Show More
@@ -1,261 +1,263 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 always(_match): |
|
38 | class always(_match): | |
39 | def __init__(self, root, cwd): |
|
39 | def __init__(self, root, cwd): | |
40 | _match.__init__(self, root, cwd, [], lambda f: True, False) |
|
40 | _match.__init__(self, root, cwd, [], lambda f: True, False) | |
41 |
|
41 | |||
42 | class never(_match): |
|
42 | class never(_match): | |
43 | def __init__(self, root, cwd): |
|
43 | def __init__(self, root, cwd): | |
44 | _match.__init__(self, root, cwd, [], lambda f: False, False) |
|
44 | _match.__init__(self, root, cwd, [], lambda f: False, False) | |
45 |
|
45 | |||
46 | class exact(_match): |
|
46 | class exact(_match): | |
47 | def __init__(self, root, cwd, files): |
|
47 | def __init__(self, root, cwd, files): | |
48 | _match.__init__(self, root, cwd, files, self.exact, False) |
|
48 | _match.__init__(self, root, cwd, files, self.exact, False) | |
49 |
|
49 | |||
50 | class match(_match): |
|
50 | class match(_match): | |
51 | def __init__(self, root, cwd, patterns, include=[], exclude=[], |
|
51 | def __init__(self, root, cwd, patterns, include=[], exclude=[], | |
52 | default='glob'): |
|
52 | default='glob'): | |
53 | f, mf, ap = _matcher(root, cwd, patterns, include, exclude, default) |
|
53 | f, mf, ap = _matcher(root, cwd, patterns, include, exclude, default) | |
54 | _match.__init__(self, root, cwd, f, mf, ap) |
|
54 | _match.__init__(self, root, cwd, f, mf, ap) | |
55 |
|
55 | |||
56 | def patkind(pat): |
|
56 | def patkind(pat): | |
57 | return _patsplit(pat, None)[0] |
|
57 | return _patsplit(pat, None)[0] | |
58 |
|
58 | |||
59 | def _patsplit(pat, default): |
|
59 | def _patsplit(pat, default): | |
60 | """Split a string into an optional pattern kind prefix and the |
|
60 | """Split a string into an optional pattern kind prefix and the | |
61 | actual pattern.""" |
|
61 | actual pattern.""" | |
62 | for prefix in 're', 'glob', 'path', 'relglob', 'relpath', 'relre': |
|
62 | if ':' in pat: | |
63 |
|
|
63 | pat, val = pat.split(':', 1) | |
|
64 | if pat in ('re', 'glob', 'path', 'relglob', 'relpath', 'relre'): | |||
|
65 | return pat, val | |||
64 | return default, pat |
|
66 | return default, pat | |
65 |
|
67 | |||
66 | def _globre(pat, head, tail): |
|
68 | def _globre(pat, head, tail): | |
67 | "convert a glob pattern into a regexp" |
|
69 | "convert a glob pattern into a regexp" | |
68 | i, n = 0, len(pat) |
|
70 | i, n = 0, len(pat) | |
69 | res = '' |
|
71 | res = '' | |
70 | group = 0 |
|
72 | group = 0 | |
71 | def peek(): return i < n and pat[i] |
|
73 | def peek(): return i < n and pat[i] | |
72 | while i < n: |
|
74 | while i < n: | |
73 | c = pat[i] |
|
75 | c = pat[i] | |
74 | i = i+1 |
|
76 | i = i+1 | |
75 | if c == '*': |
|
77 | if c == '*': | |
76 | if peek() == '*': |
|
78 | if peek() == '*': | |
77 | i += 1 |
|
79 | i += 1 | |
78 | res += '.*' |
|
80 | res += '.*' | |
79 | else: |
|
81 | else: | |
80 | res += '[^/]*' |
|
82 | res += '[^/]*' | |
81 | elif c == '?': |
|
83 | elif c == '?': | |
82 | res += '.' |
|
84 | res += '.' | |
83 | elif c == '[': |
|
85 | elif c == '[': | |
84 | j = i |
|
86 | j = i | |
85 | if j < n and pat[j] in '!]': |
|
87 | if j < n and pat[j] in '!]': | |
86 | j += 1 |
|
88 | j += 1 | |
87 | while j < n and pat[j] != ']': |
|
89 | while j < n and pat[j] != ']': | |
88 | j += 1 |
|
90 | j += 1 | |
89 | if j >= n: |
|
91 | if j >= n: | |
90 | res += '\\[' |
|
92 | res += '\\[' | |
91 | else: |
|
93 | else: | |
92 | stuff = pat[i:j].replace('\\','\\\\') |
|
94 | stuff = pat[i:j].replace('\\','\\\\') | |
93 | i = j + 1 |
|
95 | i = j + 1 | |
94 | if stuff[0] == '!': |
|
96 | if stuff[0] == '!': | |
95 | stuff = '^' + stuff[1:] |
|
97 | stuff = '^' + stuff[1:] | |
96 | elif stuff[0] == '^': |
|
98 | elif stuff[0] == '^': | |
97 | stuff = '\\' + stuff |
|
99 | stuff = '\\' + stuff | |
98 | res = '%s[%s]' % (res, stuff) |
|
100 | res = '%s[%s]' % (res, stuff) | |
99 | elif c == '{': |
|
101 | elif c == '{': | |
100 | group += 1 |
|
102 | group += 1 | |
101 | res += '(?:' |
|
103 | res += '(?:' | |
102 | elif c == '}' and group: |
|
104 | elif c == '}' and group: | |
103 | res += ')' |
|
105 | res += ')' | |
104 | group -= 1 |
|
106 | group -= 1 | |
105 | elif c == ',' and group: |
|
107 | elif c == ',' and group: | |
106 | res += '|' |
|
108 | res += '|' | |
107 | elif c == '\\': |
|
109 | elif c == '\\': | |
108 | p = peek() |
|
110 | p = peek() | |
109 | if p: |
|
111 | if p: | |
110 | i += 1 |
|
112 | i += 1 | |
111 | res += re.escape(p) |
|
113 | res += re.escape(p) | |
112 | else: |
|
114 | else: | |
113 | res += re.escape(c) |
|
115 | res += re.escape(c) | |
114 | else: |
|
116 | else: | |
115 | res += re.escape(c) |
|
117 | res += re.escape(c) | |
116 | return head + res + tail |
|
118 | return head + res + tail | |
117 |
|
119 | |||
118 | def _regex(kind, name, tail): |
|
120 | def _regex(kind, name, tail): | |
119 | '''convert a pattern into a regular expression''' |
|
121 | '''convert a pattern into a regular expression''' | |
120 | if not name: |
|
122 | if not name: | |
121 | return '' |
|
123 | return '' | |
122 | if kind == 're': |
|
124 | if kind == 're': | |
123 | return name |
|
125 | return name | |
124 | elif kind == 'path': |
|
126 | elif kind == 'path': | |
125 | return '^' + re.escape(name) + '(?:/|$)' |
|
127 | return '^' + re.escape(name) + '(?:/|$)' | |
126 | elif kind == 'relglob': |
|
128 | elif kind == 'relglob': | |
127 | return _globre(name, '(?:|.*/)', tail) |
|
129 | return _globre(name, '(?:|.*/)', tail) | |
128 | elif kind == 'relpath': |
|
130 | elif kind == 'relpath': | |
129 | return re.escape(name) + '(?:/|$)' |
|
131 | return re.escape(name) + '(?:/|$)' | |
130 | elif kind == 'relre': |
|
132 | elif kind == 'relre': | |
131 | if name.startswith('^'): |
|
133 | if name.startswith('^'): | |
132 | return name |
|
134 | return name | |
133 | return '.*' + name |
|
135 | return '.*' + name | |
134 | return _globre(name, '', tail) |
|
136 | return _globre(name, '', tail) | |
135 |
|
137 | |||
136 | def _matchfn(pats, tail): |
|
138 | def _matchfn(pats, tail): | |
137 | """build a matching function from a set of patterns""" |
|
139 | """build a matching function from a set of patterns""" | |
138 | try: |
|
140 | try: | |
139 | pat = '(?:%s)' % '|'.join([_regex(k, p, tail) for (k, p) in pats]) |
|
141 | pat = '(?:%s)' % '|'.join([_regex(k, p, tail) for (k, p) in pats]) | |
140 | if len(pat) > 20000: |
|
142 | if len(pat) > 20000: | |
141 | raise OverflowError() |
|
143 | raise OverflowError() | |
142 | return re.compile(pat).match |
|
144 | return re.compile(pat).match | |
143 | except OverflowError: |
|
145 | except OverflowError: | |
144 | # We're using a Python with a tiny regex engine and we |
|
146 | # We're using a Python with a tiny regex engine and we | |
145 | # made it explode, so we'll divide the pattern list in two |
|
147 | # made it explode, so we'll divide the pattern list in two | |
146 | # until it works |
|
148 | # until it works | |
147 | l = len(pats) |
|
149 | l = len(pats) | |
148 | if l < 2: |
|
150 | if l < 2: | |
149 | raise |
|
151 | raise | |
150 | a, b = _matchfn(pats[:l//2], tail), matchfn(pats[l//2:], tail) |
|
152 | a, b = _matchfn(pats[:l//2], tail), matchfn(pats[l//2:], tail) | |
151 | return lambda s: a(s) or b(s) |
|
153 | return lambda s: a(s) or b(s) | |
152 | except re.error: |
|
154 | except re.error: | |
153 | for k, p in pats: |
|
155 | for k, p in pats: | |
154 | try: |
|
156 | try: | |
155 | re.compile('(?:%s)' % _regex(k, p, tail)) |
|
157 | re.compile('(?:%s)' % _regex(k, p, tail)) | |
156 | except re.error: |
|
158 | except re.error: | |
157 | raise util.Abort("invalid pattern (%s): %s" % (k, p)) |
|
159 | raise util.Abort("invalid pattern (%s): %s" % (k, p)) | |
158 | raise util.Abort("invalid pattern") |
|
160 | raise util.Abort("invalid pattern") | |
159 |
|
161 | |||
160 | def _globprefix(pat): |
|
162 | def _globprefix(pat): | |
161 | '''return the non-glob prefix of a path, e.g. foo/* -> foo''' |
|
163 | '''return the non-glob prefix of a path, e.g. foo/* -> foo''' | |
162 | root = [] |
|
164 | root = [] | |
163 | for p in pat.split('/'): |
|
165 | for p in pat.split('/'): | |
164 | if '[' in p or '{' in p or '*' in p or '?' in p: |
|
166 | if '[' in p or '{' in p or '*' in p or '?' in p: | |
165 | break |
|
167 | break | |
166 | root.append(p) |
|
168 | root.append(p) | |
167 | return '/'.join(root) or '.' |
|
169 | return '/'.join(root) or '.' | |
168 |
|
170 | |||
169 | def _normalize(names, default, root, cwd): |
|
171 | def _normalize(names, default, root, cwd): | |
170 | pats = [] |
|
172 | pats = [] | |
171 | for kind, name in [_patsplit(p, default) for p in names]: |
|
173 | for kind, name in [_patsplit(p, default) for p in names]: | |
172 | if kind in ('glob', 'relpath'): |
|
174 | if kind in ('glob', 'relpath'): | |
173 | name = util.canonpath(root, cwd, name) |
|
175 | name = util.canonpath(root, cwd, name) | |
174 | elif kind in ('relglob', 'path'): |
|
176 | elif kind in ('relglob', 'path'): | |
175 | name = util.normpath(name) |
|
177 | name = util.normpath(name) | |
176 |
|
178 | |||
177 | pats.append((kind, name)) |
|
179 | pats.append((kind, name)) | |
178 | return pats |
|
180 | return pats | |
179 |
|
181 | |||
180 | def _roots(patterns): |
|
182 | def _roots(patterns): | |
181 | r = [] |
|
183 | r = [] | |
182 | for kind, name in patterns: |
|
184 | for kind, name in patterns: | |
183 | if kind == 'glob': |
|
185 | if kind == 'glob': | |
184 | r.append(_globprefix(name)) |
|
186 | r.append(_globprefix(name)) | |
185 | elif kind in ('relpath', 'path'): |
|
187 | elif kind in ('relpath', 'path'): | |
186 | r.append(name or '.') |
|
188 | r.append(name or '.') | |
187 | elif kind == 'relglob': |
|
189 | elif kind == 'relglob': | |
188 | r.append('.') |
|
190 | r.append('.') | |
189 | return r |
|
191 | return r | |
190 |
|
192 | |||
191 | def _anypats(patterns): |
|
193 | def _anypats(patterns): | |
192 | for kind, name in patterns: |
|
194 | for kind, name in patterns: | |
193 | if kind in ('glob', 're', 'relglob', 'relre'): |
|
195 | if kind in ('glob', 're', 'relglob', 'relre'): | |
194 | return True |
|
196 | return True | |
195 |
|
197 | |||
196 | def _matcher(root, cwd='', names=[], inc=[], exc=[], dflt_pat='glob'): |
|
198 | def _matcher(root, cwd='', names=[], inc=[], exc=[], dflt_pat='glob'): | |
197 | """build a function to match a set of file patterns |
|
199 | """build a function to match a set of file patterns | |
198 |
|
200 | |||
199 | arguments: |
|
201 | arguments: | |
200 | root - the canonical root of the tree you're matching against |
|
202 | root - the canonical root of the tree you're matching against | |
201 | cwd - the current working directory, if relevant |
|
203 | cwd - the current working directory, if relevant | |
202 | names - patterns to find |
|
204 | names - patterns to find | |
203 | inc - patterns to include |
|
205 | inc - patterns to include | |
204 | exc - patterns to exclude |
|
206 | exc - patterns to exclude | |
205 | dflt_pat - if a pattern in names has no explicit type, assume this one |
|
207 | dflt_pat - if a pattern in names has no explicit type, assume this one | |
206 |
|
208 | |||
207 | a pattern is one of: |
|
209 | a pattern is one of: | |
208 | 'glob:<glob>' - a glob relative to cwd |
|
210 | 'glob:<glob>' - a glob relative to cwd | |
209 | 're:<regexp>' - a regular expression |
|
211 | 're:<regexp>' - a regular expression | |
210 | 'path:<path>' - a path relative to canonroot |
|
212 | 'path:<path>' - a path relative to canonroot | |
211 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) |
|
213 | 'relglob:<glob>' - an unrooted glob (*.c matches C files in all dirs) | |
212 | 'relpath:<path>' - a path relative to cwd |
|
214 | 'relpath:<path>' - a path relative to cwd | |
213 | 'relre:<regexp>' - a regexp that doesn't have to match the start of a name |
|
215 | 'relre:<regexp>' - a regexp that doesn't have to match the start of a name | |
214 | '<something>' - one of the cases above, selected by the dflt_pat argument |
|
216 | '<something>' - one of the cases above, selected by the dflt_pat argument | |
215 |
|
217 | |||
216 | returns: |
|
218 | returns: | |
217 | a 3-tuple containing |
|
219 | a 3-tuple containing | |
218 | - list of roots (places where one should start a recursive walk of the fs); |
|
220 | - list of roots (places where one should start a recursive walk of the fs); | |
219 | this often matches the explicit non-pattern names passed in, but also |
|
221 | this often matches the explicit non-pattern names passed in, but also | |
220 | includes the initial part of glob: patterns that has no glob characters |
|
222 | includes the initial part of glob: patterns that has no glob characters | |
221 | - a bool match(filename) function |
|
223 | - a bool match(filename) function | |
222 | - a bool indicating if any patterns were passed in |
|
224 | - a bool indicating if any patterns were passed in | |
223 | """ |
|
225 | """ | |
224 |
|
226 | |||
225 | roots = [] |
|
227 | roots = [] | |
226 | anypats = bool(inc or exc) |
|
228 | anypats = bool(inc or exc) | |
227 |
|
229 | |||
228 | if names: |
|
230 | if names: | |
229 | pats = _normalize(names, dflt_pat, root, cwd) |
|
231 | pats = _normalize(names, dflt_pat, root, cwd) | |
230 | roots = _roots(pats) |
|
232 | roots = _roots(pats) | |
231 | anypats = anypats or _anypats(pats) |
|
233 | anypats = anypats or _anypats(pats) | |
232 | patmatch = _matchfn(pats, '$') |
|
234 | patmatch = _matchfn(pats, '$') | |
233 | if inc: |
|
235 | if inc: | |
234 | incmatch = _matchfn(_normalize(inc, 'glob', root, cwd), '(?:/|$)') |
|
236 | incmatch = _matchfn(_normalize(inc, 'glob', root, cwd), '(?:/|$)') | |
235 | if exc: |
|
237 | if exc: | |
236 | excmatch = _matchfn(_normalize(exc, 'glob', root, cwd), '(?:/|$)') |
|
238 | excmatch = _matchfn(_normalize(exc, 'glob', root, cwd), '(?:/|$)') | |
237 |
|
239 | |||
238 | if names: |
|
240 | if names: | |
239 | if inc: |
|
241 | if inc: | |
240 | if exc: |
|
242 | if exc: | |
241 | m = lambda f: incmatch(f) and not excmatch(f) and patmatch(f) |
|
243 | m = lambda f: incmatch(f) and not excmatch(f) and patmatch(f) | |
242 | else: |
|
244 | else: | |
243 | m = lambda f: incmatch(f) and patmatch(f) |
|
245 | m = lambda f: incmatch(f) and patmatch(f) | |
244 | else: |
|
246 | else: | |
245 | if exc: |
|
247 | if exc: | |
246 | m = lambda f: not excmatch(f) and patmatch(f) |
|
248 | m = lambda f: not excmatch(f) and patmatch(f) | |
247 | else: |
|
249 | else: | |
248 | m = patmatch |
|
250 | m = patmatch | |
249 | else: |
|
251 | else: | |
250 | if inc: |
|
252 | if inc: | |
251 | if exc: |
|
253 | if exc: | |
252 | m = lambda f: incmatch(f) and not excmatch(f) |
|
254 | m = lambda f: incmatch(f) and not excmatch(f) | |
253 | else: |
|
255 | else: | |
254 | m = incmatch |
|
256 | m = incmatch | |
255 | else: |
|
257 | else: | |
256 | if exc: |
|
258 | if exc: | |
257 | m = lambda f: not excmatch(f) |
|
259 | m = lambda f: not excmatch(f) | |
258 | else: |
|
260 | else: | |
259 | m = lambda f: True |
|
261 | m = lambda f: True | |
260 |
|
262 | |||
261 | return (roots, m, anypats) |
|
263 | return (roots, m, anypats) |
General Comments 0
You need to be logged in to leave comments.
Login now