« back to all changes in this revision
Viewing changes to bzrlib/inventory.py
- browse files at revision 423
- compare with another revision
- download diff
- download tarball
- view history from revision 423
- Committer: Martin Pool
- Date: 2005-05-10 03:55:34 UTC
- Revision ID: mbp@sourcefrog.net-20050510035534-643062e821052ac5
- Add fortune-cookie external plugin demonstration
- files added:
- .bzrignore
- bzrlib
- contrib
- contrib/bash
- contrib/zsh
- doc
- elementtree
- notes
- 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
return other
149
150
151
def __repr__(self):
152
return ("%s(%r, %r, kind=%r, parent_id=%r)"
153
% (self.__class__.__name__,
154
self.file_id,
155
self.name,
156
self.kind,
157
self.parent_id))
158
159
160
def to_element(self):
161
"""Convert to XML element"""
162
e = Element('entry')
163
164
e.set('name', self.name)
165
e.set('file_id', self.file_id)
166
e.set('kind', self.kind)
167
168
if self.text_size != None:
169
e.set('text_size', '%d' % self.text_size)
170
171
for f in ['text_id', 'text_sha1']:
172
v = getattr(self, f)
173
if v != None:
174
e.set(f, v)
175
176
# to be conservative, we don't externalize the root pointers
177
# for now, leaving them as null in the xml form. in a future
178
# version it will be implied by nested elements.
179
if self.parent_id != ROOT_ID:
180
assert isinstance(self.parent_id, basestring)
181
e.set('parent_id', self.parent_id)
182
183
e.tail = '\n'
184
185
return e
186
187
188
def from_element(cls, elt):
189
assert elt.tag == 'entry'
190
191
## original format inventories don't have a parent_id for
192
## nodes in the root directory, but it's cleaner to use one
193
## internally.
194
parent_id = elt.get('parent_id')
195
if parent_id == None:
196
parent_id = ROOT_ID
197
198
self = cls(elt.get('file_id'), elt.get('name'), elt.get('kind'), parent_id)
199
self.text_id = elt.get('text_id')
200
self.text_sha1 = elt.get('text_sha1')
201
202
## mutter("read inventoryentry: %r" % (elt.attrib))
203
204
v = elt.get('text_size')
205
self.text_size = v and int(v)
206
207
return self
208
209
210
from_element = classmethod(from_element)
211
212
def __cmp__(self, other):
213
if self is other:
214
return 0
215
if not isinstance(other, InventoryEntry):
216
return NotImplemented
217
218
return cmp(self.file_id, other.file_id) \
219
or cmp(self.name, other.name) \
220
or cmp(self.text_sha1, other.text_sha1) \
221
or cmp(self.text_size, other.text_size) \
222
or cmp(self.text_id, other.text_id) \
223
or cmp(self.parent_id, other.parent_id) \
224
or cmp(self.kind, other.kind)
225
226
227
228
class RootEntry(InventoryEntry):
229
def __init__(self, file_id):
230
self.file_id = file_id
231
self.children = {}
232
self.kind = 'root_directory'
233
self.parent_id = None
234
self.name = ''
235
236
def __cmp__(self, other):
237
if self is other:
238
return 0
239
if not isinstance(other, RootEntry):
240
return NotImplemented
241
return cmp(self.file_id, other.file_id) \
242
or cmp(self.children, other.children)
243
244
245
246
class Inventory(XMLMixin):
247
"""Inventory of versioned files in a tree.
248
249
This describes which file_id is present at each point in the tree,
250
and possibly the SHA-1 or other information about the file.
251
Entries can be looked up either by path or by file_id.
252
253
The inventory represents a typical unix file tree, with
254
directories containing files and subdirectories. We never store
255
the full path to a file, because renaming a directory implicitly
256
moves all of its contents. This class internally maintains a
257
lookup tree that allows the children under a directory to be
258
returned quickly.
259
260
InventoryEntry objects must not be modified after they are
261
inserted, other than through the Inventory API.
262
263
>>> inv = Inventory()
264
>>> inv.write_xml(sys.stdout)
265
<inventory>
266
</inventory>
267
>>> inv.add(InventoryEntry('123-123', 'hello.c', 'file', ROOT_ID))
268
>>> inv['123-123'].name
269
'hello.c'
270
271
May be treated as an iterator or set to look up file ids:
272
273
>>> bool(inv.path2id('hello.c'))
274
True
275
>>> '123-123' in inv
276
True
277
278
May also look up by name:
279
280
>>> [x[0] for x in inv.iter_entries()]
281
['hello.c']
282
283
>>> inv.write_xml(sys.stdout)
284
<inventory>
285
<entry file_id="123-123" kind="file" name="hello.c" />
286
</inventory>
287
288
"""
289
def __init__(self):
290
"""Create or read an inventory.
291
292
If a working directory is specified, the inventory is read
293
from there. If the file is specified, read from that. If not,
294
the inventory is created empty.
295
296
The inventory is created with a default root directory, with
297
an id of None.
298
"""
299
self.root = RootEntry(ROOT_ID)
300
self._byid = {self.root.file_id: self.root}
301
302
303
def __iter__(self):
304
return iter(self._byid)
305
306
307
def __len__(self):
308
"""Returns number of entries."""
309
return len(self._byid)
310
311
312
def iter_entries(self, from_dir=None):
313
"""Return (path, entry) pairs, in order by name."""
314
if from_dir == None:
315
assert self.root
316
from_dir = self.root
317
elif isinstance(from_dir, basestring):
318
from_dir = self._byid[from_dir]
319
320
kids = from_dir.children.items()
321
kids.sort()
322
for name, ie in kids:
323
yield name, ie
324
if ie.kind == 'directory':
325
for cn, cie in self.iter_entries(from_dir=ie.file_id):
326
yield os.path.join(name, cn), cie
327
328
329
330
def directories(self):
331
"""Return (path, entry) pairs for all directories.
332
"""
333
def descend(parent_ie):
334
parent_name = parent_ie.name
335
yield parent_name, parent_ie
336
337
# directory children in sorted order
338
dn = []
339
for ie in parent_ie.children.itervalues():
340
if ie.kind == 'directory':
341
dn.append((ie.name, ie))
342
dn.sort()
343
344
for name, child_ie in dn:
345
for sub_name, sub_ie in descend(child_ie):
346
yield appendpath(parent_name, sub_name), sub_ie
347
348
for name, ie in descend(self.root):
349
yield name, ie
350
351
352
353
def __contains__(self, file_id):
354
"""True if this entry contains a file with given id.
355
356
>>> inv = Inventory()
357
>>> inv.add(InventoryEntry('123', 'foo.c', 'file', ROOT_ID))
358
>>> '123' in inv
359
True
360
>>> '456' in inv
361
False
362
"""
363
return file_id in self._byid
364
365
366
def __getitem__(self, file_id):
367
"""Return the entry for given file_id.
368
369
>>> inv = Inventory()
370
>>> inv.add(InventoryEntry('123123', 'hello.c', 'file', ROOT_ID))
371
>>> inv['123123'].name
372
'hello.c'
373
"""
374
try:
375
return self._byid[file_id]
376
except KeyError:
377
if file_id == None:
378
raise BzrError("can't look up file_id None")
379
else:
380
raise BzrError("file_id {%s} not in inventory" % file_id)
381
382
383
def get_child(self, parent_id, filename):
384
return self[parent_id].children.get(filename)
385
386
387
def add(self, entry):
388
"""Add entry to inventory.
389
390
To add a file to a branch ready to be committed, use Branch.add,
391
which calls this."""
392
if entry.file_id in self._byid:
393
bailout("inventory already contains entry with id {%s}" % entry.file_id)
394
395
try:
396
parent = self._byid[entry.parent_id]
397
except KeyError:
398
bailout("parent_id {%s} not in inventory" % entry.parent_id)
399
400
if parent.children.has_key(entry.name):
401
bailout("%s is already versioned" %
402
appendpath(self.id2path(parent.file_id), entry.name))
403
404
self._byid[entry.file_id] = entry
405
parent.children[entry.name] = entry
406
407
408
def add_path(self, relpath, kind, file_id=None):
409
"""Add entry from a path.
410
411
The immediate parent must already be versioned"""
412
parts = bzrlib.osutils.splitpath(relpath)
413
if len(parts) == 0:
414
bailout("cannot re-add root of inventory")
415
416
if file_id == None:
417
file_id = bzrlib.branch.gen_file_id(relpath)
418
419
parent_id = self.path2id(parts[:-1])
420
assert parent_id != None
421
ie = InventoryEntry(file_id, parts[-1],
422
kind=kind, parent_id=parent_id)
423
return self.add(ie)
424
425
426
def __delitem__(self, file_id):
427
"""Remove entry by id.
428
429
>>> inv = Inventory()
430
>>> inv.add(InventoryEntry('123', 'foo.c', 'file', ROOT_ID))
431
>>> '123' in inv
432
True
433
>>> del inv['123']
434
>>> '123' in inv
435
False
436
"""
437
ie = self[file_id]
438
439
assert self[ie.parent_id].children[ie.name] == ie
440
441
# TODO: Test deleting all children; maybe hoist to a separate
442
# deltree method?
443
if ie.kind == 'directory':
444
for cie in ie.children.values():
445
del self[cie.file_id]
446
del ie.children
447
448
del self._byid[file_id]
449
del self[ie.parent_id].children[ie.name]
450
451
452
def id_set(self):
453
return Set(self._byid)
454
455
456
def to_element(self):
457
"""Convert to XML Element"""
458
e = Element('inventory')
459
e.text = '\n'
460
for path, ie in self.iter_entries():
461
e.append(ie.to_element())
462
return e
463
464
465
def from_element(cls, elt):
466
"""Construct from XML Element
467
468
>>> inv = Inventory()
469
>>> inv.add(InventoryEntry('foo.c-123981239', 'foo.c', 'file', ROOT_ID))
470
>>> elt = inv.to_element()
471
>>> inv2 = Inventory.from_element(elt)
472
>>> inv2 == inv
473
True
474
"""
475
assert elt.tag == 'inventory'
476
o = cls()
477
for e in elt:
478
o.add(InventoryEntry.from_element(e))
479
return o
480
481
from_element = classmethod(from_element)
482
483
484
def __cmp__(self, other):
485
"""Compare two sets by comparing their contents.
486
487
>>> i1 = Inventory()
488
>>> i2 = Inventory()
489
>>> i1 == i2
490
True
491
>>> i1.add(InventoryEntry('123', 'foo', 'file', ROOT_ID))
492
>>> i1 == i2
493
False
494
>>> i2.add(InventoryEntry('123', 'foo', 'file', ROOT_ID))
495
>>> i1 == i2
496
True
497
"""
498
if self is other:
499
return 0
500
501
if not isinstance(other, Inventory):
502
return NotImplemented
503
504
if self.id_set() ^ other.id_set():
505
return 1
506
507
for file_id in self._byid:
508
c = cmp(self[file_id], other[file_id])
509
if c: return c
510
511
return 0
512
513
514
def get_idpath(self, file_id):
515
"""Return a list of file_ids for the path to an entry.
516
517
The list contains one element for each directory followed by
518
the id of the file itself. So the length of the returned list
519
is equal to the depth of the file in the tree, counting the
520
root directory as depth 1.
521
"""
522
p = []
523
while file_id != None:
524
try:
525
ie = self._byid[file_id]
526
except KeyError:
527
bailout("file_id {%s} not found in inventory" % file_id)
528
p.insert(0, ie.file_id)
529
file_id = ie.parent_id
530
return p
531
532
533
def id2path(self, file_id):
534
"""Return as a list the path to file_id."""
535
536
# get all names, skipping root
537
p = [self[fid].name for fid in self.get_idpath(file_id)[1:]]
538
return os.sep.join(p)
539
540
541
542
def path2id(self, name):
543
"""Walk down through directories to return entry of last component.
544
545
names may be either a list of path components, or a single
546
string, in which case it is automatically split.
547
548
This returns the entry of the last component in the path,
549
which may be either a file or a directory.
550
551
Returns None iff the path is not found.
552
"""
553
if isinstance(name, types.StringTypes):
554
name = splitpath(name)
555
556
mutter("lookup path %r" % name)
557
558
parent = self.root
559
for f in name:
560
try:
561
cie = parent.children[f]
562
assert cie.name == f
563
assert cie.parent_id == parent.file_id
564
parent = cie
565
except KeyError:
566
# or raise an error?
567
return None
568
569
return parent.file_id
570
571
572
def has_filename(self, names):
573
return bool(self.path2id(names))
574
575
576
def has_id(self, file_id):
577
return self._byid.has_key(file_id)
578
579
580
def rename(self, file_id, new_parent_id, new_name):
581
"""Move a file within the inventory.
582
583
This can change either the name, or the parent, or both.
584
585
This does not move the working file."""
586
if not is_valid_name(new_name):
587
bailout("not an acceptable filename: %r" % new_name)
588
589
new_parent = self._byid[new_parent_id]
590
if new_name in new_parent.children:
591
bailout("%r already exists in %r" % (new_name, self.id2path(new_parent_id)))
592
593
new_parent_idpath = self.get_idpath(new_parent_id)
594
if file_id in new_parent_idpath:
595
bailout("cannot move directory %r into a subdirectory of itself, %r"
596
% (self.id2path(file_id), self.id2path(new_parent_id)))
597
598
file_ie = self._byid[file_id]
599
old_parent = self._byid[file_ie.parent_id]
600
601
# TODO: Don't leave things messed up if this fails
602
603
del old_parent.children[file_ie.name]
604
new_parent.children[new_name] = file_ie
605
606
file_ie.name = new_name
607
file_ie.parent_id = new_parent_id
608
609
610
611
612
_NAME_RE = re.compile(r'^[^/\\]+$')
613
614
def is_valid_name(name):
615
return bool(_NAME_RE.match(name))
Loggerhead is a web-based interface for Breezy
Version: 2.0.1