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
|
|
|
if s is not None: |
61
|
|
|
s = unicode(s, 'latin1').encode('utf-8') |
62
|
|
|
return s == ' ' or s == '\t' |
63
|
|
|
|
64
|
|
|
|
65
|
|
|
def peek(ln, pos): |
66
|
|
|
if pos < len(ln): |
67
|
|
|
return ln[pos] |
68
|
|
|
else: |
69
|
|
|
return None |
70
|
|
|
|
71
|
|
|
|
72
|
|
|
def ends_with_blank_line(block): |
73
|
|
|
""" Returns true if block ends with a blank line, |
74
|
|
|
descending if needed into lists and sublists.""" |
75
|
|
|
while block: |
76
|
|
|
if block.last_line_blank: |
77
|
|
|
return True |
78
|
|
|
if (block.t == 'list' or block.t == 'item'): |
79
|
|
|
block = block.last_child |
80
|
|
|
else: |
81
|
|
|
break |
82
|
|
|
|
83
|
|
|
return False |
84
|
|
|
|
85
|
|
|
|
86
|
|
|
def parse_list_marker(parser, container): |
87
|
|
|
""" Parse a list marker and return data on the marker (type, |
88
|
|
|
start, delimiter, bullet character, padding) or None.""" |
89
|
|
|
rest = parser.current_line[parser.next_nonspace:] |
90
|
|
|
data = { |
91
|
|
|
'type': None, |
92
|
|
|
'tight': True, # lists are tight by default |
93
|
|
|
'bullet_char': None, |
94
|
|
|
'start': None, |
95
|
|
|
'delimiter': None, |
96
|
|
|
'padding': None, |
97
|
|
|
'marker_offset': parser.indent, |
98
|
|
|
} |
99
|
|
|
m = re.search(reBulletListMarker, rest) |
100
|
|
|
m2 = re.search(reOrderedListMarker, rest) |
101
|
|
|
if m: |
102
|
|
|
data['type'] = 'bullet' |
103
|
|
|
data['bullet_char'] = m.group()[0] |
104
|
|
|
elif m2 and (container.t != 'paragraph' or m2.group(1) == '1'): |
105
|
|
|
m = m2 |
106
|
|
|
data['type'] = 'ordered' |
107
|
|
|
data['start'] = int(m.group(1)) |
108
|
|
|
data['delimiter'] = m.group(2) |
109
|
|
|
else: |
110
|
|
|
return None |
111
|
|
|
|
112
|
|
|
# make sure we have spaces after |
113
|
|
|
nextc = peek(parser.current_line, parser.next_nonspace + len(m.group())) |
114
|
|
|
if not (nextc is None or nextc == '\t' or nextc == ' '): |
115
|
|
|
return None |
116
|
|
|
|
117
|
|
|
# if it interrupts paragraph, make sure first line isn't blank |
118
|
|
|
if container.t == 'paragraph' and \ |
119
|
|
|
not re.search( |
120
|
|
|
reNonSpace, |
121
|
|
|
parser.current_line[parser.next_nonspace + len(m.group()):]): |
122
|
|
|
return None |
123
|
|
|
|
124
|
|
|
# we've got a match! advance offset and calculate padding |
125
|
|
|
parser.advance_next_nonspace() # to start of marker |
126
|
|
|
parser.advance_offset(len(m.group()), True) # to end of marker |
127
|
|
|
spaces_start_col = parser.column |
128
|
|
|
spaces_start_offset = parser.offset |
129
|
|
|
while True: |
130
|
|
|
parser.advance_offset(1, True) |
131
|
|
|
nextc = peek(parser.current_line, parser.offset) |
132
|
|
|
if parser.column - spaces_start_col < 5 and \ |
133
|
|
|
is_space_or_tab(nextc): |
134
|
|
|
pass |
135
|
|
|
else: |
136
|
|
|
break |
137
|
|
|
blank_item = peek(parser.current_line, parser.offset) is None |
138
|
|
|
spaces_after_marker = parser.column - spaces_start_col |
139
|
|
|
if spaces_after_marker >= 5 or \ |
140
|
|
|
spaces_after_marker < 1 or \ |
141
|
|
|
blank_item: |
142
|
|
|
data['padding'] = len(m.group()) + 1 |
143
|
|
|
parser.column = spaces_start_col |
144
|
|
|
parser.offset = spaces_start_offset |
145
|
|
|
if is_space_or_tab(peek(parser.current_line, parser.offset)): |
146
|
|
|
parser.advance_offset(1, True) |
147
|
|
|
else: |
148
|
|
|
data['padding'] = len(m.group()) + spaces_after_marker |
149
|
|
|
|
150
|
|
|
return data |
151
|
|
|
|
152
|
|
|
|
153
|
|
|
def lists_match(list_data, item_data): |
154
|
|
|
""" |
155
|
|
|
Returns True if the two list items are of the same type, |
156
|
|
|
with the same delimiter and bullet character. This is used |
157
|
|
|
in agglomerating list items into lists. |
158
|
|
|
""" |
159
|
|
|
return list_data.get('type') == item_data.get('type') and \ |
160
|
|
|
list_data.get('delimiter') == item_data.get('delimiter') and \ |
161
|
|
|
list_data.get('bullet_char') == item_data.get('bullet_char') |
162
|
|
|
|
163
|
|
|
|
164
|
|
|
class Block(object): |
165
|
|
|
accepts_lines = None |
166
|
|
|
|
167
|
|
|
@staticmethod |
168
|
|
|
def continue_(parser=None, container=None): |
169
|
|
|
return |
170
|
|
|
|
171
|
|
|
@staticmethod |
172
|
|
|
def finalize(parser=None, block=None): |
173
|
|
|
return |
174
|
|
|
|
175
|
|
|
@staticmethod |
176
|
|
|
def can_contain(t): |
177
|
|
|
return |
178
|
|
|
|
179
|
|
|
|
180
|
|
|
class Document(Block): |
181
|
|
|
accepts_lines = False |
182
|
|
|
|
183
|
|
|
@staticmethod |
184
|
|
|
def continue_(parser=None, container=None): |
185
|
|
|
return 0 |
186
|
|
|
|
187
|
|
|
@staticmethod |
188
|
|
|
def finalize(parser=None, block=None): |
189
|
|
|
return |
190
|
|
|
|
191
|
|
|
@staticmethod |
192
|
|
|
def can_contain(t): |
193
|
|
|
return t != 'item' |
194
|
|
|
|
195
|
|
|
|
196
|
|
|
class List(Block): |
197
|
|
|
accepts_lines = False |
198
|
|
|
|
199
|
|
|
@staticmethod |
200
|
|
|
def continue_(parser=None, container=None): |
201
|
|
|
return 0 |
202
|
|
|
|
203
|
|
|
@staticmethod |
204
|
|
|
def finalize(parser=None, block=None): |
205
|
|
|
item = block.first_child |
206
|
|
|
while item: |
207
|
|
|
# check for non-final list item ending with blank line: |
208
|
|
|
if ends_with_blank_line(item) and item.nxt: |
209
|
|
|
block.list_data['tight'] = False |
210
|
|
|
break |
211
|
|
|
# recurse into children of list item, to see if there are |
212
|
|
|
# spaces between any of them: |
213
|
|
|
subitem = item.first_child |
214
|
|
|
while subitem: |
215
|
|
|
if ends_with_blank_line(subitem) and \ |
216
|
|
|
(item.nxt or subitem.nxt): |
217
|
|
|
block.list_data['tight'] = False |
218
|
|
|
break |
219
|
|
|
subitem = subitem.nxt |
220
|
|
|
item = item.nxt |
221
|
|
|
|
222
|
|
|
@staticmethod |
223
|
|
|
def can_contain(t): |
224
|
|
|
return t == 'item' |
225
|
|
|
|
226
|
|
|
|
227
|
|
|
class BlockQuote(Block): |
228
|
|
|
accepts_lines = False |
229
|
|
|
|
230
|
|
|
@staticmethod |
231
|
|
|
def continue_(parser=None, container=None): |
232
|
|
|
ln = parser.current_line |
233
|
|
|
if not parser.indented and peek(ln, parser.next_nonspace) == '>': |
234
|
|
|
parser.advance_next_nonspace() |
235
|
|
|
parser.advance_offset(1, False) |
236
|
|
|
if is_space_or_tab(peek(ln, parser.offset)): |
237
|
|
|
parser.advance_offset(1, True) |
238
|
|
|
else: |
239
|
|
|
return 1 |
240
|
|
|
return 0 |
241
|
|
|
|
242
|
|
|
@staticmethod |
243
|
|
|
def finalize(parser=None, block=None): |
244
|
|
|
return |
245
|
|
|
|
246
|
|
|
@staticmethod |
247
|
|
|
def can_contain(t): |
248
|
|
|
return t != 'item' |
249
|
|
|
|
250
|
|
|
|
251
|
|
|
class Item(Block): |
252
|
|
|
accepts_lines = False |
253
|
|
|
|
254
|
|
|
@staticmethod |
255
|
|
|
def continue_(parser=None, container=None): |
256
|
|
|
if parser.blank: |
257
|
|
|
if container.first_child is None: |
258
|
|
|
# Blank line after empty list item |
259
|
|
|
return 1 |
260
|
|
|
else: |
261
|
|
|
parser.advance_next_nonspace() |
262
|
|
|
elif parser.indent >= (container.list_data['marker_offset'] + |
263
|
|
|
container.list_data['padding']): |
264
|
|
|
parser.advance_offset( |
265
|
|
|
container.list_data['marker_offset'] + |
266
|
|
|
container.list_data['padding'], True) |
267
|
|
|
else: |
268
|
|
|
return 1 |
269
|
|
|
return 0 |
270
|
|
|
|
271
|
|
|
@staticmethod |
272
|
|
|
def finalize(parser=None, block=None): |
273
|
|
|
return |
274
|
|
|
|
275
|
|
|
@staticmethod |
276
|
|
|
def can_contain(t): |
277
|
|
|
return t != 'item' |
278
|
|
|
|
279
|
|
|
|
280
|
|
|
class Heading(Block): |
281
|
|
|
accepts_lines = False |
282
|
|
|
|
283
|
|
|
@staticmethod |
284
|
|
|
def continue_(parser=None, container=None): |
285
|
|
|
# A heading can never container > 1 line, so fail to match: |
286
|
|
|
return 1 |
287
|
|
|
|
288
|
|
|
@staticmethod |
289
|
|
|
def finalize(parser=None, block=None): |
290
|
|
|
return |
291
|
|
|
|
292
|
|
|
@staticmethod |
293
|
|
|
def can_contain(t): |
294
|
|
|
return False |
295
|
|
|
|
296
|
|
|
|
297
|
|
|
class ThematicBreak(Block): |
298
|
|
|
accepts_lines = False |
299
|
|
|
|
300
|
|
|
@staticmethod |
301
|
|
|
def continue_(parser=None, container=None): |
302
|
|
|
# A thematic break can never container > 1 line, so fail to match: |
303
|
|
|
return 1 |
304
|
|
|
|
305
|
|
|
@staticmethod |
306
|
|
|
def finalize(parser=None, block=None): |
307
|
|
|
return |
308
|
|
|
|
309
|
|
|
@staticmethod |
310
|
|
|
def can_contain(t): |
311
|
|
|
return False |
312
|
|
|
|
313
|
|
|
|
314
|
|
|
class CodeBlock(Block): |
315
|
|
|
accepts_lines = True |
316
|
|
|
|
317
|
|
|
@staticmethod |
318
|
|
|
def continue_(parser=None, container=None): |
319
|
|
|
ln = parser.current_line |
320
|
|
|
indent = parser.indent |
321
|
|
|
if container.is_fenced: |
322
|
|
|
match = indent <= 3 and \ |
323
|
|
|
len(ln) >= parser.next_nonspace + 1 and \ |
324
|
|
|
ln[parser.next_nonspace] == container.fence_char and \ |
325
|
|
|
re.search(reClosingCodeFence, ln[parser.next_nonspace:]) |
326
|
|
|
if match and len(match.group()) >= container.fence_length: |
327
|
|
|
# closing fence - we're at end of line, so we can return |
328
|
|
|
parser.finalize(container, parser.line_number) |
329
|
|
|
return 2 |
330
|
|
|
else: |
331
|
|
|
# skip optional spaces of fence offset |
332
|
|
|
i = container.fence_offset |
333
|
|
|
while i > 0 and is_space_or_tab(peek(ln, parser.offset)): |
334
|
|
|
parser.advance_offset(1, True) |
335
|
|
|
i -= 1 |
336
|
|
|
else: |
337
|
|
|
# indented |
338
|
|
|
if indent >= CODE_INDENT: |
339
|
|
|
parser.advance_offset(CODE_INDENT, True) |
340
|
|
|
elif parser.blank: |
341
|
|
|
parser.advance_next_nonspace() |
342
|
|
|
else: |
343
|
|
|
return 1 |
344
|
|
|
return 0 |
345
|
|
|
|
346
|
|
|
@staticmethod |
347
|
|
|
def finalize(parser=None, block=None): |
348
|
|
|
if block.is_fenced: |
349
|
|
|
# first line becomes info string |
350
|
|
|
content = block.string_content |
351
|
|
|
newline_pos = content.index('\n') |
352
|
|
|
first_line = content[0:newline_pos] |
353
|
|
|
rest = content[newline_pos + 1:] |
354
|
|
|
block.info = unescape_string(first_line.strip()) |
355
|
|
|
block.literal = rest |
356
|
|
|
else: |
357
|
|
|
# indented |
358
|
|
|
block.literal = re.sub(r'(\n *)+$', '\n', block.string_content) |
359
|
|
|
|
360
|
|
|
block.string_content = None |
361
|
|
|
|
362
|
|
|
@staticmethod |
363
|
|
|
def can_contain(t): |
364
|
|
|
return False |
365
|
|
|
|
366
|
|
|
|
367
|
|
|
class HtmlBlock(Block): |
368
|
|
|
accepts_lines = True |
369
|
|
|
|
370
|
|
|
@staticmethod |
371
|
|
|
def continue_(parser=None, container=None): |
372
|
|
|
if parser.blank and (container.html_block_type == 6 or |
373
|
|
|
container.html_block_type == 7): |
374
|
|
|
return 1 |
375
|
|
|
else: |
376
|
|
|
return 0 |
377
|
|
|
|
378
|
|
|
@staticmethod |
379
|
|
|
def finalize(parser=None, block=None): |
380
|
|
|
block.literal = re.sub(r'(\n *)+$', '', block.string_content) |
381
|
|
|
# allow GC |
382
|
|
|
block.string_content = None |
383
|
|
|
|
384
|
|
|
@staticmethod |
385
|
|
|
def can_contain(t): |
386
|
|
|
return False |
387
|
|
|
|
388
|
|
|
|
389
|
|
|
class Paragraph(Block): |
390
|
|
|
accepts_lines = True |
391
|
|
|
|
392
|
|
|
@staticmethod |
393
|
|
|
def continue_(parser=None, container=None): |
394
|
|
|
return 1 if parser.blank else 0 |
395
|
|
|
|
396
|
|
|
@staticmethod |
397
|
|
|
def finalize(parser=None, block=None): |
398
|
|
|
has_reference_defs = False |
399
|
|
|
|
400
|
|
|
# try parsing the beginning as link reference definitions: |
401
|
|
|
while peek(block.string_content, 0) == '[': |
402
|
|
|
pos = parser.inline_parser.parseReference( |
403
|
|
|
block.string_content, parser.refmap) |
404
|
|
|
if not pos: |
405
|
|
|
break |
406
|
|
|
block.string_content = block.string_content[pos:] |
407
|
|
|
has_reference_defs = True |
408
|
|
|
if has_reference_defs and is_blank(block.string_content): |
409
|
|
|
block.unlink() |
410
|
|
|
|
411
|
|
|
@staticmethod |
412
|
|
|
def can_contain(t): |
413
|
|
|
return False |
414
|
|
|
|
415
|
|
|
|
416
|
|
|
class BlockStarts(object): |
417
|
|
|
"""Block start functions. |
418
|
|
|
|
419
|
|
|
Return values: |
420
|
|
|
0 = no match |
421
|
|
|
1 = matched container, keep going |
422
|
|
|
2 = matched leaf, no more block starts |
423
|
|
|
""" |
424
|
|
|
METHODS = [ |
425
|
|
|
'block_quote', |
426
|
|
|
'atx_heading', |
427
|
|
|
'fenced_code_block', |
428
|
|
|
'html_block', |
429
|
|
|
'setext_heading', |
430
|
|
|
'thematic_break', |
431
|
|
|
'list_item', |
432
|
|
|
'indented_code_block', |
433
|
|
|
] |
434
|
|
|
|
435
|
|
|
@staticmethod |
436
|
|
|
def block_quote(parser, container=None): |
437
|
|
|
if not parser.indented and \ |
438
|
|
|
peek(parser.current_line, parser.next_nonspace) == '>': |
439
|
|
|
parser.advance_next_nonspace() |
440
|
|
|
parser.advance_offset(1, False) |
441
|
|
|
# optional following space |
442
|
|
|
if is_space_or_tab(peek(parser.current_line, parser.offset)): |
443
|
|
|
parser.advance_offset(1, True) |
444
|
|
|
parser.close_unmatched_blocks() |
445
|
|
|
parser.add_child('block_quote', parser.next_nonspace) |
446
|
|
|
return 1 |
447
|
|
|
|
448
|
|
|
return 0 |
449
|
|
|
|
450
|
|
|
@staticmethod |
451
|
|
|
def atx_heading(parser, container=None): |
452
|
|
|
if not parser.indented: |
453
|
|
|
m = re.search(reATXHeadingMarker, |
454
|
|
|
parser.current_line[parser.next_nonspace:]) |
455
|
|
|
if m: |
456
|
|
|
parser.advance_next_nonspace() |
457
|
|
|
parser.advance_offset(len(m.group()), False) |
458
|
|
|
parser.close_unmatched_blocks() |
459
|
|
|
container = parser.add_child('heading', parser.next_nonspace) |
460
|
|
|
# number of #s |
461
|
|
|
container.level = len(m.group().strip()) |
462
|
|
|
# remove trailing ###s: |
463
|
|
|
container.string_content = re.sub( |
464
|
|
|
r'[ \t]+#+[ \t]*$', '', re.sub( |
465
|
|
|
r'^[ \t]*#+[ \t]*$', |
466
|
|
|
'', |
467
|
|
|
parser.current_line[parser.offset:])) |
468
|
|
|
parser.advance_offset( |
469
|
|
|
len(parser.current_line) - parser.offset, False) |
470
|
|
|
return 2 |
471
|
|
|
|
472
|
|
|
return 0 |
473
|
|
|
|
474
|
|
|
@staticmethod |
475
|
|
|
def fenced_code_block(parser, container=None): |
476
|
|
|
if not parser.indented: |
477
|
|
|
m = re.search( |
478
|
|
|
reCodeFence, |
479
|
|
|
parser.current_line[parser.next_nonspace:]) |
480
|
|
|
if m: |
481
|
|
|
fence_length = len(m.group()) |
482
|
|
|
parser.close_unmatched_blocks() |
483
|
|
|
container = parser.add_child( |
484
|
|
|
'code_block', parser.next_nonspace) |
485
|
|
|
container.is_fenced = True |
486
|
|
|
container.fence_length = fence_length |
487
|
|
|
container.fence_char = m.group()[0] |
488
|
|
|
container.fence_offset = parser.indent |
489
|
|
|
parser.advance_next_nonspace() |
490
|
|
|
parser.advance_offset(fence_length, False) |
491
|
|
|
return 2 |
492
|
|
|
|
493
|
|
|
return 0 |
494
|
|
|
|
495
|
|
|
@staticmethod |
496
|
|
|
def html_block(parser, container=None): |
497
|
|
|
if not parser.indented and \ |
498
|
|
|
peek(parser.current_line, parser.next_nonspace) == '<': |
499
|
|
|
s = parser.current_line[parser.next_nonspace:] |
500
|
|
|
|
501
|
|
|
for block_type in range(1, 8): |
502
|
|
|
if re.search(reHtmlBlockOpen[block_type], s) and \ |
503
|
|
|
(block_type < 7 or container.t != 'paragraph'): |
504
|
|
|
parser.close_unmatched_blocks() |
505
|
|
|
# We don't adjust parser.offset; |
506
|
|
|
# spaces are part of the HTML block: |
507
|
|
|
b = parser.add_child('html_block', parser.offset) |
508
|
|
|
b.html_block_type = block_type |
509
|
|
|
return 2 |
510
|
|
|
return 0 |
511
|
|
|
|
512
|
|
|
@staticmethod |
513
|
|
|
def setext_heading(parser, container=None): |
514
|
|
|
if not parser.indented and container.t == 'paragraph': |
515
|
|
|
m = re.search( |
516
|
|
|
reSetextHeadingLine, |
517
|
|
|
parser.current_line[parser.next_nonspace:]) |
518
|
|
|
if m: |
519
|
|
|
parser.close_unmatched_blocks() |
520
|
|
|
heading = Node('heading', container.sourcepos) |
521
|
|
|
heading.level = 1 if m.group()[0] == '=' else 2 |
522
|
|
|
heading.string_content = container.string_content |
523
|
|
|
container.insert_after(heading) |
524
|
|
|
container.unlink() |
525
|
|
|
parser.tip = heading |
526
|
|
|
parser.advance_offset( |
527
|
|
|
len(parser.current_line) - parser.offset, False) |
528
|
|
|
return 2 |
529
|
|
|
|
530
|
|
|
return 0 |
531
|
|
|
|
532
|
|
|
@staticmethod |
533
|
|
|
def thematic_break(parser, container=None): |
534
|
|
|
if not parser.indented and re.search( |
535
|
|
|
reThematicBreak, parser.current_line[parser.next_nonspace:]): |
536
|
|
|
parser.close_unmatched_blocks() |
537
|
|
|
parser.add_child('thematic_break', parser.next_nonspace) |
538
|
|
|
parser.advance_offset( |
539
|
|
|
len(parser.current_line) - parser.offset, False) |
540
|
|
|
return 2 |
541
|
|
|
return 0 |
542
|
|
|
|
543
|
|
|
@staticmethod |
544
|
|
|
def list_item(parser, container=None): |
545
|
|
|
if (not parser.indented or container.t == 'list'): |
546
|
|
|
data = parse_list_marker(parser, container) |
547
|
|
|
if data: |
548
|
|
|
parser.close_unmatched_blocks() |
549
|
|
|
|
550
|
|
|
# add the list if needed |
551
|
|
|
if parser.tip.t != 'list' or \ |
552
|
|
|
not lists_match(container.list_data, data): |
553
|
|
|
container = parser.add_child('list', parser.next_nonspace) |
554
|
|
|
container.list_data = data |
555
|
|
|
|
556
|
|
|
# add the list item |
557
|
|
|
container = parser.add_child('item', parser.next_nonspace) |
558
|
|
|
container.list_data = data |
559
|
|
|
return 1 |
560
|
|
|
|
561
|
|
|
return 0 |
562
|
|
|
|
563
|
|
|
@staticmethod |
564
|
|
|
def indented_code_block(parser, container=None): |
565
|
|
|
if parser.indented and \ |
566
|
|
|
parser.tip.t != 'paragraph' and \ |
567
|
|
|
not parser.blank: |
568
|
|
|
# indented code |
569
|
|
|
parser.advance_offset(CODE_INDENT, True) |
570
|
|
|
parser.close_unmatched_blocks() |
571
|
|
|
parser.add_child('code_block', parser.offset) |
572
|
|
|
return 2 |
573
|
|
|
|
574
|
|
|
return 0 |
575
|
|
|
|
576
|
|
|
|
577
|
|
|
class Parser(object): |
578
|
|
|
def __init__(self, options={}): |
579
|
|
|
self.doc = Node('document', [[1, 1], [0, 0]]) |
580
|
|
|
self.block_starts = BlockStarts() |
581
|
|
|
self.tip = self.doc |
582
|
|
|
self.oldtip = self.doc |
583
|
|
|
self.current_line = '' |
584
|
|
|
self.line_number = 0 |
585
|
|
|
self.offset = 0 |
586
|
|
|
self.column = 0 |
587
|
|
|
self.next_nonspace = 0 |
588
|
|
|
self.next_nonspace_column = 0 |
589
|
|
|
self.indent = 0 |
590
|
|
|
self.indented = False |
591
|
|
|
self.blank = False |
592
|
|
|
self.partially_consumed_tab = False |
593
|
|
|
self.all_closed = True |
594
|
|
|
self.last_matched_container = self.doc |
595
|
|
|
self.refmap = {} |
596
|
|
|
self.last_line_length = 0 |
597
|
|
|
self.inline_parser = InlineParser(options) |
598
|
|
|
self.options = options |
599
|
|
|
|
600
|
|
|
def add_line(self): |
601
|
|
|
""" Add a line to the block at the tip. We assume the tip |
602
|
|
|
can accept lines -- that check should be done before calling this.""" |
603
|
|
|
if self.partially_consumed_tab: |
604
|
|
|
# Skip over tab |
605
|
|
|
self.offset += 1 |
606
|
|
|
# Add space characters |
607
|
|
|
chars_to_tab = 4 - (self.column % 4) |
608
|
|
|
self.tip.string_content += (' ' * chars_to_tab) |
609
|
|
|
self.tip.string_content += (self.current_line[self.offset:] + '\n') |
610
|
|
|
|
611
|
|
|
def add_child(self, tag, offset): |
612
|
|
|
""" Add block of type tag as a child of the tip. If the tip can't |
613
|
|
|
accept children, close and finalize it and try its parent, |
614
|
|
|
and so on til we find a block that can accept children.""" |
615
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
616
|
|
|
to_camel_case(self.tip.t)) |
617
|
|
|
while not block_class.can_contain(tag): |
618
|
|
|
self.finalize(self.tip, self.line_number - 1) |
619
|
|
|
block_class = getattr( |
620
|
|
|
import_module('CommonMark.blocks'), |
621
|
|
|
to_camel_case(self.tip.t)) |
622
|
|
|
|
623
|
|
|
column_number = offset + 1 |
624
|
|
|
new_block = Node(tag, [[self.line_number, column_number], [0, 0]]) |
625
|
|
|
new_block.string_content = '' |
626
|
|
|
self.tip.append_child(new_block) |
627
|
|
|
self.tip = new_block |
628
|
|
|
return new_block |
629
|
|
|
|
630
|
|
|
def close_unmatched_blocks(self): |
631
|
|
|
"""Finalize and close any unmatched blocks.""" |
632
|
|
|
if not self.all_closed: |
633
|
|
|
while self.oldtip != self.last_matched_container: |
634
|
|
|
parent = self.oldtip.parent |
635
|
|
|
self.finalize(self.oldtip, self.line_number - 1) |
636
|
|
|
self.oldtip = parent |
637
|
|
|
self.all_closed = True |
638
|
|
|
|
639
|
|
|
def find_next_nonspace(self): |
640
|
|
|
current_line = self.current_line |
641
|
|
|
i = self.offset |
642
|
|
|
cols = self.column |
643
|
|
|
|
644
|
|
|
try: |
645
|
|
|
c = current_line[i] |
646
|
|
|
if c is not None: |
647
|
|
|
c = unicode(c, 'latin1').encode('utf-8') |
648
|
|
|
except IndexError: |
649
|
|
|
c = '' |
650
|
|
|
while c != '': |
651
|
|
|
if c == ' ': |
652
|
|
|
i += 1 |
653
|
|
|
cols += 1 |
654
|
|
|
elif c == '\t': |
655
|
|
|
i += 1 |
656
|
|
|
cols += (4 - (cols % 4)) |
657
|
|
|
else: |
658
|
|
|
break |
659
|
|
|
|
660
|
|
|
try: |
661
|
|
|
c = current_line[i] |
662
|
|
|
except IndexError: |
663
|
|
|
c = '' |
664
|
|
|
|
665
|
|
|
self.blank = (c == '\n' or c == '\r' or c == '') |
666
|
|
|
self.next_nonspace = i |
667
|
|
|
self.next_nonspace_column = cols |
668
|
|
|
self.indent = self.next_nonspace_column - self.column |
669
|
|
|
self.indented = self.indent >= CODE_INDENT |
670
|
|
|
|
671
|
|
|
def advance_next_nonspace(self): |
672
|
|
|
self.offset = self.next_nonspace |
673
|
|
|
self.column = self.next_nonspace_column |
674
|
|
|
self.partially_consumed_tab = False |
675
|
|
|
|
676
|
|
|
def advance_offset(self, count, columns): |
677
|
|
|
current_line = self.current_line |
678
|
|
|
try: |
679
|
|
|
c = current_line[self.offset] |
680
|
|
|
except IndexError: |
681
|
|
|
c = None |
682
|
|
|
while count > 0 and c is not None: |
683
|
|
|
if c == '\t': |
684
|
|
|
chars_to_tab = 4 - (self.column % 4) |
685
|
|
|
if columns: |
686
|
|
|
self.partially_consumed_tab = chars_to_tab > count |
687
|
|
|
chars_to_advance = min(count, chars_to_tab) |
688
|
|
|
self.column += chars_to_advance |
689
|
|
|
self.offset += 0 if self.partially_consumed_tab else 1 |
690
|
|
|
count -= chars_to_advance |
691
|
|
|
else: |
692
|
|
|
self.partially_consumed_tab = False |
693
|
|
|
self.column += chars_to_tab |
694
|
|
|
self.offset += 1 |
695
|
|
|
count -= 1 |
696
|
|
|
else: |
697
|
|
|
self.partially_consumed_tab = False |
698
|
|
|
self.offset += 1 |
699
|
|
|
# assume ascii; block starts are ascii |
700
|
|
|
self.column += 1 |
701
|
|
|
count -= 1 |
702
|
|
|
try: |
703
|
|
|
c = current_line[self.offset] |
704
|
|
|
except IndexError: |
705
|
|
|
c = None |
706
|
|
|
|
707
|
|
|
def incorporate_line(self, ln): |
708
|
|
|
"""Analyze a line of text and update the document appropriately. |
709
|
|
|
|
710
|
|
|
We parse markdown text by calling this on each line of input, |
711
|
|
|
then finalizing the document. |
712
|
|
|
""" |
713
|
|
|
all_matched = True |
714
|
|
|
|
715
|
|
|
container = self.doc |
716
|
|
|
self.oldtip = self.tip |
717
|
|
|
self.offset = 0 |
718
|
|
|
self.column = 0 |
719
|
|
|
self.blank = False |
720
|
|
|
self.partially_consumed_tab = False |
721
|
|
|
self.line_number += 1 |
722
|
|
|
|
723
|
|
|
# replace NUL characters for security |
724
|
|
|
if re.search(r'\u0000', ln) is not None: |
725
|
|
|
ln = re.sub(r'\0', '\uFFFD', ln) |
726
|
|
|
|
727
|
|
|
self.current_line = ln |
728
|
|
|
|
729
|
|
|
# For each containing block, try to parse the associated line start. |
730
|
|
|
# Bail out on failure: container will point to the last matching block. |
731
|
|
|
# Set all_matched to false if not all containers match. |
732
|
|
|
last_child = container.last_child |
733
|
|
|
while last_child and last_child.is_open: |
734
|
|
|
container = last_child |
735
|
|
|
|
736
|
|
|
self.find_next_nonspace() |
737
|
|
|
block_class = getattr( |
738
|
|
|
import_module('CommonMark.blocks'), |
739
|
|
|
to_camel_case(container.t)) |
740
|
|
|
rv = block_class.continue_(self, container) |
741
|
|
|
if rv == 0: |
742
|
|
|
# we've matched, keep going |
743
|
|
|
pass |
744
|
|
|
elif rv == 1: |
745
|
|
|
# we've failed to match a block |
746
|
|
|
all_matched = False |
747
|
|
|
elif rv == 2: |
748
|
|
|
# we've hit end of line for fenced code close and can return |
749
|
|
|
self.last_line_length = len(ln) |
750
|
|
|
return |
751
|
|
|
else: |
752
|
|
|
raise ValueError('returned illegal value, must be 0, 1, or 2') |
753
|
|
|
|
754
|
|
|
if not all_matched: |
755
|
|
|
# back up to last matching block |
756
|
|
|
container = container.parent |
757
|
|
|
break |
758
|
|
|
|
759
|
|
|
last_child = container.last_child |
760
|
|
|
|
761
|
|
|
self.all_closed = (container == self.oldtip) |
762
|
|
|
self.last_matched_container = container |
763
|
|
|
|
764
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
765
|
|
|
to_camel_case(container.t)) |
766
|
|
|
matched_leaf = container.t != 'paragraph' and block_class.accepts_lines |
767
|
|
|
starts = self.block_starts |
768
|
|
|
starts_len = len(starts.METHODS) |
769
|
|
|
# Unless last matched container is a code block, try new container |
770
|
|
|
# starts, adding children to the last matched container: |
771
|
|
|
while not matched_leaf: |
772
|
|
|
self.find_next_nonspace() |
773
|
|
|
|
774
|
|
|
# this is a little performance optimization: |
775
|
|
|
if not self.indented and \ |
776
|
|
|
not re.search(reMaybeSpecial, ln[self.next_nonspace:]): |
777
|
|
|
self.advance_next_nonspace() |
778
|
|
|
break |
779
|
|
|
|
780
|
|
|
i = 0 |
781
|
|
|
while i < starts_len: |
782
|
|
|
res = getattr(starts, starts.METHODS[i])(self, container) |
783
|
|
|
if res == 1: |
784
|
|
|
container = self.tip |
785
|
|
|
break |
786
|
|
|
elif res == 2: |
787
|
|
|
container = self.tip |
788
|
|
|
matched_leaf = True |
789
|
|
|
break |
790
|
|
|
else: |
791
|
|
|
i += 1 |
792
|
|
|
|
793
|
|
|
if i == starts_len: |
794
|
|
|
# nothing matched |
795
|
|
|
self.advance_next_nonspace() |
796
|
|
|
break |
797
|
|
|
|
798
|
|
|
# What remains at the offset is a text line. Add the text to the |
799
|
|
|
# appropriate container. |
800
|
|
|
if not self.all_closed and not self.blank and \ |
801
|
|
|
self.tip.t == 'paragraph': |
802
|
|
|
# lazy paragraph continuation |
803
|
|
|
self.add_line() |
804
|
|
|
else: |
805
|
|
|
# not a lazy continuation |
806
|
|
|
# finalize any blocks not matched |
807
|
|
|
self.close_unmatched_blocks() |
808
|
|
|
if self.blank and container.last_child: |
809
|
|
|
container.last_child.last_line_blank = True |
810
|
|
|
|
811
|
|
|
t = container.t |
812
|
|
|
|
813
|
|
|
# Block quote lines are never blank as they start with > |
814
|
|
|
# and we don't count blanks in fenced code for purposes of |
815
|
|
|
# tight/loose lists or breaking out of lists. We also |
816
|
|
|
# don't set last_line_blank on an empty list item, or if we |
817
|
|
|
# just closed a fenced block. |
818
|
|
|
last_line_blank = self.blank and \ |
819
|
|
|
not (t == 'block_quote' or |
820
|
|
|
(t == 'code_block' and container.is_fenced) or |
821
|
|
|
(t == 'item' and |
822
|
|
|
not container.first_child and |
823
|
|
|
container.sourcepos[0][0] == self.line_number)) |
824
|
|
|
|
825
|
|
|
# propagate last_line_blank up through parents: |
826
|
|
|
cont = container |
827
|
|
|
while cont: |
828
|
|
|
cont.last_line_blank = last_line_blank |
829
|
|
|
cont = cont.parent |
830
|
|
|
|
831
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
832
|
|
|
to_camel_case(t)) |
833
|
|
|
if block_class.accepts_lines: |
834
|
|
|
self.add_line() |
835
|
|
|
# if HtmlBlock, check for end condition |
836
|
|
|
if t == 'html_block' and \ |
837
|
|
|
container.html_block_type >= 1 and \ |
838
|
|
|
container.html_block_type <= 5 and \ |
839
|
|
|
re.search( |
840
|
|
|
reHtmlBlockClose[container.html_block_type], |
841
|
|
|
self.current_line[self.offset:]): |
842
|
|
|
self.finalize(container, self.line_number) |
843
|
|
|
elif self.offset < len(ln) and not self.blank: |
844
|
|
|
# create a paragraph container for one line |
845
|
|
|
container = self.add_child('paragraph', self.offset) |
846
|
|
|
self.advance_next_nonspace() |
847
|
|
|
self.add_line() |
848
|
|
|
|
849
|
|
|
self.last_line_length = len(ln) |
850
|
|
|
|
851
|
|
|
def finalize(self, block, line_number): |
852
|
|
|
""" Finalize a block. Close it and do any necessary postprocessing, |
853
|
|
|
e.g. creating string_content from strings, setting the 'tight' |
854
|
|
|
or 'loose' status of a list, and parsing the beginnings |
855
|
|
|
of paragraphs for reference definitions. Reset the tip to the |
856
|
|
|
parent of the closed block.""" |
857
|
|
|
above = block.parent |
858
|
|
|
block.is_open = False |
859
|
|
|
block.sourcepos[1] = [line_number, self.last_line_length] |
860
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
861
|
|
|
to_camel_case(block.t)) |
862
|
|
|
block_class.finalize(self, block) |
863
|
|
|
|
864
|
|
|
self.tip = above |
865
|
|
|
|
866
|
|
|
def process_inlines(self, block): |
867
|
|
|
""" |
868
|
|
|
Walk through a block & children recursively, parsing string content |
869
|
|
|
into inline content where appropriate. |
870
|
|
|
""" |
871
|
|
|
walker = block.walker() |
872
|
|
|
self.inline_parser.refmap = self.refmap |
873
|
|
|
self.inline_parser.options = self.options |
874
|
|
|
event = walker.nxt() |
875
|
|
|
while event is not None: |
876
|
|
|
node = event['node'] |
877
|
|
|
t = node.t |
878
|
|
|
if not event['entering'] and (t == 'paragraph' or t == 'heading'): |
879
|
|
|
self.inline_parser.parse(node) |
880
|
|
|
event = walker.nxt() |
881
|
|
|
|
882
|
|
|
def parse(self, my_input): |
883
|
|
|
""" The main parsing function. Returns a parsed document AST.""" |
884
|
|
|
self.doc = Node('document', [[1, 1], [0, 0]]) |
885
|
|
|
self.tip = self.doc |
886
|
|
|
self.refmap = {} |
887
|
|
|
self.line_number = 0 |
888
|
|
|
self.last_line_length = 0 |
889
|
|
|
self.offset = 0 |
890
|
|
|
self.column = 0 |
891
|
|
|
self.last_matched_container = self.doc |
892
|
|
|
self.current_line = '' |
893
|
|
|
lines = re.split(reLineEnding, my_input) |
894
|
|
|
length = len(lines) |
895
|
|
|
if len(my_input) > 0 and my_input[-1] == '\n': |
896
|
|
|
# ignore last blank line created by final newline |
897
|
|
|
length -= 1 |
898
|
|
|
for i in range(length): |
899
|
|
|
self.incorporate_line(lines[i]) |
900
|
|
|
while (self.tip): |
901
|
|
|
self.finalize(self.tip, length) |
902
|
|
|
self.process_inlines(self.doc) |
903
|
|
|
return self.doc |
904
|
|
|
|