« back to all changes in this revision
Viewing changes to bzrlib/tree.py
- browse files at revision 6
- compare with another revision
- download diff
- download tarball
- view history from revision 6
- Committer: mbp at sourcefrog
- Date: 2005-03-09 04:51:05 UTC
- Revision ID: mbp@sourcefrog.net-20050309045105-d02cd410a115da2c
import all docs from arch
- files added:
- README
- bzrlib
- doc
- files removed:
- .bzrignore
added
removed
bzrlib/tree.py
1
#! /usr/bin/env python
2
# -*- coding: UTF-8 -*-
3
4
# This program is free software; you can redistribute it and/or modify
5
# it under the terms of the GNU General Public License as published by
6
# the Free Software Foundation; either version 2 of the License, or
7
# (at your option) any later version.
8
9
# This program is distributed in the hope that it will be useful,
10
# but WITHOUT ANY WARRANTY; without even the implied warranty of
11
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12
# GNU General Public License for more details.
13
14
# You should have received a copy of the GNU General Public License
15
# along with this program; if not, write to the Free Software
16
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
18
"""Tree classes, representing directory at point in time.
19
"""
20
21
from sets import Set
22
import os.path, os, fnmatch
23
24
from inventory import Inventory
25
from trace import mutter, note
26
from osutils import pumpfile, compare_files, filesize, quotefn, sha_file, \
27
joinpath, splitpath, appendpath, isdir, isfile, file_kind
28
from errors import bailout
29
import branch
30
from stat import S_ISREG, S_ISDIR, ST_MODE, ST_SIZE
31
32
import bzrlib
33
34
class Tree:
35
"""Abstract file tree.
36
37
There are several subclasses:
38
39
* `WorkingTree` exists as files on disk editable by the user.
40
41
* `RevisionTree` is a tree as recorded at some point in the past.
42
43
* `EmptyTree`
44
45
Trees contain an `Inventory` object, and also know how to retrieve
46
file texts mentioned in the inventory, either from a working
47
directory or from a store.
48
49
It is possible for trees to contain files that are not described
50
in their inventory or vice versa; for this use `filenames()`.
51
52
Trees can be compared, etc, regardless of whether they are working
53
trees or versioned trees.
54
"""
55
56
def has_filename(self, filename):
57
"""True if the tree has given filename."""
58
raise NotImplementedError()
59
60
def has_id(self, file_id):
61
return self.inventory.has_id(file_id)
62
63
def id_set(self):
64
"""Return set of all ids in this tree."""
65
return self.inventory.id_set()
66
67
def id2path(self, file_id):
68
return self.inventory.id2path(file_id)
69
70
def _get_inventory(self):
71
return self._inventory
72
73
inventory = property(_get_inventory,
74
doc="Inventory of this Tree")
75
76
def _check_retrieved(self, ie, f):
77
# TODO: Test this check by damaging the store?
78
if ie.text_size is not None:
79
fs = filesize(f)
80
if fs != ie.text_size:
81
bailout("mismatched size for file %r in %r" % (ie.file_id, self._store),
82
["inventory expects %d bytes" % ie.text_size,
83
"file is actually %d bytes" % fs,
84
"store is probably damaged/corrupt"])
85
86
f_hash = sha_file(f)
87
f.seek(0)
88
if ie.text_sha1 != f_hash:
89
bailout("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
90
["inventory expects %s" % ie.text_sha1,
91
"file is actually %s" % f_hash,
92
"store is probably damaged/corrupt"])
93
94
95
def export(self, dest):
96
"""Export this tree to a new directory.
97
98
`dest` should not exist, and will be created holding the
99
contents of this tree.
100
101
:todo: To handle subdirectories we need to create the
102
directories first.
103
104
:note: If the export fails, the destination directory will be
105
left in a half-assed state.
106
"""
107
os.mkdir(dest)
108
mutter('export version %r' % self)
109
inv = self.inventory
110
for dp, ie in inv.iter_entries():
111
kind = ie.kind
112
fullpath = appendpath(dest, dp)
113
if kind == 'directory':
114
os.mkdir(fullpath)
115
elif kind == 'file':
116
pumpfile(self.get_file(ie.file_id), file(fullpath, 'wb'))
117
else:
118
bailout("don't know how to export {%s} of kind %r", fid, kind)
119
mutter(" export {%s} kind %s to %s" % (ie.file_id, kind, fullpath))
120
121
122
123
class WorkingTree(Tree):
124
"""Working copy tree.
125
126
The inventory is held in the `Branch` working-inventory, and the
127
files are in a directory on disk.
128
129
It is possible for a `WorkingTree` to have a filename which is
130
not listed in the Inventory and vice versa.
131
"""
132
def __init__(self, basedir, inv):
133
self._inventory = inv
134
self.basedir = basedir
135
self.path2id = inv.path2id
136
137
def __repr__(self):
138
return "<%s of %s>" % (self.__class__.__name__,
139
self.basedir)
140
141
def _rel(self, filename):
142
return os.path.join(self.basedir, filename)
143
144
def has_filename(self, filename):
145
return os.path.exists(self._rel(filename))
146
147
def get_file(self, file_id):
148
return self.get_file_byname(self.id2path(file_id))
149
150
def get_file_byname(self, filename):
151
return file(self._rel(filename), 'rb')
152
153
def _get_store_filename(self, file_id):
154
return self._rel(self.id2path(file_id))
155
156
def get_file_size(self, file_id):
157
return os.stat(self._get_store_filename(file_id))[ST_SIZE]
158
159
def get_file_sha1(self, file_id):
160
f = self.get_file(file_id)
161
return sha_file(f)
162
163
164
def file_class(self, filename):
165
if self.path2id(filename):
166
return 'V'
167
elif self.is_ignored(filename):
168
return 'I'
169
else:
170
return '?'
171
172
173
def file_kind(self, filename):
174
if isfile(self._rel(filename)):
175
return 'file'
176
elif isdir(self._rel(filename)):
177
return 'directory'
178
else:
179
return 'unknown'
180
181
182
def list_files(self):
183
"""Recursively list all files as (path, class, kind, id).
184
185
Lists, but does not descend into unversioned directories.
186
187
This does not include files that have been deleted in this
188
tree.
189
190
Skips the control directory.
191
"""
192
inv = self.inventory
193
194
def descend(from_dir, from_dir_id, dp):
195
ls = os.listdir(dp)
196
ls.sort()
197
for f in ls:
198
if bzrlib.BZRDIR == f:
199
continue
200
201
# path within tree
202
fp = appendpath(from_dir, f)
203
204
# absolute path
205
fap = appendpath(dp, f)
206
207
f_ie = inv.get_child(from_dir_id, f)
208
if f_ie:
209
c = 'V'
210
elif self.is_ignored(fp):
211
c = 'I'
212
else:
213
c = '?'
214
215
fk = file_kind(fap)
216
217
if f_ie:
218
if f_ie.kind != fk:
219
bailout("file %r entered as kind %r id %r, now of kind %r"
220
% (fap, f_ie.kind, f_ie.file_id, fk))
221
222
yield fp, c, fk, (f_ie and f_ie.file_id)
223
224
if fk != 'directory':
225
continue
226
227
if c != 'V':
228
# don't descend unversioned directories
229
continue
230
231
for ff in descend(fp, f_ie.file_id, fap):
232
yield ff
233
234
for f in descend('', None, self.basedir):
235
yield f
236
237
238
239
def unknowns(self, path='', dir_id=None):
240
"""Yield names of unknown files in this WorkingTree.
241
242
If there are any unknown directories then only the directory is
243
returned, not all its children. But if there are unknown files
244
under a versioned subdirectory, they are returned.
245
246
Currently returned depth-first, sorted by name within directories.
247
"""
248
for fpath, fclass, fkind, fid in self.list_files():
249
if fclass == '?':
250
yield fpath
251
252
253
def ignored_files(self):
254
for fpath, fclass, fkind, fid in self.list_files():
255
if fclass == 'I':
256
yield fpath
257
258
259
def get_ignore_list(self):
260
"""Return list of ignore patterns."""
261
if self.has_filename(bzrlib.IGNORE_FILENAME):
262
f = self.get_file_byname(bzrlib.IGNORE_FILENAME)
263
return [line.rstrip("\n\r") for line in f.readlines()]
264
else:
265
return bzrlib.DEFAULT_IGNORE
266
267
268
def is_ignored(self, filename):
269
"""Check whether the filename matches an ignore pattern.
270
271
Patterns containing '/' need to match the whole path; others
272
match against only the last component."""
273
## TODO: Take them from a file, not hardcoded
274
## TODO: Use extended zsh-style globs maybe?
275
## TODO: Use '**' to match directories?
276
for pat in self.get_ignore_list():
277
if '/' in pat:
278
if fnmatch.fnmatchcase(filename, pat):
279
return True
280
else:
281
if fnmatch.fnmatchcase(splitpath(filename)[-1], pat):
282
return True
283
return False
284
285
286
287
288
289
class RevisionTree(Tree):
290
"""Tree viewing a previous revision.
291
292
File text can be retrieved from the text store.
293
294
:todo: Some kind of `__repr__` method, but a good one
295
probably means knowing the branch and revision number,
296
or at least passing a description to the constructor.
297
"""
298
299
def __init__(self, store, inv):
300
self._store = store
301
self._inventory = inv
302
303
def get_file(self, file_id):
304
ie = self._inventory[file_id]
305
f = self._store[ie.text_id]
306
mutter(" get fileid{%s} from %r" % (file_id, self))
307
fs = filesize(f)
308
if ie.text_size is None:
309
note("warning: no text size recorded on %r" % ie)
310
self._check_retrieved(ie, f)
311
return f
312
313
def get_file_size(self, file_id):
314
return self._inventory[file_id].text_size
315
316
def get_file_sha1(self, file_id):
317
ie = self._inventory[file_id]
318
return ie.text_sha1
319
320
def has_filename(self, filename):
321
return bool(self.inventory.path2id(filename))
322
323
def list_files(self):
324
# The only files returned by this are those from the version
325
for path, entry in self.inventory.iter_entries():
326
yield path, 'V', entry.kind, entry.file_id
327
328
329
class EmptyTree(Tree):
330
def __init__(self):
331
self._inventory = Inventory()
332
333
def has_filename(self, filename):
334
return False
335
336
def list_files(self):
337
if False: # just to make it a generator
338
yield None
339
340
341
342
######################################################################
343
# diff
344
345
# TODO: Merge these two functions into a single one that can operate
346
# on either a whole tree or a set of files.
347
348
# TODO: Return the diff in order by filename, not by category or in
349
# random order. Can probably be done by lock-stepping through the
350
# filenames from both trees.
351
352
353
def file_status(filename, old_tree, new_tree):
354
"""Return single-letter status, old and new names for a file.
355
356
The complexity here is in deciding how to represent renames;
357
many complex cases are possible.
358
"""
359
old_inv = old_tree.inventory
360
new_inv = new_tree.inventory
361
new_id = new_inv.path2id(filename)
362
old_id = old_inv.path2id(filename)
363
364
if not new_id and not old_id:
365
# easy: doesn't exist in either; not versioned at all
366
if new_tree.is_ignored(filename):
367
return 'I', None, None
368
else:
369
return '?', None, None
370
elif new_id:
371
# There is now a file of this name, great.
372
pass
373
else:
374
# There is no longer a file of this name, but we can describe
375
# what happened to the file that used to have
376
# this name. There are two possibilities: either it was
377
# deleted entirely, or renamed.
378
assert old_id
379
if new_inv.has_id(old_id):
380
return 'X', old_inv.id2path(old_id), new_inv.id2path(old_id)
381
else:
382
return 'D', old_inv.id2path(old_id), None
383
384
# if the file_id is new in this revision, it is added
385
if new_id and not old_inv.has_id(new_id):
386
return 'A'
387
388
# if there used to be a file of this name, but that ID has now
389
# disappeared, it is deleted
390
if old_id and not new_inv.has_id(old_id):
391
return 'D'
392
393
return 'wtf?'
394
395
396
Loggerhead is a web-based interface for Breezy
Version: 2.0.1