1
# Copyright (C) 2005, 2006, 2007 Canonical Ltd
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.
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.
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
24
revision as _mod_revision,
26
from bzrlib.xml_serializer import SubElement, Element, Serializer
27
from bzrlib.inventory import ROOT_ID, Inventory, InventoryEntry
28
from bzrlib.revision import Revision
29
from bzrlib.errors import BzrError
36
"'":"'", # FIXME: overkill
43
def _ensure_utf8_re():
44
"""Make sure the _utf8_re and _unicode_re regexes have been compiled."""
45
global _utf8_re, _unicode_re
47
_utf8_re = re.compile('[&<>\'\"]|[\x80-\xff]+')
48
if _unicode_re is None:
49
_unicode_re = re.compile(u'[&<>\'\"\u0080-\uffff]')
52
def _unicode_escape_replace(match, _map=_xml_escape_map):
53
"""Replace a string of non-ascii, non XML safe characters with their escape
55
This will escape both Standard XML escapes, like <>"', etc.
56
As well as escaping non ascii characters, because ElementTree did.
57
This helps us remain compatible to older versions of bzr. We may change
58
our policy in the future, though.
60
# jam 20060816 Benchmarks show that try/KeyError is faster if you
61
# expect the entity to rarely miss. There is about a 10% difference
62
# in overall time. But if you miss frequently, then if None is much
63
# faster. For our use case, we *rarely* have a revision id, file id
64
# or path name that is unicode. So use try/KeyError.
66
return _map[match.group()]
68
return "&#%d;" % ord(match.group())
71
def _utf8_escape_replace(match, _map=_xml_escape_map):
72
"""Escape utf8 characters into XML safe ones.
74
This uses 2 tricks. It is either escaping "standard" characters, like "&<>,
75
or it is handling characters with the high-bit set. For ascii characters,
76
we just lookup the replacement in the dictionary. For everything else, we
77
decode back into Unicode, and then use the XML escape code.
80
return _map[match.group()]
82
return ''.join('&#%d;' % ord(uni_chr)
83
for uni_chr in match.group().decode('utf8'))
88
def _encode_and_escape(unicode_or_utf8_str, _map=_to_escaped_map):
89
"""Encode the string into utf8, and escape invalid XML characters"""
90
# We frequently get entities we have not seen before, so it is better
91
# to check if None, rather than try/KeyError
92
text = _map.get(unicode_or_utf8_str)
94
if unicode_or_utf8_str.__class__ == unicode:
95
# The alternative policy is to do a regular UTF8 encoding
96
# and then escape only XML meta characters.
97
# Performance is equivalent once you use cache_utf8. *However*
98
# this makes the serialized texts incompatible with old versions
99
# of bzr. So no net gain. (Perhaps the read code would handle utf8
100
# better than entity escapes, but cElementTree seems to do just fine
102
text = str(_unicode_re.sub(_unicode_escape_replace,
103
unicode_or_utf8_str)) + '"'
105
# Plain strings are considered to already be in utf-8 so we do a
106
# slightly different method for escaping.
107
text = _utf8_re.sub(_utf8_escape_replace,
108
unicode_or_utf8_str) + '"'
109
_map[unicode_or_utf8_str] = text
113
def _get_utf8_or_ascii(a_str,
114
_encode_utf8=cache_utf8.encode,
115
_get_cached_ascii=cache_utf8.get_cached_ascii):
116
"""Return a cached version of the string.
118
cElementTree will return a plain string if the XML is plain ascii. It only
119
returns Unicode when it needs to. We want to work in utf-8 strings. So if
120
cElementTree returns a plain string, we can just return the cached version.
121
If it is Unicode, then we need to encode it.
123
:param a_str: An 8-bit string or Unicode as returned by
124
cElementTree.Element.get()
125
:return: A utf-8 encoded 8-bit string.
127
# This is fairly optimized because we know what cElementTree does, this is
128
# not meant as a generic function for all cases. Because it is possible for
129
# an 8-bit string to not be ascii or valid utf8.
130
if a_str.__class__ == unicode:
131
return _encode_utf8(a_str)
133
return _get_cached_ascii(a_str)
137
"""Clean out the unicode => escaped map"""
138
_to_escaped_map.clear()
141
class Serializer_v5(Serializer):
142
"""Version 5 serializer
144
Packs objects into XML and vice versa.
150
support_altered_by_hack = True
151
# This format supports the altered-by hack that reads file ids directly out
152
# of the versionedfile, without doing XML parsing.
154
supported_kinds = set(['file', 'directory', 'symlink'])
157
def _check_revisions(self, inv):
158
"""Extension point for subclasses to check during serialisation.
160
By default no checking is done.
162
:param inv: An inventory about to be serialised, to be checked.
163
:raises: AssertionError if an error has occured.
166
def write_inventory_to_lines(self, inv):
167
"""Return a list of lines with the encoded inventory."""
168
return self.write_inventory(inv, None)
170
def write_inventory_to_string(self, inv, working=False):
171
"""Just call write_inventory with a StringIO and return the value.
173
:param working: If True skip history data - text_sha1, text_size,
174
reference_revision, symlink_target.
176
sio = cStringIO.StringIO()
177
self.write_inventory(inv, sio, working)
178
return sio.getvalue()
180
def write_inventory(self, inv, f, working=False):
181
"""Write inventory to a file.
183
:param inv: the inventory to write.
184
:param f: the file to write. (May be None if the lines are the desired
186
:param working: If True skip history data - text_sha1, text_size,
187
reference_revision, symlink_target.
188
:return: The inventory as a list of lines.
191
self._check_revisions(inv)
193
append = output.append
194
self._append_inventory_root(append, inv)
195
entries = inv.iter_entries()
197
root_path, root_ie = entries.next()
198
for path, ie in entries:
199
if ie.parent_id != self.root_id:
200
parent_str = ' parent_id="'
201
parent_id = _encode_and_escape(ie.parent_id)
205
if ie.kind == 'file':
207
executable = ' executable="yes"'
211
append('<file%s file_id="%s name="%s%s%s revision="%s '
212
'text_sha1="%s" text_size="%d" />\n' % (
213
executable, _encode_and_escape(ie.file_id),
214
_encode_and_escape(ie.name), parent_str, parent_id,
215
_encode_and_escape(ie.revision), ie.text_sha1,
218
append('<file%s file_id="%s name="%s%s%s />\n' % (
219
executable, _encode_and_escape(ie.file_id),
220
_encode_and_escape(ie.name), parent_str, parent_id))
221
elif ie.kind == 'directory':
223
append('<directory file_id="%s name="%s%s%s revision="%s '
225
_encode_and_escape(ie.file_id),
226
_encode_and_escape(ie.name),
227
parent_str, parent_id,
228
_encode_and_escape(ie.revision)))
230
append('<directory file_id="%s name="%s%s%s />\n' % (
231
_encode_and_escape(ie.file_id),
232
_encode_and_escape(ie.name),
233
parent_str, parent_id))
234
elif ie.kind == 'symlink':
236
append('<symlink file_id="%s name="%s%s%s revision="%s '
237
'symlink_target="%s />\n' % (
238
_encode_and_escape(ie.file_id),
239
_encode_and_escape(ie.name),
240
parent_str, parent_id,
241
_encode_and_escape(ie.revision),
242
_encode_and_escape(ie.symlink_target)))
244
append('<symlink file_id="%s name="%s%s%s />\n' % (
245
_encode_and_escape(ie.file_id),
246
_encode_and_escape(ie.name),
247
parent_str, parent_id))
248
elif ie.kind == 'tree-reference':
249
if ie.kind not in self.supported_kinds:
250
raise errors.UnsupportedInventoryKind(ie.kind)
252
append('<tree-reference file_id="%s name="%s%s%s '
253
'revision="%s reference_revision="%s />\n' % (
254
_encode_and_escape(ie.file_id),
255
_encode_and_escape(ie.name),
256
parent_str, parent_id,
257
_encode_and_escape(ie.revision),
258
_encode_and_escape(ie.reference_revision)))
260
append('<tree-reference file_id="%s name="%s%s%s />\n' % (
261
_encode_and_escape(ie.file_id),
262
_encode_and_escape(ie.name),
263
parent_str, parent_id))
265
raise errors.UnsupportedInventoryKind(ie.kind)
266
append('</inventory>\n')
269
# Just to keep the cache from growing without bounds
270
# but we may actually not want to do clear the cache
274
def _append_inventory_root(self, append, inv):
275
"""Append the inventory root to output."""
276
if inv.root.file_id not in (None, ROOT_ID):
277
fileid1 = ' file_id="'
278
fileid2 = _encode_and_escape(inv.root.file_id)
282
if inv.revision_id is not None:
283
revid1 = ' revision_id="'
284
revid2 = _encode_and_escape(inv.revision_id)
288
append('<inventory%s%s format="5"%s%s>\n' % (
289
fileid1, fileid2, revid1, revid2))
291
def _pack_revision(self, rev):
292
"""Revision object -> xml tree"""
293
# For the XML format, we need to write them as Unicode rather than as
294
# utf-8 strings. So that cElementTree can handle properly escaping
296
decode_utf8 = cache_utf8.decode
297
revision_id = rev.revision_id
298
if isinstance(revision_id, str):
299
revision_id = decode_utf8(revision_id)
300
root = Element('revision',
301
committer = rev.committer,
302
timestamp = '%.3f' % rev.timestamp,
303
revision_id = revision_id,
304
inventory_sha1 = rev.inventory_sha1,
307
if rev.timezone is not None:
308
root.set('timezone', str(rev.timezone))
310
msg = SubElement(root, 'message')
311
msg.text = rev.message
314
pelts = SubElement(root, 'parents')
315
pelts.tail = pelts.text = '\n'
316
for parent_id in rev.parent_ids:
317
assert isinstance(parent_id, basestring)
318
_mod_revision.check_not_reserved_id(parent_id)
319
p = SubElement(pelts, 'revision_ref')
321
if isinstance(parent_id, str):
322
parent_id = decode_utf8(parent_id)
323
p.set('revision_id', parent_id)
325
self._pack_revision_properties(rev, root)
328
def _pack_revision_properties(self, rev, under_element):
329
top_elt = SubElement(under_element, 'properties')
330
for prop_name, prop_value in sorted(rev.properties.items()):
331
assert isinstance(prop_name, basestring)
332
assert isinstance(prop_value, basestring)
333
prop_elt = SubElement(top_elt, 'property')
334
prop_elt.set('name', prop_name)
335
prop_elt.text = prop_value
339
def _unpack_inventory(self, elt, revision_id):
340
"""Construct from XML Element
342
assert elt.tag == 'inventory'
343
root_id = elt.get('file_id') or ROOT_ID
344
root_id = _get_utf8_or_ascii(root_id)
346
format = elt.get('format')
347
if format is not None:
349
raise BzrError("invalid format version %r on inventory"
351
data_revision_id = elt.get('revision_id')
352
if data_revision_id is not None:
353
revision_id = cache_utf8.encode(data_revision_id)
354
inv = Inventory(root_id, revision_id=revision_id)
356
ie = self._unpack_entry(e)
357
if ie.parent_id is None:
358
ie.parent_id = root_id
360
if revision_id is not None:
361
inv.root.revision = revision_id
364
def _unpack_entry(self, elt):
366
if not InventoryEntry.versionable_kind(kind):
367
raise AssertionError('unsupported entry kind %s' % kind)
369
get_cached = _get_utf8_or_ascii
371
parent_id = elt.get('parent_id')
372
if parent_id is not None:
373
parent_id = get_cached(parent_id)
374
file_id = get_cached(elt.get('file_id'))
376
if kind == 'directory':
377
ie = inventory.InventoryDirectory(file_id,
381
ie = inventory.InventoryFile(file_id,
384
ie.text_sha1 = elt.get('text_sha1')
385
if elt.get('executable') == 'yes':
387
v = elt.get('text_size')
388
ie.text_size = v and int(v)
389
elif kind == 'symlink':
390
ie = inventory.InventoryLink(file_id,
393
ie.symlink_target = elt.get('symlink_target')
395
raise errors.UnsupportedInventoryKind(kind)
396
revision = elt.get('revision')
397
if revision is not None:
398
revision = get_cached(revision)
399
ie.revision = revision
403
def _unpack_revision(self, elt):
404
"""XML Element -> Revision object"""
405
assert elt.tag == 'revision'
406
format = elt.get('format')
407
if format is not None:
409
raise BzrError("invalid format version %r on inventory"
411
get_cached = _get_utf8_or_ascii
412
rev = Revision(committer = elt.get('committer'),
413
timestamp = float(elt.get('timestamp')),
414
revision_id = get_cached(elt.get('revision_id')),
415
inventory_sha1 = elt.get('inventory_sha1')
417
parents = elt.find('parents') or []
419
assert p.tag == 'revision_ref', \
420
"bad parent node tag %r" % p.tag
421
rev.parent_ids.append(get_cached(p.get('revision_id')))
422
self._unpack_revision_properties(elt, rev)
423
v = elt.get('timezone')
427
rev.timezone = int(v)
428
rev.message = elt.findtext('message') # text of <message>
431
def _unpack_revision_properties(self, elt, rev):
432
"""Unpack properties onto a revision."""
433
props_elt = elt.find('properties')
434
assert len(rev.properties) == 0
437
for prop_elt in props_elt:
438
assert prop_elt.tag == 'property', \
439
"bad tag under properties list: %r" % prop_elt.tag
440
name = prop_elt.get('name')
441
value = prop_elt.text
442
# If a property had an empty value ('') cElementTree reads
443
# that back as None, convert it back to '', so that all
444
# properties have string values
447
assert name not in rev.properties, \
448
"repeated property %r" % name
449
rev.properties[name] = value
452
serializer_v5 = Serializer_v5()
added
removed
bzrlib/xml5.py
