« back to all changes in this revision
Viewing changes to bzrlib/inventory.py
- browse files at revision 1
- compare with another revision
- download diff
- download tarball
- view history from revision 1
- Committer: mbp at sourcefrog
- Date: 2005-03-09 04:08:15 UTC
- Revision ID: mbp@sourcefrog.net-20050309040815-13242001617e4a06
import from baz patch-364
added
removed
bzrlib/inventory.py
1
#! /usr/bin/env python
2
# -*- coding: UTF-8 -*-
3
4
# This program is free software; you can redistribute it and/or modify
5
# it under the terms of the GNU General Public License as published by
6
# the Free Software Foundation; either version 2 of the License, or
7
# (at your option) any later version.
8
9
# This program is distributed in the hope that it will be useful,
10
# but WITHOUT ANY WARRANTY; without even the implied warranty of
11
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12
# GNU General Public License for more details.
13
14
# You should have received a copy of the GNU General Public License
15
# along with this program; if not, write to the Free Software
16
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
18
"""Inventories map files to their name in a revision."""
19
20
21
__copyright__ = "Copyright (C) 2005 Canonical Ltd."
22
__author__ = "Martin Pool <mbp@canonical.com>"
23
24
import sys, os.path, types
25
from sets import Set
26
27
from xml import XMLMixin
28
from ElementTree import ElementTree, Element
29
from errors import bailout
30
from osutils import uuid, quotefn, splitpath, joinpath, appendpath
31
from trace import mutter
32
33
class InventoryEntry(XMLMixin):
34
"""Description of a versioned file.
35
36
An InventoryEntry has the following fields, which are also
37
present in the XML inventory-entry element:
38
39
* *file_id*
40
* *name*: (only the basename within the directory, must not
41
contain slashes)
42
* *kind*: "directory" or "file"
43
* *directory_id*: (if absent/null means the branch root directory)
44
* *text_sha1*: only for files
45
* *text_size*: in bytes, only for files
46
* *text_id*: identifier for the text version, only for files
47
48
InventoryEntries can also exist inside a WorkingTree
49
inventory, in which case they are not yet bound to a
50
particular revision of the file. In that case the text_sha1,
51
text_size and text_id are absent.
52
53
54
>>> i = Inventory()
55
>>> i.path2id('')
56
>>> i.add(InventoryEntry('123', 'src', kind='directory'))
57
>>> i.add(InventoryEntry('2323', 'hello.c', parent_id='123'))
58
>>> for j in i.iter_entries():
59
... print j
60
...
61
('src', InventoryEntry('123', 'src', kind='directory', parent_id=None))
62
('src/hello.c', InventoryEntry('2323', 'hello.c', kind='file', parent_id='123'))
63
>>> i.add(InventoryEntry('2323', 'bye.c', parent_id='123'))
64
Traceback (most recent call last):
65
...
66
BzrError: ('inventory already contains entry with id {2323}', [])
67
>>> i.add(InventoryEntry('2324', 'bye.c', parent_id='123'))
68
>>> i.add(InventoryEntry('2325', 'wibble', parent_id='123', kind='directory'))
69
>>> i.path2id('src/wibble')
70
'2325'
71
>>> '2325' in i
72
True
73
>>> i.add(InventoryEntry('2326', 'wibble.c', parent_id='2325'))
74
>>> i['2326']
75
InventoryEntry('2326', 'wibble.c', kind='file', parent_id='2325')
76
>>> for j in i.iter_entries():
77
... print j[0]
78
... assert i.path2id(j[0])
79
...
80
src
81
src/bye.c
82
src/hello.c
83
src/wibble
84
src/wibble/wibble.c
85
>>> i.id2path('2326')
86
'src/wibble/wibble.c'
87
88
:todo: Maybe also keep the full path of the entry, and the children?
89
But those depend on its position within a particular inventory, and
90
it would be nice not to need to hold the backpointer here.
91
"""
92
def __init__(self, file_id, name, kind='file', text_id=None,
93
parent_id=None):
94
"""Create an InventoryEntry
95
96
The filename must be a single component, relative to the
97
parent directory; it cannot be a whole path or relative name.
98
99
>>> e = InventoryEntry('123', 'hello.c')
100
>>> e.name
101
'hello.c'
102
>>> e.file_id
103
'123'
104
>>> e = InventoryEntry('123', 'src/hello.c')
105
Traceback (most recent call last):
106
BzrError: ("InventoryEntry name is not a simple filename: 'src/hello.c'", [])
107
"""
108
109
if len(splitpath(name)) != 1:
110
bailout('InventoryEntry name is not a simple filename: %r'
111
% name)
112
113
self.file_id = file_id
114
self.name = name
115
assert kind in ['file', 'directory']
116
self.kind = kind
117
self.text_id = text_id
118
self.parent_id = parent_id
119
self.text_sha1 = None
120
self.text_size = None
121
122
123
def copy(self):
124
other = InventoryEntry(self.file_id, self.name, self.kind,
125
self.text_id, self.parent_id)
126
other.text_sha1 = self.text_sha1
127
other.text_size = self.text_size
128
return other
129
130
131
def __repr__(self):
132
return ("%s(%r, %r, kind=%r, parent_id=%r)"
133
% (self.__class__.__name__,
134
self.file_id,
135
self.name,
136
self.kind,
137
self.parent_id))
138
139
140
def to_element(self):
141
"""Convert to XML element"""
142
e = Element('entry')
143
144
e.set('name', self.name)
145
e.set('file_id', self.file_id)
146
e.set('kind', self.kind)
147
148
if self.text_size is not None:
149
e.set('text_size', '%d' % self.text_size)
150
151
for f in ['text_id', 'text_sha1', 'parent_id']:
152
v = getattr(self, f)
153
if v is not None:
154
e.set(f, v)
155
156
e.tail = '\n'
157
158
return e
159
160
161
def from_element(cls, elt):
162
assert elt.tag == 'entry'
163
self = cls(elt.get('file_id'), elt.get('name'), elt.get('kind'))
164
self.text_id = elt.get('text_id')
165
self.text_sha1 = elt.get('text_sha1')
166
self.parent_id = elt.get('parent_id')
167
168
## mutter("read inventoryentry: %r" % (elt.attrib))
169
170
v = elt.get('text_size')
171
self.text_size = v and int(v)
172
173
return self
174
175
176
from_element = classmethod(from_element)
177
178
def __cmp__(self, other):
179
if self is other:
180
return 0
181
if not isinstance(other, InventoryEntry):
182
return NotImplemented
183
184
return cmp(self.file_id, other.file_id) \
185
or cmp(self.name, other.name) \
186
or cmp(self.text_sha1, other.text_sha1) \
187
or cmp(self.text_size, other.text_size) \
188
or cmp(self.text_id, other.text_id) \
189
or cmp(self.parent_id, other.parent_id) \
190
or cmp(self.kind, other.kind)
191
192
193
194
class Inventory(XMLMixin):
195
"""Inventory of versioned files in a tree.
196
197
An Inventory acts like a set of InventoryEntry items. You can
198
also look files up by their file_id or name.
199
200
May be read from and written to a metadata file in a tree. To
201
manipulate the inventory (for example to add a file), it is read
202
in, modified, and then written back out.
203
204
The inventory represents a typical unix file tree, with
205
directories containing files and subdirectories. We never store
206
the full path to a file, because renaming a directory implicitly
207
moves all of its contents. This class internally maintains a
208
lookup tree that allows the children under a directory to be
209
returned quickly.
210
211
InventoryEntry objects must not be modified after they are
212
inserted.
213
214
>>> inv = Inventory()
215
>>> inv.write_xml(sys.stdout)
216
<inventory>
217
</inventory>
218
>>> inv.add(InventoryEntry('123-123', 'hello.c'))
219
>>> inv['123-123'].name
220
'hello.c'
221
>>> for file_id in inv: print file_id
222
...
223
123-123
224
225
May be treated as an iterator or set to look up file ids:
226
227
>>> bool(inv.path2id('hello.c'))
228
True
229
>>> '123-123' in inv
230
True
231
232
May also look up by name:
233
234
>>> [x[0] for x in inv.iter_entries()]
235
['hello.c']
236
237
>>> inv.write_xml(sys.stdout)
238
<inventory>
239
<entry file_id="123-123" kind="file" name="hello.c" />
240
</inventory>
241
242
"""
243
244
## TODO: Clear up handling of files in subdirectories; we probably
245
## do want to be able to just look them up by name but this
246
## probably means gradually walking down the path, looking up as we go.
247
248
## TODO: Make sure only canonical filenames are stored.
249
250
## TODO: Do something sensible about the possible collisions on
251
## case-losing filesystems. Perhaps we should just always forbid
252
## such collisions.
253
254
## _tree should probably just be stored as
255
## InventoryEntry._children on each directory.
256
257
def __init__(self):
258
"""Create or read an inventory.
259
260
If a working directory is specified, the inventory is read
261
from there. If the file is specified, read from that. If not,
262
the inventory is created empty.
263
"""
264
self._byid = dict()
265
266
# _tree is indexed by parent_id; at each level a map from name
267
# to ie. The None entry is the root.
268
self._tree = {None: {}}
269
270
271
def __iter__(self):
272
return iter(self._byid)
273
274
275
def __len__(self):
276
"""Returns number of entries."""
277
return len(self._byid)
278
279
280
def iter_entries(self, parent_id=None):
281
"""Return (path, entry) pairs, in order by name."""
282
kids = self._tree[parent_id].items()
283
kids.sort()
284
for name, ie in kids:
285
yield name, ie
286
if ie.kind == 'directory':
287
for cn, cie in self.iter_entries(parent_id=ie.file_id):
288
yield joinpath([name, cn]), cie
289
290
291
def directories(self, include_root=True):
292
"""Return (path, entry) pairs for all directories.
293
"""
294
if include_root:
295
yield '', None
296
for path, entry in self.iter_entries():
297
if entry.kind == 'directory':
298
yield path, entry
299
300
301
302
def children(self, parent_id):
303
"""Return entries that are direct children of parent_id."""
304
return self._tree[parent_id]
305
306
307
308
# TODO: return all paths and entries
309
310
311
def __contains__(self, file_id):
312
"""True if this entry contains a file with given id.
313
314
>>> inv = Inventory()
315
>>> inv.add(InventoryEntry('123', 'foo.c'))
316
>>> '123' in inv
317
True
318
>>> '456' in inv
319
False
320
"""
321
return file_id in self._byid
322
323
324
def __getitem__(self, file_id):
325
"""Return the entry for given file_id.
326
327
>>> inv = Inventory()
328
>>> inv.add(InventoryEntry('123123', 'hello.c'))
329
>>> inv['123123'].name
330
'hello.c'
331
"""
332
return self._byid[file_id]
333
334
335
def add(self, entry):
336
"""Add entry to inventory.
337
338
To add a file to a branch ready to be committed, use Branch.add,
339
which calls this."""
340
if entry.file_id in self:
341
bailout("inventory already contains entry with id {%s}" % entry.file_id)
342
343
if entry.parent_id != None:
344
if entry.parent_id not in self:
345
bailout("parent_id %s of new entry not found in inventory"
346
% entry.parent_id)
347
348
if self._tree[entry.parent_id].has_key(entry.name):
349
bailout("%s is already versioned"
350
% appendpath(self.id2path(entry.parent_id), entry.name))
351
352
self._byid[entry.file_id] = entry
353
self._tree[entry.parent_id][entry.name] = entry
354
355
if entry.kind == 'directory':
356
self._tree[entry.file_id] = {}
357
358
359
def __delitem__(self, file_id):
360
"""Remove entry by id.
361
362
>>> inv = Inventory()
363
>>> inv.add(InventoryEntry('123', 'foo.c'))
364
>>> '123' in inv
365
True
366
>>> del inv['123']
367
>>> '123' in inv
368
False
369
"""
370
ie = self[file_id]
371
372
assert self._tree[ie.parent_id][ie.name] == ie
373
374
# TODO: Test deleting all children; maybe hoist to a separate
375
# deltree method?
376
if ie.kind == 'directory':
377
for cie in self._tree[file_id].values():
378
del self[cie.file_id]
379
del self._tree[file_id]
380
381
del self._byid[file_id]
382
del self._tree[ie.parent_id][ie.name]
383
384
385
def id_set(self):
386
return Set(self._byid)
387
388
389
def to_element(self):
390
"""Convert to XML Element"""
391
e = Element('inventory')
392
e.text = '\n'
393
for path, ie in self.iter_entries():
394
e.append(ie.to_element())
395
return e
396
397
398
def from_element(cls, elt):
399
"""Construct from XML Element
400
401
>>> inv = Inventory()
402
>>> inv.add(InventoryEntry('foo.c-123981239', 'foo.c'))
403
>>> elt = inv.to_element()
404
>>> inv2 = Inventory.from_element(elt)
405
>>> inv2 == inv
406
True
407
"""
408
assert elt.tag == 'inventory'
409
o = cls()
410
for e in elt:
411
o.add(InventoryEntry.from_element(e))
412
return o
413
414
from_element = classmethod(from_element)
415
416
417
def __cmp__(self, other):
418
"""Compare two sets by comparing their contents.
419
420
>>> i1 = Inventory()
421
>>> i2 = Inventory()
422
>>> i1 == i2
423
True
424
>>> i1.add(InventoryEntry('123', 'foo'))
425
>>> i1 == i2
426
False
427
>>> i2.add(InventoryEntry('123', 'foo'))
428
>>> i1 == i2
429
True
430
"""
431
if self is other:
432
return 0
433
434
if not isinstance(other, Inventory):
435
return NotImplemented
436
437
if self.id_set() ^ other.id_set():
438
return 1
439
440
for file_id in self._byid:
441
c = cmp(self[file_id], other[file_id])
442
if c: return c
443
444
return 0
445
446
447
def id2path(self, file_id):
448
"""Return as a list the path to file_id."""
449
p = []
450
while file_id != None:
451
ie = self[file_id]
452
p = [ie.name] + p
453
file_id = ie.parent_id
454
return joinpath(p)
455
456
457
458
def path2id(self, name):
459
"""Walk down through directories to return entry of last component.
460
461
names may be either a list of path components, or a single
462
string, in which case it is automatically split.
463
464
This returns the entry of the last component in the path,
465
which may be either a file or a directory.
466
"""
467
assert isinstance(name, types.StringTypes)
468
469
parent_id = None
470
for f in splitpath(name):
471
try:
472
cie = self._tree[parent_id][f]
473
assert cie.name == f
474
parent_id = cie.file_id
475
except KeyError:
476
# or raise an error?
477
return None
478
479
return parent_id
480
481
482
def get_child(self, parent_id, child_name):
483
return self._tree[parent_id].get(child_name)
484
485
486
def has_filename(self, names):
487
return bool(self.path2id(names))
488
489
490
def has_id(self, file_id):
491
assert isinstance(file_id, str)
492
return self._byid.has_key(file_id)
493
494
495
496
if __name__ == '__main__':
497
import doctest, inventory
498
doctest.testmod(inventory)
Loggerhead is a web-based interface for Breezy
Version: 2.0.1