« back to all changes in this revision
Viewing changes to bzrlib/tree.py
- browse files at revision 716
- compare with another revision
- download diff
- download tarball
- view history from revision 716
- Committer: Martin Pool
- Date: 2005-06-20 04:33:23 UTC
- Revision ID: mbp@sourcefrog.net-20050620043323-1f681c2b5b15d026
- write into store using AtomicFile
- files added:
- .rsyncexclude
- contrib
- contrib/bash
- contrib/zsh
- patches
- urlgrabber
- files removed:
- bzrlib/tests.py
- files modified:
- .bzrignore
added
removed
bzrlib/tree.py
18
18
"""
19
19
20
20
from sets import Set
21
import os.path, os, fnmatch
21
import os.path, os, fnmatch, time
22
22
23
from inventory import Inventory
24
from trace import mutter, note
25
from osutils import pumpfile, compare_files, filesize, quotefn, sha_file, \
23
from osutils import pumpfile, filesize, quotefn, sha_file, \
26
24
joinpath, splitpath, appendpath, isdir, isfile, file_kind, fingerprint_file
27
from errors import bailout
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
28
31
import branch
29
from stat import S_ISREG, S_ISDIR, ST_MODE, ST_SIZE
30
32
31
33
import bzrlib
32
34
33
class Tree:
35
exporters = {}
36
37
class Tree(object):
34
38
"""Abstract file tree.
35
39
36
40
There are several subclasses:
59
63
def has_id(self, file_id):
60
64
return self.inventory.has_id(file_id)
61
65
62
def id_set(self):
63
"""Return set of all ids in this tree."""
64
return self.inventory.id_set()
66
__contains__ = has_id
67
68
def __iter__(self):
69
return iter(self.inventory)
65
70
66
71
def id2path(self, file_id):
67
72
return self.inventory.id2path(file_id)
76
81
fp = fingerprint_file(f)
77
82
f.seek(0)
78
83
79
if ie.text_size is not None:
84
if ie.text_size != None:
80
85
if ie.text_size != fp['size']:
81
bailout("mismatched size for file %r in %r" % (ie.file_id, self._store),
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
88
"file is actually %d bytes" % fp['size'],
84
89
"store is probably damaged/corrupt"])
85
90
86
91
if ie.text_sha1 != fp['sha1']:
87
bailout("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
92
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
88
93
["inventory expects %s" % ie.text_sha1,
89
94
"file is actually %s" % fp['sha1'],
90
95
"store is probably damaged/corrupt"])
91
96
92
97
93
def export(self, dest):
94
"""Export this tree to a new directory.
95
96
`dest` should not exist, and will be created holding the
97
contents of this tree.
98
99
:todo: To handle subdirectories we need to create the
100
directories first.
101
102
:note: If the export fails, the destination directory will be
103
left in a half-assed state.
104
"""
105
os.mkdir(dest)
106
mutter('export version %r' % self)
107
inv = self.inventory
108
for dp, ie in inv.iter_entries():
109
kind = ie.kind
110
fullpath = appendpath(dest, dp)
111
if kind == 'directory':
112
os.mkdir(fullpath)
113
elif kind == 'file':
114
pumpfile(self.get_file(ie.file_id), file(fullpath, 'wb'))
115
else:
116
bailout("don't know how to export {%s} of kind %r", fid, kind)
117
mutter(" export {%s} kind %s to %s" % (ie.file_id, kind, fullpath))
118
119
120
121
class WorkingTree(Tree):
122
"""Working copy tree.
123
124
The inventory is held in the `Branch` working-inventory, and the
125
files are in a directory on disk.
126
127
It is possible for a `WorkingTree` to have a filename which is
128
not listed in the Inventory and vice versa.
129
"""
130
def __init__(self, basedir, inv):
131
self._inventory = inv
132
self.basedir = basedir
133
self.path2id = inv.path2id
134
135
def __repr__(self):
136
return "<%s of %s>" % (self.__class__.__name__,
137
self.basedir)
138
139
def abspath(self, filename):
140
return os.path.join(self.basedir, filename)
141
142
def has_filename(self, filename):
143
return os.path.exists(self.abspath(filename))
144
145
def get_file(self, file_id):
146
return self.get_file_byname(self.id2path(file_id))
147
148
def get_file_byname(self, filename):
149
return file(self.abspath(filename), 'rb')
150
151
def _get_store_filename(self, file_id):
152
return self.abspath(self.id2path(file_id))
153
154
def has_id(self, file_id):
155
# files that have been deleted are excluded
156
if not self.inventory.has_id(file_id):
157
return False
158
return os.access(self.abspath(self.inventory.id2path(file_id)), os.F_OK)
159
160
def get_file_size(self, file_id):
161
return os.stat(self._get_store_filename(file_id))[ST_SIZE]
162
163
def get_file_sha1(self, file_id):
164
f = self.get_file(file_id)
165
return sha_file(f)
166
167
168
def file_class(self, filename):
169
if self.path2id(filename):
170
return 'V'
171
elif self.is_ignored(filename):
172
return 'I'
173
else:
174
return '?'
175
176
177
def list_files(self):
178
"""Recursively list all files as (path, class, kind, id).
179
180
Lists, but does not descend into unversioned directories.
181
182
This does not include files that have been deleted in this
183
tree.
184
185
Skips the control directory.
186
"""
187
inv = self.inventory
188
189
def descend(from_dir, from_dir_id, dp):
190
ls = os.listdir(dp)
191
ls.sort()
192
for f in ls:
193
## TODO: If we find a subdirectory with its own .bzr
194
## directory, then that is a separate tree and we
195
## should exclude it.
196
if bzrlib.BZRDIR == f:
197
continue
198
199
# path within tree
200
fp = appendpath(from_dir, f)
201
202
# absolute path
203
fap = appendpath(dp, f)
204
205
f_ie = inv.get_child(from_dir_id, f)
206
if f_ie:
207
c = 'V'
208
elif self.is_ignored(fp):
209
c = 'I'
210
else:
211
c = '?'
212
213
fk = file_kind(fap)
214
215
if f_ie:
216
if f_ie.kind != fk:
217
bailout("file %r entered as kind %r id %r, now of kind %r"
218
% (fap, f_ie.kind, f_ie.file_id, fk))
219
220
yield fp, c, fk, (f_ie and f_ie.file_id)
221
222
if fk != 'directory':
223
continue
224
225
if c != 'V':
226
# don't descend unversioned directories
227
continue
228
229
for ff in descend(fp, f_ie.file_id, fap):
230
yield ff
231
232
for f in descend('', None, self.basedir):
233
yield f
234
235
236
237
def unknowns(self):
238
for subp in self.extras():
239
if not self.is_ignored(subp):
240
yield subp
241
242
243
def extras(self):
244
"""Yield all unknown files in this WorkingTree.
245
246
If there are any unknown directories then only the directory is
247
returned, not all its children. But if there are unknown files
248
under a versioned subdirectory, they are returned.
249
250
Currently returned depth-first, sorted by name within directories.
251
"""
252
## TODO: Work from given directory downwards
253
254
for path, dir_entry in self.inventory.directories():
255
mutter("search for unknowns in %r" % path)
256
dirabs = self.abspath(path)
257
if not isdir(dirabs):
258
# e.g. directory deleted
259
continue
260
261
fl = []
262
for subf in os.listdir(dirabs):
263
if (subf != '.bzr'
264
and (subf not in dir_entry.children)):
265
fl.append(subf)
266
267
fl.sort()
268
for subf in fl:
269
subp = appendpath(path, subf)
270
yield subp
271
272
273
def ignored_files(self):
274
"""Yield list of PATH, IGNORE_PATTERN"""
275
for subp in self.extras():
276
pat = self.is_ignored(subp)
277
if pat != None:
278
yield subp, pat
279
280
281
def get_ignore_list(self):
282
"""Return list of ignore patterns.
283
284
Cached in the Tree object after the first call.
285
"""
286
if hasattr(self, '_ignorelist'):
287
return self._ignorelist
288
289
l = bzrlib.DEFAULT_IGNORE[:]
290
if self.has_filename(bzrlib.IGNORE_FILENAME):
291
f = self.get_file_byname(bzrlib.IGNORE_FILENAME)
292
l.extend([line.rstrip("\n\r") for line in f.readlines()])
293
self._ignorelist = l
294
return l
295
296
297
def is_ignored(self, filename):
298
"""Check whether the filename matches an ignore pattern.
299
300
Patterns containing '/' need to match the whole path; others
301
match against only the last component.
302
303
If the file is ignored, returns the pattern which caused it to
304
be ignored, otherwise None. So this can simply be used as a
305
boolean if desired."""
306
307
## TODO: Use '**' to match directories, and other extended globbing stuff from cvs/rsync.
308
309
for pat in self.get_ignore_list():
310
if '/' in pat:
311
# as a special case, you can put ./ at the start of a pattern;
312
# this is good to match in the top-level only;
313
if pat[:2] == './':
314
newpat = pat[2:]
315
else:
316
newpat = pat
317
if fnmatch.fnmatchcase(filename, newpat):
318
return pat
319
else:
320
if fnmatch.fnmatchcase(splitpath(filename)[-1], pat):
321
return pat
322
return None
323
324
325
326
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
327
113
328
114
class RevisionTree(Tree):
329
115
"""Tree viewing a previous revision.
330
116
331
117
File text can be retrieved from the text store.
332
118
333
:todo: Some kind of `__repr__` method, but a good one
119
TODO: Some kind of `__repr__` method, but a good one
334
120
probably means knowing the branch and revision number,
335
121
or at least passing a description to the constructor.
336
122
"""
430
216
431
217
432
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