1
# Copyright (C) 2005, 2006 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
from bzrlib.xml_serializer import SubElement, Element, Serializer
25
from bzrlib.inventory import ROOT_ID, Inventory, InventoryEntry
26
from bzrlib.revision import Revision
27
from bzrlib.errors import BzrError
33
"'":"'", # FIXME: overkill
40
def _ensure_utf8_re():
41
"""Make sure the _utf8_re regex has been compiled"""
43
if _utf8_re is not None:
45
_utf8_re = re.compile(u'[&<>\'\"\u0080-\uffff]')
48
def _utf8_escape_replace(match, _map=_utf8_escape_map):
49
"""Replace a string of non-ascii, non XML safe characters with their escape
51
This will escape both Standard XML escapes, like <>"', etc.
52
As well as escaping non ascii characters, because ElementTree did.
53
This helps us remain compatible to older versions of bzr. We may change
54
our policy in the future, though.
56
# jam 20060816 Benchmarks show that try/KeyError is faster if you
57
# expect the entity to rarely miss. There is about a 10% difference
58
# in overall time. But if you miss frequently, then if None is much
59
# faster. For our use case, we *rarely* have a revision id, file id
60
# or path name that is unicode. So use try/KeyError.
62
return _map[match.group()]
64
return "&#%d;" % ord(match.group())
67
_unicode_to_escaped_map = {}
69
def _encode_and_escape(unicode_str, _map=_unicode_to_escaped_map):
70
"""Encode the string into utf8, and escape invalid XML characters"""
71
# We frequently get entities we have not seen before, so it is better
72
# to check if None, rather than try/KeyError
73
text = _map.get(unicode_str)
75
# The alternative policy is to do a regular UTF8 encoding
76
# and then escape only XML meta characters.
77
# Performance is equivalent once you use cache_utf8. *However*
78
# this makes the serialized texts incompatible with old versions
79
# of bzr. So no net gain. (Perhaps the read code would handle utf8
80
# better than entity escapes, but cElementTree seems to do just fine
82
text = str(_utf8_re.sub(_utf8_escape_replace, unicode_str)) + '"'
83
_map[unicode_str] = text
88
"""Clean out the unicode => escaped map"""
89
_unicode_to_escaped_map.clear()
92
class Serializer_v5(Serializer):
93
"""Version 5 serializer
95
Packs objects into XML and vice versa.
100
def write_inventory_to_string(self, inv):
101
"""Just call write_inventory with a StringIO and return the value"""
102
sio = cStringIO.StringIO()
103
self.write_inventory(inv, sio)
104
return sio.getvalue()
106
def write_inventory(self, inv, f):
107
"""Write inventory to a file.
109
:param inv: the inventory to write.
110
:param f: the file to write.
114
append = output.append
115
self._append_inventory_root(append, inv)
116
entries = inv.iter_entries()
118
root_path, root_ie = entries.next()
119
for path, ie in entries:
120
self._append_entry(append, ie)
121
append('</inventory>\n')
123
# Just to keep the cache from growing without bounds
124
# but we may actually not want to do clear the cache
127
def _append_inventory_root(self, append, inv):
128
"""Append the inventory root to output."""
130
if inv.root.file_id not in (None, ROOT_ID):
132
append(_encode_and_escape(inv.root.file_id))
133
append(' format="5"')
134
if inv.revision_id is not None:
135
append(' revision_id="')
136
append(_encode_and_escape(inv.revision_id))
139
def _append_entry(self, append, ie):
140
"""Convert InventoryEntry to XML element and append to output."""
141
# TODO: should just be a plain assertion
142
assert InventoryEntry.versionable_kind(ie.kind), \
143
'unsupported entry kind %s' % ie.kind
148
append(' executable="yes"')
150
append(_encode_and_escape(ie.file_id))
152
append(_encode_and_escape(ie.name))
153
if ie.parent_id != ROOT_ID:
154
assert isinstance(ie.parent_id, basestring)
155
append(' parent_id="')
156
append(_encode_and_escape(ie.parent_id))
157
if ie.revision is not None:
158
append(' revision="')
159
append(_encode_and_escape(ie.revision))
160
if ie.symlink_target is not None:
161
append(' symlink_target="')
162
append(_encode_and_escape(ie.symlink_target))
163
if ie.text_sha1 is not None:
164
append(' text_sha1="')
167
if ie.text_size is not None:
168
append(' text_size="%d"' % ie.text_size)
172
def _pack_inventory(self, inv):
173
"""Convert to XML Element"""
174
entries = inv.iter_entries()
175
e = Element('inventory',
178
path, root = entries.next()
179
if root.file_id not in (None, ROOT_ID):
180
e.set('file_id', root.file_id)
181
if inv.revision_id is not None:
182
e.set('revision_id', inv.revision_id)
183
for path, ie in entries:
184
e.append(self._pack_entry(ie))
187
def _pack_entry(self, ie):
188
"""Convert InventoryEntry to XML element"""
189
# TODO: should just be a plain assertion
190
if not InventoryEntry.versionable_kind(ie.kind):
191
raise AssertionError('unsupported entry kind %s' % ie.kind)
193
e.set('name', ie.name)
194
e.set('file_id', ie.file_id)
196
if ie.text_size != None:
197
e.set('text_size', '%d' % ie.text_size)
199
for f in ['text_sha1', 'revision', 'symlink_target']:
205
e.set('executable', 'yes')
207
# to be conservative, we don't externalize the root pointers
208
# for now, leaving them as null in the xml form. in a future
209
# version it will be implied by nested elements.
210
if ie.parent_id != ROOT_ID:
211
assert isinstance(ie.parent_id, basestring)
212
e.set('parent_id', ie.parent_id)
216
def _pack_revision(self, rev):
217
"""Revision object -> xml tree"""
218
root = Element('revision',
219
committer = rev.committer,
220
timestamp = '%.9f' % rev.timestamp,
221
revision_id = rev.revision_id,
222
inventory_sha1 = rev.inventory_sha1,
225
if rev.timezone is not None:
226
root.set('timezone', str(rev.timezone))
228
msg = SubElement(root, 'message')
229
msg.text = rev.message
232
pelts = SubElement(root, 'parents')
233
pelts.tail = pelts.text = '\n'
234
for parent_id in rev.parent_ids:
235
assert isinstance(parent_id, basestring)
236
p = SubElement(pelts, 'revision_ref')
238
p.set('revision_id', parent_id)
240
self._pack_revision_properties(rev, root)
244
def _pack_revision_properties(self, rev, under_element):
245
top_elt = SubElement(under_element, 'properties')
246
for prop_name, prop_value in sorted(rev.properties.items()):
247
assert isinstance(prop_name, basestring)
248
assert isinstance(prop_value, basestring)
249
prop_elt = SubElement(top_elt, 'property')
250
prop_elt.set('name', prop_name)
251
prop_elt.text = prop_value
256
def _unpack_inventory(self, elt):
257
"""Construct from XML Element
259
assert elt.tag == 'inventory'
260
root_id = elt.get('file_id') or ROOT_ID
261
format = elt.get('format')
262
if format is not None:
264
raise BzrError("invalid format version %r on inventory"
266
revision_id = elt.get('revision_id')
267
if revision_id is not None:
268
revision_id = cache_utf8.get_cached_unicode(revision_id)
269
inv = Inventory(root_id, revision_id=revision_id)
271
ie = self._unpack_entry(e)
272
if ie.parent_id == ROOT_ID:
273
ie.parent_id = root_id
278
def _unpack_entry(self, elt):
280
if not InventoryEntry.versionable_kind(kind):
281
raise AssertionError('unsupported entry kind %s' % kind)
283
get_cached = cache_utf8.get_cached_unicode
285
parent_id = elt.get('parent_id')
286
if parent_id == None:
288
#parent_id = get_cached(parent_id)
289
#file_id = get_cached(elt.get('file_id'))
290
file_id = elt.get('file_id')
292
if kind == 'directory':
293
ie = inventory.InventoryDirectory(file_id,
297
ie = inventory.InventoryFile(file_id,
300
ie.text_sha1 = elt.get('text_sha1')
301
if elt.get('executable') == 'yes':
303
v = elt.get('text_size')
304
ie.text_size = v and int(v)
305
elif kind == 'symlink':
306
ie = inventory.InventoryLink(file_id,
309
ie.symlink_target = elt.get('symlink_target')
311
raise BzrError("unknown kind %r" % kind)
312
revision = elt.get('revision')
313
if revision is not None:
314
revision = get_cached(revision)
315
ie.revision = revision
320
def _unpack_revision(self, elt):
321
"""XML Element -> Revision object"""
322
assert elt.tag == 'revision'
323
format = elt.get('format')
324
if format is not None:
326
raise BzrError("invalid format version %r on inventory"
328
get_cached = cache_utf8.get_cached_unicode
329
rev = Revision(committer = elt.get('committer'),
330
timestamp = float(elt.get('timestamp')),
331
revision_id = get_cached(elt.get('revision_id')),
332
inventory_sha1 = elt.get('inventory_sha1')
334
parents = elt.find('parents') or []
336
assert p.tag == 'revision_ref', \
337
"bad parent node tag %r" % p.tag
338
rev.parent_ids.append(get_cached(p.get('revision_id')))
339
self._unpack_revision_properties(elt, rev)
340
v = elt.get('timezone')
344
rev.timezone = int(v)
345
rev.message = elt.findtext('message') # text of <message>
349
def _unpack_revision_properties(self, elt, rev):
350
"""Unpack properties onto a revision."""
351
props_elt = elt.find('properties')
352
assert len(rev.properties) == 0
355
for prop_elt in props_elt:
356
assert prop_elt.tag == 'property', \
357
"bad tag under properties list: %r" % prop_elt.tag
358
name = prop_elt.get('name')
359
value = prop_elt.text
360
# If a property had an empty value ('') cElementTree reads
361
# that back as None, convert it back to '', so that all
362
# properties have string values
365
assert name not in rev.properties, \
366
"repeated property %r" % name
367
rev.properties[name] = value
370
serializer_v5 = Serializer_v5()