← Back to branch summary

~bzr-pqm/bzr/bzr.dev

« back to all changes in this revision

Viewing changes to bzrlib/xml_serializer.py

  • Committer: Andrew Bennetts
  • Date: 2010-10-08 08:15:14 UTC
  • mto: This revision was merged to the branch mainline in revision 5498.
  • Revision ID: andrew.bennetts@canonical.com-20101008081514-dviqzrdfwyzsqbz2
Split NEWS into per-release doc/en/release-notes/bzr-*.txt

Show diffs side-by-side

added added

removed removed

Lines of Context:
bzrlib/xml_serializer.py
1
 
# Copyright (C) 2005, 2006 Canonical Ltd
 
1
# Copyright (C) 2005-2010 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 import registry
26
 
from bzrlib.trace import mutter, warning
 
25
import re
 
26
 
 
27
from bzrlib.serializer import Serializer
 
28
from bzrlib.trace import mutter
27
29
 
28
30
try:
29
31
    try:
31
33
        from xml.etree.cElementTree import (ElementTree, SubElement, Element,
32
34
            XMLTreeBuilder, fromstring, tostring)
33
35
        import xml.etree as elementtree
 
36
        # Also import ElementTree module so monkey-patching below always works
 
37
        import xml.etree.ElementTree
34
38
    except ImportError:
35
39
        from cElementTree import (ElementTree, SubElement, Element,
36
40
                                  XMLTreeBuilder, fromstring, tostring)
37
 
        import elementtree
 
41
        import elementtree.ElementTree
38
42
    ParseError = SyntaxError
39
43
except ImportError:
40
44
    mutter('WARNING: using slower ElementTree; consider installing cElementTree'
49
53
from bzrlib import errors
50
54
 
51
55
 
52
 
class Serializer(object):
53
 
    """Abstract object serialize/deserialize"""
54
 
    def write_inventory(self, inv, f):
55
 
        """Write inventory to a file"""
56
 
        elt = self._pack_inventory(inv)
57
 
        self._write_element(elt, f)
58
 
 
59
 
    def write_inventory_to_string(self, inv):
60
 
        return tostring(self._pack_inventory(inv)) + '\n'
61
 
 
62
 
    def read_inventory_from_string(self, xml_string):
 
56
class XMLSerializer(Serializer):
 
57
    """Abstract XML object serialize/deserialize"""
 
58
 
 
59
    squashes_xml_invalid_characters = True
 
60
 
 
61
    def read_inventory_from_string(self, xml_string, revision_id=None,
 
62
                                   entry_cache=None, return_from_cache=False):
 
63
        """Read xml_string into an inventory object.
 
64
 
 
65
        :param xml_string: The xml to read.
 
66
        :param revision_id: If not-None, the expected revision id of the
 
67
            inventory. Some serialisers use this to set the results' root
 
68
            revision. This should be supplied for deserialising all
 
69
            from-repository inventories so that xml5 inventories that were
 
70
            serialised without a revision identifier can be given the right
 
71
            revision id (but not for working tree inventories where users can
 
72
            edit the data without triggering checksum errors or anything).
 
73
        :param entry_cache: An optional cache of InventoryEntry objects. If
 
74
            supplied we will look up entries via (file_id, revision_id) which
 
75
            should map to a valid InventoryEntry (File/Directory/etc) object.
 
76
        :param return_from_cache: Return entries directly from the cache,
 
77
            rather than copying them first. This is only safe if the caller
 
78
            promises not to mutate the returned inventory entries, but it can
 
79
            make some operations significantly faster.
 
80
        """
63
81
        try:
64
 
            return self._unpack_inventory(fromstring(xml_string))
 
82
            return self._unpack_inventory(fromstring(xml_string), revision_id,
 
83
                                          entry_cache=entry_cache,
 
84
                                          return_from_cache=return_from_cache)
65
85
        except ParseError, e:
66
86
            raise errors.UnexpectedInventoryFormat(e)
67
87
 
68
 
    def read_inventory(self, f):
 
88
    def read_inventory(self, f, revision_id=None):
69
89
        try:
70
 
            return self._unpack_inventory(self._read_element(f))
 
90
            try:
 
91
                return self._unpack_inventory(self._read_element(f),
 
92
                    revision_id=None)
 
93
            finally:
 
94
                f.close()
71
95
        except ParseError, e:
72
96
            raise errors.UnexpectedInventoryFormat(e)
73
97
 
91
115
        return ElementTree().parse(f)
92
116
 
93
117
 
94
 
# performance tuning for elementree's serialiser. This should be
95
 
# sent upstream - RBC 20060523.
96
 
# the functions here are patched into elementtree at runtime.
97
 
import re
98
 
escape_re = re.compile("[&'\"<>]")
99
 
escape_map = {
100
 
    "&":'&amp;',
101
 
    "'":"&apos;", # FIXME: overkill
102
 
    "\"":"&quot;",
103
 
    "<":"&lt;",
104
 
    ">":"&gt;",
105
 
    }
106
 
def _escape_replace(match, map=escape_map):
107
 
    return map[match.group()]
108
 
 
109
 
def _escape_attrib(text, encoding=None, replace=None):
110
 
    # escape attribute value
111
 
    try:
112
 
        if encoding:
113
 
            try:
114
 
                text = elementtree.ElementTree._encode(text, encoding)
115
 
            except UnicodeError:
116
 
                return elementtree.ElementTree._encode_entity(text)
117
 
        if replace is None:
118
 
            return escape_re.sub(_escape_replace, text)
119
 
        else:
120
 
            text = replace(text, "&", "&amp;")
121
 
            text = replace(text, "'", "&apos;") # FIXME: overkill
122
 
            text = replace(text, "\"", "&quot;")
123
 
            text = replace(text, "<", "&lt;")
124
 
            text = replace(text, ">", "&gt;")
125
 
            return text
126
 
    except (TypeError, AttributeError):
127
 
        elementtree.ElementTree._raise_serialization_error(text)
128
 
 
129
 
elementtree.ElementTree._escape_attrib = _escape_attrib
130
 
 
131
 
escape_cdata_re = re.compile("[&<>]")
132
 
escape_cdata_map = {
133
 
    "&":'&amp;',
134
 
    "<":"&lt;",
135
 
    ">":"&gt;",
136
 
    }
137
 
def _escape_cdata_replace(match, map=escape_cdata_map):
138
 
    return map[match.group()]
139
 
 
140
 
def _escape_cdata(text, encoding=None, replace=None):
141
 
    # escape character data
142
 
    try:
143
 
        if encoding:
144
 
            try:
145
 
                text = elementtree.ElementTree._encode(text, encoding)
146
 
            except UnicodeError:
147
 
                return elementtree.ElementTree._encode_entity(text)
148
 
        if replace is None:
149
 
            return escape_cdata_re.sub(_escape_cdata_replace, text)
150
 
        else:
151
 
            text = replace(text, "&", "&amp;")
152
 
            text = replace(text, "<", "&lt;")
153
 
            text = replace(text, ">", "&gt;")
154
 
            return text
155
 
    except (TypeError, AttributeError):
156
 
        elementtree.ElementTree._raise_serialization_error(text)
157
 
 
158
 
elementtree.ElementTree._escape_cdata = _escape_cdata
159
 
 
160
 
 
161
 
class SerializerRegistry(registry.Registry):
162
 
    """Registry for serializer objects"""
163
 
 
164
 
 
165
 
format_registry = SerializerRegistry()
166
 
format_registry.register_lazy('4', 'bzrlib.xml4', 'serializer_v4')
167
 
format_registry.register_lazy('5', 'bzrlib.xml5', 'serializer_v5')
168
 
format_registry.register_lazy('6', 'bzrlib.xml6', 'serializer_v6')
169
 
format_registry.register_lazy('7', 'bzrlib.xml7', 'serializer_v7')
 
118
def escape_invalid_chars(message):
 
119
    """Escape the XML-invalid characters in a commit message.
 
120
 
 
121
    :param message: Commit message to escape
 
122
    :return: tuple with escaped message and number of characters escaped
 
123
    """
 
124
    if message is None:
 
125
        return None, 0
 
126
    # Python strings can include characters that can't be
 
127
    # represented in well-formed XML; escape characters that
 
128
    # aren't listed in the XML specification
 
129
    # (http://www.w3.org/TR/REC-xml/#NT-Char).
 
130
    return re.subn(u'[^\x09\x0A\x0D\u0020-\uD7FF\uE000-\uFFFD]+',
 
131
            lambda match: match.group(0).encode('unicode_escape'),
 
132
            message)