← Back to branch summary

~bzr-pqm/bzr/bzr.dev

« back to all changes in this revision

Viewing changes to bzrlib/xml_serializer.py

  • Committer: Martin Pool
  • Date: 2009-08-20 04:53:23 UTC
  • mto: This revision was merged to the branch mainline in revision 4632.
  • Revision ID: mbp@sourcefrog.net-20090820045323-4hsicfa87pdq3l29
Correction to xdg_cache_dir and add a simple test

Show diffs side-by-side

added added

removed removed

Lines of Context:
bzrlib/xml_serializer.py
1
 
# -*- coding: UTF-8 -*-
 
1
# Copyright (C) 2005, 2006 Canonical Ltd
2
2
#
3
3
# This program is free software; you can redistribute it and/or modify
4
4
# it under the terms of the GNU General Public License as published by
12
12
#
13
13
# You should have received a copy of the GNU General Public License
14
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
 
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
16
16
 
17
17
"""XML externalization support."""
18
18
 
22
22
# importing this module is fairly slow because it has to load several
23
23
# ElementTree bits
24
24
 
 
25
from bzrlib.serializer import Serializer
25
26
from bzrlib.trace import mutter, warning
26
27
 
27
28
try:
28
 
    from cElementTree import (ElementTree, SubElement, Element,
29
 
                              XMLTreeBuilder, fromstring, tostring)
30
 
    import elementtree
 
29
    try:
 
30
        # it's in this package in python2.5
 
31
        from xml.etree.cElementTree import (ElementTree, SubElement, Element,
 
32
            XMLTreeBuilder, fromstring, tostring)
 
33
        import xml.etree as elementtree
 
34
    except ImportError:
 
35
        from cElementTree import (ElementTree, SubElement, Element,
 
36
                                  XMLTreeBuilder, fromstring, tostring)
 
37
        import elementtree.ElementTree
 
38
    ParseError = SyntaxError
31
39
except ImportError:
32
40
    mutter('WARNING: using slower ElementTree; consider installing cElementTree'
33
41
           " and make sure it's on your PYTHONPATH")
 
42
    # this copy is shipped with bzr
34
43
    from util.elementtree.ElementTree import (ElementTree, SubElement,
35
44
                                              Element, XMLTreeBuilder,
36
45
                                              fromstring, tostring)
37
46
    import util.elementtree as elementtree
38
 
 
39
 
from bzrlib.errors import BzrError
40
 
 
41
 
 
42
 
class Serializer(object):
43
 
    """Abstract object serialize/deserialize"""
44
 
    def write_inventory(self, inv, f):
45
 
        """Write inventory to a file"""
46
 
        elt = self._pack_inventory(inv)
47
 
        self._write_element(elt, f)
48
 
 
49
 
    def write_inventory_to_string(self, inv):
50
 
        return tostring(self._pack_inventory(inv)) + '\n'
51
 
 
52
 
    def read_inventory_from_string(self, xml_string):
53
 
        return self._unpack_inventory(fromstring(xml_string))
54
 
 
55
 
    def read_inventory(self, f):
56
 
        return self._unpack_inventory(self._read_element(f))
 
47
    from xml.parsers.expat import ExpatError as ParseError
 
48
 
 
49
from bzrlib import errors
 
50
 
 
51
 
 
52
class XMLSerializer(Serializer):
 
53
    """Abstract XML object serialize/deserialize"""
 
54
 
 
55
    squashes_xml_invalid_characters = True
 
56
 
 
57
    def read_inventory_from_string(self, xml_string, revision_id=None,
 
58
                                   entry_cache=None):
 
59
        """Read xml_string into an inventory object.
 
60
 
 
61
        :param xml_string: The xml to read.
 
62
        :param revision_id: If not-None, the expected revision id of the
 
63
            inventory. Some serialisers use this to set the results' root
 
64
            revision. This should be supplied for deserialising all
 
65
            from-repository inventories so that xml5 inventories that were
 
66
            serialised without a revision identifier can be given the right
 
67
            revision id (but not for working tree inventories where users can
 
68
            edit the data without triggering checksum errors or anything).
 
69
        :param entry_cache: An optional cache of InventoryEntry objects. If
 
70
            supplied we will look up entries via (file_id, revision_id) which
 
71
            should map to a valid InventoryEntry (File/Directory/etc) object.
 
72
        """
 
73
        try:
 
74
            return self._unpack_inventory(fromstring(xml_string), revision_id,
 
75
                                          entry_cache=entry_cache)
 
76
        except ParseError, e:
 
77
            raise errors.UnexpectedInventoryFormat(e)
 
78
 
 
79
    def read_inventory(self, f, revision_id=None):
 
80
        try:
 
81
            return self._unpack_inventory(self._read_element(f),
 
82
                revision_id=None)
 
83
        except ParseError, e:
 
84
            raise errors.UnexpectedInventoryFormat(e)
57
85
 
58
86
    def write_revision(self, rev, f):
59
87
        self._write_element(self._pack_revision(rev), f)
89
117
    }
90
118
def _escape_replace(match, map=escape_map):
91
119
    return map[match.group()]
92
 
 
 
120
 
93
121
def _escape_attrib(text, encoding=None, replace=None):
94
122
    # escape attribute value
95
123
    try:
120
148
    }
121
149
def _escape_cdata_replace(match, map=escape_cdata_map):
122
150
    return map[match.group()]
123
 
 
 
151
 
124
152
def _escape_cdata(text, encoding=None, replace=None):
125
153
    # escape character data
126
154
    try:
140
168
        elementtree.ElementTree._raise_serialization_error(text)
141
169
 
142
170
elementtree.ElementTree._escape_cdata = _escape_cdata
 
171
 
 
172
 
 
173
def escape_invalid_chars(message):
 
174
    """Escape the XML-invalid characters in a commit message.
 
175
 
 
176
    :param message: Commit message to escape
 
177
    :return: tuple with escaped message and number of characters escaped
 
178
    """
 
179
    if message is None:
 
180
        return None, 0
 
181
    # Python strings can include characters that can't be
 
182
    # represented in well-formed XML; escape characters that
 
183
    # aren't listed in the XML specification
 
184
    # (http://www.w3.org/TR/REC-xml/#NT-Char).
 
185
    return re.subn(u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
 
186
            lambda match: match.group(0).encode('unicode_escape'),
 
187
            message)