« back to all changes in this revision
Viewing changes to bzrlib/tree.py
- browse files at revision 702
- compare with another revision
- download diff
- download tarball
- view history from revision 702
- Committer: Martin Pool
- Date: 2005-06-20 02:03:32 UTC
- Revision ID: mbp@sourcefrog.net-20050620020332-29835ac6826eb948
todo: bzr upgrade
- 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
22
23
from inventory import Inventory
24
from trace import mutter, note
25
from osutils import pumpfile, compare_files, filesize, quotefn, sha_file, \
26
joinpath, splitpath, appendpath, isdir, isfile, file_kind
27
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
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)
73
78
doc="Inventory of this Tree")
74
79
75
80
def _check_retrieved(self, ie, f):
76
# TODO: Test this check by damaging the store?
77
if ie.text_size is not None:
78
fs = filesize(f)
79
if fs != ie.text_size:
80
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),
81
87
["inventory expects %d bytes" % ie.text_size,
82
"file is actually %d bytes" % fs,
88
"file is actually %d bytes" % fp['size'],
83
89
"store is probably damaged/corrupt"])
84
90
85
f_hash = sha_file(f)
86
f.seek(0)
87
if ie.text_sha1 != f_hash:
88
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),
89
93
["inventory expects %s" % ie.text_sha1,
90
"file is actually %s" % f_hash,
94
"file is actually %s" % fp['sha1'],
91
95
"store is probably damaged/corrupt"])
92
96
93
97
94
def export(self, dest):
95
"""Export this tree to a new directory.
96
97
`dest` should not exist, and will be created holding the
98
contents of this tree.
99
100
:todo: To handle subdirectories we need to create the
101
directories first.
102
103
:note: If the export fails, the destination directory will be
104
left in a half-assed state.
105
"""
106
os.mkdir(dest)
107
mutter('export version %r' % self)
108
inv = self.inventory
109
for dp, ie in inv.iter_entries():
110
kind = ie.kind
111
fullpath = appendpath(dest, dp)
112
if kind == 'directory':
113
os.mkdir(fullpath)
114
elif kind == 'file':
115
pumpfile(self.get_file(ie.file_id), file(fullpath, 'wb'))
116
else:
117
bailout("don't know how to export {%s} of kind %r", fid, kind)
118
mutter(" export {%s} kind %s to %s" % (ie.file_id, kind, fullpath))
119
120
121
122
class WorkingTree(Tree):
123
"""Working copy tree.
124
125
The inventory is held in the `Branch` working-inventory, and the
126
files are in a directory on disk.
127
128
It is possible for a `WorkingTree` to have a filename which is
129
not listed in the Inventory and vice versa.
130
"""
131
def __init__(self, basedir, inv):
132
self._inventory = inv
133
self.basedir = basedir
134
self.path2id = inv.path2id
135
136
def __repr__(self):
137
return "<%s of %s>" % (self.__class__.__name__,
138
self.basedir)
139
140
def abspath(self, filename):
141
return os.path.join(self.basedir, filename)
142
143
def has_filename(self, filename):
144
return os.path.exists(self.abspath(filename))
145
146
def get_file(self, file_id):
147
return self.get_file_byname(self.id2path(file_id))
148
149
def get_file_byname(self, filename):
150
return file(self.abspath(filename), 'rb')
151
152
def _get_store_filename(self, file_id):
153
return self.abspath(self.id2path(file_id))
154
155
def has_id(self, file_id):
156
# files that have been deleted are excluded
157
if not self.inventory.has_id(file_id):
158
return False
159
return os.access(self.abspath(self.inventory.id2path(file_id)), os.F_OK)
160
161
def get_file_size(self, file_id):
162
return os.stat(self._get_store_filename(file_id))[ST_SIZE]
163
164
def get_file_sha1(self, file_id):
165
f = self.get_file(file_id)
166
return sha_file(f)
167
168
169
def file_class(self, filename):
170
if self.path2id(filename):
171
return 'V'
172
elif self.is_ignored(filename):
173
return 'I'
174
else:
175
return '?'
176
177
178
def file_kind(self, filename):
179
if isfile(self.abspath(filename)):
180
return 'file'
181
elif isdir(self.abspath(filename)):
182
return 'directory'
183
else:
184
return 'unknown'
185
186
187
def list_files(self):
188
"""Recursively list all files as (path, class, kind, id).
189
190
Lists, but does not descend into unversioned directories.
191
192
This does not include files that have been deleted in this
193
tree.
194
195
Skips the control directory.
196
"""
197
inv = self.inventory
198
199
def descend(from_dir, from_dir_id, dp):
200
ls = os.listdir(dp)
201
ls.sort()
202
for f in ls:
203
if bzrlib.BZRDIR == f:
204
continue
205
206
# path within tree
207
fp = appendpath(from_dir, f)
208
209
# absolute path
210
fap = appendpath(dp, f)
211
212
f_ie = inv.get_child(from_dir_id, f)
213
if f_ie:
214
c = 'V'
215
elif self.is_ignored(fp):
216
c = 'I'
217
else:
218
c = '?'
219
220
fk = file_kind(fap)
221
222
if f_ie:
223
if f_ie.kind != fk:
224
bailout("file %r entered as kind %r id %r, now of kind %r"
225
% (fap, f_ie.kind, f_ie.file_id, fk))
226
227
yield fp, c, fk, (f_ie and f_ie.file_id)
228
229
if fk != 'directory':
230
continue
231
232
if c != 'V':
233
# don't descend unversioned directories
234
continue
235
236
for ff in descend(fp, f_ie.file_id, fap):
237
yield ff
238
239
for f in descend('', None, self.basedir):
240
yield f
241
242
243
244
def unknowns(self, path='', dir_id=None):
245
"""Yield names of unknown files in this WorkingTree.
246
247
If there are any unknown directories then only the directory is
248
returned, not all its children. But if there are unknown files
249
under a versioned subdirectory, they are returned.
250
251
Currently returned depth-first, sorted by name within directories.
252
"""
253
for fpath, fclass, fkind, fid in self.list_files():
254
if fclass == '?':
255
yield fpath
256
257
258
def ignored_files(self):
259
for fpath, fclass, fkind, fid in self.list_files():
260
if fclass == 'I':
261
yield fpath
262
263
264
def get_ignore_list(self):
265
"""Return list of ignore patterns."""
266
if self.has_filename(bzrlib.IGNORE_FILENAME):
267
f = self.get_file_byname(bzrlib.IGNORE_FILENAME)
268
return [line.rstrip("\n\r") for line in f.readlines()]
269
else:
270
return bzrlib.DEFAULT_IGNORE
271
272
273
def is_ignored(self, filename):
274
"""Check whether the filename matches an ignore pattern.
275
276
Patterns containing '/' need to match the whole path; others
277
match against only the last component."""
278
## TODO: Take them from a file, not hardcoded
279
## TODO: Use extended zsh-style globs maybe?
280
## TODO: Use '**' to match directories?
281
for pat in self.get_ignore_list():
282
if '/' in pat:
283
if fnmatch.fnmatchcase(filename, pat):
284
return True
285
else:
286
if fnmatch.fnmatchcase(splitpath(filename)[-1], pat):
287
return True
288
return False
289
290
291
292
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
293
113
294
114
class RevisionTree(Tree):
295
115
"""Tree viewing a previous revision.
296
116
297
117
File text can be retrieved from the text store.
298
118
299
:todo: Some kind of `__repr__` method, but a good one
119
TODO: Some kind of `__repr__` method, but a good one
300
120
probably means knowing the branch and revision number,
301
121
or at least passing a description to the constructor.
302
122
"""
309
129
ie = self._inventory[file_id]
310
130
f = self._store[ie.text_id]
311
131
mutter(" get fileid{%s} from %r" % (file_id, self))
312
fs = filesize(f)
313
if ie.text_size is None:
314
note("warning: no text size recorded on %r" % ie)
315
132
self._check_retrieved(ie, f)
316
133
return f
317
134
399
216
400
217
401
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