Show More
@@ -1,250 +1,254 b'' | |||||
1 | # match.py - file name matching |
|
1 | # match.py - file name matching | |
2 | # |
|
2 | # | |
3 | # Copyright 2008, 2009 Matt Mackall <mpm@selenic.com> and others |
|
3 | # Copyright 2008, 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 | import util, re |
|
8 | import util, re | |
9 |
|
9 | |||
10 | class _match(object): |
|
10 | class _match(object): | |
11 | def __init__(self, root, cwd, files, mf, ap): |
|
11 | def __init__(self, root, cwd, files, mf, ap): | |
12 | self._root = root |
|
12 | self._root = root | |
13 | self._cwd = cwd |
|
13 | self._cwd = cwd | |
14 | self._files = files |
|
14 | self._files = files | |
15 | self._fmap = set(files) |
|
15 | self._fmap = set(files) | |
16 | self.matchfn = mf |
|
16 | self.matchfn = mf | |
17 | self._anypats = ap |
|
17 | self._anypats = ap | |
18 | def __call__(self, fn): |
|
18 | def __call__(self, fn): | |
19 | return self.matchfn(fn) |
|
19 | return self.matchfn(fn) | |
20 | def __iter__(self): |
|
20 | def __iter__(self): | |
21 | for f in self._files: |
|
21 | for f in self._files: | |
22 | yield f |
|
22 | yield f | |
23 | def bad(self, f, msg): |
|
23 | def bad(self, f, msg): | |
24 | return True |
|
24 | return True | |
25 | def dir(self, f): |
|
25 | def dir(self, f): | |
26 | pass |
|
26 | pass | |
27 | def missing(self, f): |
|
27 | def missing(self, f): | |
28 | pass |
|
28 | pass | |
29 | def exact(self, f): |
|
29 | def exact(self, f): | |
30 | return f in self._fmap |
|
30 | return f in self._fmap | |
31 | def rel(self, f): |
|
31 | def rel(self, f): | |
32 | return util.pathto(self._root, self._cwd, f) |
|
32 | return util.pathto(self._root, self._cwd, f) | |
33 | def files(self): |
|
33 | def files(self): | |
34 | return self._files |
|
34 | return self._files | |
35 | def anypats(self): |
|
35 | def anypats(self): | |
36 | return self._anypats |
|
36 | return self._anypats | |
37 |
|
37 | |||
38 | class match(_match): |
|
38 | class match(_match): | |
39 | def __init__(self, root, cwd, patterns, include=[], exclude=[], |
|
39 | def __init__(self, root, cwd, patterns, include=[], exclude=[], | |
40 | default='glob'): |
|
40 | default='glob', exact=False): | |
41 | """build an object to match a set of file patterns |
|
41 | """build an object to match a set of file patterns | |
42 |
|
42 | |||
43 | arguments: |
|
43 | arguments: | |
44 | root - the canonical root of the tree you're matching against |
|
44 | root - the canonical root of the tree you're matching against | |
45 | cwd - the current working directory, if relevant |
|
45 | cwd - the current working directory, if relevant | |
46 | patterns - patterns to find |
|
46 | patterns - patterns to find | |
47 | include - patterns to include |
|
47 | include - patterns to include | |
48 | exclude - patterns to exclude |
|
48 | exclude - patterns to exclude | |
49 | default - if a pattern in names has no explicit type, assume this one |
|
49 | default - if a pattern in names has no explicit type, assume this one | |
|
50 | exact - patterns are actually literals | |||
50 |
|
51 | |||
51 | a pattern is one of: |
|
52 | a pattern is one of: | |
52 | 'glob:<glob>' - a glob relative to cwd |
|
53 | 'glob:<glob>' - a glob relative to cwd | |
53 | 're:<regexp>' - a regular expression |
|
54 | 're:<regexp>' - a regular expression | |
54 | 'path:<path>' - a path relative to canonroot |
|
55 | 'path:<path>' - a path relative to canonroot | |
55 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) |
|
56 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) | |
56 | 'relpath:<path>' - a path relative to cwd |
|
57 | 'relpath:<path>' - a path relative to cwd | |
57 | 'relre:<regexp>' - a regexp that doesn't have to match the start of a name |
|
58 | 'relre:<regexp>' - a regexp that doesn't have to match the start of a name | |
58 | '<something>' - one of the cases above, selected by the dflt_pat argument |
|
59 | '<something>' - one of the cases above, selected by the dflt_pat argument | |
59 | """ |
|
60 | """ | |
60 |
|
61 | |||
61 | roots = [] |
|
62 | roots = [] | |
62 | anypats = bool(include or exclude) |
|
63 | anypats = bool(include or exclude) | |
63 |
|
64 | |||
64 |
if |
|
65 | if include: | |
|
66 | im = _buildmatch(_normalize(include, 'glob', root, cwd), '(?:/|$)') | |||
|
67 | if exclude: | |||
|
68 | em = _buildmatch(_normalize(exclude, 'glob', root, cwd), '(?:/|$)') | |||
|
69 | if exact: | |||
|
70 | roots = patterns | |||
|
71 | pm = self.exact | |||
|
72 | elif patterns: | |||
65 | pats = _normalize(patterns, default, root, cwd) |
|
73 | pats = _normalize(patterns, default, root, cwd) | |
66 | roots = _roots(pats) |
|
74 | roots = _roots(pats) | |
67 | anypats = anypats or _anypats(pats) |
|
75 | anypats = anypats or _anypats(pats) | |
68 | pm = _buildmatch(pats, '$') |
|
76 | pm = _buildmatch(pats, '$') | |
69 | if include: |
|
|||
70 | im = _buildmatch(_normalize(include, 'glob', root, cwd), '(?:/|$)') |
|
|||
71 | if exclude: |
|
|||
72 | em = _buildmatch(_normalize(exclude, 'glob', root, cwd), '(?:/|$)') |
|
|||
73 |
|
77 | |||
74 | if patterns: |
|
78 | if patterns or exact: | |
75 | if include: |
|
79 | if include: | |
76 | if exclude: |
|
80 | if exclude: | |
77 | m = lambda f: im(f) and not em(f) and pm(f) |
|
81 | m = lambda f: im(f) and not em(f) and pm(f) | |
78 | else: |
|
82 | else: | |
79 | m = lambda f: im(f) and pm(f) |
|
83 | m = lambda f: im(f) and pm(f) | |
80 | else: |
|
84 | else: | |
81 | if exclude: |
|
85 | if exclude: | |
82 | m = lambda f: not em(f) and pm(f) |
|
86 | m = lambda f: not em(f) and pm(f) | |
83 | else: |
|
87 | else: | |
84 | m = pm |
|
88 | m = pm | |
85 | else: |
|
89 | else: | |
86 | if include: |
|
90 | if include: | |
87 | if exclude: |
|
91 | if exclude: | |
88 | m = lambda f: im(f) and not em(f) |
|
92 | m = lambda f: im(f) and not em(f) | |
89 | else: |
|
93 | else: | |
90 | m = im |
|
94 | m = im | |
91 | else: |
|
95 | else: | |
92 | if exclude: |
|
96 | if exclude: | |
93 | m = lambda f: not em(f) |
|
97 | m = lambda f: not em(f) | |
94 | else: |
|
98 | else: | |
95 | m = lambda f: True |
|
99 | m = lambda f: True | |
96 |
|
100 | |||
97 | _match.__init__(self, root, cwd, roots, m, anypats) |
|
101 | _match.__init__(self, root, cwd, roots, m, anypats) | |
98 |
|
102 | |||
99 |
class exact( |
|
103 | class exact(match): | |
100 | def __init__(self, root, cwd, files): |
|
104 | def __init__(self, root, cwd, files): | |
101 |
|
|
105 | match.__init__(self, root, cwd, files, exact = True) | |
102 |
|
106 | |||
103 | class always(match): |
|
107 | class always(match): | |
104 | def __init__(self, root, cwd): |
|
108 | def __init__(self, root, cwd): | |
105 | match.__init__(self, root, cwd, []) |
|
109 | match.__init__(self, root, cwd, []) | |
106 |
|
110 | |||
107 |
class never( |
|
111 | class never(match): | |
108 | def __init__(self, root, cwd): |
|
112 | def __init__(self, root, cwd): | |
109 |
|
|
113 | match.__init__(self, root, cwd, [], exact = True) | |
110 |
|
114 | |||
111 | def patkind(pat): |
|
115 | def patkind(pat): | |
112 | return _patsplit(pat, None)[0] |
|
116 | return _patsplit(pat, None)[0] | |
113 |
|
117 | |||
114 | def _patsplit(pat, default): |
|
118 | def _patsplit(pat, default): | |
115 | """Split a string into an optional pattern kind prefix and the |
|
119 | """Split a string into an optional pattern kind prefix and the | |
116 | actual pattern.""" |
|
120 | actual pattern.""" | |
117 | if ':' in pat: |
|
121 | if ':' in pat: | |
118 | pat, val = pat.split(':', 1) |
|
122 | pat, val = pat.split(':', 1) | |
119 | if pat in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre'): |
|
123 | if pat in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre'): | |
120 | return pat, val |
|
124 | return pat, val | |
121 | return default, pat |
|
125 | return default, pat | |
122 |
|
126 | |||
123 | def _globre(pat): |
|
127 | def _globre(pat): | |
124 | "convert a glob pattern into a regexp" |
|
128 | "convert a glob pattern into a regexp" | |
125 | i, n = 0, len(pat) |
|
129 | i, n = 0, len(pat) | |
126 | res = '' |
|
130 | res = '' | |
127 | group = 0 |
|
131 | group = 0 | |
128 | escape = re.escape |
|
132 | escape = re.escape | |
129 | def peek(): return i < n and pat[i] |
|
133 | def peek(): return i < n and pat[i] | |
130 | while i < n: |
|
134 | while i < n: | |
131 | c = pat[i] |
|
135 | c = pat[i] | |
132 | i = i+1 |
|
136 | i = i+1 | |
133 | if c not in '*?[{},\\': |
|
137 | if c not in '*?[{},\\': | |
134 | res += escape(c) |
|
138 | res += escape(c) | |
135 | elif c == '*': |
|
139 | elif c == '*': | |
136 | if peek() == '*': |
|
140 | if peek() == '*': | |
137 | i += 1 |
|
141 | i += 1 | |
138 | res += '.*' |
|
142 | res += '.*' | |
139 | else: |
|
143 | else: | |
140 | res += '[^/]*' |
|
144 | res += '[^/]*' | |
141 | elif c == '?': |
|
145 | elif c == '?': | |
142 | res += '.' |
|
146 | res += '.' | |
143 | elif c == '[': |
|
147 | elif c == '[': | |
144 | j = i |
|
148 | j = i | |
145 | if j < n and pat[j] in '!]': |
|
149 | if j < n and pat[j] in '!]': | |
146 | j += 1 |
|
150 | j += 1 | |
147 | while j < n and pat[j] != ']': |
|
151 | while j < n and pat[j] != ']': | |
148 | j += 1 |
|
152 | j += 1 | |
149 | if j >= n: |
|
153 | if j >= n: | |
150 | res += '\\[' |
|
154 | res += '\\[' | |
151 | else: |
|
155 | else: | |
152 | stuff = pat[i:j].replace('\\','\\\\') |
|
156 | stuff = pat[i:j].replace('\\','\\\\') | |
153 | i = j + 1 |
|
157 | i = j + 1 | |
154 | if stuff[0] == '!': |
|
158 | if stuff[0] == '!': | |
155 | stuff = '^' + stuff[1:] |
|
159 | stuff = '^' + stuff[1:] | |
156 | elif stuff[0] == '^': |
|
160 | elif stuff[0] == '^': | |
157 | stuff = '\\' + stuff |
|
161 | stuff = '\\' + stuff | |
158 | res = '%s[%s]' % (res, stuff) |
|
162 | res = '%s[%s]' % (res, stuff) | |
159 | elif c == '{': |
|
163 | elif c == '{': | |
160 | group += 1 |
|
164 | group += 1 | |
161 | res += '(?:' |
|
165 | res += '(?:' | |
162 | elif c == '}' and group: |
|
166 | elif c == '}' and group: | |
163 | res += ')' |
|
167 | res += ')' | |
164 | group -= 1 |
|
168 | group -= 1 | |
165 | elif c == ',' and group: |
|
169 | elif c == ',' and group: | |
166 | res += '|' |
|
170 | res += '|' | |
167 | elif c == '\\': |
|
171 | elif c == '\\': | |
168 | p = peek() |
|
172 | p = peek() | |
169 | if p: |
|
173 | if p: | |
170 | i += 1 |
|
174 | i += 1 | |
171 | res += escape(p) |
|
175 | res += escape(p) | |
172 | else: |
|
176 | else: | |
173 | res += escape(c) |
|
177 | res += escape(c) | |
174 | else: |
|
178 | else: | |
175 | res += escape(c) |
|
179 | res += escape(c) | |
176 | return res |
|
180 | return res | |
177 |
|
181 | |||
178 | def _regex(kind, name, tail): |
|
182 | def _regex(kind, name, tail): | |
179 | '''convert a pattern into a regular expression''' |
|
183 | '''convert a pattern into a regular expression''' | |
180 | if not name: |
|
184 | if not name: | |
181 | return '' |
|
185 | return '' | |
182 | if kind == 're': |
|
186 | if kind == 're': | |
183 | return name |
|
187 | return name | |
184 | elif kind == 'path': |
|
188 | elif kind == 'path': | |
185 | return '^' + re.escape(name) + '(?:/|$)' |
|
189 | return '^' + re.escape(name) + '(?:/|$)' | |
186 | elif kind == 'relglob': |
|
190 | elif kind == 'relglob': | |
187 | return '(?:|.*/)' + _globre(name) + tail |
|
191 | return '(?:|.*/)' + _globre(name) + tail | |
188 | elif kind == 'relpath': |
|
192 | elif kind == 'relpath': | |
189 | return re.escape(name) + '(?:/|$)' |
|
193 | return re.escape(name) + '(?:/|$)' | |
190 | elif kind == 'relre': |
|
194 | elif kind == 'relre': | |
191 | if name.startswith('^'): |
|
195 | if name.startswith('^'): | |
192 | return name |
|
196 | return name | |
193 | return '.*' + name |
|
197 | return '.*' + name | |
194 | return _globre(name) + tail |
|
198 | return _globre(name) + tail | |
195 |
|
199 | |||
196 | def _buildmatch(pats, tail): |
|
200 | def _buildmatch(pats, tail): | |
197 | """build a matching function from a set of patterns""" |
|
201 | """build a matching function from a set of patterns""" | |
198 | try: |
|
202 | try: | |
199 | pat = '(?:%s)' % '|'.join([_regex(k, p, tail) for (k, p) in pats]) |
|
203 | pat = '(?:%s)' % '|'.join([_regex(k, p, tail) for (k, p) in pats]) | |
200 | if len(pat) > 20000: |
|
204 | if len(pat) > 20000: | |
201 | raise OverflowError() |
|
205 | raise OverflowError() | |
202 | return re.compile(pat).match |
|
206 | return re.compile(pat).match | |
203 | except OverflowError: |
|
207 | except OverflowError: | |
204 | # We're using a Python with a tiny regex engine and we |
|
208 | # We're using a Python with a tiny regex engine and we | |
205 | # made it explode, so we'll divide the pattern list in two |
|
209 | # made it explode, so we'll divide the pattern list in two | |
206 | # until it works |
|
210 | # until it works | |
207 | l = len(pats) |
|
211 | l = len(pats) | |
208 | if l < 2: |
|
212 | if l < 2: | |
209 | raise |
|
213 | raise | |
210 | a, b = _buildmatch(pats[:l//2], tail), _buildmatch(pats[l//2:], tail) |
|
214 | a, b = _buildmatch(pats[:l//2], tail), _buildmatch(pats[l//2:], tail) | |
211 | return lambda s: a(s) or b(s) |
|
215 | return lambda s: a(s) or b(s) | |
212 | except re.error: |
|
216 | except re.error: | |
213 | for k, p in pats: |
|
217 | for k, p in pats: | |
214 | try: |
|
218 | try: | |
215 | re.compile('(?:%s)' % _regex(k, p, tail)) |
|
219 | re.compile('(?:%s)' % _regex(k, p, tail)) | |
216 | except re.error: |
|
220 | except re.error: | |
217 | raise util.Abort("invalid pattern (%s): %s" % (k, p)) |
|
221 | raise util.Abort("invalid pattern (%s): %s" % (k, p)) | |
218 | raise util.Abort("invalid pattern") |
|
222 | raise util.Abort("invalid pattern") | |
219 |
|
223 | |||
220 | def _normalize(names, default, root, cwd): |
|
224 | def _normalize(names, default, root, cwd): | |
221 | pats = [] |
|
225 | pats = [] | |
222 | for kind, name in [_patsplit(p, default) for p in names]: |
|
226 | for kind, name in [_patsplit(p, default) for p in names]: | |
223 | if kind in ('glob', 'relpath'): |
|
227 | if kind in ('glob', 'relpath'): | |
224 | name = util.canonpath(root, cwd, name) |
|
228 | name = util.canonpath(root, cwd, name) | |
225 | elif kind in ('relglob', 'path'): |
|
229 | elif kind in ('relglob', 'path'): | |
226 | name = util.normpath(name) |
|
230 | name = util.normpath(name) | |
227 |
|
231 | |||
228 | pats.append((kind, name)) |
|
232 | pats.append((kind, name)) | |
229 | return pats |
|
233 | return pats | |
230 |
|
234 | |||
231 | def _roots(patterns): |
|
235 | def _roots(patterns): | |
232 | r = [] |
|
236 | r = [] | |
233 | for kind, name in patterns: |
|
237 | for kind, name in patterns: | |
234 | if kind == 'glob': # find the non-glob prefix |
|
238 | if kind == 'glob': # find the non-glob prefix | |
235 | root = [] |
|
239 | root = [] | |
236 | for p in name.split('/'): |
|
240 | for p in name.split('/'): | |
237 | if '[' in p or '{' in p or '*' in p or '?' in p: |
|
241 | if '[' in p or '{' in p or '*' in p or '?' in p: | |
238 | break |
|
242 | break | |
239 | root.append(p) |
|
243 | root.append(p) | |
240 | r.append('/'.join(root) or '.') |
|
244 | r.append('/'.join(root) or '.') | |
241 | elif kind in ('relpath', 'path'): |
|
245 | elif kind in ('relpath', 'path'): | |
242 | r.append(name or '.') |
|
246 | r.append(name or '.') | |
243 | elif kind == 'relglob': |
|
247 | elif kind == 'relglob': | |
244 | r.append('.') |
|
248 | r.append('.') | |
245 | return r |
|
249 | return r | |
246 |
|
250 | |||
247 | def _anypats(patterns): |
|
251 | def _anypats(patterns): | |
248 | for kind, name in patterns: |
|
252 | for kind, name in patterns: | |
249 | if kind in ('glob', 're', 'relglob', 'relre'): |
|
253 | if kind in ('glob', 're', 'relglob', 'relre'): | |
250 | return True |
|
254 | return True |
General Comments 0
You need to be logged in to leave comments.
Login now