« back to all changes in this revision
Viewing changes to bzrlib/inventory.py
- browse files at revision 625
- compare with another revision
- download diff
- download tarball
- view history from revision 625
- Committer: Martin Pool
- Date: 2005-06-06 05:55:19 UTC
- Revision ID: mbp@sourcefrog.net-20050606055519-2fa201b47cefec08
- fix permissions on exported tar/zip files
- files added:
- .bzrignore
- bzrlib
- contrib
- contrib/bash
- contrib/zsh
- doc
- elementtree
- notes
- patches
- urlgrabber
- files removed:
- .bzrignore
added
removed
bzrlib/inventory.py
1
# (C) 2005 Canonical Ltd
2
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.
7
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.
12
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
16
17
18
# This should really be an id randomly assigned when the tree is
19
# created, but it's not for now.
20
ROOT_ID = "TREE_ROOT"
21
22
23
import sys, os.path, types, re
24
25
try:
26
from cElementTree import Element, ElementTree, SubElement
27
except ImportError:
28
from elementtree.ElementTree import Element, ElementTree, SubElement
29
30
from xml import XMLMixin
31
from errors import bailout, BzrError, BzrCheckError
32
33
import bzrlib
34
from bzrlib.osutils import uuid, quotefn, splitpath, joinpath, appendpath
35
from bzrlib.trace import mutter
36
37
class InventoryEntry(XMLMixin):
38
"""Description of a versioned file.
39
40
An InventoryEntry has the following fields, which are also
41
present in the XML inventory-entry element:
42
43
* *file_id*
44
* *name*: (only the basename within the directory, must not
45
contain slashes)
46
* *kind*: "directory" or "file"
47
* *directory_id*: (if absent/null means the branch root directory)
48
* *text_sha1*: only for files
49
* *text_size*: in bytes, only for files
50
* *text_id*: identifier for the text version, only for files
51
52
InventoryEntries can also exist inside a WorkingTree
53
inventory, in which case they are not yet bound to a
54
particular revision of the file. In that case the text_sha1,
55
text_size and text_id are absent.
56
57
58
>>> i = Inventory()
59
>>> i.path2id('')
60
'TREE_ROOT'
61
>>> i.add(InventoryEntry('123', 'src', 'directory', ROOT_ID))
62
>>> i.add(InventoryEntry('2323', 'hello.c', 'file', parent_id='123'))
63
>>> for j in i.iter_entries():
64
... print j
65
...
66
('src', InventoryEntry('123', 'src', kind='directory', parent_id='TREE_ROOT'))
67
('src/hello.c', InventoryEntry('2323', 'hello.c', kind='file', parent_id='123'))
68
>>> i.add(InventoryEntry('2323', 'bye.c', 'file', '123'))
69
Traceback (most recent call last):
70
...
71
BzrError: ('inventory already contains entry with id {2323}', [])
72
>>> i.add(InventoryEntry('2324', 'bye.c', 'file', '123'))
73
>>> i.add(InventoryEntry('2325', 'wibble', 'directory', '123'))
74
>>> i.path2id('src/wibble')
75
'2325'
76
>>> '2325' in i
77
True
78
>>> i.add(InventoryEntry('2326', 'wibble.c', 'file', '2325'))
79
>>> i['2326']
80
InventoryEntry('2326', 'wibble.c', kind='file', parent_id='2325')
81
>>> for j in i.iter_entries():
82
... print j[0]
83
... assert i.path2id(j[0])
84
...
85
src
86
src/bye.c
87
src/hello.c
88
src/wibble
89
src/wibble/wibble.c
90
>>> i.id2path('2326')
91
'src/wibble/wibble.c'
92
93
TODO: Maybe also keep the full path of the entry, and the children?
94
But those depend on its position within a particular inventory, and
95
it would be nice not to need to hold the backpointer here.
96
"""
97
98
# TODO: split InventoryEntry into subclasses for files,
99
# directories, etc etc.
100
101
text_sha1 = None
102
text_size = None
103
104
def __init__(self, file_id, name, kind, parent_id, text_id=None):
105
"""Create an InventoryEntry
106
107
The filename must be a single component, relative to the
108
parent directory; it cannot be a whole path or relative name.
109
110
>>> e = InventoryEntry('123', 'hello.c', 'file', ROOT_ID)
111
>>> e.name
112
'hello.c'
113
>>> e.file_id
114
'123'
115
>>> e = InventoryEntry('123', 'src/hello.c', 'file', ROOT_ID)
116
Traceback (most recent call last):
117
BzrCheckError: InventoryEntry name 'src/hello.c' is invalid
118
"""
119
if '/' in name or '\\' in name:
120
raise BzrCheckError('InventoryEntry name %r is invalid' % name)
121
122
self.file_id = file_id
123
self.name = name
124
self.kind = kind
125
self.text_id = text_id
126
self.parent_id = parent_id
127
if kind == 'directory':
128
self.children = {}
129
elif kind == 'file':
130
pass
131
else:
132
raise BzrError("unhandled entry kind %r" % kind)
133
134
135
136
def sorted_children(self):
137
l = self.children.items()
138
l.sort()
139
return l
140
141
142
def copy(self):
143
other = InventoryEntry(self.file_id, self.name, self.kind,
144
self.parent_id, text_id=self.text_id)
145
other.text_sha1 = self.text_sha1
146
other.text_size = self.text_size
147
# note that children are *not* copied; they're pulled across when
148
# others are added
149
return other
150
151
152
def __repr__(self):
153
return ("%s(%r, %r, kind=%r, parent_id=%r)"
154
% (self.__class__.__name__,
155
self.file_id,
156
self.name,
157
self.kind,
158
self.parent_id))
159
160
161
def to_element(self):
162
"""Convert to XML element"""
163
e = Element('entry')
164
165
e.set('name', self.name)
166
e.set('file_id', self.file_id)
167
e.set('kind', self.kind)
168
169
if self.text_size != None:
170
e.set('text_size', '%d' % self.text_size)
171
172
for f in ['text_id', 'text_sha1']:
173
v = getattr(self, f)
174
if v != None:
175
e.set(f, v)
176
177
# to be conservative, we don't externalize the root pointers
178
# for now, leaving them as null in the xml form. in a future
179
# version it will be implied by nested elements.
180
if self.parent_id != ROOT_ID:
181
assert isinstance(self.parent_id, basestring)
182
e.set('parent_id', self.parent_id)
183
184
e.tail = '\n'
185
186
return e
187
188
189
def from_element(cls, elt):
190
assert elt.tag == 'entry'
191
192
## original format inventories don't have a parent_id for
193
## nodes in the root directory, but it's cleaner to use one
194
## internally.
195
parent_id = elt.get('parent_id')
196
if parent_id == None:
197
parent_id = ROOT_ID
198
199
self = cls(elt.get('file_id'), elt.get('name'), elt.get('kind'), parent_id)
200
self.text_id = elt.get('text_id')
201
self.text_sha1 = elt.get('text_sha1')
202
203
## mutter("read inventoryentry: %r" % (elt.attrib))
204
205
v = elt.get('text_size')
206
self.text_size = v and int(v)
207
208
return self
209
210
211
from_element = classmethod(from_element)
212
213
def __eq__(self, other):
214
if not isinstance(other, InventoryEntry):
215
return NotImplemented
216
217
return (self.file_id == other.file_id) \
218
and (self.name == other.name) \
219
and (self.text_sha1 == other.text_sha1) \
220
and (self.text_size == other.text_size) \
221
and (self.text_id == other.text_id) \
222
and (self.parent_id == other.parent_id) \
223
and (self.kind == other.kind)
224
225
226
def __ne__(self, other):
227
return not (self == other)
228
229
def __hash__(self):
230
raise ValueError('not hashable')
231
232
233
234
class RootEntry(InventoryEntry):
235
def __init__(self, file_id):
236
self.file_id = file_id
237
self.children = {}
238
self.kind = 'root_directory'
239
self.parent_id = None
240
self.name = ''
241
242
def __eq__(self, other):
243
if not isinstance(other, RootEntry):
244
return NotImplemented
245
246
return (self.file_id == other.file_id) \
247
and (self.children == other.children)
248
249
250
251
class Inventory(XMLMixin):
252
"""Inventory of versioned files in a tree.
253
254
This describes which file_id is present at each point in the tree,
255
and possibly the SHA-1 or other information about the file.
256
Entries can be looked up either by path or by file_id.
257
258
The inventory represents a typical unix file tree, with
259
directories containing files and subdirectories. We never store
260
the full path to a file, because renaming a directory implicitly
261
moves all of its contents. This class internally maintains a
262
lookup tree that allows the children under a directory to be
263
returned quickly.
264
265
InventoryEntry objects must not be modified after they are
266
inserted, other than through the Inventory API.
267
268
>>> inv = Inventory()
269
>>> inv.write_xml(sys.stdout)
270
<inventory>
271
</inventory>
272
>>> inv.add(InventoryEntry('123-123', 'hello.c', 'file', ROOT_ID))
273
>>> inv['123-123'].name
274
'hello.c'
275
276
May be treated as an iterator or set to look up file ids:
277
278
>>> bool(inv.path2id('hello.c'))
279
True
280
>>> '123-123' in inv
281
True
282
283
May also look up by name:
284
285
>>> [x[0] for x in inv.iter_entries()]
286
['hello.c']
287
288
>>> inv.write_xml(sys.stdout)
289
<inventory>
290
<entry file_id="123-123" kind="file" name="hello.c" />
291
</inventory>
292
293
"""
294
def __init__(self):
295
"""Create or read an inventory.
296
297
If a working directory is specified, the inventory is read
298
from there. If the file is specified, read from that. If not,
299
the inventory is created empty.
300
301
The inventory is created with a default root directory, with
302
an id of None.
303
"""
304
self.root = RootEntry(ROOT_ID)
305
self._byid = {self.root.file_id: self.root}
306
307
308
def __iter__(self):
309
return iter(self._byid)
310
311
312
def __len__(self):
313
"""Returns number of entries."""
314
return len(self._byid)
315
316
317
def iter_entries(self, from_dir=None):
318
"""Return (path, entry) pairs, in order by name."""
319
if from_dir == None:
320
assert self.root
321
from_dir = self.root
322
elif isinstance(from_dir, basestring):
323
from_dir = self._byid[from_dir]
324
325
kids = from_dir.children.items()
326
kids.sort()
327
for name, ie in kids:
328
yield name, ie
329
if ie.kind == 'directory':
330
for cn, cie in self.iter_entries(from_dir=ie.file_id):
331
yield os.path.join(name, cn), cie
332
333
334
def entries(self):
335
"""Return list of (path, ie) for all entries except the root.
336
337
This may be faster than iter_entries.
338
"""
339
accum = []
340
def descend(dir_ie, dir_path):
341
kids = dir_ie.children.items()
342
kids.sort()
343
for name, ie in kids:
344
child_path = os.path.join(dir_path, name)
345
accum.append((child_path, ie))
346
if ie.kind == 'directory':
347
descend(ie, child_path)
348
349
descend(self.root, '')
350
return accum
351
352
353
def directories(self):
354
"""Return (path, entry) pairs for all directories, including the root.
355
"""
356
accum = []
357
def descend(parent_ie, parent_path):
358
accum.append((parent_path, parent_ie))
359
360
kids = [(ie.name, ie) for ie in parent_ie.children.itervalues() if ie.kind == 'directory']
361
kids.sort()
362
363
for name, child_ie in kids:
364
child_path = os.path.join(parent_path, name)
365
descend(child_ie, child_path)
366
descend(self.root, '')
367
return accum
368
369
370
371
def __contains__(self, file_id):
372
"""True if this entry contains a file with given id.
373
374
>>> inv = Inventory()
375
>>> inv.add(InventoryEntry('123', 'foo.c', 'file', ROOT_ID))
376
>>> '123' in inv
377
True
378
>>> '456' in inv
379
False
380
"""
381
return file_id in self._byid
382
383
384
def __getitem__(self, file_id):
385
"""Return the entry for given file_id.
386
387
>>> inv = Inventory()
388
>>> inv.add(InventoryEntry('123123', 'hello.c', 'file', ROOT_ID))
389
>>> inv['123123'].name
390
'hello.c'
391
"""
392
try:
393
return self._byid[file_id]
394
except KeyError:
395
if file_id == None:
396
raise BzrError("can't look up file_id None")
397
else:
398
raise BzrError("file_id {%s} not in inventory" % file_id)
399
400
401
def get_file_kind(self, file_id):
402
return self._byid[file_id].kind
403
404
def get_child(self, parent_id, filename):
405
return self[parent_id].children.get(filename)
406
407
408
def add(self, entry):
409
"""Add entry to inventory.
410
411
To add a file to a branch ready to be committed, use Branch.add,
412
which calls this."""
413
if entry.file_id in self._byid:
414
bailout("inventory already contains entry with id {%s}" % entry.file_id)
415
416
try:
417
parent = self._byid[entry.parent_id]
418
except KeyError:
419
bailout("parent_id {%s} not in inventory" % entry.parent_id)
420
421
if parent.children.has_key(entry.name):
422
bailout("%s is already versioned" %
423
appendpath(self.id2path(parent.file_id), entry.name))
424
425
self._byid[entry.file_id] = entry
426
parent.children[entry.name] = entry
427
428
429
def add_path(self, relpath, kind, file_id=None):
430
"""Add entry from a path.
431
432
The immediate parent must already be versioned"""
433
parts = bzrlib.osutils.splitpath(relpath)
434
if len(parts) == 0:
435
bailout("cannot re-add root of inventory")
436
437
if file_id == None:
438
file_id = bzrlib.branch.gen_file_id(relpath)
439
440
parent_id = self.path2id(parts[:-1])
441
assert parent_id != None
442
ie = InventoryEntry(file_id, parts[-1],
443
kind=kind, parent_id=parent_id)
444
return self.add(ie)
445
446
447
def __delitem__(self, file_id):
448
"""Remove entry by id.
449
450
>>> inv = Inventory()
451
>>> inv.add(InventoryEntry('123', 'foo.c', 'file', ROOT_ID))
452
>>> '123' in inv
453
True
454
>>> del inv['123']
455
>>> '123' in inv
456
False
457
"""
458
ie = self[file_id]
459
460
assert self[ie.parent_id].children[ie.name] == ie
461
462
# TODO: Test deleting all children; maybe hoist to a separate
463
# deltree method?
464
if ie.kind == 'directory':
465
for cie in ie.children.values():
466
del self[cie.file_id]
467
del ie.children
468
469
del self._byid[file_id]
470
del self[ie.parent_id].children[ie.name]
471
472
473
def to_element(self):
474
"""Convert to XML Element"""
475
e = Element('inventory')
476
e.text = '\n'
477
for path, ie in self.iter_entries():
478
e.append(ie.to_element())
479
return e
480
481
482
def from_element(cls, elt):
483
"""Construct from XML Element
484
485
>>> inv = Inventory()
486
>>> inv.add(InventoryEntry('foo.c-123981239', 'foo.c', 'file', ROOT_ID))
487
>>> elt = inv.to_element()
488
>>> inv2 = Inventory.from_element(elt)
489
>>> inv2 == inv
490
True
491
"""
492
assert elt.tag == 'inventory'
493
o = cls()
494
for e in elt:
495
o.add(InventoryEntry.from_element(e))
496
return o
497
498
from_element = classmethod(from_element)
499
500
501
def __eq__(self, other):
502
"""Compare two sets by comparing their contents.
503
504
>>> i1 = Inventory()
505
>>> i2 = Inventory()
506
>>> i1 == i2
507
True
508
>>> i1.add(InventoryEntry('123', 'foo', 'file', ROOT_ID))
509
>>> i1 == i2
510
False
511
>>> i2.add(InventoryEntry('123', 'foo', 'file', ROOT_ID))
512
>>> i1 == i2
513
True
514
"""
515
if not isinstance(other, Inventory):
516
return NotImplemented
517
518
if len(self._byid) != len(other._byid):
519
# shortcut: obviously not the same
520
return False
521
522
return self._byid == other._byid
523
524
525
def __ne__(self, other):
526
return not (self == other)
527
528
529
def __hash__(self):
530
raise ValueError('not hashable')
531
532
533
534
def get_idpath(self, file_id):
535
"""Return a list of file_ids for the path to an entry.
536
537
The list contains one element for each directory followed by
538
the id of the file itself. So the length of the returned list
539
is equal to the depth of the file in the tree, counting the
540
root directory as depth 1.
541
"""
542
p = []
543
while file_id != None:
544
try:
545
ie = self._byid[file_id]
546
except KeyError:
547
bailout("file_id {%s} not found in inventory" % file_id)
548
p.insert(0, ie.file_id)
549
file_id = ie.parent_id
550
return p
551
552
553
def id2path(self, file_id):
554
"""Return as a list the path to file_id."""
555
556
# get all names, skipping root
557
p = [self[fid].name for fid in self.get_idpath(file_id)[1:]]
558
return os.sep.join(p)
559
560
561
562
def path2id(self, name):
563
"""Walk down through directories to return entry of last component.
564
565
names may be either a list of path components, or a single
566
string, in which case it is automatically split.
567
568
This returns the entry of the last component in the path,
569
which may be either a file or a directory.
570
571
Returns None iff the path is not found.
572
"""
573
if isinstance(name, types.StringTypes):
574
name = splitpath(name)
575
576
mutter("lookup path %r" % name)
577
578
parent = self.root
579
for f in name:
580
try:
581
cie = parent.children[f]
582
assert cie.name == f
583
assert cie.parent_id == parent.file_id
584
parent = cie
585
except KeyError:
586
# or raise an error?
587
return None
588
589
return parent.file_id
590
591
592
def has_filename(self, names):
593
return bool(self.path2id(names))
594
595
596
def has_id(self, file_id):
597
return self._byid.has_key(file_id)
598
599
600
def rename(self, file_id, new_parent_id, new_name):
601
"""Move a file within the inventory.
602
603
This can change either the name, or the parent, or both.
604
605
This does not move the working file."""
606
if not is_valid_name(new_name):
607
bailout("not an acceptable filename: %r" % new_name)
608
609
new_parent = self._byid[new_parent_id]
610
if new_name in new_parent.children:
611
bailout("%r already exists in %r" % (new_name, self.id2path(new_parent_id)))
612
613
new_parent_idpath = self.get_idpath(new_parent_id)
614
if file_id in new_parent_idpath:
615
bailout("cannot move directory %r into a subdirectory of itself, %r"
616
% (self.id2path(file_id), self.id2path(new_parent_id)))
617
618
file_ie = self._byid[file_id]
619
old_parent = self._byid[file_ie.parent_id]
620
621
# TODO: Don't leave things messed up if this fails
622
623
del old_parent.children[file_ie.name]
624
new_parent.children[new_name] = file_ie
625
626
file_ie.name = new_name
627
file_ie.parent_id = new_parent_id
628
629
630
631
632
_NAME_RE = re.compile(r'^[^/\\]+$')
633
634
def is_valid_name(name):
635
return bool(_NAME_RE.match(name))
Loggerhead is a web-based interface for Breezy
Version: 2.0.1