Show More
@@ -1,710 +1,710 b'' | |||
|
1 | 1 | # minirst.py - minimal reStructuredText parser |
|
2 | 2 | # |
|
3 | 3 | # Copyright 2009, 2010 Matt Mackall <mpm@selenic.com> and others |
|
4 | 4 | # |
|
5 | 5 | # This software may be used and distributed according to the terms of the |
|
6 | 6 | # GNU General Public License version 2 or any later version. |
|
7 | 7 | |
|
8 | 8 | """simplified reStructuredText parser. |
|
9 | 9 | |
|
10 | 10 | This parser knows just enough about reStructuredText to parse the |
|
11 | 11 | Mercurial docstrings. |
|
12 | 12 | |
|
13 | 13 | It cheats in a major way: nested blocks are not really nested. They |
|
14 | 14 | are just indented blocks that look like they are nested. This relies |
|
15 | 15 | on the user to keep the right indentation for the blocks. |
|
16 | 16 | |
|
17 | 17 | Remember to update http://mercurial.selenic.com/wiki/HelpStyleGuide |
|
18 | 18 | when adding support for new constructs. |
|
19 | 19 | """ |
|
20 | 20 | |
|
21 | 21 | import re |
|
22 | 22 | import util, encoding |
|
23 | 23 | from i18n import _ |
|
24 | 24 | |
|
25 | 25 | import cgi |
|
26 | 26 | |
|
27 | 27 | def section(s): |
|
28 | 28 | return "%s\n%s\n\n" % (s, "\"" * encoding.colwidth(s)) |
|
29 | 29 | |
|
30 | 30 | def subsection(s): |
|
31 | 31 | return "%s\n%s\n\n" % (s, '=' * encoding.colwidth(s)) |
|
32 | 32 | |
|
33 | 33 | def subsubsection(s): |
|
34 | 34 | return "%s\n%s\n\n" % (s, "-" * encoding.colwidth(s)) |
|
35 | 35 | |
|
36 | 36 | def subsubsubsection(s): |
|
37 | 37 | return "%s\n%s\n\n" % (s, "." * encoding.colwidth(s)) |
|
38 | 38 | |
|
39 | 39 | def replace(text, substs): |
|
40 | 40 | ''' |
|
41 | 41 | Apply a list of (find, replace) pairs to a text. |
|
42 | 42 | |
|
43 | 43 | >>> replace("foo bar", [('f', 'F'), ('b', 'B')]) |
|
44 | 44 | 'Foo Bar' |
|
45 | 45 | >>> encoding.encoding = 'latin1' |
|
46 | 46 | >>> replace('\\x81\\\\', [('\\\\', '/')]) |
|
47 | 47 | '\\x81/' |
|
48 | 48 | >>> encoding.encoding = 'shiftjis' |
|
49 | 49 | >>> replace('\\x81\\\\', [('\\\\', '/')]) |
|
50 | 50 | '\\x81\\\\' |
|
51 | 51 | ''' |
|
52 | 52 | |
|
53 | 53 | # some character encodings (cp932 for Japanese, at least) use |
|
54 | 54 | # ASCII characters other than control/alphabet/digit as a part of |
|
55 | 55 | # multi-bytes characters, so direct replacing with such characters |
|
56 | 56 | # on strings in local encoding causes invalid byte sequences. |
|
57 | 57 | utext = text.decode(encoding.encoding) |
|
58 | 58 | for f, t in substs: |
|
59 | 59 | utext = utext.replace(f, t) |
|
60 | 60 | return utext.encode(encoding.encoding) |
|
61 | 61 | |
|
62 | 62 | _blockre = re.compile(r"\n(?:\s*\n)+") |
|
63 | 63 | |
|
64 | 64 | def findblocks(text): |
|
65 | 65 | """Find continuous blocks of lines in text. |
|
66 | 66 | |
|
67 | 67 | Returns a list of dictionaries representing the blocks. Each block |
|
68 | 68 | has an 'indent' field and a 'lines' field. |
|
69 | 69 | """ |
|
70 | 70 | blocks = [] |
|
71 | 71 | for b in _blockre.split(text.lstrip('\n').rstrip()): |
|
72 | 72 | lines = b.splitlines() |
|
73 | 73 | if lines: |
|
74 | 74 | indent = min((len(l) - len(l.lstrip())) for l in lines) |
|
75 | 75 | lines = [l[indent:] for l in lines] |
|
76 |
blocks.append( |
|
|
76 | blocks.append({'indent': indent, 'lines': lines}) | |
|
77 | 77 | return blocks |
|
78 | 78 | |
|
79 | 79 | def findliteralblocks(blocks): |
|
80 | 80 | """Finds literal blocks and adds a 'type' field to the blocks. |
|
81 | 81 | |
|
82 | 82 | Literal blocks are given the type 'literal', all other blocks are |
|
83 | 83 | given type the 'paragraph'. |
|
84 | 84 | """ |
|
85 | 85 | i = 0 |
|
86 | 86 | while i < len(blocks): |
|
87 | 87 | # Searching for a block that looks like this: |
|
88 | 88 | # |
|
89 | 89 | # +------------------------------+ |
|
90 | 90 | # | paragraph | |
|
91 | 91 | # | (ends with "::") | |
|
92 | 92 | # +------------------------------+ |
|
93 | 93 | # +---------------------------+ |
|
94 | 94 | # | indented literal block | |
|
95 | 95 | # +---------------------------+ |
|
96 | 96 | blocks[i]['type'] = 'paragraph' |
|
97 | 97 | if blocks[i]['lines'][-1].endswith('::') and i + 1 < len(blocks): |
|
98 | 98 | indent = blocks[i]['indent'] |
|
99 | 99 | adjustment = blocks[i + 1]['indent'] - indent |
|
100 | 100 | |
|
101 | 101 | if blocks[i]['lines'] == ['::']: |
|
102 | 102 | # Expanded form: remove block |
|
103 | 103 | del blocks[i] |
|
104 | 104 | i -= 1 |
|
105 | 105 | elif blocks[i]['lines'][-1].endswith(' ::'): |
|
106 | 106 | # Partially minimized form: remove space and both |
|
107 | 107 | # colons. |
|
108 | 108 | blocks[i]['lines'][-1] = blocks[i]['lines'][-1][:-3] |
|
109 | 109 | elif len(blocks[i]['lines']) == 1 and \ |
|
110 | 110 | blocks[i]['lines'][0].lstrip(' ').startswith('.. ') and \ |
|
111 | 111 | blocks[i]['lines'][0].find(' ', 3) == -1: |
|
112 | 112 | # directive on its own line, not a literal block |
|
113 | 113 | i += 1 |
|
114 | 114 | continue |
|
115 | 115 | else: |
|
116 | 116 | # Fully minimized form: remove just one colon. |
|
117 | 117 | blocks[i]['lines'][-1] = blocks[i]['lines'][-1][:-1] |
|
118 | 118 | |
|
119 | 119 | # List items are formatted with a hanging indent. We must |
|
120 | 120 | # correct for this here while we still have the original |
|
121 | 121 | # information on the indentation of the subsequent literal |
|
122 | 122 | # blocks available. |
|
123 | 123 | m = _bulletre.match(blocks[i]['lines'][0]) |
|
124 | 124 | if m: |
|
125 | 125 | indent += m.end() |
|
126 | 126 | adjustment -= m.end() |
|
127 | 127 | |
|
128 | 128 | # Mark the following indented blocks. |
|
129 | 129 | while i + 1 < len(blocks) and blocks[i + 1]['indent'] > indent: |
|
130 | 130 | blocks[i + 1]['type'] = 'literal' |
|
131 | 131 | blocks[i + 1]['indent'] -= adjustment |
|
132 | 132 | i += 1 |
|
133 | 133 | i += 1 |
|
134 | 134 | return blocks |
|
135 | 135 | |
|
136 | 136 | _bulletre = re.compile(r'(-|[0-9A-Za-z]+\.|\(?[0-9A-Za-z]+\)|\|) ') |
|
137 | 137 | _optionre = re.compile(r'^(-([a-zA-Z0-9]), )?(--[a-z0-9-]+)' |
|
138 | 138 | r'((.*) +)(.*)$') |
|
139 | 139 | _fieldre = re.compile(r':(?![: ])([^:]*)(?<! ):[ ]+(.*)') |
|
140 | 140 | _definitionre = re.compile(r'[^ ]') |
|
141 | 141 | _tablere = re.compile(r'(=+\s+)*=+') |
|
142 | 142 | |
|
143 | 143 | def splitparagraphs(blocks): |
|
144 | 144 | """Split paragraphs into lists.""" |
|
145 | 145 | # Tuples with (list type, item regexp, single line items?). Order |
|
146 | 146 | # matters: definition lists has the least specific regexp and must |
|
147 | 147 | # come last. |
|
148 | 148 | listtypes = [('bullet', _bulletre, True), |
|
149 | 149 | ('option', _optionre, True), |
|
150 | 150 | ('field', _fieldre, True), |
|
151 | 151 | ('definition', _definitionre, False)] |
|
152 | 152 | |
|
153 | 153 | def match(lines, i, itemre, singleline): |
|
154 | 154 | """Does itemre match an item at line i? |
|
155 | 155 | |
|
156 | 156 | A list item can be followed by an indented line or another list |
|
157 | 157 | item (but only if singleline is True). |
|
158 | 158 | """ |
|
159 | 159 | line1 = lines[i] |
|
160 | 160 | line2 = i + 1 < len(lines) and lines[i + 1] or '' |
|
161 | 161 | if not itemre.match(line1): |
|
162 | 162 | return False |
|
163 | 163 | if singleline: |
|
164 | 164 | return line2 == '' or line2[0] == ' ' or itemre.match(line2) |
|
165 | 165 | else: |
|
166 | 166 | return line2.startswith(' ') |
|
167 | 167 | |
|
168 | 168 | i = 0 |
|
169 | 169 | while i < len(blocks): |
|
170 | 170 | if blocks[i]['type'] == 'paragraph': |
|
171 | 171 | lines = blocks[i]['lines'] |
|
172 | 172 | for type, itemre, singleline in listtypes: |
|
173 | 173 | if match(lines, 0, itemre, singleline): |
|
174 | 174 | items = [] |
|
175 | 175 | for j, line in enumerate(lines): |
|
176 | 176 | if match(lines, j, itemre, singleline): |
|
177 |
items.append( |
|
|
178 |
|
|
|
177 | items.append({'type': type, 'lines': [], | |
|
178 | 'indent': blocks[i]['indent']}) | |
|
179 | 179 | items[-1]['lines'].append(line) |
|
180 | 180 | blocks[i:i + 1] = items |
|
181 | 181 | break |
|
182 | 182 | i += 1 |
|
183 | 183 | return blocks |
|
184 | 184 | |
|
185 | 185 | _fieldwidth = 14 |
|
186 | 186 | |
|
187 | 187 | def updatefieldlists(blocks): |
|
188 | 188 | """Find key for field lists.""" |
|
189 | 189 | i = 0 |
|
190 | 190 | while i < len(blocks): |
|
191 | 191 | if blocks[i]['type'] != 'field': |
|
192 | 192 | i += 1 |
|
193 | 193 | continue |
|
194 | 194 | |
|
195 | 195 | j = i |
|
196 | 196 | while j < len(blocks) and blocks[j]['type'] == 'field': |
|
197 | 197 | m = _fieldre.match(blocks[j]['lines'][0]) |
|
198 | 198 | key, rest = m.groups() |
|
199 | 199 | blocks[j]['lines'][0] = rest |
|
200 | 200 | blocks[j]['key'] = key |
|
201 | 201 | j += 1 |
|
202 | 202 | |
|
203 | 203 | i = j + 1 |
|
204 | 204 | |
|
205 | 205 | return blocks |
|
206 | 206 | |
|
207 | 207 | def updateoptionlists(blocks): |
|
208 | 208 | i = 0 |
|
209 | 209 | while i < len(blocks): |
|
210 | 210 | if blocks[i]['type'] != 'option': |
|
211 | 211 | i += 1 |
|
212 | 212 | continue |
|
213 | 213 | |
|
214 | 214 | optstrwidth = 0 |
|
215 | 215 | j = i |
|
216 | 216 | while j < len(blocks) and blocks[j]['type'] == 'option': |
|
217 | 217 | m = _optionre.match(blocks[j]['lines'][0]) |
|
218 | 218 | |
|
219 | 219 | shortoption = m.group(2) |
|
220 | 220 | group3 = m.group(3) |
|
221 | 221 | longoption = group3[2:].strip() |
|
222 | 222 | desc = m.group(6).strip() |
|
223 | 223 | longoptionarg = m.group(5).strip() |
|
224 | 224 | blocks[j]['lines'][0] = desc |
|
225 | 225 | |
|
226 | 226 | noshortop = '' |
|
227 | 227 | if not shortoption: |
|
228 | 228 | noshortop = ' ' |
|
229 | 229 | |
|
230 | 230 | opt = "%s%s" % (shortoption and "-%s " % shortoption or '', |
|
231 | 231 | ("%s--%s %s") % (noshortop, longoption, |
|
232 | 232 | longoptionarg)) |
|
233 | 233 | opt = opt.rstrip() |
|
234 | 234 | blocks[j]['optstr'] = opt |
|
235 | 235 | optstrwidth = max(optstrwidth, encoding.colwidth(opt)) |
|
236 | 236 | j += 1 |
|
237 | 237 | |
|
238 | 238 | for block in blocks[i:j]: |
|
239 | 239 | block['optstrwidth'] = optstrwidth |
|
240 | 240 | i = j + 1 |
|
241 | 241 | return blocks |
|
242 | 242 | |
|
243 | 243 | def prunecontainers(blocks, keep): |
|
244 | 244 | """Prune unwanted containers. |
|
245 | 245 | |
|
246 | 246 | The blocks must have a 'type' field, i.e., they should have been |
|
247 | 247 | run through findliteralblocks first. |
|
248 | 248 | """ |
|
249 | 249 | pruned = [] |
|
250 | 250 | i = 0 |
|
251 | 251 | while i + 1 < len(blocks): |
|
252 | 252 | # Searching for a block that looks like this: |
|
253 | 253 | # |
|
254 | 254 | # +-------+---------------------------+ |
|
255 | 255 | # | ".. container ::" type | |
|
256 | 256 | # +---+ | |
|
257 | 257 | # | blocks | |
|
258 | 258 | # +-------------------------------+ |
|
259 | 259 | if (blocks[i]['type'] == 'paragraph' and |
|
260 | 260 | blocks[i]['lines'][0].startswith('.. container::')): |
|
261 | 261 | indent = blocks[i]['indent'] |
|
262 | 262 | adjustment = blocks[i + 1]['indent'] - indent |
|
263 | 263 | containertype = blocks[i]['lines'][0][15:] |
|
264 | 264 | prune = containertype not in keep |
|
265 | 265 | if prune: |
|
266 | 266 | pruned.append(containertype) |
|
267 | 267 | |
|
268 | 268 | # Always delete "..container:: type" block |
|
269 | 269 | del blocks[i] |
|
270 | 270 | j = i |
|
271 | 271 | i -= 1 |
|
272 | 272 | while j < len(blocks) and blocks[j]['indent'] > indent: |
|
273 | 273 | if prune: |
|
274 | 274 | del blocks[j] |
|
275 | 275 | else: |
|
276 | 276 | blocks[j]['indent'] -= adjustment |
|
277 | 277 | j += 1 |
|
278 | 278 | i += 1 |
|
279 | 279 | return blocks, pruned |
|
280 | 280 | |
|
281 | 281 | _sectionre = re.compile(r"""^([-=`:.'"~^_*+#])\1+$""") |
|
282 | 282 | |
|
283 | 283 | def findtables(blocks): |
|
284 | 284 | '''Find simple tables |
|
285 | 285 | |
|
286 | 286 | Only simple one-line table elements are supported |
|
287 | 287 | ''' |
|
288 | 288 | |
|
289 | 289 | for block in blocks: |
|
290 | 290 | # Searching for a block that looks like this: |
|
291 | 291 | # |
|
292 | 292 | # === ==== === |
|
293 | 293 | # A B C |
|
294 | 294 | # === ==== === <- optional |
|
295 | 295 | # 1 2 3 |
|
296 | 296 | # x y z |
|
297 | 297 | # === ==== === |
|
298 | 298 | if (block['type'] == 'paragraph' and |
|
299 | 299 | len(block['lines']) > 2 and |
|
300 | 300 | _tablere.match(block['lines'][0]) and |
|
301 | 301 | block['lines'][0] == block['lines'][-1]): |
|
302 | 302 | block['type'] = 'table' |
|
303 | 303 | block['header'] = False |
|
304 | 304 | div = block['lines'][0] |
|
305 | 305 | |
|
306 | 306 | # column markers are ASCII so we can calculate column |
|
307 | 307 | # position in bytes |
|
308 | 308 | columns = [x for x in xrange(len(div)) |
|
309 | 309 | if div[x] == '=' and (x == 0 or div[x - 1] == ' ')] |
|
310 | 310 | rows = [] |
|
311 | 311 | for l in block['lines'][1:-1]: |
|
312 | 312 | if l == div: |
|
313 | 313 | block['header'] = True |
|
314 | 314 | continue |
|
315 | 315 | row = [] |
|
316 | 316 | # we measure columns not in bytes or characters but in |
|
317 | 317 | # colwidth which makes things tricky |
|
318 | 318 | pos = columns[0] # leading whitespace is bytes |
|
319 | 319 | for n, start in enumerate(columns): |
|
320 | 320 | if n + 1 < len(columns): |
|
321 | 321 | width = columns[n + 1] - start |
|
322 | 322 | v = encoding.getcols(l, pos, width) # gather columns |
|
323 | 323 | pos += len(v) # calculate byte position of end |
|
324 | 324 | row.append(v.strip()) |
|
325 | 325 | else: |
|
326 | 326 | row.append(l[pos:].strip()) |
|
327 | 327 | rows.append(row) |
|
328 | 328 | |
|
329 | 329 | block['table'] = rows |
|
330 | 330 | |
|
331 | 331 | return blocks |
|
332 | 332 | |
|
333 | 333 | def findsections(blocks): |
|
334 | 334 | """Finds sections. |
|
335 | 335 | |
|
336 | 336 | The blocks must have a 'type' field, i.e., they should have been |
|
337 | 337 | run through findliteralblocks first. |
|
338 | 338 | """ |
|
339 | 339 | for block in blocks: |
|
340 | 340 | # Searching for a block that looks like this: |
|
341 | 341 | # |
|
342 | 342 | # +------------------------------+ |
|
343 | 343 | # | Section title | |
|
344 | 344 | # | ------------- | |
|
345 | 345 | # +------------------------------+ |
|
346 | 346 | if (block['type'] == 'paragraph' and |
|
347 | 347 | len(block['lines']) == 2 and |
|
348 | 348 | encoding.colwidth(block['lines'][0]) == len(block['lines'][1]) and |
|
349 | 349 | _sectionre.match(block['lines'][1])): |
|
350 | 350 | block['underline'] = block['lines'][1][0] |
|
351 | 351 | block['type'] = 'section' |
|
352 | 352 | del block['lines'][1] |
|
353 | 353 | return blocks |
|
354 | 354 | |
|
355 | 355 | def inlineliterals(blocks): |
|
356 | 356 | substs = [('``', '"')] |
|
357 | 357 | for b in blocks: |
|
358 | 358 | if b['type'] in ('paragraph', 'section'): |
|
359 | 359 | b['lines'] = [replace(l, substs) for l in b['lines']] |
|
360 | 360 | return blocks |
|
361 | 361 | |
|
362 | 362 | def hgrole(blocks): |
|
363 | 363 | substs = [(':hg:`', '"hg '), ('`', '"')] |
|
364 | 364 | for b in blocks: |
|
365 | 365 | if b['type'] in ('paragraph', 'section'): |
|
366 | 366 | # Turn :hg:`command` into "hg command". This also works |
|
367 | 367 | # when there is a line break in the command and relies on |
|
368 | 368 | # the fact that we have no stray back-quotes in the input |
|
369 | 369 | # (run the blocks through inlineliterals first). |
|
370 | 370 | b['lines'] = [replace(l, substs) for l in b['lines']] |
|
371 | 371 | return blocks |
|
372 | 372 | |
|
373 | 373 | def addmargins(blocks): |
|
374 | 374 | """Adds empty blocks for vertical spacing. |
|
375 | 375 | |
|
376 | 376 | This groups bullets, options, and definitions together with no vertical |
|
377 | 377 | space between them, and adds an empty block between all other blocks. |
|
378 | 378 | """ |
|
379 | 379 | i = 1 |
|
380 | 380 | while i < len(blocks): |
|
381 | 381 | if (blocks[i]['type'] == blocks[i - 1]['type'] and |
|
382 | 382 | blocks[i]['type'] in ('bullet', 'option', 'field')): |
|
383 | 383 | i += 1 |
|
384 | 384 | elif not blocks[i - 1]['lines']: |
|
385 | 385 | # no lines in previous block, do not separate |
|
386 | 386 | i += 1 |
|
387 | 387 | else: |
|
388 |
blocks.insert(i, |
|
|
388 | blocks.insert(i, {'lines': [''], 'indent': 0, 'type': 'margin'}) | |
|
389 | 389 | i += 2 |
|
390 | 390 | return blocks |
|
391 | 391 | |
|
392 | 392 | def prunecomments(blocks): |
|
393 | 393 | """Remove comments.""" |
|
394 | 394 | i = 0 |
|
395 | 395 | while i < len(blocks): |
|
396 | 396 | b = blocks[i] |
|
397 | 397 | if b['type'] == 'paragraph' and (b['lines'][0].startswith('.. ') or |
|
398 | 398 | b['lines'] == ['..']): |
|
399 | 399 | del blocks[i] |
|
400 | 400 | if i < len(blocks) and blocks[i]['type'] == 'margin': |
|
401 | 401 | del blocks[i] |
|
402 | 402 | else: |
|
403 | 403 | i += 1 |
|
404 | 404 | return blocks |
|
405 | 405 | |
|
406 | 406 | _admonitionre = re.compile(r"\.\. (admonition|attention|caution|danger|" |
|
407 | 407 | r"error|hint|important|note|tip|warning)::", |
|
408 | 408 | flags=re.IGNORECASE) |
|
409 | 409 | |
|
410 | 410 | def findadmonitions(blocks): |
|
411 | 411 | """ |
|
412 | 412 | Makes the type of the block an admonition block if |
|
413 | 413 | the first line is an admonition directive |
|
414 | 414 | """ |
|
415 | 415 | i = 0 |
|
416 | 416 | while i < len(blocks): |
|
417 | 417 | m = _admonitionre.match(blocks[i]['lines'][0]) |
|
418 | 418 | if m: |
|
419 | 419 | blocks[i]['type'] = 'admonition' |
|
420 | 420 | admonitiontitle = blocks[i]['lines'][0][3:m.end() - 2].lower() |
|
421 | 421 | |
|
422 | 422 | firstline = blocks[i]['lines'][0][m.end() + 1:] |
|
423 | 423 | if firstline: |
|
424 | 424 | blocks[i]['lines'].insert(1, ' ' + firstline) |
|
425 | 425 | |
|
426 | 426 | blocks[i]['admonitiontitle'] = admonitiontitle |
|
427 | 427 | del blocks[i]['lines'][0] |
|
428 | 428 | i = i + 1 |
|
429 | 429 | return blocks |
|
430 | 430 | |
|
431 | 431 | _admonitiontitles = {'attention': _('Attention:'), |
|
432 | 432 | 'caution': _('Caution:'), |
|
433 | 433 | 'danger': _('!Danger!') , |
|
434 | 434 | 'error': _('Error:'), |
|
435 | 435 | 'hint': _('Hint:'), |
|
436 | 436 | 'important': _('Important:'), |
|
437 | 437 | 'note': _('Note:'), |
|
438 | 438 | 'tip': _('Tip:'), |
|
439 | 439 | 'warning': _('Warning!')} |
|
440 | 440 | |
|
441 | 441 | def formatoption(block, width): |
|
442 | 442 | desc = ' '.join(map(str.strip, block['lines'])) |
|
443 | 443 | colwidth = encoding.colwidth(block['optstr']) |
|
444 | 444 | usablewidth = width - 1 |
|
445 | 445 | hanging = block['optstrwidth'] |
|
446 | 446 | initindent = '%s%s ' % (block['optstr'], ' ' * ((hanging - colwidth))) |
|
447 | 447 | hangindent = ' ' * (encoding.colwidth(initindent) + 1) |
|
448 | 448 | return ' %s\n' % (util.wrap(desc, usablewidth, |
|
449 | 449 | initindent=initindent, |
|
450 | 450 | hangindent=hangindent)) |
|
451 | 451 | |
|
452 | 452 | def formatblock(block, width): |
|
453 | 453 | """Format a block according to width.""" |
|
454 | 454 | if width <= 0: |
|
455 | 455 | width = 78 |
|
456 | 456 | indent = ' ' * block['indent'] |
|
457 | 457 | if block['type'] == 'admonition': |
|
458 | 458 | admonition = _admonitiontitles[block['admonitiontitle']] |
|
459 | 459 | if not block['lines']: |
|
460 | 460 | return indent + admonition + '\n' |
|
461 | 461 | hang = len(block['lines'][-1]) - len(block['lines'][-1].lstrip()) |
|
462 | 462 | |
|
463 | 463 | defindent = indent + hang * ' ' |
|
464 | 464 | text = ' '.join(map(str.strip, block['lines'])) |
|
465 | 465 | return '%s\n%s\n' % (indent + admonition, |
|
466 | 466 | util.wrap(text, width=width, |
|
467 | 467 | initindent=defindent, |
|
468 | 468 | hangindent=defindent)) |
|
469 | 469 | if block['type'] == 'margin': |
|
470 | 470 | return '\n' |
|
471 | 471 | if block['type'] == 'literal': |
|
472 | 472 | indent += ' ' |
|
473 | 473 | return indent + ('\n' + indent).join(block['lines']) + '\n' |
|
474 | 474 | if block['type'] == 'section': |
|
475 | 475 | underline = encoding.colwidth(block['lines'][0]) * block['underline'] |
|
476 | 476 | return "%s%s\n%s%s\n" % (indent, block['lines'][0],indent, underline) |
|
477 | 477 | if block['type'] == 'table': |
|
478 | 478 | table = block['table'] |
|
479 | 479 | # compute column widths |
|
480 | 480 | widths = [max([encoding.colwidth(e) for e in c]) for c in zip(*table)] |
|
481 | 481 | text = '' |
|
482 | 482 | span = sum(widths) + len(widths) - 1 |
|
483 | 483 | indent = ' ' * block['indent'] |
|
484 | 484 | hang = ' ' * (len(indent) + span - widths[-1]) |
|
485 | 485 | |
|
486 | 486 | for row in table: |
|
487 | 487 | l = [] |
|
488 | 488 | for w, v in zip(widths, row): |
|
489 | 489 | pad = ' ' * (w - encoding.colwidth(v)) |
|
490 | 490 | l.append(v + pad) |
|
491 | 491 | l = ' '.join(l) |
|
492 | 492 | l = util.wrap(l, width=width, initindent=indent, hangindent=hang) |
|
493 | 493 | if not text and block['header']: |
|
494 | 494 | text = l + '\n' + indent + '-' * (min(width, span)) + '\n' |
|
495 | 495 | else: |
|
496 | 496 | text += l + "\n" |
|
497 | 497 | return text |
|
498 | 498 | if block['type'] == 'definition': |
|
499 | 499 | term = indent + block['lines'][0] |
|
500 | 500 | hang = len(block['lines'][-1]) - len(block['lines'][-1].lstrip()) |
|
501 | 501 | defindent = indent + hang * ' ' |
|
502 | 502 | text = ' '.join(map(str.strip, block['lines'][1:])) |
|
503 | 503 | return '%s\n%s\n' % (term, util.wrap(text, width=width, |
|
504 | 504 | initindent=defindent, |
|
505 | 505 | hangindent=defindent)) |
|
506 | 506 | subindent = indent |
|
507 | 507 | if block['type'] == 'bullet': |
|
508 | 508 | if block['lines'][0].startswith('| '): |
|
509 | 509 | # Remove bullet for line blocks and add no extra |
|
510 | 510 | # indention. |
|
511 | 511 | block['lines'][0] = block['lines'][0][2:] |
|
512 | 512 | else: |
|
513 | 513 | m = _bulletre.match(block['lines'][0]) |
|
514 | 514 | subindent = indent + m.end() * ' ' |
|
515 | 515 | elif block['type'] == 'field': |
|
516 | 516 | key = block['key'] |
|
517 | 517 | subindent = indent + _fieldwidth * ' ' |
|
518 | 518 | if len(key) + 2 > _fieldwidth: |
|
519 | 519 | # key too large, use full line width |
|
520 | 520 | key = key.ljust(width) |
|
521 | 521 | else: |
|
522 | 522 | # key fits within field width |
|
523 | 523 | key = key.ljust(_fieldwidth) |
|
524 | 524 | block['lines'][0] = key + block['lines'][0] |
|
525 | 525 | elif block['type'] == 'option': |
|
526 | 526 | return formatoption(block, width) |
|
527 | 527 | |
|
528 | 528 | text = ' '.join(map(str.strip, block['lines'])) |
|
529 | 529 | return util.wrap(text, width=width, |
|
530 | 530 | initindent=indent, |
|
531 | 531 | hangindent=subindent) + '\n' |
|
532 | 532 | |
|
533 | 533 | def formathtml(blocks): |
|
534 | 534 | """Format RST blocks as HTML""" |
|
535 | 535 | |
|
536 | 536 | out = [] |
|
537 | 537 | headernest = '' |
|
538 | 538 | listnest = [] |
|
539 | 539 | |
|
540 | 540 | def escape(s): |
|
541 | 541 | return cgi.escape(s, True) |
|
542 | 542 | |
|
543 | 543 | def openlist(start, level): |
|
544 | 544 | if not listnest or listnest[-1][0] != start: |
|
545 | 545 | listnest.append((start, level)) |
|
546 | 546 | out.append('<%s>\n' % start) |
|
547 | 547 | |
|
548 | 548 | blocks = [b for b in blocks if b['type'] != 'margin'] |
|
549 | 549 | |
|
550 | 550 | for pos, b in enumerate(blocks): |
|
551 | 551 | btype = b['type'] |
|
552 | 552 | level = b['indent'] |
|
553 | 553 | lines = b['lines'] |
|
554 | 554 | |
|
555 | 555 | if btype == 'admonition': |
|
556 | 556 | admonition = escape(_admonitiontitles[b['admonitiontitle']]) |
|
557 | 557 | text = escape(' '.join(map(str.strip, lines))) |
|
558 | 558 | out.append('<p>\n<b>%s</b> %s\n</p>\n' % (admonition, text)) |
|
559 | 559 | elif btype == 'paragraph': |
|
560 | 560 | out.append('<p>\n%s\n</p>\n' % escape('\n'.join(lines))) |
|
561 | 561 | elif btype == 'margin': |
|
562 | 562 | pass |
|
563 | 563 | elif btype == 'literal': |
|
564 | 564 | out.append('<pre>\n%s\n</pre>\n' % escape('\n'.join(lines))) |
|
565 | 565 | elif btype == 'section': |
|
566 | 566 | i = b['underline'] |
|
567 | 567 | if i not in headernest: |
|
568 | 568 | headernest += i |
|
569 | 569 | level = headernest.index(i) + 1 |
|
570 | 570 | out.append('<h%d>%s</h%d>\n' % (level, escape(lines[0]), level)) |
|
571 | 571 | elif btype == 'table': |
|
572 | 572 | table = b['table'] |
|
573 | 573 | out.append('<table>\n') |
|
574 | 574 | for row in table: |
|
575 | 575 | out.append('<tr>') |
|
576 | 576 | for v in row: |
|
577 | 577 | out.append('<td>') |
|
578 | 578 | out.append(escape(v)) |
|
579 | 579 | out.append('</td>') |
|
580 | 580 | out.append('\n') |
|
581 | 581 | out.pop() |
|
582 | 582 | out.append('</tr>\n') |
|
583 | 583 | out.append('</table>\n') |
|
584 | 584 | elif btype == 'definition': |
|
585 | 585 | openlist('dl', level) |
|
586 | 586 | term = escape(lines[0]) |
|
587 | 587 | text = escape(' '.join(map(str.strip, lines[1:]))) |
|
588 | 588 | out.append(' <dt>%s\n <dd>%s\n' % (term, text)) |
|
589 | 589 | elif btype == 'bullet': |
|
590 | 590 | bullet, head = lines[0].split(' ', 1) |
|
591 | 591 | if bullet == '-': |
|
592 | 592 | openlist('ul', level) |
|
593 | 593 | else: |
|
594 | 594 | openlist('ol', level) |
|
595 | 595 | out.append(' <li> %s\n' % escape(' '.join([head] + lines[1:]))) |
|
596 | 596 | elif btype == 'field': |
|
597 | 597 | openlist('dl', level) |
|
598 | 598 | key = escape(b['key']) |
|
599 | 599 | text = escape(' '.join(map(str.strip, lines))) |
|
600 | 600 | out.append(' <dt>%s\n <dd>%s\n' % (key, text)) |
|
601 | 601 | elif btype == 'option': |
|
602 | 602 | openlist('dl', level) |
|
603 | 603 | opt = escape(b['optstr']) |
|
604 | 604 | desc = escape(' '.join(map(str.strip, lines))) |
|
605 | 605 | out.append(' <dt>%s\n <dd>%s\n' % (opt, desc)) |
|
606 | 606 | |
|
607 | 607 | # close lists if indent level of next block is lower |
|
608 | 608 | if listnest: |
|
609 | 609 | start, level = listnest[-1] |
|
610 | 610 | if pos == len(blocks) - 1: |
|
611 | 611 | out.append('</%s>\n' % start) |
|
612 | 612 | listnest.pop() |
|
613 | 613 | else: |
|
614 | 614 | nb = blocks[pos + 1] |
|
615 | 615 | ni = nb['indent'] |
|
616 | 616 | if (ni < level or |
|
617 | 617 | (ni == level and |
|
618 | 618 | nb['type'] not in 'definition bullet field option')): |
|
619 | 619 | out.append('</%s>\n' % start) |
|
620 | 620 | listnest.pop() |
|
621 | 621 | |
|
622 | 622 | return ''.join(out) |
|
623 | 623 | |
|
624 | 624 | def parse(text, indent=0, keep=None): |
|
625 | 625 | """Parse text into a list of blocks""" |
|
626 | 626 | pruned = [] |
|
627 | 627 | blocks = findblocks(text) |
|
628 | 628 | for b in blocks: |
|
629 | 629 | b['indent'] += indent |
|
630 | 630 | blocks = findliteralblocks(blocks) |
|
631 | 631 | blocks = findtables(blocks) |
|
632 | 632 | blocks, pruned = prunecontainers(blocks, keep or []) |
|
633 | 633 | blocks = findsections(blocks) |
|
634 | 634 | blocks = inlineliterals(blocks) |
|
635 | 635 | blocks = hgrole(blocks) |
|
636 | 636 | blocks = splitparagraphs(blocks) |
|
637 | 637 | blocks = updatefieldlists(blocks) |
|
638 | 638 | blocks = updateoptionlists(blocks) |
|
639 | 639 | blocks = findadmonitions(blocks) |
|
640 | 640 | blocks = addmargins(blocks) |
|
641 | 641 | blocks = prunecomments(blocks) |
|
642 | 642 | return blocks, pruned |
|
643 | 643 | |
|
644 | 644 | def formatblocks(blocks, width): |
|
645 | 645 | text = ''.join(formatblock(b, width) for b in blocks) |
|
646 | 646 | return text |
|
647 | 647 | |
|
648 | 648 | def format(text, width=80, indent=0, keep=None, style='plain'): |
|
649 | 649 | """Parse and format the text according to width.""" |
|
650 | 650 | blocks, pruned = parse(text, indent, keep or []) |
|
651 | 651 | if style == 'html': |
|
652 | 652 | text = formathtml(blocks) |
|
653 | 653 | else: |
|
654 | 654 | text = ''.join(formatblock(b, width) for b in blocks) |
|
655 | 655 | if keep is None: |
|
656 | 656 | return text |
|
657 | 657 | else: |
|
658 | 658 | return text, pruned |
|
659 | 659 | |
|
660 | 660 | def getsections(blocks): |
|
661 | 661 | '''return a list of (section name, nesting level, blocks) tuples''' |
|
662 | 662 | nest = "" |
|
663 | 663 | level = 0 |
|
664 | 664 | secs = [] |
|
665 | 665 | for b in blocks: |
|
666 | 666 | if b['type'] == 'section': |
|
667 | 667 | i = b['underline'] |
|
668 | 668 | if i not in nest: |
|
669 | 669 | nest += i |
|
670 | 670 | level = nest.index(i) + 1 |
|
671 | 671 | nest = nest[:level] |
|
672 | 672 | secs.append((b['lines'][0], level, [b])) |
|
673 | 673 | else: |
|
674 | 674 | if not secs: |
|
675 | 675 | # add an initial empty section |
|
676 | 676 | secs = [('', 0, [])] |
|
677 | 677 | secs[-1][2].append(b) |
|
678 | 678 | return secs |
|
679 | 679 | |
|
680 | 680 | def decorateblocks(blocks, width): |
|
681 | 681 | '''generate a list of (section name, line text) pairs for search''' |
|
682 | 682 | lines = [] |
|
683 | 683 | for s in getsections(blocks): |
|
684 | 684 | section = s[0] |
|
685 | 685 | text = formatblocks(s[2], width) |
|
686 | 686 | lines.append([(section, l) for l in text.splitlines(True)]) |
|
687 | 687 | return lines |
|
688 | 688 | |
|
689 | 689 | def maketable(data, indent=0, header=False): |
|
690 | 690 | '''Generate an RST table for the given table data as a list of lines''' |
|
691 | 691 | |
|
692 | 692 | widths = [max(encoding.colwidth(e) for e in c) for c in zip(*data)] |
|
693 | 693 | indent = ' ' * indent |
|
694 | 694 | div = indent + ' '.join('=' * w for w in widths) + '\n' |
|
695 | 695 | |
|
696 | 696 | out = [div] |
|
697 | 697 | for row in data: |
|
698 | 698 | l = [] |
|
699 | 699 | for w, v in zip(widths, row): |
|
700 | 700 | if '\n' in v: |
|
701 | 701 | # only remove line breaks and indentation, long lines are |
|
702 | 702 | # handled by the next tool |
|
703 | 703 | v = ' '.join(e.lstrip() for e in v.split('\n')) |
|
704 | 704 | pad = ' ' * (w - encoding.colwidth(v)) |
|
705 | 705 | l.append(v + pad) |
|
706 | 706 | out.append(indent + ' '.join(l) + "\n") |
|
707 | 707 | if header and len(data) > 1: |
|
708 | 708 | out.insert(2, div) |
|
709 | 709 | out.append(div) |
|
710 | 710 | return out |
General Comments 0
You need to be logged in to leave comments.
Login now