Show More
@@ -1,221 +1,223 b'' | |||
|
1 | 1 | from __future__ import absolute_import |
|
2 | 2 | |
|
3 | 3 | import errno |
|
4 | 4 | import os |
|
5 | 5 | import posixpath |
|
6 | 6 | import stat |
|
7 | 7 | |
|
8 | 8 | from .i18n import _ |
|
9 | 9 | from . import ( |
|
10 | 10 | encoding, |
|
11 | 11 | error, |
|
12 | 12 | pycompat, |
|
13 | 13 | util, |
|
14 | 14 | ) |
|
15 | 15 | |
|
16 | 16 | def _lowerclean(s): |
|
17 | 17 | return encoding.hfsignoreclean(s.lower()) |
|
18 | 18 | |
|
19 | 19 | class pathauditor(object): |
|
20 | 20 | '''ensure that a filesystem path contains no banned components. |
|
21 | 21 | the following properties of a path are checked: |
|
22 | 22 | |
|
23 | 23 | - ends with a directory separator |
|
24 | 24 | - under top-level .hg |
|
25 | 25 | - starts at the root of a windows drive |
|
26 | 26 | - contains ".." |
|
27 | 27 | |
|
28 | 28 | More check are also done about the file system states: |
|
29 | 29 | - traverses a symlink (e.g. a/symlink_here/b) |
|
30 | 30 | - inside a nested repository (a callback can be used to approve |
|
31 | 31 | some nested repositories, e.g., subrepositories) |
|
32 | 32 | |
|
33 | 33 | The file system checks are only done when 'realfs' is set to True (the |
|
34 | 34 | default). They should be disable then we are auditing path for operation on |
|
35 | 35 | stored history. |
|
36 | 36 | |
|
37 | 37 | If 'cached' is set to True, audited paths and sub-directories are cached. |
|
38 | 38 | Be careful to not keep the cache of unmanaged directories for long because |
|
39 | 39 | audited paths may be replaced with symlinks. |
|
40 | 40 | ''' |
|
41 | 41 | |
|
42 | 42 | def __init__(self, root, callback=None, realfs=True, cached=False): |
|
43 | 43 | self.audited = set() |
|
44 | 44 | self.auditeddir = set() |
|
45 | 45 | self.root = root |
|
46 | 46 | self._realfs = realfs |
|
47 | 47 | self._cached = cached |
|
48 | 48 | self.callback = callback |
|
49 | 49 | if os.path.lexists(root) and not util.fscasesensitive(root): |
|
50 | 50 | self.normcase = util.normcase |
|
51 | 51 | else: |
|
52 | 52 | self.normcase = lambda x: x |
|
53 | 53 | |
|
54 | 54 | def __call__(self, path, mode=None): |
|
55 | 55 | '''Check the relative path. |
|
56 | 56 | path may contain a pattern (e.g. foodir/**.txt)''' |
|
57 | 57 | |
|
58 | 58 | path = util.localpath(path) |
|
59 | 59 | normpath = self.normcase(path) |
|
60 | 60 | if normpath in self.audited: |
|
61 | 61 | return |
|
62 | 62 | # AIX ignores "/" at end of path, others raise EISDIR. |
|
63 | 63 | if util.endswithsep(path): |
|
64 | 64 | raise error.Abort(_("path ends in directory separator: %s") % path) |
|
65 | 65 | parts = util.splitpath(path) |
|
66 | 66 | if (os.path.splitdrive(path)[0] |
|
67 | 67 | or _lowerclean(parts[0]) in ('.hg', '.hg.', '') |
|
68 | 68 | or os.pardir in parts): |
|
69 | 69 | raise error.Abort(_("path contains illegal component: %s") % path) |
|
70 | 70 | # Windows shortname aliases |
|
71 | 71 | for p in parts: |
|
72 | 72 | if "~" in p: |
|
73 | 73 | first, last = p.split("~", 1) |
|
74 | 74 | if last.isdigit() and first.upper() in ["HG", "HG8B6C"]: |
|
75 | 75 | raise error.Abort(_("path contains illegal component: %s") |
|
76 | 76 | % path) |
|
77 | 77 | if '.hg' in _lowerclean(path): |
|
78 | 78 | lparts = [_lowerclean(p.lower()) for p in parts] |
|
79 | 79 | for p in '.hg', '.hg.': |
|
80 | 80 | if p in lparts[1:]: |
|
81 | 81 | pos = lparts.index(p) |
|
82 | 82 | base = os.path.join(*parts[:pos]) |
|
83 | 83 | raise error.Abort(_("path '%s' is inside nested repo %r") |
|
84 | 84 | % (path, base)) |
|
85 | 85 | |
|
86 | 86 | normparts = util.splitpath(normpath) |
|
87 | 87 | assert len(parts) == len(normparts) |
|
88 | 88 | |
|
89 | 89 | parts.pop() |
|
90 | 90 | normparts.pop() |
|
91 | 91 | prefixes = [] |
|
92 | 92 | # It's important that we check the path parts starting from the root. |
|
93 | 93 | # This means we won't accidentally traverse a symlink into some other |
|
94 | 94 | # filesystem (which is potentially expensive to access). |
|
95 | 95 | for i in range(len(parts)): |
|
96 | 96 | prefix = pycompat.ossep.join(parts[:i + 1]) |
|
97 | 97 | normprefix = pycompat.ossep.join(normparts[:i + 1]) |
|
98 | 98 | if normprefix in self.auditeddir: |
|
99 | 99 | continue |
|
100 | 100 | if self._realfs: |
|
101 | 101 | self._checkfs(prefix, path) |
|
102 | 102 | prefixes.append(normprefix) |
|
103 | 103 | |
|
104 | 104 | if self._cached: |
|
105 | 105 | self.audited.add(normpath) |
|
106 | 106 | # only add prefixes to the cache after checking everything: we don't |
|
107 | 107 | # want to add "foo/bar/baz" before checking if there's a "foo/.hg" |
|
108 | 108 | self.auditeddir.update(prefixes) |
|
109 | 109 | |
|
110 | 110 | def _checkfs(self, prefix, path): |
|
111 | 111 | """raise exception if a file system backed check fails""" |
|
112 | 112 | curpath = os.path.join(self.root, prefix) |
|
113 | 113 | try: |
|
114 | 114 | st = os.lstat(curpath) |
|
115 | 115 | except OSError as err: |
|
116 | 116 | # EINVAL can be raised as invalid path syntax under win32. |
|
117 | 117 | # They must be ignored for patterns can be checked too. |
|
118 | 118 | if err.errno not in (errno.ENOENT, errno.ENOTDIR, errno.EINVAL): |
|
119 | 119 | raise |
|
120 | 120 | else: |
|
121 | 121 | if stat.S_ISLNK(st.st_mode): |
|
122 | 122 | msg = _('path %r traverses symbolic link %r') % (path, prefix) |
|
123 | 123 | raise error.Abort(msg) |
|
124 | 124 | elif (stat.S_ISDIR(st.st_mode) and |
|
125 | 125 | os.path.isdir(os.path.join(curpath, '.hg'))): |
|
126 | 126 | if not self.callback or not self.callback(curpath): |
|
127 | 127 | msg = _("path '%s' is inside nested repo %r") |
|
128 | 128 | raise error.Abort(msg % (path, prefix)) |
|
129 | 129 | |
|
130 | 130 | def check(self, path): |
|
131 | 131 | try: |
|
132 | 132 | self(path) |
|
133 | 133 | return True |
|
134 | 134 | except (OSError, error.Abort): |
|
135 | 135 | return False |
|
136 | 136 | |
|
137 | 137 | def canonpath(root, cwd, myname, auditor=None): |
|
138 | 138 | '''return the canonical path of myname, given cwd and root''' |
|
139 | 139 | if util.endswithsep(root): |
|
140 | 140 | rootsep = root |
|
141 | 141 | else: |
|
142 | 142 | rootsep = root + pycompat.ossep |
|
143 | 143 | name = myname |
|
144 | 144 | if not os.path.isabs(name): |
|
145 | 145 | name = os.path.join(root, cwd, name) |
|
146 | 146 | name = os.path.normpath(name) |
|
147 | 147 | if auditor is None: |
|
148 | 148 | auditor = pathauditor(root) |
|
149 | 149 | if name != rootsep and name.startswith(rootsep): |
|
150 | 150 | name = name[len(rootsep):] |
|
151 | 151 | auditor(name) |
|
152 | 152 | return util.pconvert(name) |
|
153 | 153 | elif name == root: |
|
154 | 154 | return '' |
|
155 | 155 | else: |
|
156 | 156 | # Determine whether `name' is in the hierarchy at or beneath `root', |
|
157 | 157 | # by iterating name=dirname(name) until that causes no change (can't |
|
158 | 158 | # check name == '/', because that doesn't work on windows). The list |
|
159 | 159 | # `rel' holds the reversed list of components making up the relative |
|
160 | 160 | # file name we want. |
|
161 | 161 | rel = [] |
|
162 | 162 | while True: |
|
163 | 163 | try: |
|
164 | 164 | s = util.samefile(name, root) |
|
165 | 165 | except OSError: |
|
166 | 166 | s = False |
|
167 | 167 | if s: |
|
168 | 168 | if not rel: |
|
169 | 169 | # name was actually the same as root (maybe a symlink) |
|
170 | 170 | return '' |
|
171 | 171 | rel.reverse() |
|
172 | 172 | name = os.path.join(*rel) |
|
173 | 173 | auditor(name) |
|
174 | 174 | return util.pconvert(name) |
|
175 | 175 | dirname, basename = util.split(name) |
|
176 | 176 | rel.append(basename) |
|
177 | 177 | if dirname == name: |
|
178 | 178 | break |
|
179 | 179 | name = dirname |
|
180 | 180 | |
|
181 | 181 | # A common mistake is to use -R, but specify a file relative to the repo |
|
182 | 182 | # instead of cwd. Detect that case, and provide a hint to the user. |
|
183 | 183 | hint = None |
|
184 | 184 | try: |
|
185 | 185 | if cwd != root: |
|
186 | 186 | canonpath(root, root, myname, auditor) |
|
187 | hint = (_("consider using '--cwd %s'") | |
|
188 | % os.path.relpath(root, cwd)) | |
|
187 | relpath = util.pathto(root, cwd, '') | |
|
188 | if relpath[-1] == pycompat.ossep: | |
|
189 | relpath = relpath[:-1] | |
|
190 | hint = (_("consider using '--cwd %s'") % relpath) | |
|
189 | 191 | except error.Abort: |
|
190 | 192 | pass |
|
191 | 193 | |
|
192 | 194 | raise error.Abort(_("%s not under root '%s'") % (myname, root), |
|
193 | 195 | hint=hint) |
|
194 | 196 | |
|
195 | 197 | def normasprefix(path): |
|
196 | 198 | '''normalize the specified path as path prefix |
|
197 | 199 | |
|
198 | 200 | Returned value can be used safely for "p.startswith(prefix)", |
|
199 | 201 | "p[len(prefix):]", and so on. |
|
200 | 202 | |
|
201 | 203 | For efficiency, this expects "path" argument to be already |
|
202 | 204 | normalized by "os.path.normpath", "os.path.realpath", and so on. |
|
203 | 205 | |
|
204 | 206 | See also issue3033 for detail about need of this function. |
|
205 | 207 | |
|
206 | 208 | >>> normasprefix(b'/foo/bar').replace(pycompat.ossep, b'/') |
|
207 | 209 | '/foo/bar/' |
|
208 | 210 | >>> normasprefix(b'/').replace(pycompat.ossep, b'/') |
|
209 | 211 | '/' |
|
210 | 212 | ''' |
|
211 | 213 | d, p = os.path.splitdrive(path) |
|
212 | 214 | if len(p) != len(pycompat.ossep): |
|
213 | 215 | return path + pycompat.ossep |
|
214 | 216 | else: |
|
215 | 217 | return path |
|
216 | 218 | |
|
217 | 219 | # forward two methods from posixpath that do what we need, but we'd |
|
218 | 220 | # rather not let our internals know that we're thinking in posix terms |
|
219 | 221 | # - instead we'll let them be oblivious. |
|
220 | 222 | join = posixpath.join |
|
221 | 223 | dirname = posixpath.dirname |
General Comments 0
You need to be logged in to leave comments.
Login now