« back to all changes in this revision
Viewing changes to bzrlib/inventory.py
- browse files at revision 542
- compare with another revision
- download diff
- download tarball
- view history from revision 542
- Committer: Martin Pool
- Date: 2005-05-19 08:31:06 UTC
- Revision ID: mbp@sourcefrog.net-20050519083106-ebe71562d3bda4a7
- fix typo
- 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 __cmp__(self, other):
214
if self is other:
215
return 0
216
if not isinstance(other, InventoryEntry):
217
return NotImplemented
218
219
return cmp(self.file_id, other.file_id) \
220
or cmp(self.name, other.name) \
221
or cmp(self.text_sha1, other.text_sha1) \
222
or cmp(self.text_size, other.text_size) \
223
or cmp(self.text_id, other.text_id) \
224
or cmp(self.parent_id, other.parent_id) \
225
or cmp(self.kind, other.kind)
226
227
228
229
class RootEntry(InventoryEntry):
230
def __init__(self, file_id):
231
self.file_id = file_id
232
self.children = {}
233
self.kind = 'root_directory'
234
self.parent_id = None
235
self.name = ''
236
237
def __cmp__(self, other):
238
if self is other:
239
return 0
240
if not isinstance(other, RootEntry):
241
return NotImplemented
242
return cmp(self.file_id, other.file_id) \
243
or cmp(self.children, other.children)
244
245
246
247
class Inventory(XMLMixin):
248
"""Inventory of versioned files in a tree.
249
250
This describes which file_id is present at each point in the tree,
251
and possibly the SHA-1 or other information about the file.
252
Entries can be looked up either by path or by file_id.
253
254
The inventory represents a typical unix file tree, with
255
directories containing files and subdirectories. We never store
256
the full path to a file, because renaming a directory implicitly
257
moves all of its contents. This class internally maintains a
258
lookup tree that allows the children under a directory to be
259
returned quickly.
260
261
InventoryEntry objects must not be modified after they are
262
inserted, other than through the Inventory API.
263
264
>>> inv = Inventory()
265
>>> inv.write_xml(sys.stdout)
266
<inventory>
267
</inventory>
268
>>> inv.add(InventoryEntry('123-123', 'hello.c', 'file', ROOT_ID))
269
>>> inv['123-123'].name
270
'hello.c'
271
272
May be treated as an iterator or set to look up file ids:
273
274
>>> bool(inv.path2id('hello.c'))
275
True
276
>>> '123-123' in inv
277
True
278
279
May also look up by name:
280
281
>>> [x[0] for x in inv.iter_entries()]
282
['hello.c']
283
284
>>> inv.write_xml(sys.stdout)
285
<inventory>
286
<entry file_id="123-123" kind="file" name="hello.c" />
287
</inventory>
288
289
"""
290
def __init__(self):
291
"""Create or read an inventory.
292
293
If a working directory is specified, the inventory is read
294
from there. If the file is specified, read from that. If not,
295
the inventory is created empty.
296
297
The inventory is created with a default root directory, with
298
an id of None.
299
"""
300
self.root = RootEntry(ROOT_ID)
301
self._byid = {self.root.file_id: self.root}
302
303
304
def __iter__(self):
305
return iter(self._byid)
306
307
308
def __len__(self):
309
"""Returns number of entries."""
310
return len(self._byid)
311
312
313
def iter_entries(self, from_dir=None):
314
"""Return (path, entry) pairs, in order by name."""
315
if from_dir == None:
316
assert self.root
317
from_dir = self.root
318
elif isinstance(from_dir, basestring):
319
from_dir = self._byid[from_dir]
320
321
kids = from_dir.children.items()
322
kids.sort()
323
for name, ie in kids:
324
yield name, ie
325
if ie.kind == 'directory':
326
for cn, cie in self.iter_entries(from_dir=ie.file_id):
327
yield os.path.join(name, cn), cie
328
329
330
331
def directories(self):
332
"""Return (path, entry) pairs for all directories.
333
"""
334
def descend(parent_ie):
335
parent_name = parent_ie.name
336
yield parent_name, parent_ie
337
338
# directory children in sorted order
339
dn = []
340
for ie in parent_ie.children.itervalues():
341
if ie.kind == 'directory':
342
dn.append((ie.name, ie))
343
dn.sort()
344
345
for name, child_ie in dn:
346
for sub_name, sub_ie in descend(child_ie):
347
yield appendpath(parent_name, sub_name), sub_ie
348
349
for name, ie in descend(self.root):
350
yield name, ie
351
352
353
354
def __contains__(self, file_id):
355
"""True if this entry contains a file with given id.
356
357
>>> inv = Inventory()
358
>>> inv.add(InventoryEntry('123', 'foo.c', 'file', ROOT_ID))
359
>>> '123' in inv
360
True
361
>>> '456' in inv
362
False
363
"""
364
return file_id in self._byid
365
366
367
def __getitem__(self, file_id):
368
"""Return the entry for given file_id.
369
370
>>> inv = Inventory()
371
>>> inv.add(InventoryEntry('123123', 'hello.c', 'file', ROOT_ID))
372
>>> inv['123123'].name
373
'hello.c'
374
"""
375
try:
376
return self._byid[file_id]
377
except KeyError:
378
if file_id == None:
379
raise BzrError("can't look up file_id None")
380
else:
381
raise BzrError("file_id {%s} not in inventory" % file_id)
382
383
384
def get_file_kind(self, file_id):
385
return self._byid[file_id].kind
386
387
def get_child(self, parent_id, filename):
388
return self[parent_id].children.get(filename)
389
390
391
def add(self, entry):
392
"""Add entry to inventory.
393
394
To add a file to a branch ready to be committed, use Branch.add,
395
which calls this."""
396
if entry.file_id in self._byid:
397
bailout("inventory already contains entry with id {%s}" % entry.file_id)
398
399
try:
400
parent = self._byid[entry.parent_id]
401
except KeyError:
402
bailout("parent_id {%s} not in inventory" % entry.parent_id)
403
404
if parent.children.has_key(entry.name):
405
bailout("%s is already versioned" %
406
appendpath(self.id2path(parent.file_id), entry.name))
407
408
self._byid[entry.file_id] = entry
409
parent.children[entry.name] = entry
410
411
412
def add_path(self, relpath, kind, file_id=None):
413
"""Add entry from a path.
414
415
The immediate parent must already be versioned"""
416
parts = bzrlib.osutils.splitpath(relpath)
417
if len(parts) == 0:
418
bailout("cannot re-add root of inventory")
419
420
if file_id == None:
421
file_id = bzrlib.branch.gen_file_id(relpath)
422
423
parent_id = self.path2id(parts[:-1])
424
assert parent_id != None
425
ie = InventoryEntry(file_id, parts[-1],
426
kind=kind, parent_id=parent_id)
427
return self.add(ie)
428
429
430
def __delitem__(self, file_id):
431
"""Remove entry by id.
432
433
>>> inv = Inventory()
434
>>> inv.add(InventoryEntry('123', 'foo.c', 'file', ROOT_ID))
435
>>> '123' in inv
436
True
437
>>> del inv['123']
438
>>> '123' in inv
439
False
440
"""
441
ie = self[file_id]
442
443
assert self[ie.parent_id].children[ie.name] == ie
444
445
# TODO: Test deleting all children; maybe hoist to a separate
446
# deltree method?
447
if ie.kind == 'directory':
448
for cie in ie.children.values():
449
del self[cie.file_id]
450
del ie.children
451
452
del self._byid[file_id]
453
del self[ie.parent_id].children[ie.name]
454
455
456
def id_set(self):
457
from bzrlib import frozenset
458
return frozenset(self._byid)
459
460
461
def to_element(self):
462
"""Convert to XML Element"""
463
e = Element('inventory')
464
e.text = '\n'
465
for path, ie in self.iter_entries():
466
e.append(ie.to_element())
467
return e
468
469
470
def from_element(cls, elt):
471
"""Construct from XML Element
472
473
>>> inv = Inventory()
474
>>> inv.add(InventoryEntry('foo.c-123981239', 'foo.c', 'file', ROOT_ID))
475
>>> elt = inv.to_element()
476
>>> inv2 = Inventory.from_element(elt)
477
>>> inv2 == inv
478
True
479
"""
480
assert elt.tag == 'inventory'
481
o = cls()
482
for e in elt:
483
o.add(InventoryEntry.from_element(e))
484
return o
485
486
from_element = classmethod(from_element)
487
488
489
def __cmp__(self, other):
490
"""Compare two sets by comparing their contents.
491
492
>>> i1 = Inventory()
493
>>> i2 = Inventory()
494
>>> i1 == i2
495
True
496
>>> i1.add(InventoryEntry('123', 'foo', 'file', ROOT_ID))
497
>>> i1 == i2
498
False
499
>>> i2.add(InventoryEntry('123', 'foo', 'file', ROOT_ID))
500
>>> i1 == i2
501
True
502
"""
503
if self is other:
504
return 0
505
506
if not isinstance(other, Inventory):
507
return NotImplemented
508
509
if self.id_set() ^ other.id_set():
510
return 1
511
512
for file_id in self._byid:
513
c = cmp(self[file_id], other[file_id])
514
if c: return c
515
516
return 0
517
518
519
def get_idpath(self, file_id):
520
"""Return a list of file_ids for the path to an entry.
521
522
The list contains one element for each directory followed by
523
the id of the file itself. So the length of the returned list
524
is equal to the depth of the file in the tree, counting the
525
root directory as depth 1.
526
"""
527
p = []
528
while file_id != None:
529
try:
530
ie = self._byid[file_id]
531
except KeyError:
532
bailout("file_id {%s} not found in inventory" % file_id)
533
p.insert(0, ie.file_id)
534
file_id = ie.parent_id
535
return p
536
537
538
def id2path(self, file_id):
539
"""Return as a list the path to file_id."""
540
541
# get all names, skipping root
542
p = [self[fid].name for fid in self.get_idpath(file_id)[1:]]
543
return os.sep.join(p)
544
545
546
547
def path2id(self, name):
548
"""Walk down through directories to return entry of last component.
549
550
names may be either a list of path components, or a single
551
string, in which case it is automatically split.
552
553
This returns the entry of the last component in the path,
554
which may be either a file or a directory.
555
556
Returns None iff the path is not found.
557
"""
558
if isinstance(name, types.StringTypes):
559
name = splitpath(name)
560
561
mutter("lookup path %r" % name)
562
563
parent = self.root
564
for f in name:
565
try:
566
cie = parent.children[f]
567
assert cie.name == f
568
assert cie.parent_id == parent.file_id
569
parent = cie
570
except KeyError:
571
# or raise an error?
572
return None
573
574
return parent.file_id
575
576
577
def has_filename(self, names):
578
return bool(self.path2id(names))
579
580
581
def has_id(self, file_id):
582
return self._byid.has_key(file_id)
583
584
585
def rename(self, file_id, new_parent_id, new_name):
586
"""Move a file within the inventory.
587
588
This can change either the name, or the parent, or both.
589
590
This does not move the working file."""
591
if not is_valid_name(new_name):
592
bailout("not an acceptable filename: %r" % new_name)
593
594
new_parent = self._byid[new_parent_id]
595
if new_name in new_parent.children:
596
bailout("%r already exists in %r" % (new_name, self.id2path(new_parent_id)))
597
598
new_parent_idpath = self.get_idpath(new_parent_id)
599
if file_id in new_parent_idpath:
600
bailout("cannot move directory %r into a subdirectory of itself, %r"
601
% (self.id2path(file_id), self.id2path(new_parent_id)))
602
603
file_ie = self._byid[file_id]
604
old_parent = self._byid[file_ie.parent_id]
605
606
# TODO: Don't leave things messed up if this fails
607
608
del old_parent.children[file_ie.name]
609
new_parent.children[new_name] = file_ie
610
611
file_ie.name = new_name
612
file_ie.parent_id = new_parent_id
613
614
615
616
617
_NAME_RE = re.compile(r'^[^/\\]+$')
618
619
def is_valid_name(name):
620
return bool(_NAME_RE.match(name))
Loggerhead is a web-based interface for Breezy
Version: 2.0.1