmarker/ext_scanners.c +0 -879
  41. data/ext/commonmarker/ext_scanners.h +0 -24
  42. data/ext/commonmarker/footnotes.c +0 -63
  43. data/ext/commonmarker/footnotes.h +0 -27
  44. data/ext/commonmarker/houdini.h +0 -57
  45. data/ext/commonmarker/houdini_href_e.c +0 -100
  46. data/ext/commonmarker/houdini_html_e.c +0 -66
  47. data/ext/commonmarker/houdini_html_u.c +0 -149
  48. data/ext/commonmarker/html.c +0 -502
  49. data/ext/commonmarker/html.h +0 -27
  50. data/ext/commonmarker/inlines.c +0 -1788
  51. data/ext/commonmarker/inlines.h +0 -29
  52. data/ext/commonmarker/iterator.c +0 -159
  53. data/ext/commonmarker/iterator.h +0 -26
  54. data/ext/commonmarker/latex.c +0 -468
  55. data/ext/commonmarker/linked_list.c +0 -37
  56. data/ext/commonmarker/man.c +0 -274
  57. data/ext/commonmarker/map.c +0 -129
  58. data/ext/commonmarker/map.h +0 -44
  59. data/ext/commonmarker/node.c +0 -1044
  60. data/ext/commonmarker/node.h +0 -166
  61. data/ext/commonmarker/parser.h +0 -59
  62. data/ext/commonmarker/plaintext.c +0 -218
  63. data/ext/commonmarker/plugin.c +0 -36
  64. data/ext/commonmarker/plugin.h +0 -34
  65. data/ext/commonmarker/references.c +0 -43
  66. data/ext/commonmarker/references.h +0 -26
  67. data/ext/commonmarker/registry.c +0 -63
  68. data/ext/commonmarker/registry.h +0 -24
  69. data/ext/commonmarker/render.c +0 -213
  70. data/ext/commonmarker/render.h +0 -62
  71. data/ext/commonmarker/scanners.c +0 -14056
  72. data/ext/commonmarker/scanners.h +0 -70
  73. data/ext/commonmarker/scanners.re +0 -341
  74. data/ext/commonmarker/strikethrough.c +0 -167
  75. data/ext/commonmarker/strikethrough.h +0 -9
  76. data/ext/commonmarker/syntax_extension.c +0 -149
  77. data/ext/commonmarker/syntax_extension.h +0 -34
  78. data/ext/commonmarker/table.c +0 -872
  79. data/ext/commonmarker/table.h +0 -12
  80. data/ext/commonmarker/tagfilter.c +0 -60
  81. data/ext/commonmarker/tagfilter.h +0 -8
  82. data/ext/commonmarker/tasklist.c +0 -156
  83. data/ext/commonmarker/tasklist.h +0 -8
  84. data/ext/commonmarker/utf8.c +0 -317
  85. data/ext/commonmarker/utf8.h +0 -35
  86. data/ext/commonmarker/xml.c +0 -182
  87. data/lib/commonmarker/node/inspect.rb +0 -47
  88. data/lib/commonmarker/node.rb +0 -83
  89. data/lib/commonmarker/renderer/html_renderer.rb +0 -256
@@ -1,1620 +0,0 @@
1
- /**
2
- * Block parsing implementation.
3
- *
4
- * For a high-level overview of the block parsing process,
5
- * see http://spec.commonmark.org/0.24/#phase-1-block-structure
6
- */
7
-
8
- #include <stdlib.h>
9
- #include <assert.h>
10
- #include <stdio.h>
11
- #include <limits.h>
12
-
13
- #include "cmark_ctype.h"
14
- #include "syntax_extension.h"
15
- #include "config.h"
16
- #include "parser.h"
17
- #include "cmark-gfm.h"
18
- #include "node.h"
19
- #include "references.h"
20
- #include "utf8.h"
21
- #include "scanners.h"
22
- #include "inlines.h"
23
- #include "houdini.h"
24
- #include "buffer.h"
25
- #include "footnotes.h"
26
-
27
- #define CODE_INDENT 4
28
- #define TAB_STOP 4
29
-
30
- /**
31
- * Very deeply nested lists can cause quadratic performance issues.
32
- * This constant is used in open_new_blocks() to limit the nesting
33
- * depth. It is unlikely that a non-contrived markdown document will
34
- * be nested this deeply.
35
- */
36
- #define MAX_LIST_DEPTH 100
37
-
38
- #ifndef MIN
39
- #define MIN(x, y) ((x < y) ? x : y)
40
- #endif
41
-
42
- #define peek_at(i, n) (i)->data[n]
43
-
44
- static bool S_last_line_blank(const cmark_node *node) {
45
- return (node->flags & CMARK_NODE__LAST_LINE_BLANK) != 0;
46
- }
47
-
48
- static bool S_last_line_checked(const cmark_node *node) {
49
- return (node->flags & CMARK_NODE__LAST_LINE_CHECKED) != 0;
50
- }
51
-
52
- static CMARK_INLINE cmark_node_type S_type(const cmark_node *node) {
53
- return (cmark_node_type)node->type;
54
- }
55
-
56
- static void S_set_last_line_blank(cmark_node *node, bool is_blank) {
57
- if (is_blank)
58
- node->flags |= CMARK_NODE__LAST_LINE_BLANK;
59
- else
60
- node->flags &= ~CMARK_NODE__LAST_LINE_BLANK;
61
- }
62
-
63
- static void S_set_last_line_checked(cmark_node *node) {
64
- node->flags |= CMARK_NODE__LAST_LINE_CHECKED;
65
- }
66
-
67
- static CMARK_INLINE bool S_is_line_end_char(char c) {
68
- return (c == '\n' || c == '\r');
69
- }
70
-
71
- static CMARK_INLINE bool S_is_space_or_tab(char c) {
72
- return (c == ' ' || c == '\t');
73
- }
74
-
75
- static void S_parser_feed(cmark_parser *parser, const unsigned char *buffer,
76
- size_t len, bool eof);
77
-
78
- static void S_process_line(cmark_parser *parser, const unsigned char *buffer,
79
- bufsize_t bytes);
80
-
81
- static cmark_node *make_block(cmark_mem *mem, cmark_node_type tag,
82
- int start_line, int start_column) {
83
- cmark_node *e;
84
-
85
- e = (cmark_node *)mem->calloc(1, sizeof(*e));
86
- cmark_strbuf_init(mem, &e->content, 32);
87
- e->type = (uint16_t)tag;
88
- e->flags = CMARK_NODE__OPEN;
89
- e->start_line = start_line;
90
- e->start_column = start_column;
91
- e->end_line = start_line;
92
-
93
- return e;
94
- }
95
-
96
- // Create a root document node.
97
- static cmark_node *make_document(cmark_mem *mem) {
98
- cmark_node *e = make_block(mem, CMARK_NODE_DOCUMENT, 1, 1);
99
- return e;
100
- }
101
-
102
- int cmark_parser_attach_syntax_extension(cmark_parser *parser,
103
- cmark_syntax_extension *extension) {
104
- parser->syntax_extensions = cmark_llist_append(parser->mem, parser->syntax_extensions, extension);
105
- if (extension->match_inline || extension->insert_inline_from_delim) {
106
- parser->inline_syntax_extensions = cmark_llist_append(
107
- parser->mem, parser->inline_syntax_extensions, extension);
108
- }
109
-
110
- return 1;
111
- }
112
-
113
- static void cmark_parser_dispose(cmark_parser *parser) {
114
- if (parser->root)
115
- cmark_node_free(parser->root);
116
-
117
- if (parser->refmap)
118
- cmark_map_free(parser->refmap);
119
- }
120
-
121
- static void cmark_parser_reset(cmark_parser *parser) {
122
- cmark_llist *saved_exts = parser->syntax_extensions;
123
- cmark_llist *saved_inline_exts = parser->inline_syntax_extensions;
124
- int saved_options = parser->options;
125
- cmark_mem *saved_mem = parser->mem;
126
-
127
- cmark_parser_dispose(parser);
128
-
129
- memset(parser, 0, sizeof(cmark_parser));
130
- parser->mem = saved_mem;
131
-
132
- cmark_strbuf_init(parser->mem, &parser->curline, 256);
133
- cmark_strbuf_init(parser->mem, &parser->linebuf, 0);
134
-
135
- cmark_node *document = make_document(parser->mem);
136
-
137
- parser->refmap = cmark_reference_map_new(parser->mem);
138
- parser->root = document;
139
- parser->current = document;
140
-
141
- parser->syntax_extensions = saved_exts;
142
- parser->inline_syntax_extensions = saved_inline_exts;
143
- parser->options = saved_options;
144
- }
145
-
146
- cmark_parser *cmark_parser_new_with_mem(int options, cmark_mem *mem) {
147
- cmark_parser *parser = (cmark_parser *)mem->calloc(1, sizeof(cmark_parser));
148
- parser->mem = mem;
149
- parser->options = options;
150
- cmark_parser_reset(parser);
151
- return parser;
152
- }
153
-
154
- cmark_parser *cmark_parser_new(int options) {
155
- extern cmark_mem CMARK_DEFAULT_MEM_ALLOCATOR;
156
- return cmark_parser_new_with_mem(options, &CMARK_DEFAULT_MEM_ALLOCATOR);
157
- }
158
-
159
- void cmark_parser_free(cmark_parser *parser) {
160
- cmark_mem *mem = parser->mem;
161
- cmark_parser_dispose(parser);
162
- cmark_strbuf_free(&parser->curline);
163
- cmark_strbuf_free(&parser->linebuf);
164
- cmark_llist_free(parser->mem, parser->syntax_extensions);
165
- cmark_llist_free(parser->mem, parser->inline_syntax_extensions);
166
- mem->free(parser);
167
- }
168
-
169
- static cmark_node *finalize(cmark_parser *parser, cmark_node *b);
170
-
171
- // Returns true if line has only space characters, else false.
172
- static bool is_blank(cmark_strbuf *s, bufsize_t offset) {
173
- while (offset < s->size) {
174
- switch (s->ptr[offset]) {
175
- case '\r':
176
- case '\n':
177
- return true;
178
- case ' ':
179
- offset++;
180
- break;
181
- case '\t':
182
- offset++;
183
- break;
184
- default:
185
- return false;
186
- }
187
- }
188
-
189
- return true;
190
- }
191
-
192
- static CMARK_INLINE bool accepts_lines(cmark_node_type block_type) {
193
- return (block_type == CMARK_NODE_PARAGRAPH ||
194
- block_type == CMARK_NODE_HEADING ||
195
- block_type == CMARK_NODE_CODE_BLOCK);
196
- }
197
-
198
- static CMARK_INLINE bool contains_inlines(cmark_node *node) {
199
- if (node->extension && node->extension->contains_inlines_func) {
200
- return node->extension->contains_inlines_func(node->extension, node) != 0;
201
- }
202
-
203
- return (node->type == CMARK_NODE_PARAGRAPH ||
204
- node->type == CMARK_NODE_HEADING);
205
- }
206
-
207
- static void add_line(cmark_node *node, cmark_chunk *ch, cmark_parser *parser) {
208
- int chars_to_tab;
209
- int i;
210
- assert(node->flags & CMARK_NODE__OPEN);
211
- if (parser->partially_consumed_tab) {
212
- parser->offset += 1; // skip over tab
213
- // add space characters:
214
- chars_to_tab = TAB_STOP - (parser->column % TAB_STOP);
215
- for (i = 0; i < chars_to_tab; i++) {
216
- cmark_strbuf_putc(&node->content, ' ');
217
- }
218
- }
219
- cmark_strbuf_put(&node->content, ch->data + parser->offset,
220
- ch->len - parser->offset);
221
- }
222
-
223
- static void remove_trailing_blank_lines(cmark_strbuf *ln) {
224
- bufsize_t i;
225
- unsigned char c;
226
-
227
- for (i = ln->size - 1; i >= 0; --i) {
228
- c = ln->ptr[i];
229
-
230
- if (c != ' ' && c != '\t' && !S_is_line_end_char(c))
231
- break;
232
- }
233
-
234
- if (i < 0) {
235
- cmark_strbuf_clear(ln);
236
- return;
237
- }
238
-
239
- for (; i < ln->size; ++i) {
240
- c = ln->ptr[i];
241
-
242
- if (!S_is_line_end_char(c))
243
- continue;
244
-
245
- cmark_strbuf_truncate(ln, i);
246
- break;
247
- }
248
- }
249
-
250
- // Check to see if a node ends with a blank line, descending
251
- // if needed into lists and sublists.
252
- static bool S_ends_with_blank_line(cmark_node *node) {
253
- if (S_last_line_checked(node)) {
254
- return(S_last_line_blank(node));
255
- } else if ((S_type(node) == CMARK_NODE_LIST ||
256
- S_type(node) == CMARK_NODE_ITEM) && node->last_child) {
257
- S_set_last_line_checked(node);
258
- return(S_ends_with_blank_line(node->last_child));
259
- } else {
260
- S_set_last_line_checked(node);
261
- return (S_last_line_blank(node));
262
- }
263
- }
264
-
265
- // returns true if content remains after link defs are resolved.
266
- static bool resolve_reference_link_definitions(
267
- cmark_parser *parser,
268
- cmark_node *b) {
269
- bufsize_t pos;
270
- cmark_strbuf *node_content = &b->content;
271
- cmark_chunk chunk = {node_content->ptr, node_content->size, 0};
272
- while (chunk.len && chunk.data[0] == '[' &&
273
- (pos = cmark_parse_reference_inline(parser->mem, &chunk,
274
- parser->refmap))) {
275
-
276
- chunk.data += pos;
277
- chunk.len -= pos;
278
- }
279
- cmark_strbuf_drop(node_content, (node_content->size - chunk.len));
280
- return !is_blank(&b->content, 0);
281
- }
282
-
283
- static cmark_node *finalize(cmark_parser *parser, cmark_node *b) {
284
- bufsize_t pos;
285
- cmark_node *item;
286
- cmark_node *subitem;
287
- cmark_node *parent;
288
- bool has_content;
289
-
290
- parent = b->parent;
291
- assert(b->flags &
292
- CMARK_NODE__OPEN); // shouldn't call finalize on closed blocks
293
- b->flags &= ~CMARK_NODE__OPEN;
294
-
295
- if (parser->curline.size == 0) {
296
- // end of input - line number has not been incremented
297
- b->end_line = parser->line_number;
298
- b->end_column = parser->last_line_length;
299
- } else if (S_type(b) == CMARK_NODE_DOCUMENT ||
300
- (S_type(b) == CMARK_NODE_CODE_BLOCK && b->as.code.fenced) ||
301
- (S_type(b) == CMARK_NODE_HEADING && b->as.heading.setext)) {
302
- b->end_line = parser->line_number;
303
- b->end_column = parser->curline.size;
304
- if (b->end_column && parser->curline.ptr[b->end_column - 1] == '\n')
305
- b->end_column -= 1;
306
- if (b->end_column && parser->curline.ptr[b->end_column - 1] == '\r')
307
- b->end_column -= 1;
308
- } else {
309
- b->end_line = parser->line_number - 1;
310
- b->end_column = parser->last_line_length;
311
- }
312
-
313
- cmark_strbuf *node_content = &b->content;
314
-
315
- switch (S_type(b)) {
316
- case CMARK_NODE_PARAGRAPH:
317
- {
318
- has_content = resolve_reference_link_definitions(parser, b);
319
- if (!has_content) {
320
- // remove blank node (former reference def)
321
- cmark_node_free(b);
322
- }
323
- break;
324
- }
325
-
326
- case CMARK_NODE_CODE_BLOCK:
327
- if (!b->as.code.fenced) { // indented code
328
- remove_trailing_blank_lines(node_content);
329
- cmark_strbuf_putc(node_content, '\n');
330
- } else {
331
- // first line of contents becomes info
332
- for (pos = 0; pos < node_content->size; ++pos) {
333
- if (S_is_line_end_char(node_content->ptr[pos]))
334
- break;
335
- }
336
- assert(pos < node_content->size);
337
-
338
- cmark_strbuf tmp = CMARK_BUF_INIT(parser->mem);
339
- houdini_unescape_html_f(&tmp, node_content->ptr, pos);
340
- cmark_strbuf_trim(&tmp);
341
- cmark_strbuf_unescape(&tmp);
342
- b->as.code.info = cmark_chunk_buf_detach(&tmp);
343
-
344
- if (node_content->ptr[pos] == '\r')
345
- pos += 1;
346
- if (node_content->ptr[pos] == '\n')
347
- pos += 1;
348
- cmark_strbuf_drop(node_content, pos);
349
- }
350
- b->as.code.literal = cmark_chunk_buf_detach(node_content);
351
- break;
352
-
353
- case CMARK_NODE_HTML_BLOCK:
354
- b->as.literal = cmark_chunk_buf_detach(node_content);
355
- break;
356
-
357
- case CMARK_NODE_LIST: // determine tight/loose status
358
- b->as.list.tight = true; // tight by default
359
- item = b->first_child;
360
-
361
- while (item) {
362
- // check for non-final non-empty list item ending with blank line:
363
- if (S_last_line_blank(item) && item->next) {
364
- b->as.list.tight = false;
365
- break;
366
- }
367
- // recurse into children of list item, to see if there are
368
- // spaces between them:
369
- subitem = item->first_child;
370
- while (subitem) {
371
- if ((item->next || subitem->next) &&
372
- S_ends_with_blank_line(subitem)) {
373
- b->as.list.tight = false;
374
- break;
375
- }
376
- subitem = subitem->next;
377
- }
378
- if (!(b->as.list.tight)) {
379
- break;
380
- }
381
- item = item->next;
382
- }
383
-
384
- break;
385
-
386
- default:
387
- break;
388
- }
389
-
390
- return parent;
391
- }
392
-
393
- // Add a node as child of another. Return pointer to child.
394
- static cmark_node *add_child(cmark_parser *parser, cmark_node *parent,
395
- cmark_node_type block_type, int start_column) {
396
- assert(parent);
397
-
398
- // if 'parent' isn't the kind of node that can accept this child,
399
- // then back up til we hit a node that can.
400
- while (!cmark_node_can_contain_type(parent, block_type)) {
401
- parent = finalize(parser, parent);
402
- }
403
-
404
- cmark_node *child =
405
- make_block(parser->mem, block_type, parser->line_number, start_column);
406
- child->parent = parent;
407
-
408
- if (parent->last_child) {
409
- parent->last_child->next = child;
410
- child->prev = parent->last_child;
411
- } else {
412
- parent->first_child = child;
413
- child->prev = NULL;
414
- }
415
- parent->last_child = child;
416
- return child;
417
- }
418
-
419
- void cmark_manage_extensions_special_characters(cmark_parser *parser, int add) {
420
- cmark_llist *tmp_ext;
421
-
422
- for (tmp_ext = parser->inline_syntax_extensions; tmp_ext; tmp_ext=tmp_ext->next) {
423
- cmark_syntax_extension *ext = (cmark_syntax_extension *) tmp_ext->data;
424
- cmark_llist *tmp_char;
425
- for (tmp_char = ext->special_inline_chars; tmp_char; tmp_char=tmp_char->next) {
426
- unsigned char c = (unsigned char)(size_t)tmp_char->data;
427
- if (add)
428
- cmark_inlines_add_special_character(c, ext->emphasis);
429
- else
430
- cmark_inlines_remove_special_character(c, ext->emphasis);
431
- }
432
- }
433
- }
434
-
435
- // Walk through node and all children, recursively, parsing
436
- // string content into inline content where appropriate.
437
- static void process_inlines(cmark_parser *parser,
438
- cmark_map *refmap, int options) {
439
- cmark_iter *iter = cmark_iter_new(parser->root);
440
- cmark_node *cur;
441
- cmark_event_type ev_type;
442
-
443
- cmark_manage_extensions_special_characters(parser, true);
444
-
445
- while ((ev_type = cmark_iter_next(iter)) != CMARK_EVENT_DONE) {
446
- cur = cmark_iter_get_node(iter);
447
- if (ev_type == CMARK_EVENT_ENTER) {
448
- if (contains_inlines(cur)) {
449
- cmark_parse_inlines(parser, cur, refmap, options);
450
- }
451
- }
452
- }
453
-
454
- cmark_manage_extensions_special_characters(parser, false);
455
-
456
- cmark_iter_free(iter);
457
- }
458
-
459
- static int sort_footnote_by_ix(const void *_a, const void *_b) {
460
- cmark_footnote *a = *(cmark_footnote **)_a;
461
- cmark_footnote *b = *(cmark_footnote **)_b;
462
- return (int)a->ix - (int)b->ix;
463
- }
464
-
465
- static void process_footnotes(cmark_parser *parser) {
466
- // * Collect definitions in a map.
467
- // * Iterate the references in the document in order, assigning indices to
468
- // definitions in the order they're seen.
469
- // * Write out the footnotes at the bottom of the document in index order.
470
-
471
- cmark_map *map = cmark_footnote_map_new(parser->mem);
472
-
473
- cmark_iter *iter = cmark_iter_new(parser->root);
474
- cmark_node *cur;
475
- cmark_event_type ev_type;
476
-
477
- while ((ev_type = cmark_iter_next(iter)) != CMARK_EVENT_DONE) {
478
- cur = cmark_iter_get_node(iter);
479
- if (ev_type == CMARK_EVENT_EXIT && cur->type == CMARK_NODE_FOOTNOTE_DEFINITION) {
480
- cmark_footnote_create(map, cur);
481
- }
482
- }
483
-
484
- cmark_iter_free(iter);
485
- iter = cmark_iter_new(parser->root);
486
- unsigned int ix = 0;
487
-
488
- while ((ev_type = cmark_iter_next(iter)) != CMARK_EVENT_DONE) {
489
- cur = cmark_iter_get_node(iter);
490
- if (ev_type == CMARK_EVENT_EXIT && cur->type == CMARK_NODE_FOOTNOTE_REFERENCE) {
491
- cmark_footnote *footnote = (cmark_footnote *)cmark_map_lookup(map, &cur->as.literal);
492
- if (footnote) {
493
- if (!footnote->ix)
494
- footnote->ix = ++ix;
495
-
496
- // store a reference to this footnote reference's footnote definition
497
- // this is used by renderers when generating label ids
498
- cur->parent_footnote_def = footnote->node;
499
-
500
- // keep track of a) count of how many times this footnote def has been
501
- // referenced, and b) which reference index this footnote ref is at.
502
- // this is used by renderers when generating links and backreferences.
503
- cur->footnote.ref_ix = ++footnote->node->footnote.def_count;
504
-
505
- char n[32];
506
- snprintf(n, sizeof(n), "%d", footnote->ix);
507
- cmark_chunk_free(parser->mem, &cur->as.literal);
508
- cmark_strbuf buf = CMARK_BUF_INIT(parser->mem);
509
- cmark_strbuf_puts(&buf, n);
510
-
511
- cur->as.literal = cmark_chunk_buf_detach(&buf);
512
- } else {
513
- cmark_node *text = (cmark_node *)parser->mem->calloc(1, sizeof(*text));
514
- cmark_strbuf_init(parser->mem, &text->content, 0);
515
- text->type = (uint16_t) CMARK_NODE_TEXT;
516
-
517
- cmark_strbuf buf = CMARK_BUF_INIT(parser->mem);
518
- cmark_strbuf_puts(&buf, "[^");
519
- cmark_strbuf_put(&buf, cur->as.literal.data, cur->as.literal.len);
520
- cmark_strbuf_putc(&buf, ']');
521
-
522
- text->as.literal = cmark_chunk_buf_detach(&buf);
523
- cmark_node_insert_after(cur, text);
524
- cmark_node_free(cur);
525
- }
526
- }
527
- }
528
-
529
- cmark_iter_free(iter);
530
-
531
- if (map->sorted) {
532
- qsort(map->sorted, map->size, sizeof(cmark_map_entry *), sort_footnote_by_ix);
533
- for (unsigned int i = 0; i < map->size; ++i) {
534
- cmark_footnote *footnote = (cmark_footnote *)map->sorted[i];
535
- if (!footnote->ix) {
536
- cmark_node_unlink(footnote->node);
537
- continue;
538
- }
539
- cmark_node_append_child(parser->root, footnote->node);
540
- footnote->node = NULL;
541
- }
542
- }
543
-
544
- cmark_unlink_footnotes_map(map);
545
- cmark_map_free(map);
546
- }
547
-
548
- // Attempts to parse a list item marker (bullet or enumerated).
549
- // On success, returns length of the marker, and populates
550
- // data with the details. On failure, returns 0.
551
- static bufsize_t parse_list_marker(cmark_mem *mem, cmark_chunk *input,
552
- bufsize_t pos, bool interrupts_paragraph,
553
- cmark_list **dataptr) {
554
- unsigned char c;
555
- bufsize_t startpos;
556
- cmark_list *data;
557
- bufsize_t i;
558
-
559
- startpos = pos;
560
- c = peek_at(input, pos);
561
-
562
- if (c == '*' || c == '-' || c == '+') {
563
- pos++;
564
- if (!cmark_isspace(peek_at(input, pos))) {
565
- return 0;
566
- }
567
-
568
- if (interrupts_paragraph) {
569
- i = pos;
570
- // require non-blank content after list marker:
571
- while (S_is_space_or_tab(peek_at(input, i))) {
572
- i++;
573
- }
574
- if (peek_at(input, i) == '\n') {
575
- return 0;
576
- }
577
- }
578
-
579
- data = (cmark_list *)mem->calloc(1, sizeof(*data));
580
- data->marker_offset = 0; // will be adjusted later
581
- data->list_type = CMARK_BULLET_LIST;
582
- data->bullet_char = c;
583
- data->start = 0;
584
- data->delimiter = CMARK_NO_DELIM;
585
- data->tight = false;
586
- } else if (cmark_isdigit(c)) {
587
- int start = 0;
588
- int digits = 0;
589
-
590
- do {
591
- start = (10 * start) + (peek_at(input, pos) - '0');
592
- pos++;
593
- digits++;
594
- // We limit to 9 digits to avoid overflow,
595
- // assuming max int is 2^31 - 1
596
- // This also seems to be the limit for 'start' in some browsers.
597
- } while (digits < 9 && cmark_isdigit(peek_at(input, pos)));
598
-
599
- if (interrupts_paragraph && start != 1) {
600
- return 0;
601
- }
602
- c = peek_at(input, pos);
603
- if (c == '.' || c == ')') {
604
- pos++;
605
- if (!cmark_isspace(peek_at(input, pos))) {
606
- return 0;
607
- }
608
- if (interrupts_paragraph) {
609
- // require non-blank content after list marker:
610
- i = pos;
611
- while (S_is_space_or_tab(peek_at(input, i))) {
612
- i++;
613
- }
614
- if (S_is_line_end_char(peek_at(input, i))) {
615
- return 0;
616
- }
617
- }
618
-
619
- data = (cmark_list *)mem->calloc(1, sizeof(*data));
620
- data->marker_offset = 0; // will be adjusted later
621
- data->list_type = CMARK_ORDERED_LIST;
622
- data->bullet_char = 0;
623
- data->start = start;
624
- data->delimiter = (c == '.' ? CMARK_PERIOD_DELIM : CMARK_PAREN_DELIM);
625
- data->tight = false;
626
- } else {
627
- return 0;
628
- }
629
- } else {
630
- return 0;
631
- }
632
-
633
- *dataptr = data;
634
- return (pos - startpos);
635
- }
636
-
637
- // Return 1 if list item belongs in list, else 0.
638
- static int lists_match(cmark_list *list_data, cmark_list *item_data) {
639
- return (list_data->list_type == item_data->list_type &&
640
- list_data->delimiter == item_data->delimiter &&
641
- // list_data->marker_offset == item_data.marker_offset &&
642
- list_data->bullet_char == item_data->bullet_char);
643
- }
644
-
645
- static cmark_node *finalize_document(cmark_parser *parser) {
646
- while (parser->current != parser->root) {
647
- parser->current = finalize(parser, parser->current);
648
- }
649
-
650
- finalize(parser, parser->root);
651
-
652
- // Limit total size of extra content created from reference links to
653
- // document size to avoid superlinear growth. Always allow 100KB.
654
- if (parser->total_size > 100000)
655
- parser->refmap->max_ref_size = parser->total_size;
656
- else
657
- parser->refmap->max_ref_size = 100000;
658
-
659
- process_inlines(parser, parser->refmap, parser->options);
660
- if (parser->options & CMARK_OPT_FOOTNOTES)
661
- process_footnotes(parser);
662
-
663
- return parser->root;
664
- }
665
-
666
- cmark_node *cmark_parse_file(FILE *f, int options) {
667
- unsigned char buffer[4096];
668
- cmark_parser *parser = cmark_parser_new(options);
669
- size_t bytes;
670
- cmark_node *document;
671
-
672
- while ((bytes = fread(buffer, 1, sizeof(buffer), f)) > 0) {
673
- bool eof = bytes < sizeof(buffer);
674
- S_parser_feed(parser, buffer, bytes, eof);
675
- if (eof) {
676
- break;
677
- }
678
- }
679
-
680
- document = cmark_parser_finish(parser);
681
- cmark_parser_free(parser);
682
- return document;
683
- }
684
-
685
- cmark_node *cmark_parse_document(const char *buffer, size_t len, int options) {
686
- cmark_parser *parser = cmark_parser_new(options);
687
- cmark_node *document;
688
-
689
- S_parser_feed(parser, (const unsigned char *)buffer, len, true);
690
-
691
- document = cmark_parser_finish(parser);
692
- cmark_parser_free(parser);
693
- return document;
694
- }
695
-
696
- void cmark_parser_feed(cmark_parser *parser, const char *buffer, size_t len) {
697
- S_parser_feed(parser, (const unsigned char *)buffer, len, false);
698
- }
699
-
700
- void cmark_parser_feed_reentrant(cmark_parser *parser, const char *buffer, size_t len) {
701
- cmark_strbuf saved_linebuf;
702
-
703
- cmark_strbuf_init(parser->mem, &saved_linebuf, 0);
704
- cmark_strbuf_puts(&saved_linebuf, cmark_strbuf_cstr(&parser->linebuf));
705
- cmark_strbuf_clear(&parser->linebuf);
706
-
707
- S_parser_feed(parser, (const unsigned char *)buffer, len, true);
708
-
709
- cmark_strbuf_sets(&parser->linebuf, cmark_strbuf_cstr(&saved_linebuf));
710
- cmark_strbuf_free(&saved_linebuf);
711
- }
712
-
713
- static void S_parser_feed(cmark_parser *parser, const unsigned char *buffer,
714
- size_t len, bool eof) {
715
- const unsigned char *end = buffer + len;
716
- static const uint8_t repl[] = {239, 191, 189};
717
-
718
- if (len > UINT_MAX - parser->total_size)
719
- parser->total_size = UINT_MAX;
720
- else
721
- parser->total_size += len;
722
-
723
- if (parser->last_buffer_ended_with_cr && *buffer == '\n') {
724
- // skip NL if last buffer ended with CR ; see #117
725
- buffer++;
726
- }
727
- parser->last_buffer_ended_with_cr = false;
728
- while (buffer < end) {
729
- const unsigned char *eol;
730
- bufsize_t chunk_len;
731
- bool process = false;
732
- for (eol = buffer; eol < end; ++eol) {
733
- if (S_is_line_end_char(*eol)) {
734
- process = true;
735
- break;
736
- }
737
- if (*eol == '\0' && eol < end) {
738
- break;
739
- }
740
- }
741
- if (eol >= end && eof) {
742
- process = true;
743
- }
744
-
745
- chunk_len = (bufsize_t)(eol - buffer);
746
- if (process) {
747
- if (parser->linebuf.size > 0) {
748
- cmark_strbuf_put(&parser->linebuf, buffer, chunk_len);
749
- S_process_line(parser, parser->linebuf.ptr, parser->linebuf.size);
750
- cmark_strbuf_clear(&parser->linebuf);
751
- } else {
752
- S_process_line(parser, buffer, chunk_len);
753
- }
754
- } else {
755
- if (eol < end && *eol == '\0') {
756
- // omit NULL byte
757
- cmark_strbuf_put(&parser->linebuf, buffer, chunk_len);
758
- // add replacement character
759
- cmark_strbuf_put(&parser->linebuf, repl, 3);
760
- } else {
761
- cmark_strbuf_put(&parser->linebuf, buffer, chunk_len);
762
- }
763
- }
764
-
765
- buffer += chunk_len;
766
- if (buffer < end) {
767
- if (*buffer == '\0') {
768
- // skip over NULL
769
- buffer++;
770
- } else {
771
- // skip over line ending characters
772
- if (*buffer == '\r') {
773
- buffer++;
774
- if (buffer == end)
775
- parser->last_buffer_ended_with_cr = true;
776
- }
777
- if (buffer < end && *buffer == '\n')
778
- buffer++;
779
- }
780
- }
781
- }
782
- }
783
-
784
- static void chop_trailing_hashtags(cmark_chunk *ch) {
785
- bufsize_t n, orig_n;
786
-
787
- cmark_chunk_rtrim(ch);
788
- orig_n = n = ch->len - 1;
789
-
790
- // if string ends in space followed by #s, remove these:
791
- while (n >= 0 && peek_at(ch, n) == '#')
792
- n--;
793
-
794
- // Check for a space before the final #s:
795
- if (n != orig_n && n >= 0 && S_is_space_or_tab(peek_at(ch, n))) {
796
- ch->len = n;
797
- cmark_chunk_rtrim(ch);
798
- }
799
- }
800
-
801
- // Check for thematic break. On failure, return 0 and update
802
- // thematic_break_kill_pos with the index at which the
803
- // parse fails. On success, return length of match.
804
- // "...three or more hyphens, asterisks,
805
- // or underscores on a line by themselves. If you wish, you may use
806
- // spaces between the hyphens or asterisks."
807
- static int S_scan_thematic_break(cmark_parser *parser, cmark_chunk *input,
808
- bufsize_t offset) {
809
- bufsize_t i;
810
- char c;
811
- char nextc = '\0';
812
- int count;
813
- i = offset;
814
- c = peek_at(input, i);
815
- if (!(c == '*' || c == '_' || c == '-')) {
816
- parser->thematic_break_kill_pos = i;
817
- return 0;
818
- }
819
- count = 1;
820
- while ((nextc = peek_at(input, ++i))) {
821
- if (nextc == c) {
822
- count++;
823
- } else if (nextc != ' ' && nextc != '\t') {
824
- break;
825
- }
826
- }
827
- if (count >= 3 && (nextc == '\r' || nextc == '\n')) {
828
- return (i - offset) + 1;
829
- } else {
830
- parser->thematic_break_kill_pos = i;
831
- return 0;
832
- }
833
- }
834
-
835
- // Find first nonspace character from current offset, setting
836
- // parser->first_nonspace, parser->first_nonspace_column,
837
- // parser->indent, and parser->blank. Does not advance parser->offset.
838
- static void S_find_first_nonspace(cmark_parser *parser, cmark_chunk *input) {
839
- char c;
840
- int chars_to_tab = TAB_STOP - (parser->column % TAB_STOP);
841
-
842
- if (parser->first_nonspace <= parser->offset) {
843
- parser->first_nonspace = parser->offset;
844
- parser->first_nonspace_column = parser->column;
845
- while ((c = peek_at(input, parser->first_nonspace))) {
846
- if (c == ' ') {
847
- parser->first_nonspace += 1;
848
- parser->first_nonspace_column += 1;
849
- chars_to_tab = chars_to_tab - 1;
850
- if (chars_to_tab == 0) {
851
- chars_to_tab = TAB_STOP;
852
- }
853
- } else if (c == '\t') {
854
- parser->first_nonspace += 1;
855
- parser->first_nonspace_column += chars_to_tab;
856
- chars_to_tab = TAB_STOP;
857
- } else {
858
- break;
859
- }
860
- }
861
- }
862
-
863
- parser->indent = parser->first_nonspace_column - parser->column;
864
- parser->blank = S_is_line_end_char(peek_at(input, parser->first_nonspace));
865
- }
866
-
867
- // Advance parser->offset and parser->column. parser->offset is the
868
- // byte position in input; parser->column is a virtual column number
869
- // that takes into account tabs. (Multibyte characters are not taken
870
- // into account, because the Markdown line prefixes we are interested in
871
- // analyzing are entirely ASCII.) The count parameter indicates
872
- // how far to advance the offset. If columns is true, then count
873
- // indicates a number of columns; otherwise, a number of bytes.
874
- // If advancing a certain number of columns partially consumes
875
- // a tab character, parser->partially_consumed_tab is set to true.
876
- static void S_advance_offset(cmark_parser *parser, cmark_chunk *input,
877
- bufsize_t count, bool columns) {
878
- char c;
879
- int chars_to_tab;
880
- int chars_to_advance;
881
- while (count > 0 && (c = peek_at(input, parser->offset))) {
882
- if (c == '\t') {
883
- chars_to_tab = TAB_STOP - (parser->column % TAB_STOP);
884
- if (columns) {
885
- parser->partially_consumed_tab = chars_to_tab > count;
886
- chars_to_advance = MIN(count, chars_to_tab);
887
- parser->column += chars_to_advance;
888
- parser->offset += (parser->partially_consumed_tab ? 0 : 1);
889
- count -= chars_to_advance;
890
- } else {
891
- parser->partially_consumed_tab = false;
892
- parser->column += chars_to_tab;
893
- parser->offset += 1;
894
- count -= 1;
895
- }
896
- } else {
897
- parser->partially_consumed_tab = false;
898
- parser->offset += 1;
899
- parser->column += 1; // assume ascii; block starts are ascii
900
- count -= 1;
901
- }
902
- }
903
- }
904
-
905
- static bool S_last_child_is_open(cmark_node *container) {
906
- return container->last_child &&
907
- (container->last_child->flags & CMARK_NODE__OPEN);
908
- }
909
-
910
- static bool parse_block_quote_prefix(cmark_parser *parser, cmark_chunk *input) {
911
- bool res = false;
912
- bufsize_t matched = 0;
913
-
914
- matched =
915
- parser->indent <= 3 && peek_at(input, parser->first_nonspace) == '>';
916
- if (matched) {
917
-
918
- S_advance_offset(parser, input, parser->indent + 1, true);
919
-
920
- if (S_is_space_or_tab(peek_at(input, parser->offset))) {
921
- S_advance_offset(parser, input, 1, true);
922
- }
923
-
924
- res = true;
925
- }
926
- return res;
927
- }
928
-
929
- static bool parse_footnote_definition_block_prefix(cmark_parser *parser, cmark_chunk *input,
930
- cmark_node *container) {
931
- if (parser->indent >= 4) {
932
- S_advance_offset(parser, input, 4, true);
933
- return true;
934
- } else if (input->len > 0 && (input->data[0] == '\n' || (input->data[0] == '\r' && input->data[1] == '\n'))) {
935
- return true;
936
- }
937
-
938
- return false;
939
- }
940
-
941
- static bool parse_node_item_prefix(cmark_parser *parser, cmark_chunk *input,
942
- cmark_node *container) {
943
- bool res = false;
944
-
945
- if (parser->indent >=
946
- container->as.list.marker_offset + container->as.list.padding) {
947
- S_advance_offset(parser, input, container->as.list.marker_offset +
948
- container->as.list.padding,
949
- true);
950
- res = true;
951
- } else if (parser->blank && container->first_child != NULL) {
952
- // if container->first_child is NULL, then the opening line
953
- // of the list item was blank after the list marker; in this
954
- // case, we are done with the list item.
955
- S_advance_offset(parser, input, parser->first_nonspace - parser->offset,
956
- false);
957
- res = true;
958
- }
959
- return res;
960
- }
961
-
962
- static bool parse_code_block_prefix(cmark_parser *parser, cmark_chunk *input,
963
- cmark_node *container,
964
- bool *should_continue) {
965
- bool res = false;
966
-
967
- if (!container->as.code.fenced) { // indented
968
- if (parser->indent >= CODE_INDENT) {
969
- S_advance_offset(parser, input, CODE_INDENT, true);
970
- res = true;
971
- } else if (parser->blank) {
972
- S_advance_offset(parser, input, parser->first_nonspace - parser->offset,
973
- false);
974
- res = true;
975
- }
976
- } else { // fenced
977
- bufsize_t matched = 0;
978
-
979
- if (parser->indent <= 3 && (peek_at(input, parser->first_nonspace) ==
980
- container->as.code.fence_char)) {
981
- matched = scan_close_code_fence(input, parser->first_nonspace);
982
- }
983
-
984
- if (matched >= container->as.code.fence_length) {
985
- // closing fence - and since we're at
986
- // the end of a line, we can stop processing it:
987
- *should_continue = false;
988
- S_advance_offset(parser, input, matched, false);
989
- parser->current = finalize(parser, container);
990
- } else {
991
- // skip opt. spaces of fence parser->offset
992
- int i = container->as.code.fence_offset;
993
-
994
- while (i > 0 && S_is_space_or_tab(peek_at(input, parser->offset))) {
995
- S_advance_offset(parser, input, 1, true);
996
- i--;
997
- }
998
- res = true;
999
- }
1000
- }
1001
-
1002
- return res;
1003
- }
1004
-
1005
- static bool parse_html_block_prefix(cmark_parser *parser,
1006
- cmark_node *container) {
1007
- bool res = false;
1008
- int html_block_type = container->as.html_block_type;
1009
-
1010
- assert(html_block_type >= 1 && html_block_type <= 7);
1011
- switch (html_block_type) {
1012
- case 1:
1013
- case 2:
1014
- case 3:
1015
- case 4:
1016
- case 5:
1017
- // these types of blocks can accept blanks
1018
- res = true;
1019
- break;
1020
- case 6:
1021
- case 7:
1022
- res = !parser->blank;
1023
- break;
1024
- }
1025
-
1026
- return res;
1027
- }
1028
-
1029
- static bool parse_extension_block(cmark_parser *parser,
1030
- cmark_node *container,
1031
- cmark_chunk *input)
1032
- {
1033
- bool res = false;
1034
-
1035
- if (container->extension->last_block_matches) {
1036
- if (container->extension->last_block_matches(
1037
- container->extension, parser, input->data, input->len, container))
1038
- res = true;
1039
- }
1040
-
1041
- return res;
1042
- }
1043
-
1044
- /**
1045
- * For each containing node, try to parse the associated line start.
1046
- *
1047
- * Will not close unmatched blocks, as we may have a lazy continuation
1048
- * line -> http://spec.commonmark.org/0.24/#lazy-continuation-line
1049
- *
1050
- * Returns: The last matching node, or NULL
1051
- */
1052
- static cmark_node *check_open_blocks(cmark_parser *parser, cmark_chunk *input,
1053
- bool *all_matched) {
1054
- bool should_continue = true;
1055
- *all_matched = false;
1056
- cmark_node *container = parser->root;
1057
- cmark_node_type cont_type;
1058
-
1059
- while (S_last_child_is_open(container)) {
1060
- container = container->last_child;
1061
- cont_type = S_type(container);
1062
-
1063
- S_find_first_nonspace(parser, input);
1064
-
1065
- if (container->extension) {
1066
- if (!parse_extension_block(parser, container, input))
1067
- goto done;
1068
- continue;
1069
- }
1070
-
1071
- switch (cont_type) {
1072
- case CMARK_NODE_BLOCK_QUOTE:
1073
- if (!parse_block_quote_prefix(parser, input))
1074
- goto done;
1075
- break;
1076
- case CMARK_NODE_ITEM:
1077
- if (!parse_node_item_prefix(parser, input, container))
1078
- goto done;
1079
- break;
1080
- case CMARK_NODE_CODE_BLOCK:
1081
- if (!parse_code_block_prefix(parser, input, container, &should_continue))
1082
- goto done;
1083
- break;
1084
- case CMARK_NODE_HEADING:
1085
- // a heading can never contain more than one line
1086
- goto done;
1087
- case CMARK_NODE_HTML_BLOCK:
1088
- if (!parse_html_block_prefix(parser, container))
1089
- goto done;
1090
- break;
1091
- case CMARK_NODE_PARAGRAPH:
1092
- if (parser->blank)
1093
- goto done;
1094
- break;
1095
- case CMARK_NODE_FOOTNOTE_DEFINITION:
1096
- if (!parse_footnote_definition_block_prefix(parser, input, container))
1097
- goto done;
1098
- break;
1099
- default:
1100
- break;
1101
- }
1102
- }
1103
-
1104
- *all_matched = true;
1105
-
1106
- done:
1107
- if (!*all_matched) {
1108
- container = container->parent; // back up to last matching node
1109
- }
1110
-
1111
- if (!should_continue) {
1112
- container = NULL;
1113
- }
1114
-
1115
- return container;
1116
- }
1117
-
1118
- static void open_new_blocks(cmark_parser *parser, cmark_node **container,
1119
- cmark_chunk *input, bool all_matched) {
1120
- bool indented;
1121
- cmark_list *data = NULL;
1122
- bool maybe_lazy = S_type(parser->current) == CMARK_NODE_PARAGRAPH;
1123
- cmark_node_type cont_type = S_type(*container);
1124
- bufsize_t matched = 0;
1125
- int lev = 0;
1126
- bool save_partially_consumed_tab;
1127
- bool has_content;
1128
- int save_offset;
1129
- int save_column;
1130
- size_t depth = 0;
1131
-
1132
- while (cont_type != CMARK_NODE_CODE_BLOCK &&
1133
- cont_type != CMARK_NODE_HTML_BLOCK) {
1134
- depth++;
1135
- S_find_first_nonspace(parser, input);
1136
- indented = parser->indent >= CODE_INDENT;
1137
-
1138
- if (!indented && peek_at(input, parser->first_nonspace) == '>') {
1139
-
1140
- bufsize_t blockquote_startpos = parser->first_nonspace;
1141
-
1142
- S_advance_offset(parser, input,
1143
- parser->first_nonspace + 1 - parser->offset, false);
1144
- // optional following character
1145
- if (S_is_space_or_tab(peek_at(input, parser->offset))) {
1146
- S_advance_offset(parser, input, 1, true);
1147
- }
1148
- *container = add_child(parser, *container, CMARK_NODE_BLOCK_QUOTE,
1149
- blockquote_startpos + 1);
1150
-
1151
- } else if (!indented && (matched = scan_atx_heading_start(
1152
- input, parser->first_nonspace))) {
1153
- bufsize_t hashpos;
1154
- int level = 0;
1155
- bufsize_t heading_startpos = parser->first_nonspace;
1156
-
1157
- S_advance_offset(parser, input,
1158
- parser->first_nonspace + matched - parser->offset,
1159
- false);
1160
- *container = add_child(parser, *container, CMARK_NODE_HEADING,
1161
- heading_startpos + 1);
1162
-
1163
- hashpos = cmark_chunk_strchr(input, '#', parser->first_nonspace);
1164
-
1165
- while (peek_at(input, hashpos) == '#') {
1166
- level++;
1167
- hashpos++;
1168
- }
1169
-
1170
- (*container)->as.heading.level = level;
1171
- (*container)->as.heading.setext = false;
1172
- (*container)->internal_offset = matched;
1173
-
1174
- } else if (!indented && (matched = scan_open_code_fence(
1175
- input, parser->first_nonspace))) {
1176
- *container = add_child(parser, *container, CMARK_NODE_CODE_BLOCK,
1177
- parser->first_nonspace + 1);
1178
- (*container)->as.code.fenced = true;
1179
- (*container)->as.code.fence_char = peek_at(input, parser->first_nonspace);
1180
- (*container)->as.code.fence_length = (matched > 255) ? 255 : (uint8_t)matched;
1181
- (*container)->as.code.fence_offset =
1182
- (int8_t)(parser->first_nonspace - parser->offset);
1183
- (*container)->as.code.info = cmark_chunk_literal("");
1184
- S_advance_offset(parser, input,
1185
- parser->first_nonspace + matched - parser->offset,
1186
- false);
1187
-
1188
- } else if (!indented && ((matched = scan_html_block_start(
1189
- input, parser->first_nonspace)) ||
1190
- (cont_type != CMARK_NODE_PARAGRAPH &&
1191
- (matched = scan_html_block_start_7(
1192
- input, parser->first_nonspace))))) {
1193
- *container = add_child(parser, *container, CMARK_NODE_HTML_BLOCK,
1194
- parser->first_nonspace + 1);
1195
- (*container)->as.html_block_type = matched;
1196
- // note, we don't adjust parser->offset because the tag is part of the
1197
- // text
1198
- } else if (!indented && cont_type == CMARK_NODE_PARAGRAPH &&
1199
- (lev =
1200
- scan_setext_heading_line(input, parser->first_nonspace))) {
1201
- // finalize paragraph, resolving reference links
1202
- has_content = resolve_reference_link_definitions(parser, *container);
1203
-
1204
- if (has_content) {
1205
-
1206
- (*container)->type = (uint16_t)CMARK_NODE_HEADING;
1207
- (*container)->as.heading.level = lev;
1208
- (*container)->as.heading.setext = true;
1209
- S_advance_offset(parser, input, input->len - 1 - parser->offset, false);
1210
- }
1211
- } else if (!indented &&
1212
- !(cont_type == CMARK_NODE_PARAGRAPH && !all_matched) &&
1213
- (parser->thematic_break_kill_pos <= parser->first_nonspace) &&
1214
- (matched = S_scan_thematic_break(parser, input, parser->first_nonspace))) {
1215
- // it's only now that we know the line is not part of a setext heading:
1216
- *container = add_child(parser, *container, CMARK_NODE_THEMATIC_BREAK,
1217
- parser->first_nonspace + 1);
1218
- S_advance_offset(parser, input, input->len - 1 - parser->offset, false);
1219
- } else if (!indented &&
1220
- parser->options & CMARK_OPT_FOOTNOTES &&
1221
- (matched = scan_footnote_definition(input, parser->first_nonspace))) {
1222
- cmark_chunk c = cmark_chunk_dup(input, parser->first_nonspace + 2, matched - 2);
1223
- cmark_chunk_to_cstr(parser->mem, &c);
1224
-
1225
- while (c.data[c.len - 1] != ']')
1226
- --c.len;
1227
- --c.len;
1228
-
1229
- S_advance_offset(parser, input, parser->first_nonspace + matched - parser->offset, false);
1230
- *container = add_child(parser, *container, CMARK_NODE_FOOTNOTE_DEFINITION, parser->first_nonspace + matched + 1);
1231
- (*container)->as.literal = c;
1232
-
1233
- (*container)->internal_offset = matched;
1234
- } else if ((!indented || cont_type == CMARK_NODE_LIST) &&
1235
- parser->indent < 4 &&
1236
- depth < MAX_LIST_DEPTH &&
1237
- (matched = parse_list_marker(
1238
- parser->mem, input, parser->first_nonspace,
1239
- (*container)->type == CMARK_NODE_PARAGRAPH, &data))) {
1240
-
1241
- // Note that we can have new list items starting with >= 4
1242
- // spaces indent, as long as the list container is still open.
1243
- int i = 0;
1244
-
1245
- // compute padding:
1246
- S_advance_offset(parser, input,
1247
- parser->first_nonspace + matched - parser->offset,
1248
- false);
1249
-
1250
- save_partially_consumed_tab = parser->partially_consumed_tab;
1251
- save_offset = parser->offset;
1252
- save_column = parser->column;
1253
-
1254
- while (parser->column - save_column <= 5 &&
1255
- S_is_space_or_tab(peek_at(input, parser->offset))) {
1256
- S_advance_offset(parser, input, 1, true);
1257
- }
1258
-
1259
- i = parser->column - save_column;
1260
- if (i >= 5 || i < 1 ||
1261
- // only spaces after list marker:
1262
- S_is_line_end_char(peek_at(input, parser->offset))) {
1263
- data->padding = matched + 1;
1264
- parser->offset = save_offset;
1265
- parser->column = save_column;
1266
- parser->partially_consumed_tab = save_partially_consumed_tab;
1267
- if (i > 0) {
1268
- S_advance_offset(parser, input, 1, true);
1269
- }
1270
- } else {
1271
- data->padding = matched + i;
1272
- }
1273
-
1274
- // check container; if it's a list, see if this list item
1275
- // can continue the list; otherwise, create a list container.
1276
-
1277
- data->marker_offset = parser->indent;
1278
-
1279
- if (cont_type != CMARK_NODE_LIST ||
1280
- !lists_match(&((*container)->as.list), data)) {
1281
- *container = add_child(parser, *container, CMARK_NODE_LIST,
1282
- parser->first_nonspace + 1);
1283
-
1284
- memcpy(&((*container)->as.list), data, sizeof(*data));
1285
- }
1286
-
1287
- // add the list item
1288
- *container = add_child(parser, *container, CMARK_NODE_ITEM,
1289
- parser->first_nonspace + 1);
1290
- /* TODO: static */
1291
- memcpy(&((*container)->as.list), data, sizeof(*data));
1292
- parser->mem->free(data);
1293
- } else if (indented && !maybe_lazy && !parser->blank) {
1294
- S_advance_offset(parser, input, CODE_INDENT, true);
1295
- *container = add_child(parser, *container, CMARK_NODE_CODE_BLOCK,
1296
- parser->offset + 1);
1297
- (*container)->as.code.fenced = false;
1298
- (*container)->as.code.fence_char = 0;
1299
- (*container)->as.code.fence_length = 0;
1300
- (*container)->as.code.fence_offset = 0;
1301
- (*container)->as.code.info = cmark_chunk_literal("");
1302
- } else {
1303
- cmark_llist *tmp;
1304
- cmark_node *new_container = NULL;
1305
-
1306
- for (tmp = parser->syntax_extensions; tmp; tmp=tmp->next) {
1307
- cmark_syntax_extension *ext = (cmark_syntax_extension *) tmp->data;
1308
-
1309
- if (ext->try_opening_block) {
1310
- new_container = ext->try_opening_block(
1311
- ext, indented, parser, *container, input->data, input->len);
1312
-
1313
- if (new_container) {
1314
- *container = new_container;
1315
- break;
1316
- }
1317
- }
1318
- }
1319
-
1320
- if (!new_container) {
1321
- break;
1322
- }
1323
- }
1324
-
1325
- if (accepts_lines(S_type(*container))) {
1326
- // if it's a line container, it can't contain other containers
1327
- break;
1328
- }
1329
-
1330
- cont_type = S_type(*container);
1331
- maybe_lazy = false;
1332
- }
1333
- }
1334
-
1335
- static void add_text_to_container(cmark_parser *parser, cmark_node *container,
1336
- cmark_node *last_matched_container,
1337
- cmark_chunk *input) {
1338
- cmark_node *tmp;
1339
- // what remains at parser->offset is a text line. add the text to the
1340
- // appropriate container.
1341
-
1342
- S_find_first_nonspace(parser, input);
1343
-
1344
- if (parser->blank && container->last_child)
1345
- S_set_last_line_blank(container->last_child, true);
1346
-
1347
- // block quote lines are never blank as they start with >
1348
- // and we don't count blanks in fenced code for purposes of tight/loose
1349
- // lists or breaking out of lists. we also don't set last_line_blank
1350
- // on an empty list item.
1351
- const cmark_node_type ctype = S_type(container);
1352
- const bool last_line_blank =
1353
- (parser->blank && ctype != CMARK_NODE_BLOCK_QUOTE &&
1354
- ctype != CMARK_NODE_HEADING && ctype != CMARK_NODE_THEMATIC_BREAK &&
1355
- !(ctype == CMARK_NODE_CODE_BLOCK && container->as.code.fenced) &&
1356
- !(ctype == CMARK_NODE_ITEM && container->first_child == NULL &&
1357
- container->start_line == parser->line_number));
1358
-
1359
- S_set_last_line_blank(container, last_line_blank);
1360
-
1361
- tmp = container;
1362
- while (tmp->parent) {
1363
- S_set_last_line_blank(tmp->parent, false);
1364
- tmp = tmp->parent;
1365
- }
1366
-
1367
- // If the last line processed belonged to a paragraph node,
1368
- // and we didn't match all of the line prefixes for the open containers,
1369
- // and we didn't start any new containers,
1370
- // and the line isn't blank,
1371
- // then treat this as a "lazy continuation line" and add it to
1372
- // the open paragraph.
1373
- if (parser->current != last_matched_container &&
1374
- container == last_matched_container && !parser->blank &&
1375
- S_type(parser->current) == CMARK_NODE_PARAGRAPH) {
1376
- add_line(parser->current, input, parser);
1377
- } else { // not a lazy continuation
1378
- // Finalize any blocks that were not matched and set cur to container:
1379
- while (parser->current != last_matched_container) {
1380
- parser->current = finalize(parser, parser->current);
1381
- assert(parser->current != NULL);
1382
- }
1383
-
1384
- if (S_type(container) == CMARK_NODE_CODE_BLOCK) {
1385
- add_line(container, input, parser);
1386
- } else if (S_type(container) == CMARK_NODE_HTML_BLOCK) {
1387
- add_line(container, input, parser);
1388
-
1389
- int matches_end_condition;
1390
- switch (container->as.html_block_type) {
1391
- case 1:
1392
- // </script>, </style>, </pre>
1393
- matches_end_condition =
1394
- scan_html_block_end_1(input, parser->first_nonspace);
1395
- break;
1396
- case 2:
1397
- // -->
1398
- matches_end_condition =
1399
- scan_html_block_end_2(input, parser->first_nonspace);
1400
- break;
1401
- case 3:
1402
- // ?>
1403
- matches_end_condition =
1404
- scan_html_block_end_3(input, parser->first_nonspace);
1405
- break;
1406
- case 4:
1407
- // >
1408
- matches_end_condition =
1409
- scan_html_block_end_4(input, parser->first_nonspace);
1410
- break;
1411
- case 5:
1412
- // ]]>
1413
- matches_end_condition =
1414
- scan_html_block_end_5(input, parser->first_nonspace);
1415
- break;
1416
- default:
1417
- matches_end_condition = 0;
1418
- break;
1419
- }
1420
-
1421
- if (matches_end_condition) {
1422
- container = finalize(parser, container);
1423
- assert(parser->current != NULL);
1424
- }
1425
- } else if (parser->blank) {
1426
- // ??? do nothing
1427
- } else if (accepts_lines(S_type(container))) {
1428
- if (S_type(container) == CMARK_NODE_HEADING &&
1429
- container->as.heading.setext == false) {
1430
- chop_trailing_hashtags(input);
1431
- }
1432
- S_advance_offset(parser, input, parser->first_nonspace - parser->offset,
1433
- false);
1434
- add_line(container, input, parser);
1435
- } else {
1436
- // create paragraph container for line
1437
- container = add_child(parser, container, CMARK_NODE_PARAGRAPH,
1438
- parser->first_nonspace + 1);
1439
- S_advance_offset(parser, input, parser->first_nonspace - parser->offset,
1440
- false);
1441
- add_line(container, input, parser);
1442
- }
1443
-
1444
- parser->current = container;
1445
- }
1446
- }
1447
-
1448
- /* See http://spec.commonmark.org/0.24/#phase-1-block-structure */
1449
- static void S_process_line(cmark_parser *parser, const unsigned char *buffer,
1450
- bufsize_t bytes) {
1451
- cmark_node *last_matched_container;
1452
- bool all_matched = true;
1453
- cmark_node *container;
1454
- cmark_chunk input;
1455
- cmark_node *current;
1456
-
1457
- cmark_strbuf_clear(&parser->curline);
1458
-
1459
- if (parser->options & CMARK_OPT_VALIDATE_UTF8)
1460
- cmark_utf8proc_check(&parser->curline, buffer, bytes);
1461
- else
1462
- cmark_strbuf_put(&parser->curline, buffer, bytes);
1463
-
1464
- bytes = parser->curline.size;
1465
-
1466
- // ensure line ends with a newline:
1467
- if (bytes == 0 || !S_is_line_end_char(parser->curline.ptr[bytes - 1]))
1468
- cmark_strbuf_putc(&parser->curline, '\n');
1469
-
1470
- parser->offset = 0;
1471
- parser->column = 0;
1472
- parser->first_nonspace = 0;
1473
- parser->first_nonspace_column = 0;
1474
- parser->thematic_break_kill_pos = 0;
1475
- parser->indent = 0;
1476
- parser->blank = false;
1477
- parser->partially_consumed_tab = false;
1478
-
1479
- input.data = parser->curline.ptr;
1480
- input.len = parser->curline.size;
1481
- input.alloc = 0;
1482
-
1483
- // Skip UTF-8 BOM.
1484
- if (parser->line_number == 0 &&
1485
- input.len >= 3 &&
1486
- memcmp(input.data, "\xef\xbb\xbf", 3) == 0)
1487
- parser->offset += 3;
1488
-
1489
- parser->line_number++;
1490
-
1491
- last_matched_container = check_open_blocks(parser, &input, &all_matched);
1492
-
1493
- if (!last_matched_container)
1494
- goto finished;
1495
-
1496
- container = last_matched_container;
1497
-
1498
- current = parser->current;
1499
-
1500
- open_new_blocks(parser, &container, &input, all_matched);
1501
-
1502
- /* parser->current might have changed if feed_reentrant was called */
1503
- if (current == parser->current)
1504
- add_text_to_container(parser, container, last_matched_container, &input);
1505
-
1506
- finished:
1507
- parser->last_line_length = input.len;
1508
- if (parser->last_line_length &&
1509
- input.data[parser->last_line_length - 1] == '\n')
1510
- parser->last_line_length -= 1;
1511
- if (parser->last_line_length &&
1512
- input.data[parser->last_line_length - 1] == '\r')
1513
- parser->last_line_length -= 1;
1514
-
1515
- cmark_strbuf_clear(&parser->curline);
1516
- }
1517
-
1518
- cmark_node *cmark_parser_finish(cmark_parser *parser) {
1519
- cmark_node *res;
1520
- cmark_llist *extensions;
1521
-
1522
- /* Parser was already finished once */
1523
- if (parser->root == NULL)
1524
- return NULL;
1525
-
1526
- if (parser->linebuf.size) {
1527
- S_process_line(parser, parser->linebuf.ptr, parser->linebuf.size);
1528
- cmark_strbuf_clear(&parser->linebuf);
1529
- }
1530
-
1531
- finalize_document(parser);
1532
-
1533
- cmark_consolidate_text_nodes(parser->root);
1534
-
1535
- cmark_strbuf_free(&parser->curline);
1536
- cmark_strbuf_free(&parser->linebuf);
1537
-
1538
- #if CMARK_DEBUG_NODES
1539
- if (cmark_node_check(parser->root, stderr)) {
1540
- abort();
1541
- }
1542
- #endif
1543
-
1544
- for (extensions = parser->syntax_extensions; extensions; extensions = extensions->next) {
1545
- cmark_syntax_extension *ext = (cmark_syntax_extension *) extensions->data;
1546
- if (ext->postprocess_func) {
1547
- cmark_node *processed = ext->postprocess_func(ext, parser, parser->root);
1548
- if (processed)
1549
- parser->root = processed;
1550
- }
1551
- }
1552
-
1553
- res = parser->root;
1554
- parser->root = NULL;
1555
-
1556
- cmark_parser_reset(parser);
1557
-
1558
- return res;
1559
- }
1560
-
1561
- int cmark_parser_get_line_number(cmark_parser *parser) {
1562
- return parser->line_number;
1563
- }
1564
-
1565
- bufsize_t cmark_parser_get_offset(cmark_parser *parser) {
1566
- return parser->offset;
1567
- }
1568
-
1569
- bufsize_t cmark_parser_get_column(cmark_parser *parser) {
1570
- return parser->column;
1571
- }
1572
-
1573
- int cmark_parser_get_first_nonspace(cmark_parser *parser) {
1574
- return parser->first_nonspace;
1575
- }
1576
-
1577
- int cmark_parser_get_first_nonspace_column(cmark_parser *parser) {
1578
- return parser->first_nonspace_column;
1579
- }
1580
-
1581
- int cmark_parser_get_indent(cmark_parser *parser) {
1582
- return parser->indent;
1583
- }
1584
-
1585
- int cmark_parser_is_blank(cmark_parser *parser) {
1586
- return parser->blank;
1587
- }
1588
-
1589
- int cmark_parser_has_partially_consumed_tab(cmark_parser *parser) {
1590
- return parser->partially_consumed_tab;
1591
- }
1592
-
1593
- int cmark_parser_get_last_line_length(cmark_parser *parser) {
1594
- return parser->last_line_length;
1595
- }
1596
-
1597
- cmark_node *cmark_parser_add_child(cmark_parser *parser,
1598
- cmark_node *parent,
1599
- cmark_node_type block_type,
1600
- int start_column) {
1601
- return add_child(parser, parent, block_type, start_column);
1602
- }
1603
-
1604
- void cmark_parser_advance_offset(cmark_parser *parser,
1605
- const char *input,
1606
- int count,
1607
- int columns) {
1608
- cmark_chunk input_chunk = cmark_chunk_literal(input);
1609
-
1610
- S_advance_offset(parser, &input_chunk, count, columns != 0);
1611
- }
1612
-
1613
- void cmark_parser_set_backslash_ispunct_func(cmark_parser *parser,
1614
- cmark_ispunct_func func) {
1615
- parser->backslash_ispunct = func;
1616
- }
1617
-
1618
- cmark_llist *cmark_parser_get_syntax_extensions(cmark_parser *parser) {
1619
- return parser->syntax_extensions;
1620
- }