« back to all changes in this revision
Viewing changes to bzrlib/tree.py
- browse files at revision 736
- compare with another revision
- download diff
- download tarball
- view history from revision 736
- Committer: Martin Pool
- Date: 2005-06-22 06:37:43 UTC
- Revision ID: mbp@sourcefrog.net-20050622063743-e395f04c4db8977f
- move old blackbox code from testbzr into bzrlib.selftest.blackbox
- files added:
- .rsyncexclude
- bzrlib/selftest
- contrib
- contrib/bash
- contrib/zsh
- elementtree
- patches
- urlgrabber
- files removed:
- bzrlib/tests.py
- files modified:
- .bzrignore
added
removed
bzrlib/tree.py
1
#! /usr/bin/env python
2
# -*- coding: UTF-8 -*-
1
# Copyright (C) 2005 Canonical Ltd
3
2
4
3
# This program is free software; you can redistribute it and/or modify
5
4
# it under the terms of the GNU General Public License as published by
19
18
"""
20
19
21
20
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
21
import os.path, os, fnmatch, time
22
23
from osutils import pumpfile, filesize, quotefn, sha_file, \
24
joinpath, splitpath, appendpath, isdir, isfile, file_kind, fingerprint_file
25
import errno
26
from stat import S_ISREG, S_ISDIR, ST_MODE, ST_SIZE
27
28
from bzrlib.inventory import Inventory
29
from bzrlib.trace import mutter, note
30
from bzrlib.errors import BzrError
29
31
import branch
30
from stat import S_ISREG, S_ISDIR, ST_MODE, ST_SIZE
31
32
32
33
import bzrlib
33
34
34
class Tree:
35
exporters = {}
36
37
class Tree(object):
35
38
"""Abstract file tree.
36
39
37
40
There are several subclasses:
60
63
def has_id(self, file_id):
61
64
return self.inventory.has_id(file_id)
62
65
63
def id_set(self):
64
"""Return set of all ids in this tree."""
65
return self.inventory.id_set()
66
__contains__ = has_id
67
68
def __iter__(self):
69
return iter(self.inventory)
66
70
67
71
def id2path(self, file_id):
68
72
return self.inventory.id2path(file_id)
74
78
doc="Inventory of this Tree")
75
79
76
80
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),
81
fp = fingerprint_file(f)
82
f.seek(0)
83
84
if ie.text_size != None:
85
if ie.text_size != fp['size']:
86
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
82
87
["inventory expects %d bytes" % ie.text_size,
83
"file is actually %d bytes" % fs,
88
"file is actually %d bytes" % fp['size'],
84
89
"store is probably damaged/corrupt"])
85
90
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),
91
if ie.text_sha1 != fp['sha1']:
92
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
90
93
["inventory expects %s" % ie.text_sha1,
91
"file is actually %s" % f_hash,
94
"file is actually %s" % fp['sha1'],
92
95
"store is probably damaged/corrupt"])
93
96
94
97
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 has_id(self, file_id):
157
# files that have been deleted are excluded
158
if not self.inventory.has_id(file_id):
159
return False
160
return os.access(self._rel(self.inventory.id2path(file_id)), os.F_OK)
161
162
def get_file_size(self, file_id):
163
return os.stat(self._get_store_filename(file_id))[ST_SIZE]
164
165
def get_file_sha1(self, file_id):
166
f = self.get_file(file_id)
167
return sha_file(f)
168
169
170
def file_class(self, filename):
171
if self.path2id(filename):
172
return 'V'
173
elif self.is_ignored(filename):
174
return 'I'
175
else:
176
return '?'
177
178
179
def file_kind(self, filename):
180
if isfile(self._rel(filename)):
181
return 'file'
182
elif isdir(self._rel(filename)):
183
return 'directory'
184
else:
185
return 'unknown'
186
187
188
def list_files(self):
189
"""Recursively list all files as (path, class, kind, id).
190
191
Lists, but does not descend into unversioned directories.
192
193
This does not include files that have been deleted in this
194
tree.
195
196
Skips the control directory.
197
"""
198
inv = self.inventory
199
200
def descend(from_dir, from_dir_id, dp):
201
ls = os.listdir(dp)
202
ls.sort()
203
for f in ls:
204
if bzrlib.BZRDIR == f:
205
continue
206
207
# path within tree
208
fp = appendpath(from_dir, f)
209
210
# absolute path
211
fap = appendpath(dp, f)
212
213
f_ie = inv.get_child(from_dir_id, f)
214
if f_ie:
215
c = 'V'
216
elif self.is_ignored(fp):
217
c = 'I'
218
else:
219
c = '?'
220
221
fk = file_kind(fap)
222
223
if f_ie:
224
if f_ie.kind != fk:
225
bailout("file %r entered as kind %r id %r, now of kind %r"
226
% (fap, f_ie.kind, f_ie.file_id, fk))
227
228
yield fp, c, fk, (f_ie and f_ie.file_id)
229
230
if fk != 'directory':
231
continue
232
233
if c != 'V':
234
# don't descend unversioned directories
235
continue
236
237
for ff in descend(fp, f_ie.file_id, fap):
238
yield ff
239
240
for f in descend('', None, self.basedir):
241
yield f
242
243
244
245
def unknowns(self, path='', dir_id=None):
246
"""Yield names of unknown files in this WorkingTree.
247
248
If there are any unknown directories then only the directory is
249
returned, not all its children. But if there are unknown files
250
under a versioned subdirectory, they are returned.
251
252
Currently returned depth-first, sorted by name within directories.
253
"""
254
for fpath, fclass, fkind, fid in self.list_files():
255
if fclass == '?':
256
yield fpath
257
258
259
def ignored_files(self):
260
for fpath, fclass, fkind, fid in self.list_files():
261
if fclass == 'I':
262
yield fpath
263
264
265
def get_ignore_list(self):
266
"""Return list of ignore patterns."""
267
if self.has_filename(bzrlib.IGNORE_FILENAME):
268
f = self.get_file_byname(bzrlib.IGNORE_FILENAME)
269
return [line.rstrip("\n\r") for line in f.readlines()]
270
else:
271
return bzrlib.DEFAULT_IGNORE
272
273
274
def is_ignored(self, filename):
275
"""Check whether the filename matches an ignore pattern.
276
277
Patterns containing '/' need to match the whole path; others
278
match against only the last component."""
279
## TODO: Take them from a file, not hardcoded
280
## TODO: Use extended zsh-style globs maybe?
281
## TODO: Use '**' to match directories?
282
for pat in self.get_ignore_list():
283
if '/' in pat:
284
if fnmatch.fnmatchcase(filename, pat):
285
return True
286
else:
287
if fnmatch.fnmatchcase(splitpath(filename)[-1], pat):
288
return True
289
return False
290
291
292
293
98
def print_file(self, fileid):
99
"""Print file with id `fileid` to stdout."""
100
import sys
101
pumpfile(self.get_file(fileid), sys.stdout)
102
103
104
def export(self, dest, format='dir'):
105
"""Export this tree."""
106
try:
107
exporter = exporters[format]
108
except KeyError:
109
raise BzrCommandError("export format %r not supported" % format)
110
exporter(self, dest)
111
112
294
113
295
114
class RevisionTree(Tree):
296
115
"""Tree viewing a previous revision.
297
116
298
117
File text can be retrieved from the text store.
299
118
300
:todo: Some kind of `__repr__` method, but a good one
119
TODO: Some kind of `__repr__` method, but a good one
301
120
probably means knowing the branch and revision number,
302
121
or at least passing a description to the constructor.
303
122
"""
310
129
ie = self._inventory[file_id]
311
130
f = self._store[ie.text_id]
312
131
mutter(" get fileid{%s} from %r" % (file_id, self))
313
fs = filesize(f)
314
if ie.text_size is None:
315
note("warning: no text size recorded on %r" % ie)
316
132
self._check_retrieved(ie, f)
317
133
return f
318
134
400
216
401
217
402
218
219
def find_renames(old_inv, new_inv):
220
for file_id in old_inv:
221
if file_id not in new_inv:
222
continue
223
old_name = old_inv.id2path(file_id)
224
new_name = new_inv.id2path(file_id)
225
if old_name != new_name:
226
yield (old_name, new_name)
227
228
229
230
######################################################################
231
# export
232
233
def dir_exporter(tree, dest):
234
"""Export this tree to a new directory.
235
236
`dest` should not exist, and will be created holding the
237
contents of this tree.
238
239
TODO: To handle subdirectories we need to create the
240
directories first.
241
242
:note: If the export fails, the destination directory will be
243
left in a half-assed state.
244
"""
245
os.mkdir(dest)
246
mutter('export version %r' % tree)
247
inv = tree.inventory
248
for dp, ie in inv.iter_entries():
249
kind = ie.kind
250
fullpath = appendpath(dest, dp)
251
if kind == 'directory':
252
os.mkdir(fullpath)
253
elif kind == 'file':
254
pumpfile(tree.get_file(ie.file_id), file(fullpath, 'wb'))
255
else:
256
raise BzrError("don't know how to export {%s} of kind %r" % (ie.file_id, kind))
257
mutter(" export {%s} kind %s to %s" % (ie.file_id, kind, fullpath))
258
exporters['dir'] = dir_exporter
259
260
try:
261
import tarfile
262
except ImportError:
263
pass
264
else:
265
def tar_exporter(tree, dest, compression=None):
266
"""Export this tree to a new tar file.
267
268
`dest` will be created holding the contents of this tree; if it
269
already exists, it will be clobbered, like with "tar -c".
270
"""
271
now = time.time()
272
compression = str(compression or '')
273
try:
274
ball = tarfile.open(dest, 'w:' + compression)
275
except tarfile.CompressionError, e:
276
raise BzrError(str(e))
277
mutter('export version %r' % tree)
278
inv = tree.inventory
279
for dp, ie in inv.iter_entries():
280
mutter(" export {%s} kind %s to %s" % (ie.file_id, ie.kind, dest))
281
item = tarfile.TarInfo(dp)
282
# TODO: would be cool to actually set it to the timestamp of the
283
# revision it was last changed
284
item.mtime = now
285
if ie.kind == 'directory':
286
item.type = tarfile.DIRTYPE
287
fileobj = None
288
item.name += '/'
289
item.size = 0
290
item.mode = 0755
291
elif ie.kind == 'file':
292
item.type = tarfile.REGTYPE
293
fileobj = tree.get_file(ie.file_id)
294
item.size = _find_file_size(fileobj)
295
item.mode = 0644
296
else:
297
raise BzrError("don't know how to export {%s} of kind %r" %
298
(ie.file_id, ie.kind))
299
300
ball.addfile(item, fileobj)
301
ball.close()
302
exporters['tar'] = tar_exporter
303
304
def tgz_exporter(tree, dest):
305
tar_exporter(tree, dest, compression='gz')
306
exporters['tgz'] = tgz_exporter
307
308
def tbz_exporter(tree, dest):
309
tar_exporter(tree, dest, compression='bz2')
310
exporters['tbz2'] = tbz_exporter
311
312
313
def _find_file_size(fileobj):
314
offset = fileobj.tell()
315
try:
316
fileobj.seek(0, 2)
317
size = fileobj.tell()
318
except TypeError:
319
# gzip doesn't accept second argument to seek()
320
fileobj.seek(0)
321
size = 0
322
while True:
323
nread = len(fileobj.read())
324
if nread == 0:
325
break
326
size += nread
327
fileobj.seek(offset)
328
return size
Loggerhead is a web-based interface for Breezy
Version: 2.0.1