1
# Copyright (C) 2005 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
"""Tree classes, representing directory at point in time.
21
from cStringIO import StringIO
24
from bzrlib.trace import mutter, note
25
from bzrlib.errors import BzrError, BzrCheckError
26
from bzrlib.inventory import Inventory
27
from bzrlib.osutils import fingerprint_file
31
"""Abstract file tree.
33
There are several subclasses:
35
* `WorkingTree` exists as files on disk editable by the user.
37
* `RevisionTree` is a tree as recorded at some point in the past.
41
Trees contain an `Inventory` object, and also know how to retrieve
42
file texts mentioned in the inventory, either from a working
43
directory or from a store.
45
It is possible for trees to contain files that are not described
46
in their inventory or vice versa; for this use `filenames()`.
48
Trees can be compared, etc, regardless of whether they are working
49
trees or versioned trees.
52
def has_filename(self, filename):
53
"""True if the tree has given filename."""
54
raise NotImplementedError()
56
def has_id(self, file_id):
57
return self.inventory.has_id(file_id)
59
def has_or_had_id(self, file_id):
60
if file_id == self.inventory.root.file_id:
62
return self.inventory.has_id(file_id)
67
return iter(self.inventory)
69
def id2path(self, file_id):
70
return self.inventory.id2path(file_id)
72
def kind(self, file_id):
73
raise NotImplementedError("subclasses must implement kind")
75
def _get_inventory(self):
76
return self._inventory
78
def get_file_by_path(self, path):
79
return self.get_file(self._inventory.path2id(path))
81
inventory = property(_get_inventory,
82
doc="Inventory of this Tree")
84
def _check_retrieved(self, ie, f):
87
fp = fingerprint_file(f)
90
if ie.text_size != None:
91
if ie.text_size != fp['size']:
92
raise BzrError("mismatched size for file %r in %r" % (ie.file_id, self._store),
93
["inventory expects %d bytes" % ie.text_size,
94
"file is actually %d bytes" % fp['size'],
95
"store is probably damaged/corrupt"])
97
if ie.text_sha1 != fp['sha1']:
98
raise BzrError("wrong SHA-1 for file %r in %r" % (ie.file_id, self._store),
99
["inventory expects %s" % ie.text_sha1,
100
"file is actually %s" % fp['sha1'],
101
"store is probably damaged/corrupt"])
104
def print_file(self, file_id):
105
"""Print file with id `file_id` to stdout."""
107
sys.stdout.write(self.get_file_text(file_id))
115
def filter_unversioned_files(self, paths):
116
"""Filter out paths that are not versioned.
118
:return: set of paths.
120
# NB: we specifically *don't* call self.has_filename, because for
121
# WorkingTrees that can indicate files that exist on disk but that
123
pred = self.inventory.has_filename
124
return set((p for p in paths if not pred(p)))
127
class RevisionTree(Tree):
128
"""Tree viewing a previous revision.
130
File text can be retrieved from the text store.
132
TODO: Some kind of `__repr__` method, but a good one
133
probably means knowing the branch and revision number,
134
or at least passing a description to the constructor.
137
def __init__(self, branch, inv, revision_id):
138
self._branch = branch
139
self._weave_store = branch.weave_store
140
self._inventory = inv
141
self._revision_id = revision_id
143
def get_weave(self, file_id):
144
return self._weave_store.get_weave(file_id,
145
self._branch.get_transaction())
147
def get_file_lines(self, file_id):
148
ie = self._inventory[file_id]
149
weave = self.get_weave(file_id)
150
return weave.get_lines(ie.revision)
152
def get_file_text(self, file_id):
153
return ''.join(self.get_file_lines(file_id))
155
def get_file(self, file_id):
156
return StringIO(self.get_file_text(file_id))
158
def get_file_size(self, file_id):
159
return self._inventory[file_id].text_size
161
def get_file_sha1(self, file_id):
162
ie = self._inventory[file_id]
163
if ie.kind == "file":
166
def is_executable(self, file_id):
167
ie = self._inventory[file_id]
168
if ie.kind != "file":
170
return self._inventory[file_id].executable
172
def has_filename(self, filename):
173
return bool(self.inventory.path2id(filename))
175
def has_file_or_id(self, filename, file_id):
176
"""Do we have this filename or file_id?
178
Some trees are faster to lookup by filename, some
179
are faster by id, if we have both, figure out which
182
# Default to has_id, since that is best for Revision trees
183
return self.has_id(file_id)
185
def list_files(self):
186
# The only files returned by this are those from the version
187
for path, entry in self.inventory.iter_entries():
188
yield path, 'V', entry.kind, entry.file_id, entry
190
def get_symlink_target(self, file_id):
191
ie = self._inventory[file_id]
192
return ie.symlink_target;
194
def kind(self, file_id):
195
return self._inventory[file_id].kind
198
self._branch.lock_read()
201
self._branch.unlock()
204
class EmptyTree(Tree):
206
self._inventory = Inventory()
208
def get_symlink_target(self, file_id):
211
def has_filename(self, filename):
214
def kind(self, file_id):
215
assert self._inventory[file_id].kind == "root_directory"
216
return "root_directory"
218
def list_files(self):
221
def __contains__(self, file_id):
222
return file_id in self._inventory
224
def get_file_sha1(self, file_id):
225
assert self._inventory[file_id].kind == "root_directory"
229
######################################################################
232
# TODO: Merge these two functions into a single one that can operate
233
# on either a whole tree or a set of files.
235
# TODO: Return the diff in order by filename, not by category or in
236
# random order. Can probably be done by lock-stepping through the
237
# filenames from both trees.
240
def file_status(filename, old_tree, new_tree):
241
"""Return single-letter status, old and new names for a file.
243
The complexity here is in deciding how to represent renames;
244
many complex cases are possible.
246
old_inv = old_tree.inventory
247
new_inv = new_tree.inventory
248
new_id = new_inv.path2id(filename)
249
old_id = old_inv.path2id(filename)
251
if not new_id and not old_id:
252
# easy: doesn't exist in either; not versioned at all
253
if new_tree.is_ignored(filename):
254
return 'I', None, None
256
return '?', None, None
258
# There is now a file of this name, great.
261
# There is no longer a file of this name, but we can describe
262
# what happened to the file that used to have
263
# this name. There are two possibilities: either it was
264
# deleted entirely, or renamed.
266
if new_inv.has_id(old_id):
267
return 'X', old_inv.id2path(old_id), new_inv.id2path(old_id)
269
return 'D', old_inv.id2path(old_id), None
271
# if the file_id is new in this revision, it is added
272
if new_id and not old_inv.has_id(new_id):
275
# if there used to be a file of this name, but that ID has now
276
# disappeared, it is deleted
277
if old_id and not new_inv.has_id(old_id):
284
def find_renames(old_inv, new_inv):
285
for file_id in old_inv:
286
if file_id not in new_inv:
288
old_name = old_inv.id2path(file_id)
289
new_name = new_inv.id2path(file_id)
290
if old_name != new_name:
291
yield (old_name, new_name)