« back to all changes in this revision
Viewing changes to bzrlib/inventory.py
- browse files at revision 520
- compare with another revision
- download diff
- download tarball
- view history from revision 520
- Committer: Martin Pool
- Date: 2005-05-17 06:56:16 UTC
- Revision ID: mbp@sourcefrog.net-20050517065616-6f23381d6184a8aa
- add space for un-merged patches
- 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
from sets import Set
25
26
try:
27
from cElementTree import Element, ElementTree, SubElement
28
except ImportError:
29
from elementtree.ElementTree import Element, ElementTree, SubElement
30
31
from xml import XMLMixin
32
from errors import bailout, BzrError, BzrCheckError
33
34
import bzrlib
35
from bzrlib.osutils import uuid, quotefn, splitpath, joinpath, appendpath
36
from bzrlib.trace import mutter
37
38
class InventoryEntry(XMLMixin):
39
"""Description of a versioned file.
40
41
An InventoryEntry has the following fields, which are also
42
present in the XML inventory-entry element:
43
44
* *file_id*
45
* *name*: (only the basename within the directory, must not
46
contain slashes)
47
* *kind*: "directory" or "file"
48
* *directory_id*: (if absent/null means the branch root directory)
49
* *text_sha1*: only for files
50
* *text_size*: in bytes, only for files
51
* *text_id*: identifier for the text version, only for files
52
53
InventoryEntries can also exist inside a WorkingTree
54
inventory, in which case they are not yet bound to a
55
particular revision of the file. In that case the text_sha1,
56
text_size and text_id are absent.
57
58
59
>>> i = Inventory()
60
>>> i.path2id('')
61
'TREE_ROOT'
62
>>> i.add(InventoryEntry('123', 'src', 'directory', ROOT_ID))
63
>>> i.add(InventoryEntry('2323', 'hello.c', 'file', parent_id='123'))
64
>>> for j in i.iter_entries():
65
... print j
66
...
67
('src', InventoryEntry('123', 'src', kind='directory', parent_id='TREE_ROOT'))
68
('src/hello.c', InventoryEntry('2323', 'hello.c', kind='file', parent_id='123'))
69
>>> i.add(InventoryEntry('2323', 'bye.c', 'file', '123'))
70
Traceback (most recent call last):
71
...
72
BzrError: ('inventory already contains entry with id {2323}', [])
73
>>> i.add(InventoryEntry('2324', 'bye.c', 'file', '123'))
74
>>> i.add(InventoryEntry('2325', 'wibble', 'directory', '123'))
75
>>> i.path2id('src/wibble')
76
'2325'
77
>>> '2325' in i
78
True
79
>>> i.add(InventoryEntry('2326', 'wibble.c', 'file', '2325'))
80
>>> i['2326']
81
InventoryEntry('2326', 'wibble.c', kind='file', parent_id='2325')
82
>>> for j in i.iter_entries():
83
... print j[0]
84
... assert i.path2id(j[0])
85
...
86
src
87
src/bye.c
88
src/hello.c
89
src/wibble
90
src/wibble/wibble.c
91
>>> i.id2path('2326')
92
'src/wibble/wibble.c'
93
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.
97
"""
98
99
# TODO: split InventoryEntry into subclasses for files,
100
# directories, etc etc.
101
102
text_sha1 = None
103
text_size = None
104
105
def __init__(self, file_id, name, kind, parent_id, text_id=None):
106
"""Create an InventoryEntry
107
108
The filename must be a single component, relative to the
109
parent directory; it cannot be a whole path or relative name.
110
111
>>> e = InventoryEntry('123', 'hello.c', 'file', ROOT_ID)
112
>>> e.name
113
'hello.c'
114
>>> e.file_id
115
'123'
116
>>> e = InventoryEntry('123', 'src/hello.c', 'file', ROOT_ID)
117
Traceback (most recent call last):
118
BzrCheckError: InventoryEntry name 'src/hello.c' is invalid
119
"""
120
if '/' in name or '\\' in name:
121
raise BzrCheckError('InventoryEntry name %r is invalid' % name)
122
123
self.file_id = file_id
124
self.name = name
125
self.kind = kind
126
self.text_id = text_id
127
self.parent_id = parent_id
128
if kind == 'directory':
129
self.children = {}
130
elif kind == 'file':
131
pass
132
else:
133
raise BzrError("unhandled entry kind %r" % kind)
134
135
136
137
def sorted_children(self):
138
l = self.children.items()
139
l.sort()
140
return l
141
142
143
def copy(self):
144
other = InventoryEntry(self.file_id, self.name, self.kind,
145
self.parent_id, text_id=self.text_id)
146
other.text_sha1 = self.text_sha1
147
other.text_size = self.text_size
148
# note that children are *not* copied; they're pulled across when
149
# others are added
150
return other
151
152
153
def __repr__(self):
154
return ("%s(%r, %r, kind=%r, parent_id=%r)"
155
% (self.__class__.__name__,
156
self.file_id,
157
self.name,
158
self.kind,
159
self.parent_id))
160
161
162
def to_element(self):
163
"""Convert to XML element"""
164
e = Element('entry')
165
166
e.set('name', self.name)
167
e.set('file_id', self.file_id)
168
e.set('kind', self.kind)
169
170
if self.text_size != None:
171
e.set('text_size', '%d' % self.text_size)
172
173
for f in ['text_id', 'text_sha1']:
174
v = getattr(self, f)
175
if v != None:
176
e.set(f, v)
177
178
# to be conservative, we don't externalize the root pointers
179
# for now, leaving them as null in the xml form. in a future
180
# version it will be implied by nested elements.
181
if self.parent_id != ROOT_ID:
182
assert isinstance(self.parent_id, basestring)
183
e.set('parent_id', self.parent_id)
184
185
e.tail = '\n'
186
187
return e
188
189
190
def from_element(cls, elt):
191
assert elt.tag == 'entry'
192
193
## original format inventories don't have a parent_id for
194
## nodes in the root directory, but it's cleaner to use one
195
## internally.
196
parent_id = elt.get('parent_id')
197
if parent_id == None:
198
parent_id = ROOT_ID
199
200
self = cls(elt.get('file_id'), elt.get('name'), elt.get('kind'), parent_id)
201
self.text_id = elt.get('text_id')
202
self.text_sha1 = elt.get('text_sha1')
203
204
## mutter("read inventoryentry: %r" % (elt.attrib))
205
206
v = elt.get('text_size')
207
self.text_size = v and int(v)
208
209
return self
210
211
212
from_element = classmethod(from_element)
213
214
def __cmp__(self, other):
215
if self is other:
216
return 0
217
if not isinstance(other, InventoryEntry):
218
return NotImplemented
219
220
return cmp(self.file_id, other.file_id) \
221
or cmp(self.name, other.name) \
222
or cmp(self.text_sha1, other.text_sha1) \
223
or cmp(self.text_size, other.text_size) \
224
or cmp(self.text_id, other.text_id) \
225
or cmp(self.parent_id, other.parent_id) \
226
or cmp(self.kind, other.kind)
227
228
229
230
class RootEntry(InventoryEntry):
231
def __init__(self, file_id):
232
self.file_id = file_id
233
self.children = {}
234
self.kind = 'root_directory'
235
self.parent_id = None
236
self.name = ''
237
238
def __cmp__(self, other):
239
if self is other:
240
return 0
241
if not isinstance(other, RootEntry):
242
return NotImplemented
243
return cmp(self.file_id, other.file_id) \
244
or cmp(self.children, other.children)
245
246
247
248
class Inventory(XMLMixin):
249
"""Inventory of versioned files in a tree.
250
251
This describes which file_id is present at each point in the tree,
252
and possibly the SHA-1 or other information about the file.
253
Entries can be looked up either by path or by file_id.
254
255
The inventory represents a typical unix file tree, with
256
directories containing files and subdirectories. We never store
257
the full path to a file, because renaming a directory implicitly
258
moves all of its contents. This class internally maintains a
259
lookup tree that allows the children under a directory to be
260
returned quickly.
261
262
InventoryEntry objects must not be modified after they are
263
inserted, other than through the Inventory API.
264
265
>>> inv = Inventory()
266
>>> inv.write_xml(sys.stdout)
267
<inventory>
268
</inventory>
269
>>> inv.add(InventoryEntry('123-123', 'hello.c', 'file', ROOT_ID))
270
>>> inv['123-123'].name
271
'hello.c'
272
273
May be treated as an iterator or set to look up file ids:
274
275
>>> bool(inv.path2id('hello.c'))
276
True
277
>>> '123-123' in inv
278
True
279
280
May also look up by name:
281
282
>>> [x[0] for x in inv.iter_entries()]
283
['hello.c']
284
285
>>> inv.write_xml(sys.stdout)
286
<inventory>
287
<entry file_id="123-123" kind="file" name="hello.c" />
288
</inventory>
289
290
"""
291
def __init__(self):
292
"""Create or read an inventory.
293
294
If a working directory is specified, the inventory is read
295
from there. If the file is specified, read from that. If not,
296
the inventory is created empty.
297
298
The inventory is created with a default root directory, with
299
an id of None.
300
"""
301
self.root = RootEntry(ROOT_ID)
302
self._byid = {self.root.file_id: self.root}
303
304
305
def __iter__(self):
306
return iter(self._byid)
307
308
309
def __len__(self):
310
"""Returns number of entries."""
311
return len(self._byid)
312
313
314
def iter_entries(self, from_dir=None):
315
"""Return (path, entry) pairs, in order by name."""
316
if from_dir == None:
317
assert self.root
318
from_dir = self.root
319
elif isinstance(from_dir, basestring):
320
from_dir = self._byid[from_dir]
321
322
kids = from_dir.children.items()
323
kids.sort()
324
for name, ie in kids:
325
yield name, ie
326
if ie.kind == 'directory':
327
for cn, cie in self.iter_entries(from_dir=ie.file_id):
328
yield os.path.join(name, cn), cie
329
330
331
332
def directories(self):
333
"""Return (path, entry) pairs for all directories.
334
"""
335
def descend(parent_ie):
336
parent_name = parent_ie.name
337
yield parent_name, parent_ie
338
339
# directory children in sorted order
340
dn = []
341
for ie in parent_ie.children.itervalues():
342
if ie.kind == 'directory':
343
dn.append((ie.name, ie))
344
dn.sort()
345
346
for name, child_ie in dn:
347
for sub_name, sub_ie in descend(child_ie):
348
yield appendpath(parent_name, sub_name), sub_ie
349
350
for name, ie in descend(self.root):
351
yield name, ie
352
353
354
355
def __contains__(self, file_id):
356
"""True if this entry contains a file with given id.
357
358
>>> inv = Inventory()
359
>>> inv.add(InventoryEntry('123', 'foo.c', 'file', ROOT_ID))
360
>>> '123' in inv
361
True
362
>>> '456' in inv
363
False
364
"""
365
return file_id in self._byid
366
367
368
def __getitem__(self, file_id):
369
"""Return the entry for given file_id.
370
371
>>> inv = Inventory()
372
>>> inv.add(InventoryEntry('123123', 'hello.c', 'file', ROOT_ID))
373
>>> inv['123123'].name
374
'hello.c'
375
"""
376
try:
377
return self._byid[file_id]
378
except KeyError:
379
if file_id == None:
380
raise BzrError("can't look up file_id None")
381
else:
382
raise BzrError("file_id {%s} not in inventory" % file_id)
383
384
385
def get_file_kind(self, file_id):
386
return self._byid[file_id].kind
387
388
def get_child(self, parent_id, filename):
389
return self[parent_id].children.get(filename)
390
391
392
def add(self, entry):
393
"""Add entry to inventory.
394
395
To add a file to a branch ready to be committed, use Branch.add,
396
which calls this."""
397
if entry.file_id in self._byid:
398
bailout("inventory already contains entry with id {%s}" % entry.file_id)
399
400
try:
401
parent = self._byid[entry.parent_id]
402
except KeyError:
403
bailout("parent_id {%s} not in inventory" % entry.parent_id)
404
405
if parent.children.has_key(entry.name):
406
bailout("%s is already versioned" %
407
appendpath(self.id2path(parent.file_id), entry.name))
408
409
self._byid[entry.file_id] = entry
410
parent.children[entry.name] = entry
411
412
413
def add_path(self, relpath, kind, file_id=None):
414
"""Add entry from a path.
415
416
The immediate parent must already be versioned"""
417
parts = bzrlib.osutils.splitpath(relpath)
418
if len(parts) == 0:
419
bailout("cannot re-add root of inventory")
420
421
if file_id == None:
422
file_id = bzrlib.branch.gen_file_id(relpath)
423
424
parent_id = self.path2id(parts[:-1])
425
assert parent_id != None
426
ie = InventoryEntry(file_id, parts[-1],
427
kind=kind, parent_id=parent_id)
428
return self.add(ie)
429
430
431
def __delitem__(self, file_id):
432
"""Remove entry by id.
433
434
>>> inv = Inventory()
435
>>> inv.add(InventoryEntry('123', 'foo.c', 'file', ROOT_ID))
436
>>> '123' in inv
437
True
438
>>> del inv['123']
439
>>> '123' in inv
440
False
441
"""
442
ie = self[file_id]
443
444
assert self[ie.parent_id].children[ie.name] == ie
445
446
# TODO: Test deleting all children; maybe hoist to a separate
447
# deltree method?
448
if ie.kind == 'directory':
449
for cie in ie.children.values():
450
del self[cie.file_id]
451
del ie.children
452
453
del self._byid[file_id]
454
del self[ie.parent_id].children[ie.name]
455
456
457
def id_set(self):
458
return Set(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