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
|
|
|
try: |
63
|
|
|
s = unicode(s, 'latin1').encode('utf-8') |
64
|
|
|
except Exception: |
65
|
|
|
s = str(s); |
66
|
|
|
return s == ' ' or s == '\t' |
67
|
|
|
|
68
|
|
|
|
69
|
|
|
def peek(ln, pos): |
70
|
|
|
if pos < len(ln): |
71
|
|
|
return ln[pos] |
72
|
|
|
else: |
73
|
|
|
return None |
74
|
|
|
|
75
|
|
|
|
76
|
|
|
def ends_with_blank_line(block): |
77
|
|
|
""" Returns true if block ends with a blank line, |
78
|
|
|
descending if needed into lists and sublists.""" |
79
|
|
|
while block: |
80
|
|
|
if block.last_line_blank: |
81
|
|
|
return True |
82
|
|
|
if (block.t == 'list' or block.t == 'item'): |
83
|
|
|
block = block.last_child |
84
|
|
|
else: |
85
|
|
|
break |
86
|
|
|
|
87
|
|
|
return False |
88
|
|
|
|
89
|
|
|
|
90
|
|
|
def parse_list_marker(parser, container): |
91
|
|
|
""" Parse a list marker and return data on the marker (type, |
92
|
|
|
start, delimiter, bullet character, padding) or None.""" |
93
|
|
|
rest = parser.current_line[parser.next_nonspace:] |
94
|
|
|
data = { |
95
|
|
|
'type': None, |
96
|
|
|
'tight': True, # lists are tight by default |
97
|
|
|
'bullet_char': None, |
98
|
|
|
'start': None, |
99
|
|
|
'delimiter': None, |
100
|
|
|
'padding': None, |
101
|
|
|
'marker_offset': parser.indent, |
102
|
|
|
} |
103
|
|
|
m = re.search(reBulletListMarker, rest) |
104
|
|
|
m2 = re.search(reOrderedListMarker, rest) |
105
|
|
|
if m: |
106
|
|
|
data['type'] = 'bullet' |
107
|
|
|
data['bullet_char'] = m.group()[0] |
108
|
|
|
elif m2 and (container.t != 'paragraph' or m2.group(1) == '1'): |
109
|
|
|
m = m2 |
110
|
|
|
data['type'] = 'ordered' |
111
|
|
|
data['start'] = int(m.group(1)) |
112
|
|
|
data['delimiter'] = m.group(2) |
113
|
|
|
else: |
114
|
|
|
return None |
115
|
|
|
|
116
|
|
|
# make sure we have spaces after |
117
|
|
|
nextc = peek(parser.current_line, parser.next_nonspace + len(m.group())) |
118
|
|
|
if not (nextc is None or nextc == '\t' or nextc == ' '): |
119
|
|
|
return None |
120
|
|
|
|
121
|
|
|
# if it interrupts paragraph, make sure first line isn't blank |
122
|
|
|
if container.t == 'paragraph' and \ |
123
|
|
|
not re.search( |
124
|
|
|
reNonSpace, |
125
|
|
|
parser.current_line[parser.next_nonspace + len(m.group()):]): |
126
|
|
|
return None |
127
|
|
|
|
128
|
|
|
# we've got a match! advance offset and calculate padding |
129
|
|
|
parser.advance_next_nonspace() # to start of marker |
130
|
|
|
parser.advance_offset(len(m.group()), True) # to end of marker |
131
|
|
|
spaces_start_col = parser.column |
132
|
|
|
spaces_start_offset = parser.offset |
133
|
|
|
while True: |
134
|
|
|
parser.advance_offset(1, True) |
135
|
|
|
nextc = peek(parser.current_line, parser.offset) |
136
|
|
|
if parser.column - spaces_start_col < 5 and \ |
137
|
|
|
is_space_or_tab(nextc): |
138
|
|
|
pass |
139
|
|
|
else: |
140
|
|
|
break |
141
|
|
|
blank_item = peek(parser.current_line, parser.offset) is None |
142
|
|
|
spaces_after_marker = parser.column - spaces_start_col |
143
|
|
|
if spaces_after_marker >= 5 or \ |
144
|
|
|
spaces_after_marker < 1 or \ |
145
|
|
|
blank_item: |
146
|
|
|
data['padding'] = len(m.group()) + 1 |
147
|
|
|
parser.column = spaces_start_col |
148
|
|
|
parser.offset = spaces_start_offset |
149
|
|
|
if is_space_or_tab(peek(parser.current_line, parser.offset)): |
150
|
|
|
parser.advance_offset(1, True) |
151
|
|
|
else: |
152
|
|
|
data['padding'] = len(m.group()) + spaces_after_marker |
153
|
|
|
|
154
|
|
|
return data |
155
|
|
|
|
156
|
|
|
|
157
|
|
|
def lists_match(list_data, item_data): |
158
|
|
|
""" |
159
|
|
|
Returns True if the two list items are of the same type, |
160
|
|
|
with the same delimiter and bullet character. This is used |
161
|
|
|
in agglomerating list items into lists. |
162
|
|
|
""" |
163
|
|
|
return list_data.get('type') == item_data.get('type') and \ |
164
|
|
|
list_data.get('delimiter') == item_data.get('delimiter') and \ |
165
|
|
|
list_data.get('bullet_char') == item_data.get('bullet_char') |
166
|
|
|
|
167
|
|
|
|
168
|
|
|
class Block(object): |
169
|
|
|
accepts_lines = None |
170
|
|
|
|
171
|
|
|
@staticmethod |
172
|
|
|
def continue_(parser=None, container=None): |
173
|
|
|
return |
174
|
|
|
|
175
|
|
|
@staticmethod |
176
|
|
|
def finalize(parser=None, block=None): |
177
|
|
|
return |
178
|
|
|
|
179
|
|
|
@staticmethod |
180
|
|
|
def can_contain(t): |
181
|
|
|
return |
182
|
|
|
|
183
|
|
|
|
184
|
|
|
class Document(Block): |
185
|
|
|
accepts_lines = False |
186
|
|
|
|
187
|
|
|
@staticmethod |
188
|
|
|
def continue_(parser=None, container=None): |
189
|
|
|
return 0 |
190
|
|
|
|
191
|
|
|
@staticmethod |
192
|
|
|
def finalize(parser=None, block=None): |
193
|
|
|
return |
194
|
|
|
|
195
|
|
|
@staticmethod |
196
|
|
|
def can_contain(t): |
197
|
|
|
return t != 'item' |
198
|
|
|
|
199
|
|
|
|
200
|
|
|
class List(Block): |
201
|
|
|
accepts_lines = False |
202
|
|
|
|
203
|
|
|
@staticmethod |
204
|
|
|
def continue_(parser=None, container=None): |
205
|
|
|
return 0 |
206
|
|
|
|
207
|
|
|
@staticmethod |
208
|
|
|
def finalize(parser=None, block=None): |
209
|
|
|
item = block.first_child |
210
|
|
|
while item: |
211
|
|
|
# check for non-final list item ending with blank line: |
212
|
|
|
if ends_with_blank_line(item) and item.nxt: |
213
|
|
|
block.list_data['tight'] = False |
214
|
|
|
break |
215
|
|
|
# recurse into children of list item, to see if there are |
216
|
|
|
# spaces between any of them: |
217
|
|
|
subitem = item.first_child |
218
|
|
|
while subitem: |
219
|
|
|
if ends_with_blank_line(subitem) and \ |
220
|
|
|
(item.nxt or subitem.nxt): |
221
|
|
|
block.list_data['tight'] = False |
222
|
|
|
break |
223
|
|
|
subitem = subitem.nxt |
224
|
|
|
item = item.nxt |
225
|
|
|
|
226
|
|
|
@staticmethod |
227
|
|
|
def can_contain(t): |
228
|
|
|
return t == 'item' |
229
|
|
|
|
230
|
|
|
|
231
|
|
|
class BlockQuote(Block): |
232
|
|
|
accepts_lines = False |
233
|
|
|
|
234
|
|
|
@staticmethod |
235
|
|
|
def continue_(parser=None, container=None): |
236
|
|
|
ln = parser.current_line |
237
|
|
|
if not parser.indented and peek(ln, parser.next_nonspace) == '>': |
238
|
|
|
parser.advance_next_nonspace() |
239
|
|
|
parser.advance_offset(1, False) |
240
|
|
|
if is_space_or_tab(peek(ln, parser.offset)): |
241
|
|
|
parser.advance_offset(1, True) |
242
|
|
|
else: |
243
|
|
|
return 1 |
244
|
|
|
return 0 |
245
|
|
|
|
246
|
|
|
@staticmethod |
247
|
|
|
def finalize(parser=None, block=None): |
248
|
|
|
return |
249
|
|
|
|
250
|
|
|
@staticmethod |
251
|
|
|
def can_contain(t): |
252
|
|
|
return t != 'item' |
253
|
|
|
|
254
|
|
|
|
255
|
|
|
class Item(Block): |
256
|
|
|
accepts_lines = False |
257
|
|
|
|
258
|
|
|
@staticmethod |
259
|
|
|
def continue_(parser=None, container=None): |
260
|
|
|
if parser.blank: |
261
|
|
|
if container.first_child is None: |
262
|
|
|
# Blank line after empty list item |
263
|
|
|
return 1 |
264
|
|
|
else: |
265
|
|
|
parser.advance_next_nonspace() |
266
|
|
|
elif parser.indent >= (container.list_data['marker_offset'] + |
267
|
|
|
container.list_data['padding']): |
268
|
|
|
parser.advance_offset( |
269
|
|
|
container.list_data['marker_offset'] + |
270
|
|
|
container.list_data['padding'], True) |
271
|
|
|
else: |
272
|
|
|
return 1 |
273
|
|
|
return 0 |
274
|
|
|
|
275
|
|
|
@staticmethod |
276
|
|
|
def finalize(parser=None, block=None): |
277
|
|
|
return |
278
|
|
|
|
279
|
|
|
@staticmethod |
280
|
|
|
def can_contain(t): |
281
|
|
|
return t != 'item' |
282
|
|
|
|
283
|
|
|
|
284
|
|
|
class Heading(Block): |
285
|
|
|
accepts_lines = False |
286
|
|
|
|
287
|
|
|
@staticmethod |
288
|
|
|
def continue_(parser=None, container=None): |
289
|
|
|
# A heading can never container > 1 line, so fail to match: |
290
|
|
|
return 1 |
291
|
|
|
|
292
|
|
|
@staticmethod |
293
|
|
|
def finalize(parser=None, block=None): |
294
|
|
|
return |
295
|
|
|
|
296
|
|
|
@staticmethod |
297
|
|
|
def can_contain(t): |
298
|
|
|
return False |
299
|
|
|
|
300
|
|
|
|
301
|
|
|
class ThematicBreak(Block): |
302
|
|
|
accepts_lines = False |
303
|
|
|
|
304
|
|
|
@staticmethod |
305
|
|
|
def continue_(parser=None, container=None): |
306
|
|
|
# A thematic break can never container > 1 line, so fail to match: |
307
|
|
|
return 1 |
308
|
|
|
|
309
|
|
|
@staticmethod |
310
|
|
|
def finalize(parser=None, block=None): |
311
|
|
|
return |
312
|
|
|
|
313
|
|
|
@staticmethod |
314
|
|
|
def can_contain(t): |
315
|
|
|
return False |
316
|
|
|
|
317
|
|
|
|
318
|
|
|
class CodeBlock(Block): |
319
|
|
|
accepts_lines = True |
320
|
|
|
|
321
|
|
|
@staticmethod |
322
|
|
|
def continue_(parser=None, container=None): |
323
|
|
|
ln = parser.current_line |
324
|
|
|
indent = parser.indent |
325
|
|
|
if container.is_fenced: |
326
|
|
|
match = indent <= 3 and \ |
327
|
|
|
len(ln) >= parser.next_nonspace + 1 and \ |
328
|
|
|
ln[parser.next_nonspace] == container.fence_char and \ |
329
|
|
|
re.search(reClosingCodeFence, ln[parser.next_nonspace:]) |
330
|
|
|
if match and len(match.group()) >= container.fence_length: |
331
|
|
|
# closing fence - we're at end of line, so we can return |
332
|
|
|
parser.finalize(container, parser.line_number) |
333
|
|
|
return 2 |
334
|
|
|
else: |
335
|
|
|
# skip optional spaces of fence offset |
336
|
|
|
i = container.fence_offset |
337
|
|
|
while i > 0 and is_space_or_tab(peek(ln, parser.offset)): |
338
|
|
|
parser.advance_offset(1, True) |
339
|
|
|
i -= 1 |
340
|
|
|
else: |
341
|
|
|
# indented |
342
|
|
|
if indent >= CODE_INDENT: |
343
|
|
|
parser.advance_offset(CODE_INDENT, True) |
344
|
|
|
elif parser.blank: |
345
|
|
|
parser.advance_next_nonspace() |
346
|
|
|
else: |
347
|
|
|
return 1 |
348
|
|
|
return 0 |
349
|
|
|
|
350
|
|
|
@staticmethod |
351
|
|
|
def finalize(parser=None, block=None): |
352
|
|
|
if block.is_fenced: |
353
|
|
|
# first line becomes info string |
354
|
|
|
content = block.string_content |
355
|
|
|
newline_pos = content.index('\n') |
356
|
|
|
first_line = content[0:newline_pos] |
357
|
|
|
rest = content[newline_pos + 1:] |
358
|
|
|
block.info = unescape_string(first_line.strip()) |
359
|
|
|
block.literal = rest |
360
|
|
|
else: |
361
|
|
|
# indented |
362
|
|
|
block.literal = re.sub(r'(\n *)+$', '\n', block.string_content) |
363
|
|
|
|
364
|
|
|
block.string_content = None |
365
|
|
|
|
366
|
|
|
@staticmethod |
367
|
|
|
def can_contain(t): |
368
|
|
|
return False |
369
|
|
|
|
370
|
|
|
|
371
|
|
|
class HtmlBlock(Block): |
372
|
|
|
accepts_lines = True |
373
|
|
|
|
374
|
|
|
@staticmethod |
375
|
|
|
def continue_(parser=None, container=None): |
376
|
|
|
if parser.blank and (container.html_block_type == 6 or |
377
|
|
|
container.html_block_type == 7): |
378
|
|
|
return 1 |
379
|
|
|
else: |
380
|
|
|
return 0 |
381
|
|
|
|
382
|
|
|
@staticmethod |
383
|
|
|
def finalize(parser=None, block=None): |
384
|
|
|
block.literal = re.sub(r'(\n *)+$', '', block.string_content) |
385
|
|
|
# allow GC |
386
|
|
|
block.string_content = None |
387
|
|
|
|
388
|
|
|
@staticmethod |
389
|
|
|
def can_contain(t): |
390
|
|
|
return False |
391
|
|
|
|
392
|
|
|
|
393
|
|
|
class Paragraph(Block): |
394
|
|
|
accepts_lines = True |
395
|
|
|
|
396
|
|
|
@staticmethod |
397
|
|
|
def continue_(parser=None, container=None): |
398
|
|
|
return 1 if parser.blank else 0 |
399
|
|
|
|
400
|
|
|
@staticmethod |
401
|
|
|
def finalize(parser=None, block=None): |
402
|
|
|
has_reference_defs = False |
403
|
|
|
|
404
|
|
|
# try parsing the beginning as link reference definitions: |
405
|
|
|
while peek(block.string_content, 0) == '[': |
406
|
|
|
pos = parser.inline_parser.parseReference( |
407
|
|
|
block.string_content, parser.refmap) |
408
|
|
|
if not pos: |
409
|
|
|
break |
410
|
|
|
block.string_content = block.string_content[pos:] |
411
|
|
|
has_reference_defs = True |
412
|
|
|
if has_reference_defs and is_blank(block.string_content): |
413
|
|
|
block.unlink() |
414
|
|
|
|
415
|
|
|
@staticmethod |
416
|
|
|
def can_contain(t): |
417
|
|
|
return False |
418
|
|
|
|
419
|
|
|
|
420
|
|
|
class BlockStarts(object): |
421
|
|
|
"""Block start functions. |
422
|
|
|
|
423
|
|
|
Return values: |
424
|
|
|
0 = no match |
425
|
|
|
1 = matched container, keep going |
426
|
|
|
2 = matched leaf, no more block starts |
427
|
|
|
""" |
428
|
|
|
METHODS = [ |
429
|
|
|
'block_quote', |
430
|
|
|
'atx_heading', |
431
|
|
|
'fenced_code_block', |
432
|
|
|
'html_block', |
433
|
|
|
'setext_heading', |
434
|
|
|
'thematic_break', |
435
|
|
|
'list_item', |
436
|
|
|
'indented_code_block', |
437
|
|
|
] |
438
|
|
|
|
439
|
|
|
@staticmethod |
440
|
|
|
def block_quote(parser, container=None): |
441
|
|
|
if not parser.indented and \ |
442
|
|
|
peek(parser.current_line, parser.next_nonspace) == '>': |
443
|
|
|
parser.advance_next_nonspace() |
444
|
|
|
parser.advance_offset(1, False) |
445
|
|
|
# optional following space |
446
|
|
|
if is_space_or_tab(peek(parser.current_line, parser.offset)): |
447
|
|
|
parser.advance_offset(1, True) |
448
|
|
|
parser.close_unmatched_blocks() |
449
|
|
|
parser.add_child('block_quote', parser.next_nonspace) |
450
|
|
|
return 1 |
451
|
|
|
|
452
|
|
|
return 0 |
453
|
|
|
|
454
|
|
|
@staticmethod |
455
|
|
|
def atx_heading(parser, container=None): |
456
|
|
|
if not parser.indented: |
457
|
|
|
m = re.search(reATXHeadingMarker, |
458
|
|
|
parser.current_line[parser.next_nonspace:]) |
459
|
|
|
if m: |
460
|
|
|
parser.advance_next_nonspace() |
461
|
|
|
parser.advance_offset(len(m.group()), False) |
462
|
|
|
parser.close_unmatched_blocks() |
463
|
|
|
container = parser.add_child('heading', parser.next_nonspace) |
464
|
|
|
# number of #s |
465
|
|
|
container.level = len(m.group().strip()) |
466
|
|
|
# remove trailing ###s: |
467
|
|
|
container.string_content = re.sub( |
468
|
|
|
r'[ \t]+#+[ \t]*$', '', re.sub( |
469
|
|
|
r'^[ \t]*#+[ \t]*$', |
470
|
|
|
'', |
471
|
|
|
parser.current_line[parser.offset:])) |
472
|
|
|
parser.advance_offset( |
473
|
|
|
len(parser.current_line) - parser.offset, False) |
474
|
|
|
return 2 |
475
|
|
|
|
476
|
|
|
return 0 |
477
|
|
|
|
478
|
|
|
@staticmethod |
479
|
|
|
def fenced_code_block(parser, container=None): |
480
|
|
|
if not parser.indented: |
481
|
|
|
m = re.search( |
482
|
|
|
reCodeFence, |
483
|
|
|
parser.current_line[parser.next_nonspace:]) |
484
|
|
|
if m: |
485
|
|
|
fence_length = len(m.group()) |
486
|
|
|
parser.close_unmatched_blocks() |
487
|
|
|
container = parser.add_child( |
488
|
|
|
'code_block', parser.next_nonspace) |
489
|
|
|
container.is_fenced = True |
490
|
|
|
container.fence_length = fence_length |
491
|
|
|
container.fence_char = m.group()[0] |
492
|
|
|
container.fence_offset = parser.indent |
493
|
|
|
parser.advance_next_nonspace() |
494
|
|
|
parser.advance_offset(fence_length, False) |
495
|
|
|
return 2 |
496
|
|
|
|
497
|
|
|
return 0 |
498
|
|
|
|
499
|
|
|
@staticmethod |
500
|
|
|
def html_block(parser, container=None): |
501
|
|
|
if not parser.indented and \ |
502
|
|
|
peek(parser.current_line, parser.next_nonspace) == '<': |
503
|
|
|
s = parser.current_line[parser.next_nonspace:] |
504
|
|
|
|
505
|
|
|
for block_type in range(1, 8): |
506
|
|
|
if re.search(reHtmlBlockOpen[block_type], s) and \ |
507
|
|
|
(block_type < 7 or container.t != 'paragraph'): |
508
|
|
|
parser.close_unmatched_blocks() |
509
|
|
|
# We don't adjust parser.offset; |
510
|
|
|
# spaces are part of the HTML block: |
511
|
|
|
b = parser.add_child('html_block', parser.offset) |
512
|
|
|
b.html_block_type = block_type |
513
|
|
|
return 2 |
514
|
|
|
return 0 |
515
|
|
|
|
516
|
|
|
@staticmethod |
517
|
|
|
def setext_heading(parser, container=None): |
518
|
|
|
if not parser.indented and container.t == 'paragraph': |
519
|
|
|
m = re.search( |
520
|
|
|
reSetextHeadingLine, |
521
|
|
|
parser.current_line[parser.next_nonspace:]) |
522
|
|
|
if m: |
523
|
|
|
parser.close_unmatched_blocks() |
524
|
|
|
heading = Node('heading', container.sourcepos) |
525
|
|
|
heading.level = 1 if m.group()[0] == '=' else 2 |
526
|
|
|
heading.string_content = container.string_content |
527
|
|
|
container.insert_after(heading) |
528
|
|
|
container.unlink() |
529
|
|
|
parser.tip = heading |
530
|
|
|
parser.advance_offset( |
531
|
|
|
len(parser.current_line) - parser.offset, False) |
532
|
|
|
return 2 |
533
|
|
|
|
534
|
|
|
return 0 |
535
|
|
|
|
536
|
|
|
@staticmethod |
537
|
|
|
def thematic_break(parser, container=None): |
538
|
|
|
if not parser.indented and re.search( |
539
|
|
|
reThematicBreak, parser.current_line[parser.next_nonspace:]): |
540
|
|
|
parser.close_unmatched_blocks() |
541
|
|
|
parser.add_child('thematic_break', parser.next_nonspace) |
542
|
|
|
parser.advance_offset( |
543
|
|
|
len(parser.current_line) - parser.offset, False) |
544
|
|
|
return 2 |
545
|
|
|
return 0 |
546
|
|
|
|
547
|
|
|
@staticmethod |
548
|
|
|
def list_item(parser, container=None): |
549
|
|
|
if (not parser.indented or container.t == 'list'): |
550
|
|
|
data = parse_list_marker(parser, container) |
551
|
|
|
if data: |
552
|
|
|
parser.close_unmatched_blocks() |
553
|
|
|
|
554
|
|
|
# add the list if needed |
555
|
|
|
if parser.tip.t != 'list' or \ |
556
|
|
|
not lists_match(container.list_data, data): |
557
|
|
|
container = parser.add_child('list', parser.next_nonspace) |
558
|
|
|
container.list_data = data |
559
|
|
|
|
560
|
|
|
# add the list item |
561
|
|
|
container = parser.add_child('item', parser.next_nonspace) |
562
|
|
|
container.list_data = data |
563
|
|
|
return 1 |
564
|
|
|
|
565
|
|
|
return 0 |
566
|
|
|
|
567
|
|
|
@staticmethod |
568
|
|
|
def indented_code_block(parser, container=None): |
569
|
|
|
if parser.indented and \ |
570
|
|
|
parser.tip.t != 'paragraph' and \ |
571
|
|
|
not parser.blank: |
572
|
|
|
# indented code |
573
|
|
|
parser.advance_offset(CODE_INDENT, True) |
574
|
|
|
parser.close_unmatched_blocks() |
575
|
|
|
parser.add_child('code_block', parser.offset) |
576
|
|
|
return 2 |
577
|
|
|
|
578
|
|
|
return 0 |
579
|
|
|
|
580
|
|
|
|
581
|
|
|
class Parser(object): |
582
|
|
|
def __init__(self, options={}): |
583
|
|
|
self.doc = Node('document', [[1, 1], [0, 0]]) |
584
|
|
|
self.block_starts = BlockStarts() |
585
|
|
|
self.tip = self.doc |
586
|
|
|
self.oldtip = self.doc |
587
|
|
|
self.current_line = '' |
588
|
|
|
self.line_number = 0 |
589
|
|
|
self.offset = 0 |
590
|
|
|
self.column = 0 |
591
|
|
|
self.next_nonspace = 0 |
592
|
|
|
self.next_nonspace_column = 0 |
593
|
|
|
self.indent = 0 |
594
|
|
|
self.indented = False |
595
|
|
|
self.blank = False |
596
|
|
|
self.partially_consumed_tab = False |
597
|
|
|
self.all_closed = True |
598
|
|
|
self.last_matched_container = self.doc |
599
|
|
|
self.refmap = {} |
600
|
|
|
self.last_line_length = 0 |
601
|
|
|
self.inline_parser = InlineParser(options) |
602
|
|
|
self.options = options |
603
|
|
|
|
604
|
|
|
def add_line(self): |
605
|
|
|
""" Add a line to the block at the tip. We assume the tip |
606
|
|
|
can accept lines -- that check should be done before calling this.""" |
607
|
|
|
if self.partially_consumed_tab: |
608
|
|
|
# Skip over tab |
609
|
|
|
self.offset += 1 |
610
|
|
|
# Add space characters |
611
|
|
|
chars_to_tab = 4 - (self.column % 4) |
612
|
|
|
self.tip.string_content += (' ' * chars_to_tab) |
613
|
|
|
self.tip.string_content += (self.current_line[self.offset:] + '\n') |
614
|
|
|
|
615
|
|
|
def add_child(self, tag, offset): |
616
|
|
|
""" Add block of type tag as a child of the tip. If the tip can't |
617
|
|
|
accept children, close and finalize it and try its parent, |
618
|
|
|
and so on til we find a block that can accept children.""" |
619
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
620
|
|
|
to_camel_case(self.tip.t)) |
621
|
|
|
while not block_class.can_contain(tag): |
622
|
|
|
self.finalize(self.tip, self.line_number - 1) |
623
|
|
|
block_class = getattr( |
624
|
|
|
import_module('CommonMark.blocks'), |
625
|
|
|
to_camel_case(self.tip.t)) |
626
|
|
|
|
627
|
|
|
column_number = offset + 1 |
628
|
|
|
new_block = Node(tag, [[self.line_number, column_number], [0, 0]]) |
629
|
|
|
new_block.string_content = '' |
630
|
|
|
self.tip.append_child(new_block) |
631
|
|
|
self.tip = new_block |
632
|
|
|
return new_block |
633
|
|
|
|
634
|
|
|
def close_unmatched_blocks(self): |
635
|
|
|
"""Finalize and close any unmatched blocks.""" |
636
|
|
|
if not self.all_closed: |
637
|
|
|
while self.oldtip != self.last_matched_container: |
638
|
|
|
parent = self.oldtip.parent |
639
|
|
|
self.finalize(self.oldtip, self.line_number - 1) |
640
|
|
|
self.oldtip = parent |
641
|
|
|
self.all_closed = True |
642
|
|
|
|
643
|
|
|
def find_next_nonspace(self): |
644
|
|
|
current_line = self.current_line |
645
|
|
|
i = self.offset |
646
|
|
|
cols = self.column |
647
|
|
|
|
648
|
|
|
try: |
649
|
|
|
c = current_line[i] |
650
|
|
|
if c is not None: |
651
|
|
|
# c = unicode(c, 'latin1').encode('utf-8') |
652
|
|
|
try: |
653
|
|
|
c = unicode(c, 'latin1').encode('utf-8') |
654
|
|
|
except Exception: |
655
|
|
|
c = str(c); |
656
|
|
|
except IndexError: |
657
|
|
|
c = '' |
658
|
|
|
while c != '': |
659
|
|
|
if c == ' ': |
660
|
|
|
i += 1 |
661
|
|
|
cols += 1 |
662
|
|
|
elif c == '\t': |
663
|
|
|
i += 1 |
664
|
|
|
cols += (4 - (cols % 4)) |
665
|
|
|
else: |
666
|
|
|
break |
667
|
|
|
|
668
|
|
|
try: |
669
|
|
|
c = current_line[i] |
670
|
|
|
except IndexError: |
671
|
|
|
c = '' |
672
|
|
|
|
673
|
|
|
self.blank = (c == '\n' or c == '\r' or c == '') |
674
|
|
|
self.next_nonspace = i |
675
|
|
|
self.next_nonspace_column = cols |
676
|
|
|
self.indent = self.next_nonspace_column - self.column |
677
|
|
|
self.indented = self.indent >= CODE_INDENT |
678
|
|
|
|
679
|
|
|
def advance_next_nonspace(self): |
680
|
|
|
self.offset = self.next_nonspace |
681
|
|
|
self.column = self.next_nonspace_column |
682
|
|
|
self.partially_consumed_tab = False |
683
|
|
|
|
684
|
|
|
def advance_offset(self, count, columns): |
685
|
|
|
current_line = self.current_line |
686
|
|
|
try: |
687
|
|
|
c = current_line[self.offset] |
688
|
|
|
except IndexError: |
689
|
|
|
c = None |
690
|
|
|
while count > 0 and c is not None: |
691
|
|
|
if c == '\t': |
692
|
|
|
chars_to_tab = 4 - (self.column % 4) |
693
|
|
|
if columns: |
694
|
|
|
self.partially_consumed_tab = chars_to_tab > count |
695
|
|
|
chars_to_advance = min(count, chars_to_tab) |
696
|
|
|
self.column += chars_to_advance |
697
|
|
|
self.offset += 0 if self.partially_consumed_tab else 1 |
698
|
|
|
count -= chars_to_advance |
699
|
|
|
else: |
700
|
|
|
self.partially_consumed_tab = False |
701
|
|
|
self.column += chars_to_tab |
702
|
|
|
self.offset += 1 |
703
|
|
|
count -= 1 |
704
|
|
|
else: |
705
|
|
|
self.partially_consumed_tab = False |
706
|
|
|
self.offset += 1 |
707
|
|
|
# assume ascii; block starts are ascii |
708
|
|
|
self.column += 1 |
709
|
|
|
count -= 1 |
710
|
|
|
try: |
711
|
|
|
c = current_line[self.offset] |
712
|
|
|
except IndexError: |
713
|
|
|
c = None |
714
|
|
|
|
715
|
|
|
def incorporate_line(self, ln): |
716
|
|
|
"""Analyze a line of text and update the document appropriately. |
717
|
|
|
|
718
|
|
|
We parse markdown text by calling this on each line of input, |
719
|
|
|
then finalizing the document. |
720
|
|
|
""" |
721
|
|
|
all_matched = True |
722
|
|
|
|
723
|
|
|
container = self.doc |
724
|
|
|
self.oldtip = self.tip |
725
|
|
|
self.offset = 0 |
726
|
|
|
self.column = 0 |
727
|
|
|
self.blank = False |
728
|
|
|
self.partially_consumed_tab = False |
729
|
|
|
self.line_number += 1 |
730
|
|
|
|
731
|
|
|
# replace NUL characters for security |
732
|
|
|
if re.search(r'\u0000', ln) is not None: |
733
|
|
|
ln = re.sub(r'\0', '\uFFFD', ln) |
734
|
|
|
|
735
|
|
|
self.current_line = ln |
736
|
|
|
|
737
|
|
|
# For each containing block, try to parse the associated line start. |
738
|
|
|
# Bail out on failure: container will point to the last matching block. |
739
|
|
|
# Set all_matched to false if not all containers match. |
740
|
|
|
last_child = container.last_child |
741
|
|
|
while last_child and last_child.is_open: |
742
|
|
|
container = last_child |
743
|
|
|
|
744
|
|
|
self.find_next_nonspace() |
745
|
|
|
block_class = getattr( |
746
|
|
|
import_module('CommonMark.blocks'), |
747
|
|
|
to_camel_case(container.t)) |
748
|
|
|
rv = block_class.continue_(self, container) |
749
|
|
|
if rv == 0: |
750
|
|
|
# we've matched, keep going |
751
|
|
|
pass |
752
|
|
|
elif rv == 1: |
753
|
|
|
# we've failed to match a block |
754
|
|
|
all_matched = False |
755
|
|
|
elif rv == 2: |
756
|
|
|
# we've hit end of line for fenced code close and can return |
757
|
|
|
self.last_line_length = len(ln) |
758
|
|
|
return |
759
|
|
|
else: |
760
|
|
|
raise ValueError('returned illegal value, must be 0, 1, or 2') |
761
|
|
|
|
762
|
|
|
if not all_matched: |
763
|
|
|
# back up to last matching block |
764
|
|
|
container = container.parent |
765
|
|
|
break |
766
|
|
|
|
767
|
|
|
last_child = container.last_child |
768
|
|
|
|
769
|
|
|
self.all_closed = (container == self.oldtip) |
770
|
|
|
self.last_matched_container = container |
771
|
|
|
|
772
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
773
|
|
|
to_camel_case(container.t)) |
774
|
|
|
matched_leaf = container.t != 'paragraph' and block_class.accepts_lines |
775
|
|
|
starts = self.block_starts |
776
|
|
|
starts_len = len(starts.METHODS) |
777
|
|
|
# Unless last matched container is a code block, try new container |
778
|
|
|
# starts, adding children to the last matched container: |
779
|
|
|
while not matched_leaf: |
780
|
|
|
self.find_next_nonspace() |
781
|
|
|
|
782
|
|
|
# this is a little performance optimization: |
783
|
|
|
if not self.indented and \ |
784
|
|
|
not re.search(reMaybeSpecial, ln[self.next_nonspace:]): |
785
|
|
|
self.advance_next_nonspace() |
786
|
|
|
break |
787
|
|
|
|
788
|
|
|
i = 0 |
789
|
|
|
while i < starts_len: |
790
|
|
|
res = getattr(starts, starts.METHODS[i])(self, container) |
791
|
|
|
if res == 1: |
792
|
|
|
container = self.tip |
793
|
|
|
break |
794
|
|
|
elif res == 2: |
795
|
|
|
container = self.tip |
796
|
|
|
matched_leaf = True |
797
|
|
|
break |
798
|
|
|
else: |
799
|
|
|
i += 1 |
800
|
|
|
|
801
|
|
|
if i == starts_len: |
802
|
|
|
# nothing matched |
803
|
|
|
self.advance_next_nonspace() |
804
|
|
|
break |
805
|
|
|
|
806
|
|
|
# What remains at the offset is a text line. Add the text to the |
807
|
|
|
# appropriate container. |
808
|
|
|
if not self.all_closed and not self.blank and \ |
809
|
|
|
self.tip.t == 'paragraph': |
810
|
|
|
# lazy paragraph continuation |
811
|
|
|
self.add_line() |
812
|
|
|
else: |
813
|
|
|
# not a lazy continuation |
814
|
|
|
# finalize any blocks not matched |
815
|
|
|
self.close_unmatched_blocks() |
816
|
|
|
if self.blank and container.last_child: |
817
|
|
|
container.last_child.last_line_blank = True |
818
|
|
|
|
819
|
|
|
t = container.t |
820
|
|
|
|
821
|
|
|
# Block quote lines are never blank as they start with > |
822
|
|
|
# and we don't count blanks in fenced code for purposes of |
823
|
|
|
# tight/loose lists or breaking out of lists. We also |
824
|
|
|
# don't set last_line_blank on an empty list item, or if we |
825
|
|
|
# just closed a fenced block. |
826
|
|
|
last_line_blank = self.blank and \ |
827
|
|
|
not (t == 'block_quote' or |
828
|
|
|
(t == 'code_block' and container.is_fenced) or |
829
|
|
|
(t == 'item' and |
830
|
|
|
not container.first_child and |
831
|
|
|
container.sourcepos[0][0] == self.line_number)) |
832
|
|
|
|
833
|
|
|
# propagate last_line_blank up through parents: |
834
|
|
|
cont = container |
835
|
|
|
while cont: |
836
|
|
|
cont.last_line_blank = last_line_blank |
837
|
|
|
cont = cont.parent |
838
|
|
|
|
839
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
840
|
|
|
to_camel_case(t)) |
841
|
|
|
if block_class.accepts_lines: |
842
|
|
|
self.add_line() |
843
|
|
|
# if HtmlBlock, check for end condition |
844
|
|
|
if t == 'html_block' and \ |
845
|
|
|
container.html_block_type >= 1 and \ |
846
|
|
|
container.html_block_type <= 5 and \ |
847
|
|
|
re.search( |
848
|
|
|
reHtmlBlockClose[container.html_block_type], |
849
|
|
|
self.current_line[self.offset:]): |
850
|
|
|
self.finalize(container, self.line_number) |
851
|
|
|
elif self.offset < len(ln) and not self.blank: |
852
|
|
|
# create a paragraph container for one line |
853
|
|
|
container = self.add_child('paragraph', self.offset) |
854
|
|
|
self.advance_next_nonspace() |
855
|
|
|
self.add_line() |
856
|
|
|
|
857
|
|
|
self.last_line_length = len(ln) |
858
|
|
|
|
859
|
|
|
def finalize(self, block, line_number): |
860
|
|
|
""" Finalize a block. Close it and do any necessary postprocessing, |
861
|
|
|
e.g. creating string_content from strings, setting the 'tight' |
862
|
|
|
or 'loose' status of a list, and parsing the beginnings |
863
|
|
|
of paragraphs for reference definitions. Reset the tip to the |
864
|
|
|
parent of the closed block.""" |
865
|
|
|
above = block.parent |
866
|
|
|
block.is_open = False |
867
|
|
|
block.sourcepos[1] = [line_number, self.last_line_length] |
868
|
|
|
block_class = getattr(import_module('CommonMark.blocks'), |
869
|
|
|
to_camel_case(block.t)) |
870
|
|
|
block_class.finalize(self, block) |
871
|
|
|
|
872
|
|
|
self.tip = above |
873
|
|
|
|
874
|
|
|
def process_inlines(self, block): |
875
|
|
|
""" |
876
|
|
|
Walk through a block & children recursively, parsing string content |
877
|
|
|
into inline content where appropriate. |
878
|
|
|
""" |
879
|
|
|
walker = block.walker() |
880
|
|
|
self.inline_parser.refmap = self.refmap |
881
|
|
|
self.inline_parser.options = self.options |
882
|
|
|
event = walker.nxt() |
883
|
|
|
while event is not None: |
884
|
|
|
node = event['node'] |
885
|
|
|
t = node.t |
886
|
|
|
if not event['entering'] and (t == 'paragraph' or t == 'heading'): |
887
|
|
|
self.inline_parser.parse(node) |
888
|
|
|
event = walker.nxt() |
889
|
|
|
|
890
|
|
|
def parse(self, my_input): |
891
|
|
|
""" The main parsing function. Returns a parsed document AST.""" |
892
|
|
|
self.doc = Node('document', [[1, 1], [0, 0]]) |
893
|
|
|
self.tip = self.doc |
894
|
|
|
self.refmap = {} |
895
|
|
|
self.line_number = 0 |
896
|
|
|
self.last_line_length = 0 |
897
|
|
|
self.offset = 0 |
898
|
|
|
self.column = 0 |
899
|
|
|
self.last_matched_container = self.doc |
900
|
|
|
self.current_line = '' |
901
|
|
|
lines = re.split(reLineEnding, my_input) |
902
|
|
|
length = len(lines) |
903
|
|
|
if len(my_input) > 0 and my_input[-1] == '\n': |
904
|
|
|
# ignore last blank line created by final newline |
905
|
|
|
length -= 1 |
906
|
|
|
for i in range(length): |
907
|
|
|
self.incorporate_line(lines[i]) |
908
|
|
|
while (self.tip): |
909
|
|
|
self.finalize(self.tip, length) |
910
|
|
|
self.process_inlines(self.doc) |
911
|
|
|
return self.doc |
912
|
|
|
|