summaryrefslogtreecommitdiff
path: root/src/blocks.c
blob: 9970cc95e459b15e9e8cbeb2886427072563293a (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
/**
 * Block parsing implementation.
 *
 * For a high-level overview of the block parsing process,
 * see http://spec.commonmark.org/0.24/#phase-1-block-structure
 */

#include <stdlib.h>
#include <assert.h>
#include <stdio.h>

#include "cmark_ctype.h"
#include "config.h"
#include "parser.h"
#include "cmark.h"
#include "node.h"
#include "references.h"
#include "utf8.h"
#include "scanners.h"
#include "inlines.h"
#include "houdini.h"
#include "buffer.h"

#define CODE_INDENT 4
#define TAB_STOP 4

#ifndef MIN
#define MIN(x, y) ((x < y) ? x : y)
#endif

#define peek_at(i, n) (i)->data[n]

static bool S_last_line_blank(const cmark_node *node) {
  return (node->flags & CMARK_NODE__LAST_LINE_BLANK) != 0;
}

static bool S_last_line_checked(const cmark_node *node) {
  return (node->flags & CMARK_NODE__LAST_LINE_CHECKED) != 0;
}

static CMARK_INLINE cmark_node_type S_type(const cmark_node *node) {
  return (cmark_node_type)node->type;
}

static void S_set_last_line_blank(cmark_node *node, bool is_blank) {
  if (is_blank)
    node->flags |= CMARK_NODE__LAST_LINE_BLANK;
  else
    node->flags &= ~CMARK_NODE__LAST_LINE_BLANK;
}

static void S_set_last_line_checked(cmark_node *node) {
  node->flags |= CMARK_NODE__LAST_LINE_CHECKED;
}

static CMARK_INLINE bool S_is_line_end_char(char c) {
  return (c == '\n' || c == '\r');
}

static CMARK_INLINE bool S_is_space_or_tab(char c) {
  return (c == ' ' || c == '\t');
}

static void S_parser_feed(cmark_parser *parser, const unsigned char *buffer,
                          size_t len, bool eof);

static void S_process_line(cmark_parser *parser, const unsigned char *buffer,
                           bufsize_t bytes);

static cmark_node *make_block(cmark_mem *mem, cmark_node_type tag,
                              int start_line, int start_column) {
  cmark_node *e;

  e = (cmark_node *)mem->calloc(1, sizeof(*e));
  cmark_strbuf_init(mem, &e->content, 32);
  e->type = (uint16_t)tag;
  e->flags = CMARK_NODE__OPEN;
  e->start_line = start_line;
  e->start_column = start_column;
  e->end_line = start_line;

  return e;
}

// Create a root document node.
static cmark_node *make_document(cmark_mem *mem) {
  cmark_node *e = make_block(mem, CMARK_NODE_DOCUMENT, 1, 1);
  return e;
}

cmark_parser *cmark_parser_new_with_mem(int options, cmark_mem *mem) {
  cmark_parser *parser = (cmark_parser *)mem->calloc(1, sizeof(cmark_parser));
  parser->mem = mem;

  cmark_node *document = make_document(mem);

  cmark_strbuf_init(mem, &parser->curline, 256);
  cmark_strbuf_init(mem, &parser->linebuf, 0);

  parser->refmap = cmark_reference_map_new(mem);
  parser->root = document;
  parser->current = document;
  parser->line_number = 0;
  parser->offset = 0;
  parser->column = 0;
  parser->first_nonspace = 0;
  parser->first_nonspace_column = 0;
  parser->thematic_break_kill_pos = 0;
  parser->indent = 0;
  parser->blank = false;
  parser->partially_consumed_tab = false;
  parser->last_line_length = 0;
  parser->options = options;
  parser->last_buffer_ended_with_cr = false;

  return parser;
}

cmark_parser *cmark_parser_new(int options) {
  extern cmark_mem DEFAULT_MEM_ALLOCATOR;
  return cmark_parser_new_with_mem(options, &DEFAULT_MEM_ALLOCATOR);
}

void cmark_parser_free(cmark_parser *parser) {
  cmark_mem *mem = parser->mem;
  cmark_strbuf_free(&parser->curline);
  cmark_strbuf_free(&parser->linebuf);
  cmark_reference_map_free(parser->refmap);
  mem->free(parser);
}

static cmark_node *finalize(cmark_parser *parser, cmark_node *b);

// Returns true if line has only space characters, else false.
static bool is_blank(cmark_strbuf *s, bufsize_t offset) {
  while (offset < s->size) {
    switch (s->ptr[offset]) {
    case '\r':
    case '\n':
      return true;
    case ' ':
      offset++;
      break;
    case '\t':
      offset++;
      break;
    default:
      return false;
    }
  }

  return true;
}

static CMARK_INLINE bool can_contain(cmark_node_type parent_type,
                                     cmark_node_type child_type) {
  return (parent_type == CMARK_NODE_DOCUMENT ||
          parent_type == CMARK_NODE_BLOCK_QUOTE ||
          parent_type == CMARK_NODE_ITEM ||
          (parent_type == CMARK_NODE_LIST && child_type == CMARK_NODE_ITEM));
}

static CMARK_INLINE bool accepts_lines(cmark_node_type block_type) {
  return (block_type == CMARK_NODE_PARAGRAPH ||
          block_type == CMARK_NODE_HEADING ||
          block_type == CMARK_NODE_CODE_BLOCK);
}

static CMARK_INLINE bool contains_inlines(cmark_node_type block_type) {
  return (block_type == CMARK_NODE_PARAGRAPH ||
          block_type == CMARK_NODE_HEADING);
}

static void add_line(cmark_node *node, cmark_chunk *ch, cmark_parser *parser) {
  int chars_to_tab;
  int i;
  assert(node->flags & CMARK_NODE__OPEN);
  if (parser->partially_consumed_tab) {
    parser->offset += 1; // skip over tab
    // add space characters:
    chars_to_tab = TAB_STOP - (parser->column % TAB_STOP);
    for (i = 0; i < chars_to_tab; i++) {
      cmark_strbuf_putc(&node->content, ' ');
    }
  }
  cmark_strbuf_put(&node->content, ch->data + parser->offset,
                   ch->len - parser->offset);
}

static void remove_trailing_blank_lines(cmark_strbuf *ln) {
  bufsize_t i;
  unsigned char c;

  for (i = ln->size - 1; i >= 0; --i) {
    c = ln->ptr[i];

    if (c != ' ' && c != '\t' && !S_is_line_end_char(c))
      break;
  }

  if (i < 0) {
    cmark_strbuf_clear(ln);
    return;
  }

  for (; i < ln->size; ++i) {
    c = ln->ptr[i];

    if (!S_is_line_end_char(c))
      continue;

    cmark_strbuf_truncate(ln, i);
    break;
  }
}

// Check to see if a node ends with a blank line, descending
// if needed into lists and sublists.
static bool S_ends_with_blank_line(cmark_node *node) {
  if (S_last_line_checked(node)) {
    return(S_last_line_blank(node));
  } else if ((S_type(node) == CMARK_NODE_LIST ||
              S_type(node) == CMARK_NODE_ITEM) && node->last_child) {
    S_set_last_line_checked(node);
    return(S_ends_with_blank_line(node->last_child));
  } else {
    S_set_last_line_checked(node);
    return (S_last_line_blank(node));
  }
}

// returns true if content remains after link defs are resolved.
static bool resolve_reference_link_definitions(
		cmark_parser *parser,
                cmark_node *b) {
  bufsize_t pos;
  cmark_strbuf *node_content = &b->content;
  cmark_chunk chunk = {node_content->ptr, node_content->size, 0};
  while (chunk.len && chunk.data[0] == '[' &&
         (pos = cmark_parse_reference_inline(parser->mem, &chunk,
					     parser->refmap))) {

    chunk.data += pos;
    chunk.len -= pos;
  }
  cmark_strbuf_drop(node_content, (node_content->size - chunk.len));
  return !is_blank(&b->content, 0);
}

static cmark_node *finalize(cmark_parser *parser, cmark_node *b) {
  bufsize_t pos;
  cmark_node *item;
  cmark_node *subitem;
  cmark_node *parent;
  bool has_content;

  parent = b->parent;
  assert(b->flags &
         CMARK_NODE__OPEN); // shouldn't call finalize on closed blocks
  b->flags &= ~CMARK_NODE__OPEN;

  if (parser->curline.size == 0) {
    // end of input - line number has not been incremented
    b->end_line = parser->line_number;
    b->end_column = parser->last_line_length;
  } else if (S_type(b) == CMARK_NODE_DOCUMENT ||
             (S_type(b) == CMARK_NODE_CODE_BLOCK && b->as.code.fenced) ||
             (S_type(b) == CMARK_NODE_HEADING && b->as.heading.setext)) {
    b->end_line = parser->line_number;
    b->end_column = parser->curline.size;
    if (b->end_column && parser->curline.ptr[b->end_column - 1] == '\n')
      b->end_column -= 1;
    if (b->end_column && parser->curline.ptr[b->end_column - 1] == '\r')
      b->end_column -= 1;
  } else {
    b->end_line = parser->line_number - 1;
    b->end_column = parser->last_line_length;
  }

  cmark_strbuf *node_content = &b->content;

  switch (S_type(b)) {
  case CMARK_NODE_PARAGRAPH:
  {
    has_content = resolve_reference_link_definitions(parser, b);
    if (!has_content) {
      // remove blank node (former reference def)
      cmark_node_free(b);
    }
    break;
  }

  case CMARK_NODE_CODE_BLOCK:
    if (!b->as.code.fenced) { // indented code
      remove_trailing_blank_lines(node_content);
      cmark_strbuf_putc(node_content, '\n');
    } else {
      // first line of contents becomes info
      for (pos = 0; pos < node_content->size; ++pos) {
        if (S_is_line_end_char(node_content->ptr[pos]))
          break;
      }
      assert(pos < node_content->size);

      if (pos == 0) {
        b->as.code.info = NULL;
      } else {
        cmark_strbuf tmp = CMARK_BUF_INIT(parser->mem);
        houdini_unescape_html_f(&tmp, node_content->ptr, pos);
        cmark_strbuf_trim(&tmp);
        cmark_strbuf_unescape(&tmp);
        b->as.code.info = cmark_strbuf_detach(&tmp);
      }

      if (node_content->ptr[pos] == '\r')
        pos += 1;
      if (node_content->ptr[pos] == '\n')
        pos += 1;
      cmark_strbuf_drop(node_content, pos);
    }
    b->as.code.literal = cmark_strbuf_detach(node_content);
    break;

  case CMARK_NODE_HTML_BLOCK:
    b->as.literal.len = node_content->size;
    b->as.literal.data = cmark_strbuf_detach(node_content);
    break;

  case CMARK_NODE_LIST:      // determine tight/loose status
    b->as.list.tight = true; // tight by default
    item = b->first_child;

    while (item) {
      // check for non-final non-empty list item ending with blank line:
      if (S_last_line_blank(item) && item->next) {
        b->as.list.tight = false;
        break;
      }
      // recurse into children of list item, to see if there are
      // spaces between them:
      subitem = item->first_child;
      while (subitem) {
        if ((item->next || subitem->next) &&
            S_ends_with_blank_line(subitem)) {
          b->as.list.tight = false;
          break;
        }
        subitem = subitem->next;
      }
      if (!(b->as.list.tight)) {
        break;
      }
      item = item->next;
    }

    break;

  default:
    break;
  }

  return parent;
}

// Add a node as child of another.  Return pointer to child.
static cmark_node *add_child(cmark_parser *parser, cmark_node *parent,
                             cmark_node_type block_type, int start_column) {
  assert(parent);

  // if 'parent' isn't the kind of node that can accept this child,
  // then back up til we hit a node that can.
  while (!can_contain(S_type(parent), block_type)) {
    parent = finalize(parser, parent);
  }

  cmark_node *child =
      make_block(parser->mem, block_type, parser->line_number, start_column);
  child->parent = parent;

  if (parent->last_child) {
    parent->last_child->next = child;
    child->prev = parent->last_child;
  } else {
    parent->first_child = child;
    child->prev = NULL;
  }
  parent->last_child = child;
  return child;
}

// Walk through node and all children, recursively, parsing
// string content into inline content where appropriate.
static void process_inlines(cmark_mem *mem, cmark_node *root,
                            cmark_reference_map *refmap, int options) {
  cmark_iter *iter = cmark_iter_new(root);
  cmark_node *cur;
  cmark_event_type ev_type;

  while ((ev_type = cmark_iter_next(iter)) != CMARK_EVENT_DONE) {
    cur = cmark_iter_get_node(iter);
    if (ev_type == CMARK_EVENT_ENTER) {
      if (contains_inlines(S_type(cur))) {
        cmark_parse_inlines(mem, cur, refmap, options);
      }
    }
  }

  cmark_iter_free(iter);
}

// Attempts to parse a list item marker (bullet or enumerated).
// On success, returns length of the marker, and populates
// data with the details.  On failure, returns 0.
static bufsize_t parse_list_marker(cmark_mem *mem, cmark_chunk *input,
                                   bufsize_t pos, bool interrupts_paragraph,
                                   cmark_list **dataptr) {
  unsigned char c;
  bufsize_t startpos;
  cmark_list *data;
  bufsize_t i;

  startpos = pos;
  c = peek_at(input, pos);

  if (c == '*' || c == '-' || c == '+') {
    pos++;
    if (!cmark_isspace(peek_at(input, pos))) {
      return 0;
    }

    if (interrupts_paragraph) {
      i = pos;
      // require non-blank content after list marker:
      while (S_is_space_or_tab(peek_at(input, i))) {
        i++;
      }
      if (peek_at(input, i) == '\n') {
        return 0;
      }
    }

    data = (cmark_list *)mem->calloc(1, sizeof(*data));
    data->marker_offset = 0; // will be adjusted later
    data->list_type = CMARK_BULLET_LIST;
    data->bullet_char = c;
    data->start = 0;
    data->delimiter = CMARK_NO_DELIM;
    data->tight = false;
  } else if (cmark_isdigit(c)) {
    int start = 0;
    int digits = 0;

    do {
      start = (10 * start) + (peek_at(input, pos) - '0');
      pos++;
      digits++;
      // We limit to 9 digits to avoid overflow,
      // assuming max int is 2^31 - 1
      // This also seems to be the limit for 'start' in some browsers.
    } while (digits < 9 && cmark_isdigit(peek_at(input, pos)));

    if (interrupts_paragraph && start != 1) {
      return 0;
    }
    c = peek_at(input, pos);
    if (c == '.' || c == ')') {
      pos++;
      if (!cmark_isspace(peek_at(input, pos))) {
        return 0;
      }
      if (interrupts_paragraph) {
        // require non-blank content after list marker:
        i = pos;
        while (S_is_space_or_tab(peek_at(input, i))) {
          i++;
        }
        if (S_is_line_end_char(peek_at(input, i))) {
          return 0;
        }
      }

      data = (cmark_list *)mem->calloc(1, sizeof(*data));
      data->marker_offset = 0; // will be adjusted later
      data->list_type = CMARK_ORDERED_LIST;
      data->bullet_char = 0;
      data->start = start;
      data->delimiter = (c == '.' ? CMARK_PERIOD_DELIM : CMARK_PAREN_DELIM);
      data->tight = false;
    } else {
      return 0;
    }
  } else {
    return 0;
  }

  *dataptr = data;
  return (pos - startpos);
}

// Return 1 if list item belongs in list, else 0.
static int lists_match(cmark_list *list_data, cmark_list *item_data) {
  return (list_data->list_type == item_data->list_type &&
          list_data->delimiter == item_data->delimiter &&
          // list_data->marker_offset == item_data.marker_offset &&
          list_data->bullet_char == item_data->bullet_char);
}

static cmark_node *finalize_document(cmark_parser *parser) {
  while (parser->current != parser->root) {
    parser->current = finalize(parser, parser->current);
  }

  finalize(parser, parser->root);
  process_inlines(parser->mem, parser->root, parser->refmap, parser->options);

  return parser->root;
}

cmark_node *cmark_parse_file(FILE *f, int options) {
  unsigned char buffer[4096];
  cmark_parser *parser = cmark_parser_new(options);
  size_t bytes;
  cmark_node *document;

  while ((bytes = fread(buffer, 1, sizeof(buffer), f)) > 0) {
    bool eof = bytes < sizeof(buffer);
    S_parser_feed(parser, buffer, bytes, eof);
    if (eof) {
      break;
    }
  }

  document = cmark_parser_finish(parser);
  cmark_parser_free(parser);
  return document;
}

cmark_node *cmark_parse_document(const char *buffer, size_t len, int options) {
  cmark_parser *parser = cmark_parser_new(options);
  cmark_node *document;

  S_parser_feed(parser, (const unsigned char *)buffer, len, true);

  document = cmark_parser_finish(parser);
  cmark_parser_free(parser);
  return document;
}

void cmark_parser_feed(cmark_parser *parser, const char *buffer, size_t len) {
  S_parser_feed(parser, (const unsigned char *)buffer, len, false);
}

static void S_parser_feed(cmark_parser *parser, const unsigned char *buffer,
                          size_t len, bool eof) {
  const unsigned char *end = buffer + len;
  static const uint8_t repl[] = {239, 191, 189};

  if (parser->last_buffer_ended_with_cr && *buffer == '\n') {
    // skip NL if last buffer ended with CR ; see #117
    buffer++;
  }
  parser->last_buffer_ended_with_cr = false;
  while (buffer < end) {
    const unsigned char *eol;
    bufsize_t chunk_len;
    bool process = false;
    for (eol = buffer; eol < end; ++eol) {
      if (S_is_line_end_char(*eol)) {
        process = true;
        break;
      }
      if (*eol == '\0' && eol < end) {
        break;
      }
    }
    if (eol >= end && eof) {
      process = true;
    }

    chunk_len = (eol - buffer);
    if (process) {
      if (parser->linebuf.size > 0) {
        cmark_strbuf_put(&parser->linebuf, buffer, chunk_len);
        S_process_line(parser, parser->linebuf.ptr, parser->linebuf.size);
        cmark_strbuf_clear(&parser->linebuf);
      } else {
        S_process_line(parser, buffer, chunk_len);
      }
    } else {
      if (eol < end && *eol == '\0') {
        // omit NULL byte
        cmark_strbuf_put(&parser->linebuf, buffer, chunk_len);
        // add replacement character
        cmark_strbuf_put(&parser->linebuf, repl, 3);
      } else {
        cmark_strbuf_put(&parser->linebuf, buffer, chunk_len);
      }
    }

    buffer += chunk_len;
    if (buffer < end) {
      if (*buffer == '\0') {
        // skip over NULL
        buffer++;
      } else {
        // skip over line ending characters
        if (*buffer == '\r') {
          buffer++;
          if (buffer == end)
            parser->last_buffer_ended_with_cr = true;
        }
        if (buffer < end && *buffer == '\n')
          buffer++;
      }
    }
  }
}

static void chop_trailing_hashtags(cmark_chunk *ch) {
  bufsize_t n, orig_n;

  cmark_chunk_rtrim(ch);
  orig_n = n = ch->len - 1;

  // if string ends in space followed by #s, remove these:
  while (n >= 0 && peek_at(ch, n) == '#')
    n--;

  // Check for a space before the final #s:
  if (n != orig_n && n >= 0 && S_is_space_or_tab(peek_at(ch, n))) {
    ch->len = n;
    cmark_chunk_rtrim(ch);
  }
}

// Check for thematic break.  On failure, return 0 and update
// thematic_break_kill_pos with the index at which the
// parse fails.  On success, return length of match.
// "...three or more hyphens, asterisks,
// or underscores on a line by themselves. If you wish, you may use
// spaces between the hyphens or asterisks."
static int S_scan_thematic_break(cmark_parser *parser, cmark_chunk *input,
                                 bufsize_t offset) {
  bufsize_t i;
  char c;
  char nextc = '\0';
  int count;
  i = offset;
  c = peek_at(input, i);
  if (!(c == '*' || c == '_' || c == '-')) {
    parser->thematic_break_kill_pos = i;
    return 0;
  }
  count = 1;
  while ((nextc = peek_at(input, ++i))) {
    if (nextc == c) {
      count++;
    } else if (nextc != ' ' && nextc != '\t') {
      break;
    }
  }
  if (count >= 3 && (nextc == '\r' || nextc == '\n')) {
    return (i - offset) + 1;
  } else {
    parser->thematic_break_kill_pos = i;
    return 0;
  }
}

// Find first nonspace character from current offset, setting
// parser->first_nonspace, parser->first_nonspace_column,
// parser->indent, and parser->blank. Does not advance parser->offset.
static void S_find_first_nonspace(cmark_parser *parser, cmark_chunk *input) {
  char c;
  int chars_to_tab = TAB_STOP - (parser->column % TAB_STOP);

  if (parser->first_nonspace <= parser->offset) {
    parser->first_nonspace = parser->offset;
    parser->first_nonspace_column = parser->column;
    while ((c = peek_at(input, parser->first_nonspace))) {
      if (c == ' ') {
        parser->first_nonspace += 1;
        parser->first_nonspace_column += 1;
        chars_to_tab = chars_to_tab - 1;
        if (chars_to_tab == 0) {
          chars_to_tab = TAB_STOP;
        }
      } else if (c == '\t') {
        parser->first_nonspace += 1;
        parser->first_nonspace_column += chars_to_tab;
        chars_to_tab = TAB_STOP;
      } else {
        break;
      }
    }
  }

  parser->indent = parser->first_nonspace_column - parser->column;
  parser->blank = S_is_line_end_char(peek_at(input, parser->first_nonspace));
}

// Advance parser->offset and parser->column.  parser->offset is the
// byte position in input; parser->column is a virtual column number
// that takes into account tabs. (Multibyte characters are not taken
// into account, because the Markdown line prefixes we are interested in
// analyzing are entirely ASCII.)  The count parameter indicates
// how far to advance the offset.  If columns is true, then count
// indicates a number of columns; otherwise, a number of bytes.
// If advancing a certain number of columns partially consumes
// a tab character, parser->partially_consumed_tab is set to true.
static void S_advance_offset(cmark_parser *parser, cmark_chunk *input,
                             bufsize_t count, bool columns) {
  char c;
  int chars_to_tab;
  int chars_to_advance;
  while (count > 0 && (c = peek_at(input, parser->offset))) {
    if (c == '\t') {
      chars_to_tab = TAB_STOP - (parser->column % TAB_STOP);
      if (columns) {
        parser->partially_consumed_tab = chars_to_tab > count;
        chars_to_advance = MIN(count, chars_to_tab);
        parser->column += chars_to_advance;
        parser->offset += (parser->partially_consumed_tab ? 0 : 1);
        count -= chars_to_advance;
      } else {
        parser->partially_consumed_tab = false;
        parser->column += chars_to_tab;
        parser->offset += 1;
        count -= 1;
      }
    } else {
      parser->partially_consumed_tab = false;
      parser->offset += 1;
      parser->column += 1; // assume ascii; block starts are ascii
      count -= 1;
    }
  }
}

static bool S_last_child_is_open(cmark_node *container) {
  return container->last_child &&
         (container->last_child->flags & CMARK_NODE__OPEN);
}

static bool parse_block_quote_prefix(cmark_parser *parser, cmark_chunk *input) {
  bool res = false;
  bufsize_t matched = 0;

  matched =
      parser->indent <= 3 && peek_at(input, parser->first_nonspace) == '>';
  if (matched) {

    S_advance_offset(parser, input, parser->indent + 1, true);

    if (S_is_space_or_tab(peek_at(input, parser->offset))) {
      S_advance_offset(parser, input, 1, true);
    }

    res = true;
  }
  return res;
}

static bool parse_node_item_prefix(cmark_parser *parser, cmark_chunk *input,
                                   cmark_node *container) {
  bool res = false;

  if (parser->indent >=
      container->as.list.marker_offset + container->as.list.padding) {
    S_advance_offset(parser, input, container->as.list.marker_offset +
                                        container->as.list.padding,
                     true);
    res = true;
  } else if (parser->blank && container->first_child != NULL) {
    // if container->first_child is NULL, then the opening line
    // of the list item was blank after the list marker; in this
    // case, we are done with the list item.
    S_advance_offset(parser, input, parser->first_nonspace - parser->offset,
                     false);
    res = true;
  }
  return res;
}

static bool parse_code_block_prefix(cmark_parser *parser, cmark_chunk *input,
                                    cmark_node *container,
                                    bool *should_continue) {
  bool res = false;

  if (!container->as.code.fenced) { // indented
    if (parser->indent >= CODE_INDENT) {
      S_advance_offset(parser, input, CODE_INDENT, true);
      res = true;
    } else if (parser->blank) {
      S_advance_offset(parser, input, parser->first_nonspace - parser->offset,
                       false);
      res = true;
    }
  } else { // fenced
    bufsize_t matched = 0;

    if (parser->indent <= 3 && (peek_at(input, parser->first_nonspace) ==
                                container->as.code.fence_char)) {
      matched = scan_close_code_fence(input, parser->first_nonspace);
    }

    if (matched >= container->as.code.fence_length) {
      // closing fence - and since we're at
      // the end of a line, we can stop processing it:
      *should_continue = false;
      S_advance_offset(parser, input, matched, false);
      parser->current = finalize(parser, container);
    } else {
      // skip opt. spaces of fence parser->offset
      int i = container->as.code.fence_offset;

      while (i > 0 && S_is_space_or_tab(peek_at(input, parser->offset))) {
        S_advance_offset(parser, input, 1, true);
        i--;
      }
      res = true;
    }
  }

  return res;
}

static bool parse_html_block_prefix(cmark_parser *parser,
                                    cmark_node *container) {
  bool res = false;
  int html_block_type = container->as.html_block_type;

  assert(html_block_type >= 1 && html_block_type <= 7);
  switch (html_block_type) {
  case 1:
  case 2:
  case 3:
  case 4:
  case 5:
    // these types of blocks can accept blanks
    res = true;
    break;
  case 6:
  case 7:
    res = !parser->blank;
    break;
  }

  return res;
}

/**
 * For each containing node, try to parse the associated line start.
 *
 * Will not close unmatched blocks, as we may have a lazy continuation
 * line -> http://spec.commonmark.org/0.24/#lazy-continuation-line
 *
 * Returns: The last matching node, or NULL
 */
static cmark_node *check_open_blocks(cmark_parser *parser, cmark_chunk *input,
                                     bool *all_matched) {
  bool should_continue = true;
  *all_matched = false;
  cmark_node *container = parser->root;
  cmark_node_type cont_type;

  while (S_last_child_is_open(container)) {
    container = container->last_child;
    cont_type = S_type(container);

    S_find_first_nonspace(parser, input);

    switch (cont_type) {
    case CMARK_NODE_BLOCK_QUOTE:
      if (!parse_block_quote_prefix(parser, input))
        goto done;
      break;
    case CMARK_NODE_ITEM:
      if (!parse_node_item_prefix(parser, input, container))
        goto done;
      break;
    case CMARK_NODE_CODE_BLOCK:
      if (!parse_code_block_prefix(parser, input, container, &should_continue))
        goto done;
      break;
    case CMARK_NODE_HEADING:
      // a heading can never contain more than one line
      goto done;
    case CMARK_NODE_HTML_BLOCK:
      if (!parse_html_block_prefix(parser, container))
        goto done;
      break;
    case CMARK_NODE_PARAGRAPH:
      if (parser->blank)
        goto done;
      break;
    default:
      break;
    }
  }

  *all_matched = true;

done:
  if (!*all_matched) {
    container = container->parent; // back up to last matching node
  }

  if (!should_continue) {
    container = NULL;
  }

  return container;
}

static void open_new_blocks(cmark_parser *parser, cmark_node **container,
                            cmark_chunk *input, bool all_matched) {
  bool indented;
  cmark_list *data = NULL;
  bool maybe_lazy = S_type(parser->current) == CMARK_NODE_PARAGRAPH;
  cmark_node_type cont_type = S_type(*container);
  bufsize_t matched = 0;
  int lev = 0;
  bool save_partially_consumed_tab;
  bool has_content;
  int save_offset;
  int save_column;

  while (cont_type != CMARK_NODE_CODE_BLOCK &&
         cont_type != CMARK_NODE_HTML_BLOCK) {

    S_find_first_nonspace(parser, input);
    indented = parser->indent >= CODE_INDENT;

    if (!indented && peek_at(input, parser->first_nonspace) == '>') {

      bufsize_t blockquote_startpos = parser->first_nonspace;

      S_advance_offset(parser, input,
                       parser->first_nonspace + 1 - parser->offset, false);
      // optional following character
      if (S_is_space_or_tab(peek_at(input, parser->offset))) {
        S_advance_offset(parser, input, 1, true);
      }
      *container = add_child(parser, *container, CMARK_NODE_BLOCK_QUOTE,
                             blockquote_startpos + 1);

    } else if (!indented && (matched = scan_atx_heading_start(
                                 input, parser->first_nonspace))) {
      bufsize_t hashpos;
      int level = 0;
      bufsize_t heading_startpos = parser->first_nonspace;

      S_advance_offset(parser, input,
                       parser->first_nonspace + matched - parser->offset,
                       false);
      *container = add_child(parser, *container, CMARK_NODE_HEADING,
                             heading_startpos + 1);

      hashpos = cmark_chunk_strchr(input, '#', parser->first_nonspace);

      while (peek_at(input, hashpos) == '#') {
        level++;
        hashpos++;
      }

      (*container)->as.heading.level = level;
      (*container)->as.heading.setext = false;
      (*container)->internal_offset = matched;

    } else if (!indented && (matched = scan_open_code_fence(
                                 input, parser->first_nonspace))) {
      *container = add_child(parser, *container, CMARK_NODE_CODE_BLOCK,
                             parser->first_nonspace + 1);
      (*container)->as.code.fenced = true;
      (*container)->as.code.fence_char = peek_at(input, parser->first_nonspace);
      (*container)->as.code.fence_length = (matched > 255) ? 255 : matched;
      (*container)->as.code.fence_offset =
          (int8_t)(parser->first_nonspace - parser->offset);
      (*container)->as.code.info = NULL;
      S_advance_offset(parser, input,
                       parser->first_nonspace + matched - parser->offset,
                       false);

    } else if (!indented && ((matched = scan_html_block_start(
                                  input, parser->first_nonspace)) ||
                             (cont_type != CMARK_NODE_PARAGRAPH &&
                              (matched = scan_html_block_start_7(
                                   input, parser->first_nonspace))))) {
      *container = add_child(parser, *container, CMARK_NODE_HTML_BLOCK,
                             parser->first_nonspace + 1);
      (*container)->as.html_block_type = matched;
      // note, we don't adjust parser->offset because the tag is part of the
      // text
    } else if (!indented && cont_type == CMARK_NODE_PARAGRAPH &&
               (lev =
                    scan_setext_heading_line(input, parser->first_nonspace))) {
      // finalize paragraph, resolving reference links
      has_content = resolve_reference_link_definitions(parser, *container);

      if (has_content) {

        (*container)->type = (uint16_t)CMARK_NODE_HEADING;
        (*container)->as.heading.level = lev;
        (*container)->as.heading.setext = true;
        S_advance_offset(parser, input, input->len - 1 - parser->offset, false);
      }
    } else if (!indented &&
               !(cont_type == CMARK_NODE_PARAGRAPH && !all_matched) &&
	       (parser->thematic_break_kill_pos <= parser->first_nonspace) &&
               (matched = S_scan_thematic_break(parser, input, parser->first_nonspace))) {
      // it's only now that we know the line is not part of a setext heading:
      *container = add_child(parser, *container, CMARK_NODE_THEMATIC_BREAK,
                             parser->first_nonspace + 1);
      S_advance_offset(parser, input, input->len - 1 - parser->offset, false);
    } else if ((!indented || cont_type == CMARK_NODE_LIST) &&
	       parser->indent < 4 &&
               (matched = parse_list_marker(
                    parser->mem, input, parser->first_nonspace,
                    (*container)->type == CMARK_NODE_PARAGRAPH, &data))) {

      // Note that we can have new list items starting with >= 4
      // spaces indent, as long as the list container is still open.
      int i = 0;

      // compute padding:
      S_advance_offset(parser, input,
                       parser->first_nonspace + matched - parser->offset,
                       false);

      save_partially_consumed_tab = parser->partially_consumed_tab;
      save_offset = parser->offset;
      save_column = parser->column;

      while (parser->column - save_column <= 5 &&
             S_is_space_or_tab(peek_at(input, parser->offset))) {
        S_advance_offset(parser, input, 1, true);
      }

      i = parser->column - save_column;
      if (i >= 5 || i < 1 ||
          // only spaces after list marker:
          S_is_line_end_char(peek_at(input, parser->offset))) {
        data->padding = matched + 1;
        parser->offset = save_offset;
        parser->column = save_column;
        parser->partially_consumed_tab = save_partially_consumed_tab;
        if (i > 0) {
          S_advance_offset(parser, input, 1, true);
        }
      } else {
        data->padding = matched + i;
      }

      // check container; if it's a list, see if this list item
      // can continue the list; otherwise, create a list container.

      data->marker_offset = parser->indent;

      if (cont_type != CMARK_NODE_LIST ||
          !lists_match(&((*container)->as.list), data)) {
        *container = add_child(parser, *container, CMARK_NODE_LIST,
                               parser->first_nonspace + 1);

        memcpy(&((*container)->as.list), data, sizeof(*data));
      }

      // add the list item
      *container = add_child(parser, *container, CMARK_NODE_ITEM,
                             parser->first_nonspace + 1);
      /* TODO: static */
      memcpy(&((*container)->as.list), data, sizeof(*data));
      parser->mem->free(data);
    } else if (indented && !maybe_lazy && !parser->blank) {
      S_advance_offset(parser, input, CODE_INDENT, true);
      *container = add_child(parser, *container, CMARK_NODE_CODE_BLOCK,
                             parser->offset + 1);
      (*container)->as.code.fenced = false;
      (*container)->as.code.fence_char = 0;
      (*container)->as.code.fence_length = 0;
      (*container)->as.code.fence_offset = 0;
      (*container)->as.code.info = NULL;

    } else {
      break;
    }

    if (accepts_lines(S_type(*container))) {
      // if it's a line container, it can't contain other containers
      break;
    }

    cont_type = S_type(*container);
    maybe_lazy = false;
  }
}

static void add_text_to_container(cmark_parser *parser, cmark_node *container,
                                  cmark_node *last_matched_container,
                                  cmark_chunk *input) {
  cmark_node *tmp;
  // what remains at parser->offset is a text line.  add the text to the
  // appropriate container.

  S_find_first_nonspace(parser, input);

  if (parser->blank && container->last_child)
    S_set_last_line_blank(container->last_child, true);

  // block quote lines are never blank as they start with >
  // and we don't count blanks in fenced code for purposes of tight/loose
  // lists or breaking out of lists.  we also don't set last_line_blank
  // on an empty list item.
  const cmark_node_type ctype = S_type(container);
  const bool last_line_blank =
      (parser->blank && ctype != CMARK_NODE_BLOCK_QUOTE &&
       ctype != CMARK_NODE_HEADING && ctype != CMARK_NODE_THEMATIC_BREAK &&
       !(ctype == CMARK_NODE_CODE_BLOCK && container->as.code.fenced) &&
       !(ctype == CMARK_NODE_ITEM && container->first_child == NULL &&
         container->start_line == parser->line_number));

  S_set_last_line_blank(container, last_line_blank);

  tmp = container;
  while (tmp->parent) {
    S_set_last_line_blank(tmp->parent, false);
    tmp = tmp->parent;
  }

  // If the last line processed belonged to a paragraph node,
  // and we didn't match all of the line prefixes for the open containers,
  // and we didn't start any new containers,
  // and the line isn't blank,
  // then treat this as a "lazy continuation line" and add it to
  // the open paragraph.
  if (parser->current != last_matched_container &&
      container == last_matched_container && !parser->blank &&
      S_type(parser->current) == CMARK_NODE_PARAGRAPH) {
    add_line(parser->current, input, parser);
  } else { // not a lazy continuation
    // Finalize any blocks that were not matched and set cur to container:
    while (parser->current != last_matched_container) {
      parser->current = finalize(parser, parser->current);
      assert(parser->current != NULL);
    }

    if (S_type(container) == CMARK_NODE_CODE_BLOCK) {
      add_line(container, input, parser);
    } else if (S_type(container) == CMARK_NODE_HTML_BLOCK) {
      add_line(container, input, parser);

      int matches_end_condition;
      switch (container->as.html_block_type) {
      case 1:
        // </script>, </style>, </pre>
        matches_end_condition =
            scan_html_block_end_1(input, parser->first_nonspace);
        break;
      case 2:
        // -->
        matches_end_condition =
            scan_html_block_end_2(input, parser->first_nonspace);
        break;
      case 3:
        // ?>
        matches_end_condition =
            scan_html_block_end_3(input, parser->first_nonspace);
        break;
      case 4:
        // >
        matches_end_condition =
            scan_html_block_end_4(input, parser->first_nonspace);
        break;
      case 5:
        // ]]>
        matches_end_condition =
            scan_html_block_end_5(input, parser->first_nonspace);
        break;
      default:
        matches_end_condition = 0;
        break;
      }

      if (matches_end_condition) {
        container = finalize(parser, container);
        assert(parser->current != NULL);
      }
    } else if (parser->blank) {
      // ??? do nothing
    } else if (accepts_lines(S_type(container))) {
      if (S_type(container) == CMARK_NODE_HEADING &&
          container->as.heading.setext == false) {
        chop_trailing_hashtags(input);
      }
      S_advance_offset(parser, input, parser->first_nonspace - parser->offset,
                       false);
      add_line(container, input, parser);
    } else {
      // create paragraph container for line
      container = add_child(parser, container, CMARK_NODE_PARAGRAPH,
                            parser->first_nonspace + 1);
      S_advance_offset(parser, input, parser->first_nonspace - parser->offset,
                       false);
      add_line(container, input, parser);
    }

    parser->current = container;
  }
}

/* See http://spec.commonmark.org/0.24/#phase-1-block-structure */
static void S_process_line(cmark_parser *parser, const unsigned char *buffer,
                           bufsize_t bytes) {
  cmark_node *last_matched_container;
  bool all_matched = true;
  cmark_node *container;
  cmark_chunk input;

  if (parser->options & CMARK_OPT_VALIDATE_UTF8)
    cmark_utf8proc_check(&parser->curline, buffer, bytes);
  else
    cmark_strbuf_put(&parser->curline, buffer, bytes);

  bytes = parser->curline.size;

  // ensure line ends with a newline:
  if (bytes == 0 || !S_is_line_end_char(parser->curline.ptr[bytes - 1]))
    cmark_strbuf_putc(&parser->curline, '\n');

  parser->offset = 0;
  parser->column = 0;
  parser->first_nonspace = 0;
  parser->first_nonspace_column = 0;
  parser->thematic_break_kill_pos = 0;
  parser->indent = 0;
  parser->blank = false;
  parser->partially_consumed_tab = false;

  input.data = parser->curline.ptr;
  input.len = parser->curline.size;
  input.alloc = 0;

  parser->line_number++;

  last_matched_container = check_open_blocks(parser, &input, &all_matched);

  if (!last_matched_container)
    goto finished;

  container = last_matched_container;

  open_new_blocks(parser, &container, &input, all_matched);

  add_text_to_container(parser, container, last_matched_container, &input);

finished:
  parser->last_line_length = input.len;
  if (parser->last_line_length &&
      input.data[parser->last_line_length - 1] == '\n')
    parser->last_line_length -= 1;
  if (parser->last_line_length &&
      input.data[parser->last_line_length - 1] == '\r')
    parser->last_line_length -= 1;

  cmark_strbuf_clear(&parser->curline);
}

cmark_node *cmark_parser_finish(cmark_parser *parser) {
  if (parser->linebuf.size) {
    S_process_line(parser, parser->linebuf.ptr, parser->linebuf.size);
    cmark_strbuf_clear(&parser->linebuf);
  }

  finalize_document(parser);

  cmark_consolidate_text_nodes(parser->root);

  cmark_strbuf_free(&parser->curline);

#if CMARK_DEBUG_NODES
  if (cmark_node_check(parser->root, stderr)) {
    abort();
  }
#endif
  return parser->root;
}