/brz/remove-bazaar

To get this branch, use:
bzr branch http://gegoxaren.bato24.eu/bzr/brz/remove-bazaar

« back to all changes in this revision

Viewing changes to bzrlib/xml_serializer.py

  • Committer: Marien Zwart
  • Date: 2006-09-22 00:51:36 UTC
  • mto: (2029.1.4 python25)
  • mto: This revision was merged to the branch mainline in revision 2030.
  • Revision ID: marienz@gentoo.org-20060922005136-6b5e0b856a4baff3
Handle the different exception (non-c)ElementTree raises.

Show diffs side-by-side

added added

removed removed

Lines of Context:
bzrlib/xml_serializer.py
 
1
# -*- coding: UTF-8 -*-
 
2
#
 
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.
 
7
#
 
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.
 
12
#
 
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
 
16
 
 
17
"""XML externalization support."""
 
18
 
 
19
# "XML is like violence: if it doesn't solve your problem, you aren't
 
20
# using enough of it." -- various
 
21
 
 
22
# importing this module is fairly slow because it has to load several
 
23
# ElementTree bits
 
24
 
 
25
from bzrlib.trace import mutter, warning
 
26
 
 
27
try:
 
28
    from cElementTree import (ElementTree, SubElement, Element,
 
29
                              XMLTreeBuilder, fromstring, tostring)
 
30
    import elementtree
 
31
    ParseError = SyntaxError
 
32
except ImportError:
 
33
    mutter('WARNING: using slower ElementTree; consider installing cElementTree'
 
34
           " and make sure it's on your PYTHONPATH")
 
35
    from util.elementtree.ElementTree import (ElementTree, SubElement,
 
36
                                              Element, XMLTreeBuilder,
 
37
                                              fromstring, tostring)
 
38
    import util.elementtree as elementtree
 
39
    from xml.parsers.expat import ExpatError as ParseError
 
40
 
 
41
from bzrlib import errors
 
42
 
 
43
 
 
44
class Serializer(object):
 
45
    """Abstract object serialize/deserialize"""
 
46
    def write_inventory(self, inv, f):
 
47
        """Write inventory to a file"""
 
48
        elt = self._pack_inventory(inv)
 
49
        self._write_element(elt, f)
 
50
 
 
51
    def write_inventory_to_string(self, inv):
 
52
        return tostring(self._pack_inventory(inv)) + '\n'
 
53
 
 
54
    def read_inventory_from_string(self, xml_string):
 
55
        try:
 
56
            return self._unpack_inventory(fromstring(xml_string))
 
57
        except ParseError, e:
 
58
            raise errors.UnexpectedInventoryFormat(e)
 
59
 
 
60
    def read_inventory(self, f):
 
61
        try:
 
62
            return self._unpack_inventory(self._read_element(f))
 
63
        except ParseError, e:
 
64
            raise errors.UnexpectedInventoryFormat(e)
 
65
 
 
66
    def write_revision(self, rev, f):
 
67
        self._write_element(self._pack_revision(rev), f)
 
68
 
 
69
    def write_revision_to_string(self, rev):
 
70
        return tostring(self._pack_revision(rev)) + '\n'
 
71
 
 
72
    def read_revision(self, f):
 
73
        return self._unpack_revision(self._read_element(f))
 
74
 
 
75
    def read_revision_from_string(self, xml_string):
 
76
        return self._unpack_revision(fromstring(xml_string))
 
77
 
 
78
    def _write_element(self, elt, f):
 
79
        ElementTree(elt).write(f, 'utf-8')
 
80
        f.write('\n')
 
81
 
 
82
    def _read_element(self, f):
 
83
        return ElementTree().parse(f)
 
84
 
 
85
 
 
86
# performance tuning for elementree's serialiser. This should be
 
87
# sent upstream - RBC 20060523.
 
88
# the functions here are patched into elementtree at runtime.
 
89
import re
 
90
escape_re = re.compile("[&'\"<>]")
 
91
escape_map = {
 
92
    "&":'&amp;',
 
93
    "'":"&apos;", # FIXME: overkill
 
94
    "\"":"&quot;",
 
95
    "<":"&lt;",
 
96
    ">":"&gt;",
 
97
    }
 
98
def _escape_replace(match, map=escape_map):
 
99
    return map[match.group()]
 
100
 
 
101
def _escape_attrib(text, encoding=None, replace=None):
 
102
    # escape attribute value
 
103
    try:
 
104
        if encoding:
 
105
            try:
 
106
                text = elementtree.ElementTree._encode(text, encoding)
 
107
            except UnicodeError:
 
108
                return elementtree.ElementTree._encode_entity(text)
 
109
        if replace is None:
 
110
            return escape_re.sub(_escape_replace, text)
 
111
        else:
 
112
            text = replace(text, "&", "&amp;")
 
113
            text = replace(text, "'", "&apos;") # FIXME: overkill
 
114
            text = replace(text, "\"", "&quot;")
 
115
            text = replace(text, "<", "&lt;")
 
116
            text = replace(text, ">", "&gt;")
 
117
            return text
 
118
    except (TypeError, AttributeError):
 
119
        elementtree.ElementTree._raise_serialization_error(text)
 
120
 
 
121
elementtree.ElementTree._escape_attrib = _escape_attrib
 
122
 
 
123
escape_cdata_re = re.compile("[&<>]")
 
124
escape_cdata_map = {
 
125
    "&":'&amp;',
 
126
    "<":"&lt;",
 
127
    ">":"&gt;",
 
128
    }
 
129
def _escape_cdata_replace(match, map=escape_cdata_map):
 
130
    return map[match.group()]
 
131
 
 
132
def _escape_cdata(text, encoding=None, replace=None):
 
133
    # escape character data
 
134
    try:
 
135
        if encoding:
 
136
            try:
 
137
                text = elementtree.ElementTree._encode(text, encoding)
 
138
            except UnicodeError:
 
139
                return elementtree.ElementTree._encode_entity(text)
 
140
        if replace is None:
 
141
            return escape_cdata_re.sub(_escape_cdata_replace, text)
 
142
        else:
 
143
            text = replace(text, "&", "&amp;")
 
144
            text = replace(text, "<", "&lt;")
 
145
            text = replace(text, ">", "&gt;")
 
146
            return text
 
147
    except (TypeError, AttributeError):
 
148
        elementtree.ElementTree._raise_serialization_error(text)
 
149
 
 
150
elementtree.ElementTree._escape_cdata = _escape_cdata