Show More
@@ -1,280 +1,277 b'' | |||||
1 | # minirst.py - minimal reStructuredText parser |
|
1 | # minirst.py - minimal reStructuredText parser | |
2 | # |
|
2 | # | |
3 | # Copyright 2009 Matt Mackall <mpm@selenic.com> and others |
|
3 | # Copyright 2009 Matt Mackall <mpm@selenic.com> and others | |
4 | # |
|
4 | # | |
5 | # This software may be used and distributed according to the terms of the |
|
5 | # This software may be used and distributed according to the terms of the | |
6 | # GNU General Public License version 2, incorporated herein by reference. |
|
6 | # GNU General Public License version 2, incorporated herein by reference. | |
7 |
|
7 | |||
8 | """simplified reStructuredText parser. |
|
8 | """simplified reStructuredText parser. | |
9 |
|
9 | |||
10 | This parser knows just enough about reStructuredText to parse the |
|
10 | This parser knows just enough about reStructuredText to parse the | |
11 | Mercurial docstrings. |
|
11 | Mercurial docstrings. | |
12 |
|
12 | |||
13 | It cheats in a major way: nested blocks are not really nested. They |
|
13 | It cheats in a major way: nested blocks are not really nested. They | |
14 | are just indented blocks that look like they are nested. This relies |
|
14 | are just indented blocks that look like they are nested. This relies | |
15 | on the user to keep the right indentation for the blocks. |
|
15 | on the user to keep the right indentation for the blocks. | |
16 |
|
16 | |||
17 | It only supports a small subset of reStructuredText: |
|
17 | It only supports a small subset of reStructuredText: | |
18 |
|
18 | |||
19 | - sections |
|
19 | - sections | |
20 |
|
20 | |||
21 | - paragraphs |
|
21 | - paragraphs | |
22 |
|
22 | |||
23 | - literal blocks |
|
23 | - literal blocks | |
24 |
|
24 | |||
25 | - definition lists |
|
25 | - definition lists | |
26 |
|
26 | |||
27 | - bullet lists (items must start with '-') |
|
27 | - bullet lists (items must start with '-') | |
28 |
|
28 | |||
29 | - enumerated lists (no autonumbering) |
|
29 | - enumerated lists (no autonumbering) | |
30 |
|
30 | |||
31 | - field lists (colons cannot be escaped) |
|
31 | - field lists (colons cannot be escaped) | |
32 |
|
32 | |||
33 | - option lists (supports only long options without arguments) |
|
33 | - option lists (supports only long options without arguments) | |
34 |
|
34 | |||
35 | - inline literals (no other inline markup is not recognized) |
|
35 | - inline literals (no other inline markup is not recognized) | |
36 | """ |
|
36 | """ | |
37 |
|
37 | |||
38 | import re, sys, textwrap |
|
38 | import re, sys, textwrap | |
39 |
|
39 | |||
40 |
|
40 | |||
41 | def findblocks(text): |
|
41 | def findblocks(text): | |
42 | """Find continuous blocks of lines in text. |
|
42 | """Find continuous blocks of lines in text. | |
43 |
|
43 | |||
44 | Returns a list of dictionaries representing the blocks. Each block |
|
44 | Returns a list of dictionaries representing the blocks. Each block | |
45 | has an 'indent' field and a 'lines' field. |
|
45 | has an 'indent' field and a 'lines' field. | |
46 | """ |
|
46 | """ | |
47 | blocks = [[]] |
|
47 | blocks = [[]] | |
48 | lines = text.splitlines() |
|
48 | lines = text.splitlines() | |
49 | for line in lines: |
|
49 | for line in lines: | |
50 | if line.strip(): |
|
50 | if line.strip(): | |
51 | blocks[-1].append(line) |
|
51 | blocks[-1].append(line) | |
52 | elif blocks[-1]: |
|
52 | elif blocks[-1]: | |
53 | blocks.append([]) |
|
53 | blocks.append([]) | |
54 | if not blocks[-1]: |
|
54 | if not blocks[-1]: | |
55 | del blocks[-1] |
|
55 | del blocks[-1] | |
56 |
|
56 | |||
57 | for i, block in enumerate(blocks): |
|
57 | for i, block in enumerate(blocks): | |
58 | indent = min((len(l) - len(l.lstrip())) for l in block) |
|
58 | indent = min((len(l) - len(l.lstrip())) for l in block) | |
59 | blocks[i] = dict(indent=indent, lines=[l[indent:] for l in block]) |
|
59 | blocks[i] = dict(indent=indent, lines=[l[indent:] for l in block]) | |
60 | return blocks |
|
60 | return blocks | |
61 |
|
61 | |||
62 |
|
62 | |||
63 | def findliteralblocks(blocks): |
|
63 | def findliteralblocks(blocks): | |
64 | """Finds literal blocks and adds a 'type' field to the blocks. |
|
64 | """Finds literal blocks and adds a 'type' field to the blocks. | |
65 |
|
65 | |||
66 | Literal blocks are given the type 'literal', all other blocks are |
|
66 | Literal blocks are given the type 'literal', all other blocks are | |
67 | given type the 'paragraph'. |
|
67 | given type the 'paragraph'. | |
68 | """ |
|
68 | """ | |
69 | i = 0 |
|
69 | i = 0 | |
70 | while i < len(blocks): |
|
70 | while i < len(blocks): | |
71 | # Searching for a block that looks like this: |
|
71 | # Searching for a block that looks like this: | |
72 | # |
|
72 | # | |
73 | # +------------------------------+ |
|
73 | # +------------------------------+ | |
74 | # | paragraph | |
|
74 | # | paragraph | | |
75 | # | (ends with "::") | |
|
75 | # | (ends with "::") | | |
76 | # +------------------------------+ |
|
76 | # +------------------------------+ | |
77 | # +---------------------------+ |
|
77 | # +---------------------------+ | |
78 | # | indented literal block | |
|
78 | # | indented literal block | | |
79 | # +---------------------------+ |
|
79 | # +---------------------------+ | |
80 | blocks[i]['type'] = 'paragraph' |
|
80 | blocks[i]['type'] = 'paragraph' | |
81 | if blocks[i]['lines'][-1].endswith('::') and i+1 < len(blocks): |
|
81 | if blocks[i]['lines'][-1].endswith('::') and i+1 < len(blocks): | |
82 | indent = blocks[i]['indent'] |
|
82 | indent = blocks[i]['indent'] | |
83 | adjustment = blocks[i+1]['indent'] - indent |
|
83 | adjustment = blocks[i+1]['indent'] - indent | |
84 |
|
84 | |||
85 | if blocks[i]['lines'] == ['::']: |
|
85 | if blocks[i]['lines'] == ['::']: | |
86 | # Expanded form: remove block |
|
86 | # Expanded form: remove block | |
87 | del blocks[i] |
|
87 | del blocks[i] | |
88 | i -= 1 |
|
88 | i -= 1 | |
89 | elif blocks[i]['lines'][-1].endswith(' ::'): |
|
89 | elif blocks[i]['lines'][-1].endswith(' ::'): | |
90 | # Partially minimized form: remove space and both |
|
90 | # Partially minimized form: remove space and both | |
91 | # colons. |
|
91 | # colons. | |
92 | blocks[i]['lines'][-1] = blocks[i]['lines'][-1][:-3] |
|
92 | blocks[i]['lines'][-1] = blocks[i]['lines'][-1][:-3] | |
93 | else: |
|
93 | else: | |
94 | # Fully minimized form: remove just one colon. |
|
94 | # Fully minimized form: remove just one colon. | |
95 | blocks[i]['lines'][-1] = blocks[i]['lines'][-1][:-1] |
|
95 | blocks[i]['lines'][-1] = blocks[i]['lines'][-1][:-1] | |
96 |
|
96 | |||
97 | # List items are formatted with a hanging indent. We must |
|
97 | # List items are formatted with a hanging indent. We must | |
98 | # correct for this here while we still have the original |
|
98 | # correct for this here while we still have the original | |
99 | # information on the indentation of the subsequent literal |
|
99 | # information on the indentation of the subsequent literal | |
100 | # blocks available. |
|
100 | # blocks available. | |
101 | m = _bulletre.match(blocks[i]['lines'][0]) |
|
101 | m = _bulletre.match(blocks[i]['lines'][0]) | |
102 | if m: |
|
102 | if m: | |
103 | indent += m.end() |
|
103 | indent += m.end() | |
104 | adjustment -= m.end() |
|
104 | adjustment -= m.end() | |
105 |
|
105 | |||
106 | # Mark the following indented blocks. |
|
106 | # Mark the following indented blocks. | |
107 | while i+1 < len(blocks) and blocks[i+1]['indent'] > indent: |
|
107 | while i+1 < len(blocks) and blocks[i+1]['indent'] > indent: | |
108 | blocks[i+1]['type'] = 'literal' |
|
108 | blocks[i+1]['type'] = 'literal' | |
109 | blocks[i+1]['indent'] -= adjustment |
|
109 | blocks[i+1]['indent'] -= adjustment | |
110 | i += 1 |
|
110 | i += 1 | |
111 | i += 1 |
|
111 | i += 1 | |
112 | return blocks |
|
112 | return blocks | |
113 |
|
113 | |||
114 | _bulletre = re.compile(r'(-|[0-9A-Za-z]+\.|\(?[0-9A-Za-z]+\)) ') |
|
114 | _bulletre = re.compile(r'(-|[0-9A-Za-z]+\.|\(?[0-9A-Za-z]+\)) ') | |
115 | _optionre = re.compile(r'^(--[a-z-]+)((?:[ =][a-zA-Z][\w-]*)? +)(.*)$') |
|
115 | _optionre = re.compile(r'^(--[a-z-]+)((?:[ =][a-zA-Z][\w-]*)? +)(.*)$') | |
116 | _fieldre = re.compile(r':(?![: ])([^:]*)(?<! ):( +)(.*)') |
|
116 | _fieldre = re.compile(r':(?![: ])([^:]*)(?<! ):( +)(.*)') | |
117 | _definitionre = re.compile(r'[^ ]') |
|
117 | _definitionre = re.compile(r'[^ ]') | |
118 |
|
118 | |||
119 | def splitparagraphs(blocks): |
|
119 | def splitparagraphs(blocks): | |
120 | """Split paragraphs into lists.""" |
|
120 | """Split paragraphs into lists.""" | |
121 | # Tuples with (list type, item regexp, single line items?). Order |
|
121 | # Tuples with (list type, item regexp, single line items?). Order | |
122 | # matters: definition lists has the least specific regexp and must |
|
122 | # matters: definition lists has the least specific regexp and must | |
123 | # come last. |
|
123 | # come last. | |
124 | listtypes = [('bullet', _bulletre, True), |
|
124 | listtypes = [('bullet', _bulletre, True), | |
125 | ('option', _optionre, True), |
|
125 | ('option', _optionre, True), | |
126 | ('field', _fieldre, True), |
|
126 | ('field', _fieldre, True), | |
127 | ('definition', _definitionre, False)] |
|
127 | ('definition', _definitionre, False)] | |
128 |
|
128 | |||
129 | def match(lines, i, itemre, singleline): |
|
129 | def match(lines, i, itemre, singleline): | |
130 | """Does itemre match an item at line i? |
|
130 | """Does itemre match an item at line i? | |
131 |
|
131 | |||
132 | A list item can be followed by an idented line or another list |
|
132 | A list item can be followed by an idented line or another list | |
133 | item (but only if singleline is True). |
|
133 | item (but only if singleline is True). | |
134 | """ |
|
134 | """ | |
135 | line1 = lines[i] |
|
135 | line1 = lines[i] | |
136 | line2 = i+1 < len(lines) and lines[i+1] or '' |
|
136 | line2 = i+1 < len(lines) and lines[i+1] or '' | |
137 | if not itemre.match(line1): |
|
137 | if not itemre.match(line1): | |
138 | return False |
|
138 | return False | |
139 | if singleline: |
|
139 | if singleline: | |
140 | return line2 == '' or line2[0] == ' ' or itemre.match(line2) |
|
140 | return line2 == '' or line2[0] == ' ' or itemre.match(line2) | |
141 | else: |
|
141 | else: | |
142 | return line2.startswith(' ') |
|
142 | return line2.startswith(' ') | |
143 |
|
143 | |||
144 | i = 0 |
|
144 | i = 0 | |
145 | while i < len(blocks): |
|
145 | while i < len(blocks): | |
146 | if blocks[i]['type'] == 'paragraph': |
|
146 | if blocks[i]['type'] == 'paragraph': | |
147 | lines = blocks[i]['lines'] |
|
147 | lines = blocks[i]['lines'] | |
148 | for type, itemre, singleline in listtypes: |
|
148 | for type, itemre, singleline in listtypes: | |
149 | if match(lines, 0, itemre, singleline): |
|
149 | if match(lines, 0, itemre, singleline): | |
150 | items = [] |
|
150 | items = [] | |
151 | for j, line in enumerate(lines): |
|
151 | for j, line in enumerate(lines): | |
152 | if match(lines, j, itemre, singleline): |
|
152 | if match(lines, j, itemre, singleline): | |
153 | items.append(dict(type=type, lines=[], |
|
153 | items.append(dict(type=type, lines=[], | |
154 | indent=blocks[i]['indent'])) |
|
154 | indent=blocks[i]['indent'])) | |
155 | items[-1]['lines'].append(line) |
|
155 | items[-1]['lines'].append(line) | |
156 | blocks[i:i+1] = items |
|
156 | blocks[i:i+1] = items | |
157 | break |
|
157 | break | |
158 | i += 1 |
|
158 | i += 1 | |
159 | return blocks |
|
159 | return blocks | |
160 |
|
160 | |||
161 |
|
161 | |||
162 | def findsections(blocks): |
|
162 | def findsections(blocks): | |
163 | """Finds sections. |
|
163 | """Finds sections. | |
164 |
|
164 | |||
165 | The blocks must have a 'type' field, i.e., they should have been |
|
165 | The blocks must have a 'type' field, i.e., they should have been | |
166 | run through findliteralblocks first. |
|
166 | run through findliteralblocks first. | |
167 | """ |
|
167 | """ | |
168 | for block in blocks: |
|
168 | for block in blocks: | |
169 | # Searching for a block that looks like this: |
|
169 | # Searching for a block that looks like this: | |
170 | # |
|
170 | # | |
171 | # +------------------------------+ |
|
171 | # +------------------------------+ | |
172 | # | Section title | |
|
172 | # | Section title | | |
173 | # | ------------- | |
|
173 | # | ------------- | | |
174 | # +------------------------------+ |
|
174 | # +------------------------------+ | |
175 | if (block['type'] == 'paragraph' and |
|
175 | if (block['type'] == 'paragraph' and | |
176 | len(block['lines']) == 2 and |
|
176 | len(block['lines']) == 2 and | |
177 | block['lines'][1] == '-' * len(block['lines'][0])): |
|
177 | block['lines'][1] == '-' * len(block['lines'][0])): | |
178 | block['type'] = 'section' |
|
178 | block['type'] = 'section' | |
179 | return blocks |
|
179 | return blocks | |
180 |
|
180 | |||
181 |
|
181 | |||
182 | def inlineliterals(blocks): |
|
182 | def inlineliterals(blocks): | |
183 | for b in blocks: |
|
183 | for b in blocks: | |
184 | if b['type'] == 'paragraph': |
|
184 | if b['type'] == 'paragraph': | |
185 | b['lines'] = [l.replace('``', '"') for l in b['lines']] |
|
185 | b['lines'] = [l.replace('``', '"') for l in b['lines']] | |
186 | return blocks |
|
186 | return blocks | |
187 |
|
187 | |||
188 |
|
188 | |||
189 | def addmargins(blocks): |
|
189 | def addmargins(blocks): | |
190 | """Adds empty blocks for vertical spacing. |
|
190 | """Adds empty blocks for vertical spacing. | |
191 |
|
191 | |||
192 | This groups bullets, options, and definitions together with no vertical |
|
192 | This groups bullets, options, and definitions together with no vertical | |
193 | space between them, and adds an empty block between all other blocks. |
|
193 | space between them, and adds an empty block between all other blocks. | |
194 | """ |
|
194 | """ | |
195 | i = 1 |
|
195 | i = 1 | |
196 | while i < len(blocks): |
|
196 | while i < len(blocks): | |
197 | if (blocks[i]['type'] == blocks[i-1]['type'] and |
|
197 | if (blocks[i]['type'] == blocks[i-1]['type'] and | |
198 | blocks[i]['type'] in ('bullet', 'option', 'field', 'definition')): |
|
198 | blocks[i]['type'] in ('bullet', 'option', 'field', 'definition')): | |
199 | i += 1 |
|
199 | i += 1 | |
200 | else: |
|
200 | else: | |
201 | blocks.insert(i, dict(lines=[''], indent=0, type='margin')) |
|
201 | blocks.insert(i, dict(lines=[''], indent=0, type='margin')) | |
202 | i += 2 |
|
202 | i += 2 | |
203 | return blocks |
|
203 | return blocks | |
204 |
|
204 | |||
205 |
|
205 | |||
206 | def formatblock(block, width): |
|
206 | def formatblock(block, width): | |
207 | """Format a block according to width.""" |
|
207 | """Format a block according to width.""" | |
208 | if width <= 0: |
|
208 | if width <= 0: | |
209 | width = 78 |
|
209 | width = 78 | |
210 | indent = ' ' * block['indent'] |
|
210 | indent = ' ' * block['indent'] | |
211 | if block['type'] == 'margin': |
|
211 | if block['type'] == 'margin': | |
212 | return '' |
|
212 | return '' | |
213 | if block['type'] == 'literal': |
|
213 | if block['type'] == 'literal': | |
214 | indent += ' ' |
|
214 | indent += ' ' | |
215 | return indent + ('\n' + indent).join(block['lines']) |
|
215 | return indent + ('\n' + indent).join(block['lines']) | |
216 | if block['type'] == 'section': |
|
216 | if block['type'] == 'section': | |
217 | return indent + ('\n' + indent).join(block['lines']) |
|
217 | return indent + ('\n' + indent).join(block['lines']) | |
218 | if block['type'] == 'definition': |
|
218 | if block['type'] == 'definition': | |
219 | term = indent + block['lines'][0] |
|
219 | term = indent + block['lines'][0] | |
220 | hang = len(block['lines'][-1]) - len(block['lines'][-1].lstrip()) |
|
220 | hang = len(block['lines'][-1]) - len(block['lines'][-1].lstrip()) | |
221 | defindent = indent + hang * ' ' |
|
221 | defindent = indent + hang * ' ' | |
222 | text = ' '.join(map(str.strip, block['lines'][1:])) |
|
222 | text = ' '.join(map(str.strip, block['lines'][1:])) | |
223 | return "%s\n%s" % (term, textwrap.fill(text, width=width, |
|
223 | return "%s\n%s" % (term, textwrap.fill(text, width=width, | |
224 | initial_indent=defindent, |
|
224 | initial_indent=defindent, | |
225 | subsequent_indent=defindent)) |
|
225 | subsequent_indent=defindent)) | |
226 | initindent = subindent = indent |
|
226 | initindent = subindent = indent | |
227 | if block['type'] == 'bullet': |
|
227 | if block['type'] == 'bullet': | |
228 | m = _bulletre.match(block['lines'][0]) |
|
228 | m = _bulletre.match(block['lines'][0]) | |
229 | if m: |
|
229 | subindent = indent + m.end() * ' ' | |
230 | subindent = indent + m.end() * ' ' |
|
|||
231 | elif block['type'] == 'field': |
|
230 | elif block['type'] == 'field': | |
232 | m = _fieldre.match(block['lines'][0]) |
|
231 | m = _fieldre.match(block['lines'][0]) | |
233 | if m: |
|
232 | key, spaces, rest = m.groups() | |
234 | key, spaces, rest = m.groups() |
|
233 | # Turn ":foo: bar" into "foo bar". | |
235 | # Turn ":foo: bar" into "foo bar". |
|
234 | block['lines'][0] = '%s %s%s' % (key, spaces, rest) | |
236 | block['lines'][0] = '%s %s%s' % (key, spaces, rest) |
|
235 | subindent = indent + (2 + len(key) + len(spaces)) * ' ' | |
237 | subindent = indent + (2 + len(key) + len(spaces)) * ' ' |
|
|||
238 | elif block['type'] == 'option': |
|
236 | elif block['type'] == 'option': | |
239 | m = _optionre.match(block['lines'][0]) |
|
237 | m = _optionre.match(block['lines'][0]) | |
240 | if m: |
|
238 | option, arg, rest = m.groups() | |
241 | option, arg, rest = m.groups() |
|
239 | subindent = indent + (len(option) + len(arg)) * ' ' | |
242 | subindent = indent + (len(option) + len(arg)) * ' ' |
|
|||
243 |
|
240 | |||
244 | text = ' '.join(map(str.strip, block['lines'])) |
|
241 | text = ' '.join(map(str.strip, block['lines'])) | |
245 | return textwrap.fill(text, width=width, |
|
242 | return textwrap.fill(text, width=width, | |
246 | initial_indent=initindent, |
|
243 | initial_indent=initindent, | |
247 | subsequent_indent=subindent) |
|
244 | subsequent_indent=subindent) | |
248 |
|
245 | |||
249 |
|
246 | |||
250 | def format(text, width, indent=0): |
|
247 | def format(text, width, indent=0): | |
251 | """Parse and format the text according to width.""" |
|
248 | """Parse and format the text according to width.""" | |
252 | blocks = findblocks(text) |
|
249 | blocks = findblocks(text) | |
253 | for b in blocks: |
|
250 | for b in blocks: | |
254 | b['indent'] += indent |
|
251 | b['indent'] += indent | |
255 | blocks = findliteralblocks(blocks) |
|
252 | blocks = findliteralblocks(blocks) | |
256 | blocks = inlineliterals(blocks) |
|
253 | blocks = inlineliterals(blocks) | |
257 | blocks = splitparagraphs(blocks) |
|
254 | blocks = splitparagraphs(blocks) | |
258 | blocks = findsections(blocks) |
|
255 | blocks = findsections(blocks) | |
259 | blocks = addmargins(blocks) |
|
256 | blocks = addmargins(blocks) | |
260 | return '\n'.join(formatblock(b, width) for b in blocks) |
|
257 | return '\n'.join(formatblock(b, width) for b in blocks) | |
261 |
|
258 | |||
262 |
|
259 | |||
263 | if __name__ == "__main__": |
|
260 | if __name__ == "__main__": | |
264 | from pprint import pprint |
|
261 | from pprint import pprint | |
265 |
|
262 | |||
266 | def debug(func, blocks): |
|
263 | def debug(func, blocks): | |
267 | blocks = func(blocks) |
|
264 | blocks = func(blocks) | |
268 | print "*** after %s:" % func.__name__ |
|
265 | print "*** after %s:" % func.__name__ | |
269 | pprint(blocks) |
|
266 | pprint(blocks) | |
270 |
|
267 | |||
271 | return blocks |
|
268 | return blocks | |
272 |
|
269 | |||
273 | text = open(sys.argv[1]).read() |
|
270 | text = open(sys.argv[1]).read() | |
274 | blocks = debug(findblocks, text) |
|
271 | blocks = debug(findblocks, text) | |
275 | blocks = debug(findliteralblocks, blocks) |
|
272 | blocks = debug(findliteralblocks, blocks) | |
276 | blocks = debug(inlineliterals, blocks) |
|
273 | blocks = debug(inlineliterals, blocks) | |
277 | blocks = debug(splitparagraphs, blocks) |
|
274 | blocks = debug(splitparagraphs, blocks) | |
278 | blocks = debug(findsections, blocks) |
|
275 | blocks = debug(findsections, blocks) | |
279 | blocks = debug(addmargins, blocks) |
|
276 | blocks = debug(addmargins, blocks) | |
280 | print '\n'.join(formatblock(b, 30) for b in blocks) |
|
277 | print '\n'.join(formatblock(b, 30) for b in blocks) |
General Comments 0
You need to be logged in to leave comments.
Login now