1
|
|
|
from __future__ import absolute_import, unicode_literals |
2
|
|
|
|
3
|
|
|
import re |
4
|
|
|
from importlib import import_module |
5
|
|
|
from CommonMark import common |
6
|
|
|
from CommonMark.common import unescape_string |
7
|
|
|
from CommonMark.inlines import InlineParser |
8
|
|
|
from CommonMark.node import Node |
9
|
|
|
from CommonMark.utils import to_camel_case |
10
|
|
|
|
11
|
|
|
|
12
|
|
|
CODE_INDENT = 4 |
13
|
|
|
reHtmlBlockOpen = [ |
14
|
|
|
re.compile(r'.'), # dummy for 0 |
15
|
|
|
re.compile(r'^<(?:script|pre|style)(?:\s|>|$)', re.IGNORECASE), |
16
|
|
|
re.compile(r'^<!--'), |
17
|
|
|
re.compile(r'^<[?]'), |
18
|
|
|
re.compile(r'^<![A-Z]'), |
19
|
|
|
re.compile(r'^<!\[CDATA\['), |
20
|
|
|
re.compile( |
21
|
|
|
r'^<[/]?(?:address|article|aside|base|basefont|blockquote|body|' |
22
|
|
|
r'caption|center|col|colgroup|dd|details|dialog|dir|div|dl|dt|' |
23
|
|
|
r'fieldset|figcaption|figure|footer|form|frame|frameset|h1|head|' |
24
|
|
|
r'header|hr|html|iframe|legend|li|link|main|menu|menuitem|meta|' |
25
|
|
|
r'nav|noframes|ol|optgroup|option|p|param|section|source|title|' |
26
|
|
|
r'summary|table|tbody|td|tfoot|th|thead|title|tr|track|ul)' |
27
|
|
|
r'(?:\s|[/]?[>]|$)', |
28
|
|
|
re.IGNORECASE), |
29
|
|
|
re.compile( |
30
|
|
|
'^(?:' + common.OPENTAG + '|' + common.CLOSETAG + ')\\s*$', |
31
|
|
|
re.IGNORECASE), |
32
|
|
|
] |
33
|
|
|
reHtmlBlockClose = [ |
34
|
|
|
re.compile(r'.'), # dummy for 0 |
35
|
|
|
re.compile(r'<\/(?:script|pre|style)>', re.IGNORECASE), |
36
|
|
|
re.compile(r'-->'), |
37
|
|
|
re.compile(r'\?>'), |
38
|
|
|
re.compile(r'>'), |
39
|
|
|
re.compile(r'\]\]>'), |
40
|
|
|
] |
41
|
|
|
reThematicBreak = re.compile( |
42
|
|
|
r'^(?:(?:\*[ \t]*){3,}|(?:_[ \t]*){3,}|(?:-[ \t]*){3,})[ \t]*$') |
43
|
|
|
reMaybeSpecial = re.compile(r'^[#`~*+_=<>0-9-]') |
44
|
|
|
reNonSpace = re.compile(r'[^ \t\f\v\r\n]') |
45
|
|
|
reBulletListMarker = re.compile(r'^[*+-]') |
46
|
|
|
reOrderedListMarker = re.compile(r'^(\d{1,9})([.)])') |
47
|
|
|
reATXHeadingMarker = re.compile(r'^#{1,6}(?:[ \t]+|$)') |
48
|
|
|
reCodeFence = re.compile(r'^`{3,}(?!.*`)|^~{3,}(?!.*~)') |
49
|
|
|
reClosingCodeFence = re.compile(r'^(?:`{3,}|~{3,})(?= *$)') |
50
|
|
|
reSetextHeadingLine = re.compile(r'^(?:=+|-+)[ \t]*$') |
51
|
|
|
reLineEnding = re.compile(r'\r\n|\n|\r') |
52
|
|
|
|
53
|
|
|
|
54
|
|
|
def is_blank(s): |
55
|
|
|
"""Returns True if string contains only space characters.""" |
56
|
|
|
return re.search(reNonSpace, s) is None |
57
|
|
|
|
58
|
|
|
|
59
|
|
|
def is_space_or_tab(s): |
60
|
|
|
return s == ' ' or s == '\t' |
61
|
|
|
|
62
|
|
|
|
63
|
|
|
def peek(ln, pos): |
64
|
|
|
if pos < len(ln): |
65
|
|
|
return ln[pos] |
66
|
|
|
else: |
67
|
|
|
return None |
68
|
|
|
|
69
|
|
|
|
70
|
|
|
def ends_with_blank_line(block): |
71
|
|
|
""" Returns true if block ends with a blank line, |
72
|
|
|
descending if needed into lists and sublists.""" |
73
|
|
|
while block: |
74
|
|
|
if block.last_line_blank: |
75
|
|
|
return True |
76
|
|
|
if (block.t == 'list' or block.t == 'item'): |
77
|
|
|
block = block.last_child |
78
|
|
|
else: |
79
|
|
|
break |
80
|
|
|
|
81
|
|
|
return False |
82
|
|
|
|
83
|
|
|
|
84
|
|
|
def parse_list_marker(parser, container): |
85
|
|
|
""" Parse a list marker and return data on the marker (type, |
86
|
|
|
start, delimiter, bullet character, padding) or None.""" |
87
|
|
|
rest = parser.current_line[parser.next_nonspace:] |
88
|
|
|
data = { |
89
|
|
|
'type': None, |
90
|
|
|
'tight': True, # lists are tight by default |
91
|
|
|
'bullet_char': None, |
92
|
|
|
'start': None, |
93
|
|
|
'delimiter': None, |
94
|
|
|
'padding': None, |
95
|
|
|
'marker_offset': parser.indent, |
96
|
|
|
} |
97
|
|
|
m = re.search(reBulletListMarker, rest) |
98
|
|
|
m2 = re.search(reOrderedListMarker, rest) |
99
|
|
|
if m: |
100
|
|
|
data['type'] = 'bullet' |
101
|
|
|
data['bullet_char'] = m.group()[0] |
102
|
|
|
elif m2 and (container.t != 'paragraph' or m2.group(1) == '1'): |
103
|
|
|
m = m2 |
104
|
|
|
data['type'] = 'ordered' |
105
|
|
|
data['start'] = int(m.group(1)) |
106
|
|
|
data['delimiter'] = m.group(2) |
107
|
|
|
else: |
108
|
|
|
return None |
109
|
|
|
|
110
|
|
|
# make sure we have spaces after |
111
|
|
|
nextc = peek(parser.current_line, parser.next_nonspace + len(m.group())) |
112
|
|
|
if not (nextc is None or nextc == '\t' or nextc == ' '): |
113
|
|
|
return None |
114
|
|
|
|
115
|
|
|
# if it interrupts paragraph, make sure first line isn't blank |
116
|
|
|
if container.t == 'paragraph' and \ |
117
|
|
|
not re.search( |
118
|
|
|
reNonSpace, |
119
|
|
|
parser.current_line[parser.next_nonspace + len(m.group()):]): |
120
|
|
|
return None |
121
|
|
|
|
122
|
|
|
# we've got a match! advance offset and calculate padding |
123
|
|
|
parser.advance_next_nonspace() # to start of marker |
124
|
|
|
parser.advance_offset(len(m.group()), True) # to end of marker |
125
|
|
|
spaces_start_col = parser.column |
126
|
|
|
spaces_start_offset = parser.offset |
127
|
|
|
while True: |
128
|
|
|
parser.advance_offset(1, True) |
129
|
|
|
nextc = peek(parser.current_line, parser.offset) |
130
|
|
|
if parser.column - spaces_start_col < 5 and \ |
131
|
|
|
is_space_or_tab(nextc): |
132
|
|
|
pass |
133
|
|
|
else: |
134
|
|
|
break |
135
|
|
|
blank_item = peek(parser.current_line, parser.offset) is None |
136
|
|
|
spaces_after_marker = parser.column - spaces_start_col |
137
|
|
|
if spaces_after_marker >= 5 or \ |
138
|
|
|
spaces_after_marker < 1 or \ |
139
|
|
|
blank_item: |
140
|
|
|
data['padding'] = len(m.group()) + 1 |
141
|
|
|
parser.column = spaces_start_col |
142
|
|
|
parser.offset = spaces_start_offset |
143
|
|
|
if is_space_or_tab(peek(parser.current_line, parser.offset)): |
144
|
|
|
parser.advance_offset(1, True) |
145
|
|
|
else: |
146
|
|
|
data['padding'] = len(m.group()) + spaces_after_marker |
147
|
|
|
|
148
|
|
|
return data |
149
|
|
|
|
150
|
|
|
|
151
|
|
|
def lists_match(list_data, item_data): |
152
|
|
|
""" |
153
|
|
|
Returns True if the two list items are of the same type, |
154
|
|
|
with the same delimiter and bullet character. This is used |
155
|
|
|
in agglomerating list items into lists. |
156
|
|
|
""" |
157
|
|
|
return list_data.get('type') == item_data.get('type') and \ |
158
|
|
|
list_data.get('delimiter') == item_data.get('delimiter') and \ |
159
|
|
|
list_data.get('bullet_char') == item_data.get('bullet_char') |
160
|
|
|
|
161
|
|
|
|
162
|
|
|
class Block(object): |
163
|
|
|
accepts_lines = None |
164
|
|
|
|
165
|
|
|
@staticmethod |
166
|
|
|
def continue_(parser=None, container=None): |
167
|
|
|
return |
168
|
|
|
|
169
|
|
|
@staticmethod |
170
|
|
|
def finalize(parser=None, block=None): |
171
|
|
|
return |
172
|
|
|
|
173
|
|
|
@staticmethod |
174
|
|
|
def can_contain(t): |
175
|
|
|
return |
176
|
|
|
|
177
|
|
|
|
178
|
|
|
class Document(Block): |
179
|
|
|
accepts_lines = False |
180
|
|
|
|
181
|
|
|
@staticmethod |
182
|
|
|
def continue_(parser=None, container=None): |
183
|
|
|
return 0 |
184
|
|
|
|
185
|
|
|
@staticmethod |
186
|
|
|
def finalize(parser=None, block=None): |
187
|
|
|
return |
188
|
|
|
|
189
|
|
|
@staticmethod |
190
|
|
|
def can_contain(t): |
191
|
|
|
return t != 'item' |
192
|
|
|
|
193
|
|
|
|
194
|
|
|
class List(Block): |
195
|
|
|
accepts_lines = False |
196
|
|
|
|
197
|
|
|
@staticmethod |
198
|
|
|
def continue_(parser=None, container=None): |
199
|
|
|
return 0 |
200
|
|
|
|
201
|
|
|
@staticmethod |
202
|
|
|
def finalize(parser=None, block=None): |
203
|
|
|
item = block.first_child |
204
|
|
|
while item: |
205
|
|
|
# check for non-final list item ending with blank line: |
206
|
|
|
if ends_with_blank_line(item) and item.nxt: |
207
|
|
|
block.list_data['tight'] = False |
208
|
|
|
break |
209
|
|
|
# recurse into children of list item, to see if there are |
210
|
|
|
# spaces between any of them: |
211
|
|
|
subitem = item.first_child |
212
|
|
|
while subitem: |
213
|
|
|
if ends_with_blank_line(subitem) and \ |
214
|
|
|
(item.nxt or subitem.nxt): |
215
|
|
|
block.list_data['tight'] = False |
216
|
|
|
break |
217
|
|
|
subitem = subitem.nxt |
218
|
|
|
item = item.nxt |
219
|
|
|
|
220
|
|
|
@staticmethod |
221
|
|
|
def can_contain(t): |
222
|
|
|
return t == 'item' |
223
|
|
|
|
224
|
|
|
|
225
|
|
|
class BlockQuote(Block): |
226
|
|
|
accepts_lines = False |
227
|
|
|
|
228
|
|
|
@staticmethod |
229
|
|
|
def continue_(parser=None, container=None): |
230
|
|
|
ln = parser.current_line |
231
|
|
|
if not parser.indented and peek(ln, parser.next_nonspace) == '>': |
232
|
|
|
parser.advance_next_nonspace() |
233
|
|
|
parser.advance_offset(1, False) |
234
|
|
|
if is_space_or_tab(peek(ln, parser.offset)): |
235
|
|
|
parser.advance_offset(1, True) |
236
|
|
|
else: |
237
|
|
|
return 1 |
238
|
|
|
return 0 |
239
|
|
|
|
240
|
|
|
@staticmethod |
241
|
|
|
def finalize(parser=None, block=None): |
242
|
|
|
return |
243
|
|
|
|
244
|
|
|
@staticmethod |
245
|
|
|
def can_contain(t): |
246
|
|
|
return t != 'item' |
247
|
|
|
|
248
|
|
|
|
249
|
|
|
class Item(Block): |
250
|
|
|
accepts_lines = False |
251
|
|
|
|
252
|
|
|
@staticmethod |
253
|
|
|
def continue_(parser=None, container=None): |
254
|
|
|
if parser.blank: |
255
|
|
|
if container.first_child is None: |
256
|
|
|
# Blank line after empty list item |
257
|
|
|
return 1 |
258
|
|
|
else: |
259
|
|
|
parser.advance_next_nonspace() |
260
|
|
|
elif parser.indent >= (container.list_data['marker_offset'] + |
261
|
|
|
container.list_data['padding']): |
262
|
|
|
parser.advance_offset( |
263
|
|
|
container.list_data['marker_offset'] + |
264
|
|
|
container.list_data['padding'], True) |
265
|
|
|
else: |
266
|
|
|
return 1 |
267
|
|
|
return 0 |
268
|
|
|
|
269
|
|
|
@staticmethod |
270
|
|
|
def finalize(parser=None, block=None): |
271
|
|
|
return |
272
|
|
|
|
273
|
|
|
@staticmethod |
274
|
|
|
def can_contain(t): |
275
|
|
|
return t != 'item' |
276
|
|
|
|
277
|
|
|
|
278
|
|
|
class Heading(Block): |
279
|
|
|
accepts_lines = False |
280
|
|
|
|
281
|
|
|
@staticmethod |
282
|
|
|
def continue_(parser=None, container=None): |
283
|
|
|
# A heading can never container > 1 line, so fail to match: |
284
|
|
|
return 1 |
285
|
|
|
|
286
|
|
|
@staticmethod |
287
|
|
|
def finalize(parser=None, block=None): |
288
|
|
|
return |
289
|
|
|
|
290
|
|
|
@staticmethod |
291
|
|
|
def can_contain(t): |
292
|
|
|
return False |
293
|
|
|
|
294
|
|
|
|
295
|
|
|
class ThematicBreak(Block): |
296
|
|
|
accepts_lines = False |
297
|
|
|
|
298
|
|
|
@staticmethod |
299
|
|
|
def continue_(parser=None, container=None): |
300
|
|
|
# A thematic break can never container > 1 line, so fail to match: |
301
|
|
|
return 1 |
302
|
|
|
|
303
|
|
|
@staticmethod |
304
|
|
|
def finalize(parser=None, block=None): |
305
|
|
|
return |
306
|
|
|
|
307
|
|
|
@staticmethod |
308
|
|
|
def can_contain(t): |
309
|
|
|
return False |
310
|
|
|
|
311
|
|
|
|
312
|
|
|
class CodeBlock(Block): |
313
|
|
|
accepts_lines = True |
314
|
|
|
|
315
|
|
|
@staticmethod |
316
|
|
|
def continue_(parser=None, container=None): |
317
|
|
|
ln = parser.current_line |
318
|
|
|
indent = parser.indent |
319
|
|
|
if container.is_fenced: |
320
|
|
|
match = indent <= 3 and \ |
321
|
|
|
len(ln) >= parser.next_nonspace + 1 and \ |
322
|
|
|
ln[parser.next_nonspace] == container.fence_char and \ |
323
|
|
|
re.search(reClosingCodeFence, ln[parser.next_nonspace:]) |
324
|
|
|
if match and len(match.group()) >= container.fence_length: |
325
|
|
|
# closing fence - we're at end of line, so we can return |
326
|
|
|
parser.finalize(container, parser.line_number) |
327
|
|
|
return 2 |
328
|
|
|
else: |
329
|
|
|
# skip optional spaces of fence offset |
330
|
|
|
i = container.fence_offset |
331
|
|
|
while i > 0 and is_space_or_tab(peek(ln, parser.offset)): |
332
|
|
|
parser.advance_offset(1, True) |
333
|
|
|
i -= 1 |
334
|
|
|
else: |
335
|
|
|
# indented |
336
|
|
|
if indent >= CODE_INDENT: |
337
|
|
|
parser.advance_offset(CODE_INDENT, True) |
338
|
|
|
elif parser.blank: |
339
|
|
|
parser.advance_next_nonspace() |
340
|
|
|
else: |
341
|
|
|
return 1 |
342
|
|
|
return 0 |
343
|
|
|
|
344
|
|
|
@staticmethod |
345
|
|
|
def finalize(parser=None, block=None): |
346
|
|
|
if block.is_fenced: |
347
|
|
|
# first line becomes info string |
348
|
|
|
content = block.string_content |
349
|
|
|
newline_pos = content.index('\n') |
350
|
|
|
first_line = content[0:newline_pos] |
351
|
|
|
rest = content[newline_pos + 1:] |
352
|
|
|
block.info = unescape_string(first_line.strip()) |
353
|
|
|
block.literal = rest |
354
|
|
|
else: |
355
|
|
|
# indented |
356
|
|
|
block.literal = re.sub(r'(\n *)+$', '\n', block.string_content) |
357
|
|
|
|
358
|
|
|
block.string_content = None |
359
|
|
|
|
360
|
|
|
@staticmethod |
361
|
|
|
def can_contain(t): |
362
|
|
|
return False |
363
|
|
|
|
364
|
|
|
|
365
|
|
|
class HtmlBlock(Block): |
366
|
|
|
accepts_lines = True |
367
|
|
|
|
368
|
|
|
@staticmethod |
369
|
|
|
def continue_(parser=None, container=None): |
370
|
|
|
if parser.blank and (container.html_block_type == 6 or |
371
|
|
|
container.html_block_type == 7): |
372
|
|
|
return 1 |
373
|
|
|
else: |
374
|
|
|
return 0 |
375
|
|
|
|
376
|
|
|
@staticmethod |
377
|
|
|
def finalize(parser=None, block=None): |
378
|
|
|
block.literal = re.sub(r'(\n *)+$', '', block.string_content) |
379
|
|
|
# allow GC |
380
|
|
|
block.string_content = None |
381
|
|
|
|
382
|
|
|
@staticmethod |
383
|
|
|
def can_contain(t): |
384
|
|
|
return False |
385
|
|
|
|
386
|
|
|
|
387
|
|
|
class Paragraph(Block): |
388
|
|
|
accepts_lines = True |
389
|
|
|
|
390
|
|
|
@staticmethod |
391
|
|
|
def continue_(parser=None, container=None): |
392
|
|
|
return 1 if parser.blank else 0 |
393
|
|
|
|
394
|
|
|
@staticmethod |
395
|
|
|
def finalize(parser=None, block=None): |
396
|
|
|
has_reference_defs = False |
397
|
|
|
|
398
|
|
|
# try parsing the beginning as link reference definitions: |
399
|
|
|
while peek(block.string_content, 0) == '[': |
400
|
|
|
pos = parser.inline_parser.parseReference( |
401
|
|
|
block.string_content, parser.refmap) |
402
|
|
|
if not pos: |
403
|
|
|
break |
404
|
|
|
block.string_content = block.string_content[pos:] |
405
|
|
|
has_reference_defs = True |
406
|
|
|
if has_reference_defs and is_blank(block.string_content): |
407
|
|
|
block.unlink() |
408
|
|
|
|
409
|
|
|
@staticmethod |
410
|
|
|
def can_contain(t): |
411
|
|
|
return False |
412
|
|
|
|
413
|
|
|
|
414
|
|
|
class BlockStarts(object): |
415
|
|
|
"""Block start functions. |
416
|
|
|
|
417
|
|
|
Return values: |
418
|
|
|
0 = no match |
419
|
|
|
1 = matched container, keep going |
420
|
|
|
2 = matched leaf, no more block starts |
421
|
|
|
""" |
422
|
|
|
METHODS = [ |
423
|
|
|
'block_quote', |
424
|
|
|
'atx_heading', |
425
|
|
|
'fenced_code_block', |
426
|
|
|
'html_block', |
427
|
|
|
'setext_heading', |
428
|
|
|
'thematic_break', |
429
|
|
|
'list_item', |
430
|
|
|
'indented_code_block', |
431
|
|
|
] |
432
|
|
|
|
433
|
|
|
@staticmethod |
434
|
|
|
def block_quote(parser, container=None): |
435
|
|
|
if not parser.indented and \ |
436
|
|
|
peek(parser.current_line, parser.next_nonspace) == '>': |
437
|
|
|
parser.advance_next_nonspace() |
438
|
|
|
parser.advance_offset(1, False) |
439
|
|
|
# optional following space |
440
|
|
|
if is_space_or_tab(peek(parser.current_line, parser.offset)): |
441
|
|
|
parser.advance_offset(1, True) |
442
|
|
|
parser.close_unmatched_blocks() |
443
|
|
|
parser.add_child('block_quote', parser.next_nonspace) |
444
|
|
|
return 1 |
445
|
|
|
|
446
|
|
|
return 0 |
447
|
|
|
|
448
|
|
|
@staticmethod |
449
|
|
|
def atx_heading(parser, container=None): |
450
|
|
|
if not parser.indented: |
451
|
|
|
m = re.search(reATXHeadingMarker, |
452
|
|
|
parser.current_line[parser.next_nonspace:]) |
453
|
|
|
if m: |
454
|
|
|
parser.advance_next_nonspace() |
455
|
|
|
parser.advance_offset(len(m.group()), False) |
456
|
|
|
parser.close_unmatched_blocks() |
457
|
|
|
container = parser.add_child('heading', parser.next_nonspace) |
458
|
|
|
# number of #s |
459
|
|
|
container.level = len(m.group().strip()) |
460
|
|
|
# remove trailing ###s: |
461
|
|
|
container.string_content = re.sub( |
462
|
|
|
r'[ \t]+#+[ \t]*$', '', re.sub( |
463
|
|
|
r'^[ \t]*#+[ \t]*$', |
464
|
|
|
'', |
465
|
|
|
parser.current_line[parser.offset:])) |
466
|
|
|
parser.advance_offset( |
467
|
|
|
len(parser.current_line) - parser.offset, False) |
468
|
|
|
return 2 |
469
|
|
|
|
470
|
|
|
return 0 |
471
|
|
|
|
472
|
|
|
@staticmethod |
473
|
|
|
def fenced_code_block(parser, container=None): |
474
|
|
|
if not parser.indented: |
475
|
|
|
m = re.search( |
476
|
|
|
reCodeFence, |
477
|
|
|
parser.current_line[parser.next_nonspace:]) |
478
|
|
|
if m: |
479
|
|
|
fence_length = len(m.group()) |
480
|
|
|
parser.close_unmatched_blocks() |
481
|
|
|
container = parser.add_child( |
482
|
|
|
'code_block', parser.next_nonspace) |
483
|
|
|
container.is_fenced = True |
484
|
|
|
container.fence_length = fence_length |
485
|
|
|
container.fence_char = m.group()[0] |
486
|
|
|
container.fence_offset = parser.indent |
487
|
|
|
parser.advance_next_nonspace() |
488
|
|
|
parser.advance_offset(fence_length, False) |
489
|
|
|
return 2 |
490
|
|
|
|
491
|
|
|
return 0 |
492
|
|
|
|
493
|
|
|
@staticmethod |
494
|
|
|
def html_block(parser, container=None): |
495
|
|
|
if not parser.indented and \ |
496
|
|
|
peek(parser.current_line, parser.next_nonspace) == '<': |
497
|
|
|
s = parser.current_line[parser.next_nonspace:] |
498
|
|
|
|
499
|
|
|
for block_type in range(1, 8): |
500
|
|
|
if re.search(reHtmlBlockOpen[block_type], s) and \ |
501
|
|
|
(block_type < 7 or container.t != 'paragraph'): |
502
|
|
|
parser.close_unmatched_blocks() |
503
|
|
|
# We don't adjust parser.offset; |
504
|
|
|
# spaces are part of the HTML block: |
505
|
|
|
b = parser.add_child('html_block', parser.offset) |
506
|
|
|
b.html_block_type = block_type |
507
|
|
|
return 2 |
508
|
|
|
return 0 |
509
|
|
|
|
510
|
|
|
@staticmethod |
511
|
|
|
def setext_heading(parser, container=None): |
512
|
|
|
if not parser.indented and container.t == 'paragraph': |
513
|
|
|
m = re.search( |
514
|
|
|
reSetextHeadingLine, |
515
|
|
|
parser.current_line[parser.next_nonspace:]) |
516
|
|
|
if m: |
517
|
|
|
parser.close_unmatched_blocks() |
518
|
|
|
heading = Node('heading', container.sourcepos) |
519
|
|
|
heading.level = 1 if m.group()[0] == '=' else 2 |
520
|
|
|
heading.string_content = container.string_content |
521
|
|
|
container.insert_after(heading) |
522
|
|
|
container.unlink() |
523
|
|
|
parser.tip = heading |
524
|
|
|
parser.advance_offset( |
525
|
|
|
len(parser.current_line) - parser.offset, False) |
526
|
|
|
return 2 |
527
|
|
|
|
528
|
|
|
return 0 |
529
|
|
|
|
530
|
|
|
@staticmethod |
531
|
|
|
def thematic_break(parser, container=None): |
532
|
|
|
if not parser.indented and re.search( |
533
|
|
|
reThematicBreak, parser.current_line[parser.next_nonspace:]): |
534
|
|
|
parser.close_unmatched_blocks() |
535
|
|
|
parser.add_child('thematic_break', parser.next_nonspace) |
536
|
|
|
parser.advance_offset( |
537
|
|
|
len(parser.current_line) - parser.offset, False) |
538
|
|
|
return 2 |
539
|
|
|
return 0 |
540
|
|
|
|
541
|
|
|
@staticmethod |
542
|
|
|
def list_item(parser, container=None): |
543
|
|
|
if (not parser.indented or container.t == 'list'): |
544
|
|
|
data = parse_list_marker(parser, container) |
545
|
|
|
if data: |
546
|
|
|
parser.close_unmatched_blocks() |
547
|
|
|
|
548
|
|
|
# add the list if needed |
549
|
|
|
if parser.tip.t != 'list' or \ |
550
|
|
|
not lists_match(container.list_data, data): |
551
|
|
|
container = parser.add_child('list', parser.next_nonspace) |
552
|
|
|
container.list_data = data |
553
|
|
|
|
554
|
|
|
# add the list item |
555
|
|
|
container = parser.add_child('item', parser.next_nonspace) |
556
|
|
|
container.list_data = data |
557
|
|
|
return 1 |
558
|
|
|
|
559
|
|
|
return 0 |
560
|
|
|
|
561
|
|
|
@staticmethod |
562
|
|
|
def indented_code_block(parser, container=None): |
563
|
|
|
if parser.indented and \ |
564
|
|
|
parser.tip.t != 'paragraph' and \ |
565
|
|
|
not parser.blank: |
566
|
|
|
# indented code |
567
|
|
|
parser.advance_offset(CODE_INDENT, True) |
568
|
|
|
parser.close_unmatched_blocks() |
569
|
|
|
parser.add_child('code_block', parser.offset) |
570
|
|
|
return 2 |
571
|
|
|
|
572
|
|
|
return 0 |
573
|
|
|
|
574
|
|
|
|
575
|
|
|
class Parser(object): |
576
|
|
|
def __init__(self, options={}): |
577
|
|
|
self.doc = Node('document', [[1, 1], [0, 0]]) |
578
|
|
|
self.block_starts = BlockStarts() |
579
|
|
|
self.tip = self.doc |
580
|
|
|
self.oldtip = self.doc |
581
|
|
|
self.current_line = '' |
582
|
|
|
self.line_number = 0 |
583
|
|
|
self.offset = 0 |
584
|
|
|
self.column = 0 |
585
|
|
|
self.next_nonspace = 0 |
586
|
|
|
self.next_nonspace_column = 0 |
587
|
|
|
self.indent = 0 |
588
|
|
|
self.indented = False |
589
|
|
|
self.blank = False |
590
|
|
|
self.partially_consumed_tab = False |
591
|
|
|
self.all_closed = True |
592
|
|
|
self.last_matched_container = self.doc |
593
|
|
|
self.refmap = {} |
594
|
|
|
self.last_line_length = 0 |
595
|
|
|
self.inline_parser = InlineParser(options) |
596
|
|
|
self.options = options |
597
|
|
|
|
598
|
|
|
def add_line(self): |
599
|
|
|
""" Add a line to the block at the tip. We assume the tip |
600
|
|
|
can accept lines -- that check should be done before calling this.""" |
601
|
|
|
if self.partially_consumed_tab: |
602
|
|
|
# Skip over tab |
603
|
|
|
self.offset += 1 |
604
|
|
|
# Add space characters |
605
|
|
|
chars_to_tab = 4 - (self.column % 4) |
606
|
|
|
self.tip.string_content += (' ' * chars_to_tab) |
607
|
|
|
self.tip.string_content += (self.current_line[self.offset:] + '\n') |
608
|
|
|
|
609
|
|
|
def add_child(self, tag, offset): |
610
|
|
|
""" Add block of type tag as a child of the tip. If the tip can't |
611
|
|
|
accept children, close and finalize it and try its parent, |
612
|
|
|
and so on til we find a block that can accept children.""" |
613
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
614
|
|
|
to_camel_case(self.tip.t)) |
615
|
|
|
while not block_class.can_contain(tag): |
616
|
|
|
self.finalize(self.tip, self.line_number - 1) |
617
|
|
|
block_class = getattr( |
618
|
|
|
import_module('CommonMark.blocks'), |
619
|
|
|
to_camel_case(self.tip.t)) |
620
|
|
|
|
621
|
|
|
column_number = offset + 1 |
622
|
|
|
new_block = Node(tag, [[self.line_number, column_number], [0, 0]]) |
623
|
|
|
new_block.string_content = '' |
624
|
|
|
self.tip.append_child(new_block) |
625
|
|
|
self.tip = new_block |
626
|
|
|
return new_block |
627
|
|
|
|
628
|
|
|
def close_unmatched_blocks(self): |
629
|
|
|
"""Finalize and close any unmatched blocks.""" |
630
|
|
|
if not self.all_closed: |
631
|
|
|
while self.oldtip != self.last_matched_container: |
632
|
|
|
parent = self.oldtip.parent |
633
|
|
|
self.finalize(self.oldtip, self.line_number - 1) |
634
|
|
|
self.oldtip = parent |
635
|
|
|
self.all_closed = True |
636
|
|
|
|
637
|
|
|
def find_next_nonspace(self): |
638
|
|
|
current_line = self.current_line |
639
|
|
|
i = self.offset |
640
|
|
|
cols = self.column |
641
|
|
|
|
642
|
|
|
try: |
643
|
|
|
c = current_line[i] |
644
|
|
|
except IndexError: |
645
|
|
|
c = '' |
646
|
|
|
while c != '': |
647
|
|
|
if c == ' ': |
648
|
|
|
i += 1 |
649
|
|
|
cols += 1 |
650
|
|
|
elif c == '\t': |
651
|
|
|
i += 1 |
652
|
|
|
cols += (4 - (cols % 4)) |
653
|
|
|
else: |
654
|
|
|
break |
655
|
|
|
|
656
|
|
|
try: |
657
|
|
|
c = current_line[i] |
658
|
|
|
except IndexError: |
659
|
|
|
c = '' |
660
|
|
|
|
661
|
|
|
self.blank = (c == '\n' or c == '\r' or c == '') |
662
|
|
|
self.next_nonspace = i |
663
|
|
|
self.next_nonspace_column = cols |
664
|
|
|
self.indent = self.next_nonspace_column - self.column |
665
|
|
|
self.indented = self.indent >= CODE_INDENT |
666
|
|
|
|
667
|
|
|
def advance_next_nonspace(self): |
668
|
|
|
self.offset = self.next_nonspace |
669
|
|
|
self.column = self.next_nonspace_column |
670
|
|
|
self.partially_consumed_tab = False |
671
|
|
|
|
672
|
|
|
def advance_offset(self, count, columns): |
673
|
|
|
current_line = self.current_line |
674
|
|
|
try: |
675
|
|
|
c = current_line[self.offset] |
676
|
|
|
except IndexError: |
677
|
|
|
c = None |
678
|
|
|
while count > 0 and c is not None: |
679
|
|
|
if c == '\t': |
680
|
|
|
chars_to_tab = 4 - (self.column % 4) |
681
|
|
|
if columns: |
682
|
|
|
self.partially_consumed_tab = chars_to_tab > count |
683
|
|
|
chars_to_advance = min(count, chars_to_tab) |
684
|
|
|
self.column += chars_to_advance |
685
|
|
|
self.offset += 0 if self.partially_consumed_tab else 1 |
686
|
|
|
count -= chars_to_advance |
687
|
|
|
else: |
688
|
|
|
self.partially_consumed_tab = False |
689
|
|
|
self.column += chars_to_tab |
690
|
|
|
self.offset += 1 |
691
|
|
|
count -= 1 |
692
|
|
|
else: |
693
|
|
|
self.partially_consumed_tab = False |
694
|
|
|
self.offset += 1 |
695
|
|
|
# assume ascii; block starts are ascii |
696
|
|
|
self.column += 1 |
697
|
|
|
count -= 1 |
698
|
|
|
try: |
699
|
|
|
c = current_line[self.offset] |
700
|
|
|
except IndexError: |
701
|
|
|
c = None |
702
|
|
|
|
703
|
|
|
def incorporate_line(self, ln): |
704
|
|
|
"""Analyze a line of text and update the document appropriately. |
705
|
|
|
|
706
|
|
|
We parse markdown text by calling this on each line of input, |
707
|
|
|
then finalizing the document. |
708
|
|
|
""" |
709
|
|
|
all_matched = True |
710
|
|
|
|
711
|
|
|
container = self.doc |
712
|
|
|
self.oldtip = self.tip |
713
|
|
|
self.offset = 0 |
714
|
|
|
self.column = 0 |
715
|
|
|
self.blank = False |
716
|
|
|
self.partially_consumed_tab = False |
717
|
|
|
self.line_number += 1 |
718
|
|
|
|
719
|
|
|
# replace NUL characters for security |
720
|
|
|
if re.search(r'\u0000', ln) is not None: |
721
|
|
|
ln = re.sub(r'\0', '\uFFFD', ln) |
722
|
|
|
|
723
|
|
|
self.current_line = ln |
724
|
|
|
|
725
|
|
|
# For each containing block, try to parse the associated line start. |
726
|
|
|
# Bail out on failure: container will point to the last matching block. |
727
|
|
|
# Set all_matched to false if not all containers match. |
728
|
|
|
last_child = container.last_child |
729
|
|
|
while last_child and last_child.is_open: |
730
|
|
|
container = last_child |
731
|
|
|
|
732
|
|
|
self.find_next_nonspace() |
733
|
|
|
block_class = getattr( |
734
|
|
|
import_module('CommonMark.blocks'), |
735
|
|
|
to_camel_case(container.t)) |
736
|
|
|
rv = block_class.continue_(self, container) |
737
|
|
|
if rv == 0: |
738
|
|
|
# we've matched, keep going |
739
|
|
|
pass |
740
|
|
|
elif rv == 1: |
741
|
|
|
# we've failed to match a block |
742
|
|
|
all_matched = False |
743
|
|
|
elif rv == 2: |
744
|
|
|
# we've hit end of line for fenced code close and can return |
745
|
|
|
self.last_line_length = len(ln) |
746
|
|
|
return |
747
|
|
|
else: |
748
|
|
|
raise ValueError('returned illegal value, must be 0, 1, or 2') |
749
|
|
|
|
750
|
|
|
if not all_matched: |
751
|
|
|
# back up to last matching block |
752
|
|
|
container = container.parent |
753
|
|
|
break |
754
|
|
|
|
755
|
|
|
last_child = container.last_child |
756
|
|
|
|
757
|
|
|
self.all_closed = (container == self.oldtip) |
758
|
|
|
self.last_matched_container = container |
759
|
|
|
|
760
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
761
|
|
|
to_camel_case(container.t)) |
762
|
|
|
matched_leaf = container.t != 'paragraph' and block_class.accepts_lines |
763
|
|
|
starts = self.block_starts |
764
|
|
|
starts_len = len(starts.METHODS) |
765
|
|
|
# Unless last matched container is a code block, try new container |
766
|
|
|
# starts, adding children to the last matched container: |
767
|
|
|
while not matched_leaf: |
768
|
|
|
self.find_next_nonspace() |
769
|
|
|
|
770
|
|
|
# this is a little performance optimization: |
771
|
|
|
if not self.indented and \ |
772
|
|
|
not re.search(reMaybeSpecial, ln[self.next_nonspace:]): |
773
|
|
|
self.advance_next_nonspace() |
774
|
|
|
break |
775
|
|
|
|
776
|
|
|
i = 0 |
777
|
|
|
while i < starts_len: |
778
|
|
|
res = getattr(starts, starts.METHODS[i])(self, container) |
779
|
|
|
if res == 1: |
780
|
|
|
container = self.tip |
781
|
|
|
break |
782
|
|
|
elif res == 2: |
783
|
|
|
container = self.tip |
784
|
|
|
matched_leaf = True |
785
|
|
|
break |
786
|
|
|
else: |
787
|
|
|
i += 1 |
788
|
|
|
|
789
|
|
|
if i == starts_len: |
790
|
|
|
# nothing matched |
791
|
|
|
self.advance_next_nonspace() |
792
|
|
|
break |
793
|
|
|
|
794
|
|
|
# What remains at the offset is a text line. Add the text to the |
795
|
|
|
# appropriate container. |
796
|
|
|
if not self.all_closed and not self.blank and \ |
797
|
|
|
self.tip.t == 'paragraph': |
798
|
|
|
# lazy paragraph continuation |
799
|
|
|
self.add_line() |
800
|
|
|
else: |
801
|
|
|
# not a lazy continuation |
802
|
|
|
# finalize any blocks not matched |
803
|
|
|
self.close_unmatched_blocks() |
804
|
|
|
if self.blank and container.last_child: |
805
|
|
|
container.last_child.last_line_blank = True |
806
|
|
|
|
807
|
|
|
t = container.t |
808
|
|
|
|
809
|
|
|
# Block quote lines are never blank as they start with > |
810
|
|
|
# and we don't count blanks in fenced code for purposes of |
811
|
|
|
# tight/loose lists or breaking out of lists. We also |
812
|
|
|
# don't set last_line_blank on an empty list item, or if we |
813
|
|
|
# just closed a fenced block. |
814
|
|
|
last_line_blank = self.blank and \ |
815
|
|
|
not (t == 'block_quote' or |
816
|
|
|
(t == 'code_block' and container.is_fenced) or |
817
|
|
|
(t == 'item' and |
818
|
|
|
not container.first_child and |
819
|
|
|
container.sourcepos[0][0] == self.line_number)) |
820
|
|
|
|
821
|
|
|
# propagate last_line_blank up through parents: |
822
|
|
|
cont = container |
823
|
|
|
while cont: |
824
|
|
|
cont.last_line_blank = last_line_blank |
825
|
|
|
cont = cont.parent |
826
|
|
|
|
827
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
828
|
|
|
to_camel_case(t)) |
829
|
|
|
if block_class.accepts_lines: |
830
|
|
|
self.add_line() |
831
|
|
|
# if HtmlBlock, check for end condition |
832
|
|
|
if t == 'html_block' and \ |
833
|
|
|
container.html_block_type >= 1 and \ |
834
|
|
|
container.html_block_type <= 5 and \ |
835
|
|
|
re.search( |
836
|
|
|
reHtmlBlockClose[container.html_block_type], |
837
|
|
|
self.current_line[self.offset:]): |
838
|
|
|
self.finalize(container, self.line_number) |
839
|
|
|
elif self.offset < len(ln) and not self.blank: |
840
|
|
|
# create a paragraph container for one line |
841
|
|
|
container = self.add_child('paragraph', self.offset) |
842
|
|
|
self.advance_next_nonspace() |
843
|
|
|
self.add_line() |
844
|
|
|
|
845
|
|
|
self.last_line_length = len(ln) |
846
|
|
|
|
847
|
|
|
def finalize(self, block, line_number): |
848
|
|
|
""" Finalize a block. Close it and do any necessary postprocessing, |
849
|
|
|
e.g. creating string_content from strings, setting the 'tight' |
850
|
|
|
or 'loose' status of a list, and parsing the beginnings |
851
|
|
|
of paragraphs for reference definitions. Reset the tip to the |
852
|
|
|
parent of the closed block.""" |
853
|
|
|
above = block.parent |
854
|
|
|
block.is_open = False |
855
|
|
|
block.sourcepos[1] = [line_number, self.last_line_length] |
856
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
857
|
|
|
to_camel_case(block.t)) |
858
|
|
|
block_class.finalize(self, block) |
859
|
|
|
|
860
|
|
|
self.tip = above |
861
|
|
|
|
862
|
|
|
def process_inlines(self, block): |
863
|
|
|
""" |
864
|
|
|
Walk through a block & children recursively, parsing string content |
865
|
|
|
into inline content where appropriate. |
866
|
|
|
""" |
867
|
|
|
walker = block.walker() |
868
|
|
|
self.inline_parser.refmap = self.refmap |
869
|
|
|
self.inline_parser.options = self.options |
870
|
|
|
event = walker.nxt() |
871
|
|
|
while event is not None: |
872
|
|
|
node = event['node'] |
873
|
|
|
t = node.t |
874
|
|
|
if not event['entering'] and (t == 'paragraph' or t == 'heading'): |
875
|
|
|
self.inline_parser.parse(node) |
876
|
|
|
event = walker.nxt() |
877
|
|
|
|
878
|
|
|
def parse(self, my_input): |
879
|
|
|
""" The main parsing function. Returns a parsed document AST.""" |
880
|
|
|
self.doc = Node('document', [[1, 1], [0, 0]]) |
881
|
|
|
self.tip = self.doc |
882
|
|
|
self.refmap = {} |
883
|
|
|
self.line_number = 0 |
884
|
|
|
self.last_line_length = 0 |
885
|
|
|
self.offset = 0 |
886
|
|
|
self.column = 0 |
887
|
|
|
self.last_matched_container = self.doc |
888
|
|
|
self.current_line = '' |
889
|
|
|
lines = re.split(reLineEnding, my_input) |
890
|
|
|
length = len(lines) |
891
|
|
|
if len(my_input) > 0 and my_input[-1] == '\n': |
892
|
|
|
# ignore last blank line created by final newline |
893
|
|
|
length -= 1 |
894
|
|
|
for i in range(length): |
895
|
|
|
self.incorporate_line(lines[i]) |
896
|
|
|
while (self.tip): |
897
|
|
|
self.finalize(self.tip, length) |
898
|
|
|
self.process_inlines(self.doc) |
899
|
|
|
return self.doc |
900
|
|
|
|