1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
|
# Copyright (C) 2005, 2006, 2007, 2008 Canonical Ltd
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
import cStringIO
import re
from bzrlib import (
cache_utf8,
errors,
fifo_cache,
inventory,
revision as _mod_revision,
trace,
)
from bzrlib.xml_serializer import SubElement, Element, Serializer
from bzrlib.inventory import ROOT_ID, Inventory, InventoryEntry
from bzrlib.revision import Revision
from bzrlib.errors import BzrError
_utf8_re = None
_unicode_re = None
_xml_escape_map = {
"&":'&',
"'":"'", # FIXME: overkill
"\"":""",
"<":"<",
">":">",
}
# A cache of InventoryEntry objects
_entry_cache = fifo_cache.FIFOCache(10*1024)
def _ensure_utf8_re():
"""Make sure the _utf8_re and _unicode_re regexes have been compiled."""
global _utf8_re, _unicode_re
if _utf8_re is None:
_utf8_re = re.compile('[&<>\'\"]|[\x80-\xff]+')
if _unicode_re is None:
_unicode_re = re.compile(u'[&<>\'\"\u0080-\uffff]')
def _unicode_escape_replace(match, _map=_xml_escape_map):
"""Replace a string of non-ascii, non XML safe characters with their escape
This will escape both Standard XML escapes, like <>"', etc.
As well as escaping non ascii characters, because ElementTree did.
This helps us remain compatible to older versions of bzr. We may change
our policy in the future, though.
"""
# jam 20060816 Benchmarks show that try/KeyError is faster if you
# expect the entity to rarely miss. There is about a 10% difference
# in overall time. But if you miss frequently, then if None is much
# faster. For our use case, we *rarely* have a revision id, file id
# or path name that is unicode. So use try/KeyError.
try:
return _map[match.group()]
except KeyError:
return "&#%d;" % ord(match.group())
def _utf8_escape_replace(match, _map=_xml_escape_map):
"""Escape utf8 characters into XML safe ones.
This uses 2 tricks. It is either escaping "standard" characters, like "&<>,
or it is handling characters with the high-bit set. For ascii characters,
we just lookup the replacement in the dictionary. For everything else, we
decode back into Unicode, and then use the XML escape code.
"""
try:
return _map[match.group()]
except KeyError:
return ''.join('&#%d;' % ord(uni_chr)
for uni_chr in match.group().decode('utf8'))
_to_escaped_map = {}
def _encode_and_escape(unicode_or_utf8_str, _map=_to_escaped_map):
"""Encode the string into utf8, and escape invalid XML characters"""
# We frequently get entities we have not seen before, so it is better
# to check if None, rather than try/KeyError
text = _map.get(unicode_or_utf8_str)
if text is None:
if unicode_or_utf8_str.__class__ == unicode:
# The alternative policy is to do a regular UTF8 encoding
# and then escape only XML meta characters.
# Performance is equivalent once you use cache_utf8. *However*
# this makes the serialized texts incompatible with old versions
# of bzr. So no net gain. (Perhaps the read code would handle utf8
# better than entity escapes, but cElementTree seems to do just fine
# either way)
text = str(_unicode_re.sub(_unicode_escape_replace,
unicode_or_utf8_str)) + '"'
else:
# Plain strings are considered to already be in utf-8 so we do a
# slightly different method for escaping.
text = _utf8_re.sub(_utf8_escape_replace,
unicode_or_utf8_str) + '"'
_map[unicode_or_utf8_str] = text
return text
def _get_utf8_or_ascii(a_str,
_encode_utf8=cache_utf8.encode,
_get_cached_ascii=cache_utf8.get_cached_ascii):
"""Return a cached version of the string.
cElementTree will return a plain string if the XML is plain ascii. It only
returns Unicode when it needs to. We want to work in utf-8 strings. So if
cElementTree returns a plain string, we can just return the cached version.
If it is Unicode, then we need to encode it.
:param a_str: An 8-bit string or Unicode as returned by
cElementTree.Element.get()
:return: A utf-8 encoded 8-bit string.
"""
# This is fairly optimized because we know what cElementTree does, this is
# not meant as a generic function for all cases. Because it is possible for
# an 8-bit string to not be ascii or valid utf8.
if a_str.__class__ == unicode:
return _encode_utf8(a_str)
else:
return _get_cached_ascii(a_str)
def _clear_cache():
"""Clean out the unicode => escaped map"""
_to_escaped_map.clear()
class Serializer_v8(Serializer):
"""This serialiser adds rich roots.
Its revision format number matches its inventory number.
"""
__slots__ = []
root_id = None
support_altered_by_hack = True
# This format supports the altered-by hack that reads file ids directly out
# of the versionedfile, without doing XML parsing.
supported_kinds = set(['file', 'directory', 'symlink'])
format_num = '8'
revision_format_num = None
def _check_revisions(self, inv):
"""Extension point for subclasses to check during serialisation.
:param inv: An inventory about to be serialised, to be checked.
:raises: AssertionError if an error has occured.
"""
if inv.revision_id is None:
raise AssertionError()
if inv.root.revision is None:
raise AssertionError()
def write_inventory_to_lines(self, inv):
"""Return a list of lines with the encoded inventory."""
return self.write_inventory(inv, None)
def write_inventory_to_string(self, inv, working=False):
"""Just call write_inventory with a StringIO and return the value.
:param working: If True skip history data - text_sha1, text_size,
reference_revision, symlink_target.
"""
sio = cStringIO.StringIO()
self.write_inventory(inv, sio, working)
return sio.getvalue()
def write_inventory(self, inv, f, working=False):
"""Write inventory to a file.
:param inv: the inventory to write.
:param f: the file to write. (May be None if the lines are the desired
output).
:param working: If True skip history data - text_sha1, text_size,
reference_revision, symlink_target.
:return: The inventory as a list of lines.
"""
_ensure_utf8_re()
self._check_revisions(inv)
output = []
append = output.append
self._append_inventory_root(append, inv)
entries = inv.iter_entries()
# Skip the root
root_path, root_ie = entries.next()
for path, ie in entries:
if ie.parent_id != self.root_id:
parent_str = ' parent_id="'
parent_id = _encode_and_escape(ie.parent_id)
else:
parent_str = ''
parent_id = ''
if ie.kind == 'file':
if ie.executable:
executable = ' executable="yes"'
else:
executable = ''
if not working:
append('<file%s file_id="%s name="%s%s%s revision="%s '
'text_sha1="%s" text_size="%d" />\n' % (
executable, _encode_and_escape(ie.file_id),
_encode_and_escape(ie.name), parent_str, parent_id,
_encode_and_escape(ie.revision), ie.text_sha1,
ie.text_size))
else:
append('<file%s file_id="%s name="%s%s%s />\n' % (
executable, _encode_and_escape(ie.file_id),
_encode_and_escape(ie.name), parent_str, parent_id))
elif ie.kind == 'directory':
if not working:
append('<directory file_id="%s name="%s%s%s revision="%s '
'/>\n' % (
_encode_and_escape(ie.file_id),
_encode_and_escape(ie.name),
parent_str, parent_id,
_encode_and_escape(ie.revision)))
else:
append('<directory file_id="%s name="%s%s%s />\n' % (
_encode_and_escape(ie.file_id),
_encode_and_escape(ie.name),
parent_str, parent_id))
elif ie.kind == 'symlink':
if not working:
append('<symlink file_id="%s name="%s%s%s revision="%s '
'symlink_target="%s />\n' % (
_encode_and_escape(ie.file_id),
_encode_and_escape(ie.name),
parent_str, parent_id,
_encode_and_escape(ie.revision),
_encode_and_escape(ie.symlink_target)))
else:
append('<symlink file_id="%s name="%s%s%s />\n' % (
_encode_and_escape(ie.file_id),
_encode_and_escape(ie.name),
parent_str, parent_id))
elif ie.kind == 'tree-reference':
if ie.kind not in self.supported_kinds:
raise errors.UnsupportedInventoryKind(ie.kind)
if not working:
append('<tree-reference file_id="%s name="%s%s%s '
'revision="%s reference_revision="%s />\n' % (
_encode_and_escape(ie.file_id),
_encode_and_escape(ie.name),
parent_str, parent_id,
_encode_and_escape(ie.revision),
_encode_and_escape(ie.reference_revision)))
else:
append('<tree-reference file_id="%s name="%s%s%s />\n' % (
_encode_and_escape(ie.file_id),
_encode_and_escape(ie.name),
parent_str, parent_id))
else:
raise errors.UnsupportedInventoryKind(ie.kind)
append('</inventory>\n')
if f is not None:
f.writelines(output)
# Just to keep the cache from growing without bounds
# but we may actually not want to do clear the cache
#_clear_cache()
return output
def _append_inventory_root(self, append, inv):
"""Append the inventory root to output."""
if inv.revision_id is not None:
revid1 = ' revision_id="'
revid2 = _encode_and_escape(inv.revision_id)
else:
revid1 = ""
revid2 = ""
append('<inventory format="%s"%s%s>\n' % (
self.format_num, revid1, revid2))
append('<directory file_id="%s name="%s revision="%s />\n' % (
_encode_and_escape(inv.root.file_id),
_encode_and_escape(inv.root.name),
_encode_and_escape(inv.root.revision)))
def _pack_revision(self, rev):
"""Revision object -> xml tree"""
# For the XML format, we need to write them as Unicode rather than as
# utf-8 strings. So that cElementTree can handle properly escaping
# them.
decode_utf8 = cache_utf8.decode
revision_id = rev.revision_id
if isinstance(revision_id, str):
revision_id = decode_utf8(revision_id)
format_num = self.format_num
if self.revision_format_num is not None:
format_num = self.revision_format_num
root = Element('revision',
committer = rev.committer,
timestamp = '%.3f' % rev.timestamp,
revision_id = revision_id,
inventory_sha1 = rev.inventory_sha1,
format=format_num,
)
if rev.timezone is not None:
root.set('timezone', str(rev.timezone))
root.text = '\n'
msg = SubElement(root, 'message')
msg.text = rev.message
msg.tail = '\n'
if rev.parent_ids:
pelts = SubElement(root, 'parents')
pelts.tail = pelts.text = '\n'
for parent_id in rev.parent_ids:
_mod_revision.check_not_reserved_id(parent_id)
p = SubElement(pelts, 'revision_ref')
p.tail = '\n'
if isinstance(parent_id, str):
parent_id = decode_utf8(parent_id)
p.set('revision_id', parent_id)
if rev.properties:
self._pack_revision_properties(rev, root)
return root
def _pack_revision_properties(self, rev, under_element):
top_elt = SubElement(under_element, 'properties')
for prop_name, prop_value in sorted(rev.properties.items()):
prop_elt = SubElement(top_elt, 'property')
prop_elt.set('name', prop_name)
prop_elt.text = prop_value
prop_elt.tail = '\n'
top_elt.tail = '\n'
def _unpack_inventory(self, elt, revision_id=None):
"""Construct from XML Element"""
if elt.tag != 'inventory':
raise errors.UnexpectedInventoryFormat('Root tag is %r' % elt.tag)
format = elt.get('format')
if format != self.format_num:
raise errors.UnexpectedInventoryFormat('Invalid format version %r'
% format)
revision_id = elt.get('revision_id')
if revision_id is not None:
revision_id = cache_utf8.encode(revision_id)
inv = inventory.Inventory(root_id=None, revision_id=revision_id)
for e in elt:
ie = self._unpack_entry(e)
inv.add(ie)
return inv
def _unpack_entry(self, elt):
get_cached = _get_utf8_or_ascii
elt_get = elt.get
file_id = elt_get('file_id')
revision = elt_get('revision')
# Check and see if we have already unpacked this exact entry
# Some timings for "repo.revision_trees(last_100_revs)"
# bzr mysql
# unmodified 4.1s 40.8s
# using lru 3.5s
# using fifo 2.83s 29.1s
# lru._cache 2.8s
# dict 2.75s 26.8s
# inv.add 2.5s 26.0s
# no_copy 2.00s 20.5s
# no_c,dict 1.95s 18.0s
# Note that a cache of 10k nodes is more than sufficient to hold all of
# the inventory for the last 100 revs for bzr, but not for mysql (20k
# is enough for mysql, which saves the same 2s as using a dict)
# Breakdown of mysql using time.clock()
# 4.1s 2 calls to element.get for file_id, revision_id
# 4.5s cache_hit lookup
# 7.1s InventoryFile.copy()
# 2.4s InventoryDirectory.copy()
# 0.4s decoding unique entries
# 1.6s decoding entries after FIFO fills up
# 0.8s Adding nodes to FIFO (including flushes)
# 0.1s cache miss lookups
# Using an LRU cache
# 4.1s 2 calls to element.get for file_id, revision_id
# 9.9s cache_hit lookup
# 10.8s InventoryEntry.copy()
# 0.3s cache miss lookus
# 1.2s decoding entries
# 1.0s adding nodes to LRU
key = (file_id, revision)
try:
# We copy it, because some operatations may mutate it
cached_ie = _entry_cache[key]
except KeyError:
pass
else:
# Only copying directory entries drops us 2.85s => 2.35s
# if cached_ie.kind == 'directory':
# return cached_ie.copy()
# return cached_ie
return cached_ie.copy()
kind = elt.tag
if not InventoryEntry.versionable_kind(kind):
raise AssertionError('unsupported entry kind %s' % kind)
file_id = get_cached(file_id)
if revision is not None:
revision = get_cached(revision)
parent_id = elt_get('parent_id')
if parent_id is not None:
parent_id = get_cached(parent_id)
if kind == 'directory':
ie = inventory.InventoryDirectory(file_id,
elt_get('name'),
parent_id)
elif kind == 'file':
ie = inventory.InventoryFile(file_id,
elt_get('name'),
parent_id)
ie.text_sha1 = elt_get('text_sha1')
if elt_get('executable') == 'yes':
ie.executable = True
v = elt_get('text_size')
ie.text_size = v and int(v)
elif kind == 'symlink':
ie = inventory.InventoryLink(file_id,
elt_get('name'),
parent_id)
ie.symlink_target = elt_get('symlink_target')
else:
raise errors.UnsupportedInventoryKind(kind)
ie.revision = revision
if revision is not None:
_entry_cache[key] = ie
return ie
def _unpack_revision(self, elt):
"""XML Element -> Revision object"""
format = elt.get('format')
format_num = self.format_num
if self.revision_format_num is not None:
format_num = self.revision_format_num
if format is not None:
if format != format_num:
raise BzrError("invalid format version %r on revision"
% format)
get_cached = _get_utf8_or_ascii
rev = Revision(committer = elt.get('committer'),
timestamp = float(elt.get('timestamp')),
revision_id = get_cached(elt.get('revision_id')),
inventory_sha1 = elt.get('inventory_sha1')
)
parents = elt.find('parents') or []
for p in parents:
rev.parent_ids.append(get_cached(p.get('revision_id')))
self._unpack_revision_properties(elt, rev)
v = elt.get('timezone')
if v is None:
rev.timezone = 0
else:
rev.timezone = int(v)
rev.message = elt.findtext('message') # text of <message>
return rev
def _unpack_revision_properties(self, elt, rev):
"""Unpack properties onto a revision."""
props_elt = elt.find('properties')
if not props_elt:
return
for prop_elt in props_elt:
if prop_elt.tag != 'property':
raise AssertionError(
"bad tag under properties list: %r" % prop_elt.tag)
name = prop_elt.get('name')
value = prop_elt.text
# If a property had an empty value ('') cElementTree reads
# that back as None, convert it back to '', so that all
# properties have string values
if value is None:
value = ''
if name in rev.properties:
raise AssertionError("repeated property %r" % name)
rev.properties[name] = value
serializer_v8 = Serializer_v8()
|