1
|
|
|
from __future__ import unicode_literals |
2
|
|
|
from __future__ import absolute_import |
3
|
|
|
from . import util |
4
|
|
|
from . import odict |
5
|
|
|
from . import inlinepatterns |
6
|
|
|
|
7
|
|
|
|
8
|
|
|
def build_treeprocessors(md_instance, **kwargs): |
9
|
|
|
""" Build the default treeprocessors for Markdown. """ |
10
|
|
|
treeprocessors = odict.OrderedDict() |
11
|
|
|
treeprocessors["inline"] = InlineProcessor(md_instance) |
12
|
|
|
treeprocessors["prettify"] = PrettifyTreeprocessor(md_instance) |
13
|
|
|
return treeprocessors |
14
|
|
|
|
15
|
|
|
|
16
|
|
|
def isString(s): |
17
|
|
|
""" Check if it's string """ |
18
|
|
|
if not isinstance(s, util.AtomicString): |
19
|
|
|
return isinstance(s, util.string_type) |
20
|
|
|
return False |
21
|
|
|
|
22
|
|
|
|
23
|
|
|
class Treeprocessor(util.Processor): |
24
|
|
|
""" |
25
|
|
|
Treeprocessors are run on the ElementTree object before serialization. |
26
|
|
|
|
27
|
|
|
Each Treeprocessor implements a "run" method that takes a pointer to an |
28
|
|
|
ElementTree, modifies it as necessary and returns an ElementTree |
29
|
|
|
object. |
30
|
|
|
|
31
|
|
|
Treeprocessors must extend markdown.Treeprocessor. |
32
|
|
|
|
33
|
|
|
""" |
34
|
|
|
def run(self, root): |
35
|
|
|
""" |
36
|
|
|
Subclasses of Treeprocessor should implement a `run` method, which |
37
|
|
|
takes a root ElementTree. This method can return another ElementTree |
38
|
|
|
object, and the existing root ElementTree will be replaced, or it can |
39
|
|
|
modify the current tree and return None. |
40
|
|
|
""" |
41
|
|
|
pass # pragma: no cover |
42
|
|
|
|
43
|
|
|
|
44
|
|
|
class InlineProcessor(Treeprocessor): |
45
|
|
|
""" |
46
|
|
|
A Treeprocessor that traverses a tree, applying inline patterns. |
47
|
|
|
""" |
48
|
|
|
|
49
|
|
|
def __init__(self, md): |
50
|
|
|
self.__placeholder_prefix = util.INLINE_PLACEHOLDER_PREFIX |
51
|
|
|
self.__placeholder_suffix = util.ETX |
52
|
|
|
self.__placeholder_length = 4 + len(self.__placeholder_prefix) \ |
53
|
|
|
+ len(self.__placeholder_suffix) |
54
|
|
|
self.__placeholder_re = util.INLINE_PLACEHOLDER_RE |
55
|
|
|
self.markdown = md |
56
|
|
|
self.inlinePatterns = md.inlinePatterns |
57
|
|
|
|
58
|
|
|
def __makePlaceholder(self, type): |
59
|
|
|
""" Generate a placeholder """ |
60
|
|
|
id = "%04d" % len(self.stashed_nodes) |
61
|
|
|
hash = util.INLINE_PLACEHOLDER % id |
62
|
|
|
return hash, id |
63
|
|
|
|
64
|
|
|
def __findPlaceholder(self, data, index): |
65
|
|
|
""" |
66
|
|
|
Extract id from data string, start from index |
67
|
|
|
|
68
|
|
|
Keyword arguments: |
69
|
|
|
|
70
|
|
|
* data: string |
71
|
|
|
* index: index, from which we start search |
72
|
|
|
|
73
|
|
|
Returns: placeholder id and string index, after the found placeholder. |
74
|
|
|
|
75
|
|
|
""" |
76
|
|
|
m = self.__placeholder_re.search(data, index) |
77
|
|
|
if m: |
78
|
|
|
return m.group(1), m.end() |
79
|
|
|
else: |
80
|
|
|
return None, index + 1 |
81
|
|
|
|
82
|
|
|
def __stashNode(self, node, type): |
83
|
|
|
""" Add node to stash """ |
84
|
|
|
placeholder, id = self.__makePlaceholder(type) |
85
|
|
|
self.stashed_nodes[id] = node |
86
|
|
|
return placeholder |
87
|
|
|
|
88
|
|
|
def __handleInline(self, data, patternIndex=0): |
89
|
|
|
""" |
90
|
|
|
Process string with inline patterns and replace it |
91
|
|
|
with placeholders |
92
|
|
|
|
93
|
|
|
Keyword arguments: |
94
|
|
|
|
95
|
|
|
* data: A line of Markdown text |
96
|
|
|
* patternIndex: The index of the inlinePattern to start with |
97
|
|
|
|
98
|
|
|
Returns: String with placeholders. |
99
|
|
|
|
100
|
|
|
""" |
101
|
|
|
if not isinstance(data, util.AtomicString): |
102
|
|
|
startIndex = 0 |
103
|
|
|
while patternIndex < len(self.inlinePatterns): |
104
|
|
|
data, matched, startIndex = self.__applyPattern( |
105
|
|
|
self.inlinePatterns.value_for_index(patternIndex), |
106
|
|
|
data, patternIndex, startIndex) |
107
|
|
|
if not matched: |
108
|
|
|
patternIndex += 1 |
109
|
|
|
return data |
110
|
|
|
|
111
|
|
|
def __processElementText(self, node, subnode, isText=True): |
112
|
|
|
""" |
113
|
|
|
Process placeholders in Element.text or Element.tail |
114
|
|
|
of Elements popped from self.stashed_nodes. |
115
|
|
|
|
116
|
|
|
Keywords arguments: |
117
|
|
|
|
118
|
|
|
* node: parent node |
119
|
|
|
* subnode: processing node |
120
|
|
|
* isText: bool variable, True - it's text, False - it's tail |
121
|
|
|
|
122
|
|
|
Returns: None |
123
|
|
|
|
124
|
|
|
""" |
125
|
|
|
if isText: |
126
|
|
|
text = subnode.text |
127
|
|
|
subnode.text = None |
128
|
|
|
else: |
129
|
|
|
text = subnode.tail |
130
|
|
|
subnode.tail = None |
131
|
|
|
|
132
|
|
|
childResult = self.__processPlaceholders(text, subnode, isText) |
133
|
|
|
|
134
|
|
|
if not isText and node is not subnode: |
135
|
|
|
pos = list(node).index(subnode) + 1 |
136
|
|
|
else: |
137
|
|
|
pos = 0 |
138
|
|
|
|
139
|
|
|
childResult.reverse() |
140
|
|
|
for newChild in childResult: |
141
|
|
|
node.insert(pos, newChild) |
142
|
|
|
|
143
|
|
|
def __processPlaceholders(self, data, parent, isText=True): |
144
|
|
|
""" |
145
|
|
|
Process string with placeholders and generate ElementTree tree. |
146
|
|
|
|
147
|
|
|
Keyword arguments: |
148
|
|
|
|
149
|
|
|
* data: string with placeholders instead of ElementTree elements. |
150
|
|
|
* parent: Element, which contains processing inline data |
151
|
|
|
|
152
|
|
|
Returns: list with ElementTree elements with applied inline patterns. |
153
|
|
|
|
154
|
|
|
""" |
155
|
|
|
def linkText(text): |
156
|
|
|
if text: |
157
|
|
|
if result: |
158
|
|
|
if result[-1].tail: |
159
|
|
|
result[-1].tail += text |
160
|
|
|
else: |
161
|
|
|
result[-1].tail = text |
162
|
|
|
elif not isText: |
163
|
|
|
if parent.tail: |
164
|
|
|
parent.tail += text |
165
|
|
|
else: |
166
|
|
|
parent.tail = text |
167
|
|
|
else: |
168
|
|
|
if parent.text: |
169
|
|
|
parent.text += text |
170
|
|
|
else: |
171
|
|
|
parent.text = text |
172
|
|
|
result = [] |
173
|
|
|
strartIndex = 0 |
174
|
|
|
while data: |
175
|
|
|
index = data.find(self.__placeholder_prefix, strartIndex) |
176
|
|
|
if index != -1: |
177
|
|
|
id, phEndIndex = self.__findPlaceholder(data, index) |
178
|
|
|
|
179
|
|
|
if id in self.stashed_nodes: |
180
|
|
|
node = self.stashed_nodes.get(id) |
181
|
|
|
|
182
|
|
|
if index > 0: |
183
|
|
|
text = data[strartIndex:index] |
184
|
|
|
linkText(text) |
185
|
|
|
|
186
|
|
|
if not isString(node): # it's Element |
187
|
|
|
for child in [node] + list(node): |
188
|
|
|
if child.tail: |
189
|
|
|
if child.tail.strip(): |
190
|
|
|
self.__processElementText( |
191
|
|
|
node, child, False |
192
|
|
|
) |
193
|
|
|
if child.text: |
194
|
|
|
if child.text.strip(): |
195
|
|
|
self.__processElementText(child, child) |
196
|
|
|
else: # it's just a string |
197
|
|
|
linkText(node) |
198
|
|
|
strartIndex = phEndIndex |
199
|
|
|
continue |
200
|
|
|
|
201
|
|
|
strartIndex = phEndIndex |
202
|
|
|
result.append(node) |
203
|
|
|
|
204
|
|
|
else: # wrong placeholder |
205
|
|
|
end = index + len(self.__placeholder_prefix) |
206
|
|
|
linkText(data[strartIndex:end]) |
207
|
|
|
strartIndex = end |
208
|
|
|
else: |
209
|
|
|
text = data[strartIndex:] |
210
|
|
|
if isinstance(data, util.AtomicString): |
211
|
|
|
# We don't want to loose the AtomicString |
212
|
|
|
text = util.AtomicString(text) |
213
|
|
|
linkText(text) |
214
|
|
|
data = "" |
215
|
|
|
|
216
|
|
|
return result |
217
|
|
|
|
218
|
|
|
def __applyPattern(self, pattern, data, patternIndex, startIndex=0): |
219
|
|
|
""" |
220
|
|
|
Check if the line fits the pattern, create the necessary |
221
|
|
|
elements, add it to stashed_nodes. |
222
|
|
|
|
223
|
|
|
Keyword arguments: |
224
|
|
|
|
225
|
|
|
* data: the text to be processed |
226
|
|
|
* pattern: the pattern to be checked |
227
|
|
|
* patternIndex: index of current pattern |
228
|
|
|
* startIndex: string index, from which we start searching |
229
|
|
|
|
230
|
|
|
Returns: String with placeholders instead of ElementTree elements. |
231
|
|
|
|
232
|
|
|
""" |
233
|
|
|
match = pattern.getCompiledRegExp().match(data[startIndex:]) |
234
|
|
|
leftData = data[:startIndex] |
235
|
|
|
|
236
|
|
|
if not match: |
237
|
|
|
return data, False, 0 |
238
|
|
|
|
239
|
|
|
node = pattern.handleMatch(match) |
240
|
|
|
|
241
|
|
|
if node is None: |
242
|
|
|
return data, True, len(leftData)+match.span(len(match.groups()))[0] |
243
|
|
|
|
244
|
|
|
if not isString(node): |
245
|
|
|
if not isinstance(node.text, util.AtomicString): |
246
|
|
|
# We need to process current node too |
247
|
|
|
for child in [node] + list(node): |
248
|
|
|
if not isString(node): |
249
|
|
|
if child.text: |
250
|
|
|
child.text = self.__handleInline( |
251
|
|
|
child.text, patternIndex + 1 |
252
|
|
|
) |
253
|
|
|
if child.tail: |
254
|
|
|
child.tail = self.__handleInline( |
255
|
|
|
child.tail, patternIndex |
256
|
|
|
) |
257
|
|
|
|
258
|
|
|
placeholder = self.__stashNode(node, pattern.type()) |
259
|
|
|
|
260
|
|
|
return "%s%s%s%s" % (leftData, |
261
|
|
|
match.group(1), |
262
|
|
|
placeholder, match.groups()[-1]), True, 0 |
263
|
|
|
|
264
|
|
|
def run(self, tree): |
265
|
|
|
"""Apply inline patterns to a parsed Markdown tree. |
266
|
|
|
|
267
|
|
|
Iterate over ElementTree, find elements with inline tag, apply inline |
268
|
|
|
patterns and append newly created Elements to tree. If you don't |
269
|
|
|
want to process your data with inline paterns, instead of normal |
270
|
|
|
string, use subclass AtomicString: |
271
|
|
|
|
272
|
|
|
node.text = markdown.AtomicString("This will not be processed.") |
273
|
|
|
|
274
|
|
|
Arguments: |
275
|
|
|
|
276
|
|
|
* tree: ElementTree object, representing Markdown tree. |
277
|
|
|
|
278
|
|
|
Returns: ElementTree object with applied inline patterns. |
279
|
|
|
|
280
|
|
|
""" |
281
|
|
|
self.stashed_nodes = {} |
282
|
|
|
|
283
|
|
|
stack = [tree] |
284
|
|
|
|
285
|
|
|
while stack: |
286
|
|
|
currElement = stack.pop() |
287
|
|
|
insertQueue = [] |
288
|
|
|
for child in currElement: |
289
|
|
|
if child.text and not isinstance( |
290
|
|
|
child.text, util.AtomicString |
291
|
|
|
): |
292
|
|
|
text = child.text |
293
|
|
|
child.text = None |
294
|
|
|
lst = self.__processPlaceholders( |
295
|
|
|
self.__handleInline(text), child |
296
|
|
|
) |
297
|
|
|
stack += lst |
298
|
|
|
insertQueue.append((child, lst)) |
299
|
|
|
if child.tail: |
300
|
|
|
tail = self.__handleInline(child.tail) |
301
|
|
|
dumby = util.etree.Element('d') |
302
|
|
|
child.tail = None |
303
|
|
|
tailResult = self.__processPlaceholders(tail, dumby, False) |
304
|
|
|
if dumby.tail: |
305
|
|
|
child.tail = dumby.tail |
306
|
|
|
pos = list(currElement).index(child) + 1 |
307
|
|
|
tailResult.reverse() |
308
|
|
|
for newChild in tailResult: |
309
|
|
|
currElement.insert(pos, newChild) |
310
|
|
|
if len(child): |
311
|
|
|
stack.append(child) |
312
|
|
|
|
313
|
|
|
for element, lst in insertQueue: |
314
|
|
|
if self.markdown.enable_attributes: |
315
|
|
|
if element.text and isString(element.text): |
316
|
|
|
element.text = inlinepatterns.handleAttributes( |
317
|
|
|
element.text, element |
318
|
|
|
) |
319
|
|
|
i = 0 |
320
|
|
|
for newChild in lst: |
321
|
|
|
if self.markdown.enable_attributes: |
322
|
|
|
# Processing attributes |
323
|
|
|
if newChild.tail and isString(newChild.tail): |
324
|
|
|
newChild.tail = inlinepatterns.handleAttributes( |
325
|
|
|
newChild.tail, element |
326
|
|
|
) |
327
|
|
|
if newChild.text and isString(newChild.text): |
328
|
|
|
newChild.text = inlinepatterns.handleAttributes( |
329
|
|
|
newChild.text, newChild |
330
|
|
|
) |
331
|
|
|
element.insert(i, newChild) |
332
|
|
|
i += 1 |
333
|
|
|
return tree |
334
|
|
|
|
335
|
|
|
|
336
|
|
|
class PrettifyTreeprocessor(Treeprocessor): |
337
|
|
|
""" Add linebreaks to the html document. """ |
338
|
|
|
|
339
|
|
|
def _prettifyETree(self, elem): |
340
|
|
|
""" Recursively add linebreaks to ElementTree children. """ |
341
|
|
|
|
342
|
|
|
i = "\n" |
343
|
|
|
if util.isBlockLevel(elem.tag) and elem.tag not in ['code', 'pre']: |
344
|
|
|
if (not elem.text or not elem.text.strip()) \ |
345
|
|
|
and len(elem) and util.isBlockLevel(elem[0].tag): |
346
|
|
|
elem.text = i |
347
|
|
|
for e in elem: |
348
|
|
|
if util.isBlockLevel(e.tag): |
349
|
|
|
self._prettifyETree(e) |
350
|
|
|
if not elem.tail or not elem.tail.strip(): |
351
|
|
|
elem.tail = i |
352
|
|
|
if not elem.tail or not elem.tail.strip(): |
353
|
|
|
elem.tail = i |
354
|
|
|
|
355
|
|
|
def run(self, root): |
356
|
|
|
""" Add linebreaks to ElementTree root object. """ |
357
|
|
|
|
358
|
|
|
self._prettifyETree(root) |
359
|
|
|
# Do <br />'s seperately as they are often in the middle of |
360
|
|
|
# inline content and missed by _prettifyETree. |
361
|
|
|
brs = root.iter('br') |
362
|
|
|
for br in brs: |
363
|
|
|
if not br.tail or not br.tail.strip(): |
364
|
|
|
br.tail = '\n' |
365
|
|
|
else: |
366
|
|
|
br.tail = '\n%s' % br.tail |
367
|
|
|
# Clean up extra empty lines at end of code blocks. |
368
|
|
|
pres = root.iter('pre') |
369
|
|
|
for pre in pres: |
370
|
|
|
if len(pre) and pre[0].tag == 'code': |
371
|
|
|
pre[0].text = util.AtomicString(pre[0].text.rstrip() + '\n') |
372
|
|
|
|