1
# (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
"""Inventories map files to their name in a revision."""
19
# TODO: Maybe store inventory_id in the file? Not really needed.
21
__copyright__ = "Copyright (C) 2005 Canonical Ltd."
22
__author__ = "Martin Pool <mbp@canonical.com>"
24
import sys, os.path, types
28
from cElementTree import Element, ElementTree, SubElement
30
from elementtree.ElementTree import Element, ElementTree, SubElement
32
from xml import XMLMixin
33
from errors import bailout
36
from bzrlib.osutils import uuid, quotefn, splitpath, joinpath, appendpath
37
from bzrlib.trace import mutter
39
class InventoryEntry(XMLMixin):
40
"""Description of a versioned file.
42
An InventoryEntry has the following fields, which are also
43
present in the XML inventory-entry element:
46
* *name*: (only the basename within the directory, must not
48
* *kind*: "directory" or "file"
49
* *directory_id*: (if absent/null means the branch root directory)
50
* *text_sha1*: only for files
51
* *text_size*: in bytes, only for files
52
* *text_id*: identifier for the text version, only for files
54
InventoryEntries can also exist inside a WorkingTree
55
inventory, in which case they are not yet bound to a
56
particular revision of the file. In that case the text_sha1,
57
text_size and text_id are absent.
62
>>> i.add(InventoryEntry('123', 'src', kind='directory'))
63
>>> i.add(InventoryEntry('2323', 'hello.c', parent_id='123'))
64
>>> for j in i.iter_entries():
67
('src', InventoryEntry('123', 'src', kind='directory', parent_id=None))
68
('src/hello.c', InventoryEntry('2323', 'hello.c', kind='file', parent_id='123'))
69
>>> i.add(InventoryEntry('2323', 'bye.c', parent_id='123'))
70
Traceback (most recent call last):
72
BzrError: ('inventory already contains entry with id {2323}', [])
73
>>> i.add(InventoryEntry('2324', 'bye.c', parent_id='123'))
74
>>> i.add(InventoryEntry('2325', 'wibble', parent_id='123', kind='directory'))
75
>>> i.path2id('src/wibble')
79
>>> i.add(InventoryEntry('2326', 'wibble.c', parent_id='2325'))
81
InventoryEntry('2326', 'wibble.c', kind='file', parent_id='2325')
82
>>> for j in i.iter_entries():
84
... assert i.path2id(j[0])
94
:todo: Maybe also keep the full path of the entry, and the children?
95
But those depend on its position within a particular inventory, and
96
it would be nice not to need to hold the backpointer here.
98
def __init__(self, file_id, name, kind='file', text_id=None,
100
"""Create an InventoryEntry
102
The filename must be a single component, relative to the
103
parent directory; it cannot be a whole path or relative name.
105
>>> e = InventoryEntry('123', 'hello.c')
110
>>> e = InventoryEntry('123', 'src/hello.c')
111
Traceback (most recent call last):
112
BzrError: ("InventoryEntry name is not a simple filename: 'src/hello.c'", [])
115
if len(splitpath(name)) != 1:
116
bailout('InventoryEntry name is not a simple filename: %r'
119
self.file_id = file_id
121
assert kind in ['file', 'directory']
123
self.text_id = text_id
124
self.parent_id = parent_id
125
self.text_sha1 = None
126
self.text_size = None
127
if kind == 'directory':
132
other = InventoryEntry(self.file_id, self.name, self.kind,
133
self.text_id, self.parent_id)
134
other.text_sha1 = self.text_sha1
135
other.text_size = self.text_size
140
return ("%s(%r, %r, kind=%r, parent_id=%r)"
141
% (self.__class__.__name__,
148
def to_element(self):
149
"""Convert to XML element"""
152
e.set('name', self.name)
153
e.set('file_id', self.file_id)
154
e.set('kind', self.kind)
156
if self.text_size is not None:
157
e.set('text_size', '%d' % self.text_size)
159
for f in ['text_id', 'text_sha1', 'parent_id']:
169
def from_element(cls, elt):
170
assert elt.tag == 'entry'
171
self = cls(elt.get('file_id'), elt.get('name'), elt.get('kind'))
172
self.text_id = elt.get('text_id')
173
self.text_sha1 = elt.get('text_sha1')
174
self.parent_id = elt.get('parent_id')
176
## mutter("read inventoryentry: %r" % (elt.attrib))
178
v = elt.get('text_size')
179
self.text_size = v and int(v)
184
from_element = classmethod(from_element)
186
def __cmp__(self, other):
189
if not isinstance(other, InventoryEntry):
190
return NotImplemented
192
return cmp(self.file_id, other.file_id) \
193
or cmp(self.name, other.name) \
194
or cmp(self.text_sha1, other.text_sha1) \
195
or cmp(self.text_size, other.text_size) \
196
or cmp(self.text_id, other.text_id) \
197
or cmp(self.parent_id, other.parent_id) \
198
or cmp(self.kind, other.kind)
202
class Inventory(XMLMixin):
203
"""Inventory of versioned files in a tree.
205
An Inventory acts like a set of InventoryEntry items. You can
206
also look files up by their file_id or name.
208
May be read from and written to a metadata file in a tree. To
209
manipulate the inventory (for example to add a file), it is read
210
in, modified, and then written back out.
212
The inventory represents a typical unix file tree, with
213
directories containing files and subdirectories. We never store
214
the full path to a file, because renaming a directory implicitly
215
moves all of its contents. This class internally maintains a
216
lookup tree that allows the children under a directory to be
219
InventoryEntry objects must not be modified after they are
222
>>> inv = Inventory()
223
>>> inv.write_xml(sys.stdout)
226
>>> inv.add(InventoryEntry('123-123', 'hello.c'))
227
>>> inv['123-123'].name
230
May be treated as an iterator or set to look up file ids:
232
>>> bool(inv.path2id('hello.c'))
237
May also look up by name:
239
>>> [x[0] for x in inv.iter_entries()]
242
>>> inv.write_xml(sys.stdout)
244
<entry file_id="123-123" kind="file" name="hello.c" />
249
## TODO: Make sure only canonical filenames are stored.
251
## TODO: Do something sensible about the possible collisions on
252
## case-losing filesystems. Perhaps we should just always forbid
255
## TODO: No special cases for root, rather just give it a file id
256
## like everything else.
258
## TODO: Probably change XML serialization to use nesting
261
"""Create or read an inventory.
263
If a working directory is specified, the inventory is read
264
from there. If the file is specified, read from that. If not,
265
the inventory is created empty.
267
self._root = InventoryEntry(None, '', kind='directory')
268
self._byid = {None: self._root}
272
return iter(self._byid)
276
"""Returns number of entries."""
277
return len(self._byid)
280
def iter_entries(self, parent_id=None):
281
"""Return (path, entry) pairs, in order by name."""
282
kids = self[parent_id].children.items()
284
for name, ie in kids:
286
if ie.kind == 'directory':
287
for cn, cie in self.iter_entries(parent_id=ie.file_id):
288
yield joinpath([name, cn]), cie
291
def directories(self, include_root=True):
292
"""Return (path, entry) pairs for all directories.
296
for path, entry in self.iter_entries():
297
if entry.kind == 'directory':
302
def __contains__(self, file_id):
303
"""True if this entry contains a file with given id.
305
>>> inv = Inventory()
306
>>> inv.add(InventoryEntry('123', 'foo.c'))
312
return file_id in self._byid
315
def __getitem__(self, file_id):
316
"""Return the entry for given file_id.
318
>>> inv = Inventory()
319
>>> inv.add(InventoryEntry('123123', 'hello.c'))
320
>>> inv['123123'].name
323
return self._byid[file_id]
326
def get_child(self, parent_id, filename):
327
if parent_id == None:
328
return self._root.children.get(filename)
330
return self[parent_id].children.get(filename)
333
def add(self, entry):
334
"""Add entry to inventory.
336
To add a file to a branch ready to be committed, use Branch.add,
338
if entry.file_id in self:
339
bailout("inventory already contains entry with id {%s}" % entry.file_id)
341
if entry.parent_id != None:
342
if entry.parent_id not in self:
343
bailout("parent_id %s of new entry not found in inventory"
345
# TODO: parent must be a directory
347
if self[entry.parent_id].children.has_key(entry.name):
348
bailout("%s is already versioned"
349
% appendpath(self.id2path(entry.parent_id), entry.name))
351
self._byid[entry.file_id] = entry
352
self[entry.parent_id].children[entry.name] = entry
355
def add_path(self, relpath, kind, file_id=None):
356
"""Add entry from a path.
358
The immediate parent must already be versioned"""
359
parts = bzrlib.osutils.splitpath(relpath)
361
bailout("cannot re-add root of inventory")
364
file_id = bzrlib.branch.gen_file_id(relpath)
366
parent_id = self.path2id(parts[:-1])
367
ie = InventoryEntry(file_id, parts[-1],
368
kind=kind, parent_id=parent_id)
372
def __delitem__(self, file_id):
373
"""Remove entry by id.
375
>>> inv = Inventory()
376
>>> inv.add(InventoryEntry('123', 'foo.c'))
385
assert self[ie.parent_id].children[ie.name] == ie
387
# TODO: Test deleting all children; maybe hoist to a separate
389
if ie.kind == 'directory':
390
for cie in ie.children.values():
391
del self[cie.file_id]
394
del self._byid[file_id]
395
del self[ie.parent_id].children[ie.name]
399
return Set(self._byid)
402
def to_element(self):
403
"""Convert to XML Element"""
404
e = Element('inventory')
406
for path, ie in self.iter_entries():
407
e.append(ie.to_element())
411
def from_element(cls, elt):
412
"""Construct from XML Element
414
>>> inv = Inventory()
415
>>> inv.add(InventoryEntry('foo.c-123981239', 'foo.c'))
416
>>> elt = inv.to_element()
417
>>> inv2 = Inventory.from_element(elt)
421
assert elt.tag == 'inventory'
424
o.add(InventoryEntry.from_element(e))
427
from_element = classmethod(from_element)
430
def __cmp__(self, other):
431
"""Compare two sets by comparing their contents.
437
>>> i1.add(InventoryEntry('123', 'foo'))
440
>>> i2.add(InventoryEntry('123', 'foo'))
447
if not isinstance(other, Inventory):
448
return NotImplemented
450
if self.id_set() ^ other.id_set():
453
for file_id in self._byid:
454
c = cmp(self[file_id], other[file_id])
460
def id2path(self, file_id):
461
"""Return as a list the path to file_id."""
463
while file_id != None:
466
file_id = ie.parent_id
471
def path2id(self, name):
472
"""Walk down through directories to return entry of last component.
474
names may be either a list of path components, or a single
475
string, in which case it is automatically split.
477
This returns the entry of the last component in the path,
478
which may be either a file or a directory.
480
if isinstance(name, types.StringTypes):
481
name = splitpath(name)
486
cie = parent.children[f]
493
return parent.file_id
496
def has_filename(self, names):
497
return bool(self.path2id(names))
500
def has_id(self, file_id):
501
assert isinstance(file_id, str)
502
return self._byid.has_key(file_id)
508
if __name__ == '__main__':
509
import doctest, inventory
510
doctest.testmod(inventory)
added
removed
bzrlib/inventory.py
