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., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
15
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
17
"""XML externalization support."""
33
30
from xml.etree.cElementTree import (ElementTree, SubElement, Element,
34
31
XMLTreeBuilder, fromstring, tostring)
35
32
import xml.etree as elementtree
36
# Also import ElementTree module so monkey-patching below always works
37
import xml.etree.ElementTree
38
33
except ImportError:
39
34
from cElementTree import (ElementTree, SubElement, Element,
40
35
XMLTreeBuilder, fromstring, tostring)
41
import elementtree.ElementTree
42
37
ParseError = SyntaxError
43
38
except ImportError:
44
39
mutter('WARNING: using slower ElementTree; consider installing cElementTree'
53
48
from bzrlib import errors
56
class XMLSerializer(Serializer):
57
"""Abstract XML object serialize/deserialize"""
59
squashes_xml_invalid_characters = True
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.
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.
51
class Serializer(object):
52
"""Abstract object serialize/deserialize"""
53
def write_inventory(self, inv, f):
54
"""Write inventory to a file"""
55
elt = self._pack_inventory(inv)
56
self._write_element(elt, f)
58
def write_inventory_to_string(self, inv):
59
return tostring(self._pack_inventory(inv)) + '\n'
61
def read_inventory_from_string(self, xml_string):
82
return self._unpack_inventory(fromstring(xml_string), revision_id,
83
entry_cache=entry_cache,
84
return_from_cache=return_from_cache)
63
return self._unpack_inventory(fromstring(xml_string))
85
64
except ParseError, e:
86
65
raise errors.UnexpectedInventoryFormat(e)
88
def read_inventory(self, f, revision_id=None):
67
def read_inventory(self, f):
91
return self._unpack_inventory(self._read_element(f),
69
return self._unpack_inventory(self._read_element(f))
95
70
except ParseError, e:
96
71
raise errors.UnexpectedInventoryFormat(e)
115
90
return ElementTree().parse(f)
118
def escape_invalid_chars(message):
119
"""Escape the XML-invalid characters in a commit message.
121
:param message: Commit message to escape
122
:return: tuple with escaped message and number of characters escaped
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'),
93
# performance tuning for elementree's serialiser. This should be
94
# sent upstream - RBC 20060523.
95
# the functions here are patched into elementtree at runtime.
97
escape_re = re.compile("[&'\"<>]")
100
"'":"'", # FIXME: overkill
105
def _escape_replace(match, map=escape_map):
106
return map[match.group()]
108
def _escape_attrib(text, encoding=None, replace=None):
109
# escape attribute value
113
text = elementtree.ElementTree._encode(text, encoding)
115
return elementtree.ElementTree._encode_entity(text)
117
return escape_re.sub(_escape_replace, text)
119
text = replace(text, "&", "&")
120
text = replace(text, "'", "'") # FIXME: overkill
121
text = replace(text, "\"", """)
122
text = replace(text, "<", "<")
123
text = replace(text, ">", ">")
125
except (TypeError, AttributeError):
126
elementtree.ElementTree._raise_serialization_error(text)
128
elementtree.ElementTree._escape_attrib = _escape_attrib
130
escape_cdata_re = re.compile("[&<>]")
136
def _escape_cdata_replace(match, map=escape_cdata_map):
137
return map[match.group()]
139
def _escape_cdata(text, encoding=None, replace=None):
140
# escape character data
144
text = elementtree.ElementTree._encode(text, encoding)
146
return elementtree.ElementTree._encode_entity(text)
148
return escape_cdata_re.sub(_escape_cdata_replace, text)
150
text = replace(text, "&", "&")
151
text = replace(text, "<", "<")
152
text = replace(text, ">", ">")
154
except (TypeError, AttributeError):
155
elementtree.ElementTree._raise_serialization_error(text)
157
elementtree.ElementTree._escape_cdata = _escape_cdata