« back to all changes in this revision
Viewing changes to bzrlib/repository.py
- browse files at revision 5582.10.64
- compare with another revision
- compare with revision 5401
- download diff
- download tarball
- view history from revision 5582.10.64
- Committer: Jelmer Vernooij
- Date: 2011-02-21 23:43:10 UTC
- mfrom: (5675.2.1 repository-format-deprecation)
- mto: (5582.12.2 weave-plugin)
- mto: This revision was merged to the branch mainline in revision 5718.
- Revision ID: jelmer@samba.org-20110221234310-tpda0135m9dibmyh
Merge repository-format-deprecation branch.
added
removed
bzrlib/repository.py
34
34
graph,
35
35
inventory,
36
36
inventory_delta,
37
lazy_regex,
38
37
lockable_files,
39
38
lockdir,
40
39
lru_cache,
988
987
# in a Repository class subclass rather than to override
989
988
# get_commit_builder.
990
989
_commit_builder_class = CommitBuilder
991
# The search regex used by xml based repositories to determine what things
992
# where changed in a single commit.
993
_file_ids_altered_regex = lazy_regex.lazy_compile(
994
r'file_id="(?P<file_id>[^"]+)"'
995
r'.* revision="(?P<revision_id>[^"]+)"'
996
)
997
990
998
991
def abort_write_group(self, suppress_errors=False):
999
992
"""Commit the contents accrued within the current write group.
2063
2056
w = self.inventories
2064
2057
pb = ui.ui_factory.nested_progress_bar()
2065
2058
try:
2066
return self._find_text_key_references_from_xml_inventory_lines(
2059
return self._serializer._find_text_key_references(
2067
2060
w.iter_lines_added_or_present_in_keys(revision_keys, pb=pb))
2068
2061
finally:
2069
2062
pb.finished()
2070
2063
2071
def _find_text_key_references_from_xml_inventory_lines(self,
2072
line_iterator):
2073
"""Core routine for extracting references to texts from inventories.
2074
2075
This performs the translation of xml lines to revision ids.
2076
2077
:param line_iterator: An iterator of lines, origin_version_id
2078
:return: A dictionary mapping text keys ((fileid, revision_id) tuples)
2079
to whether they were referred to by the inventory of the
2080
revision_id that they contain. Note that if that revision_id was
2081
not part of the line_iterator's output then False will be given -
2082
even though it may actually refer to that key.
2083
"""
2084
if not self._serializer.support_altered_by_hack:
2085
raise AssertionError(
2086
"_find_text_key_references_from_xml_inventory_lines only "
2087
"supported for branches which store inventory as unnested xml"
2088
", not on %r" % self)
2089
result = {}
2090
2091
# this code needs to read every new line in every inventory for the
2092
# inventories [revision_ids]. Seeing a line twice is ok. Seeing a line
2093
# not present in one of those inventories is unnecessary but not
2094
# harmful because we are filtering by the revision id marker in the
2095
# inventory lines : we only select file ids altered in one of those
2096
# revisions. We don't need to see all lines in the inventory because
2097
# only those added in an inventory in rev X can contain a revision=X
2098
# line.
2099
unescape_revid_cache = {}
2100
unescape_fileid_cache = {}
2101
2102
# jam 20061218 In a big fetch, this handles hundreds of thousands
2103
# of lines, so it has had a lot of inlining and optimizing done.
2104
# Sorry that it is a little bit messy.
2105
# Move several functions to be local variables, since this is a long
2106
# running loop.
2107
search = self._file_ids_altered_regex.search
2108
unescape = _unescape_xml
2109
setdefault = result.setdefault
2110
for line, line_key in line_iterator:
2111
match = search(line)
2112
if match is None:
2113
continue
2114
# One call to match.group() returning multiple items is quite a
2115
# bit faster than 2 calls to match.group() each returning 1
2116
file_id, revision_id = match.group('file_id', 'revision_id')
2117
2118
# Inlining the cache lookups helps a lot when you make 170,000
2119
# lines and 350k ids, versus 8.4 unique ids.
2120
# Using a cache helps in 2 ways:
2121
# 1) Avoids unnecessary decoding calls
2122
# 2) Re-uses cached strings, which helps in future set and
2123
# equality checks.
2124
# (2) is enough that removing encoding entirely along with
2125
# the cache (so we are using plain strings) results in no
2126
# performance improvement.
2127
try:
2128
revision_id = unescape_revid_cache[revision_id]
2129
except KeyError:
2130
unescaped = unescape(revision_id)
2131
unescape_revid_cache[revision_id] = unescaped
2132
revision_id = unescaped
2133
2134
# Note that unconditionally unescaping means that we deserialise
2135
# every fileid, which for general 'pull' is not great, but we don't
2136
# really want to have some many fulltexts that this matters anyway.
2137
# RBC 20071114.
2138
try:
2139
file_id = unescape_fileid_cache[file_id]
2140
except KeyError:
2141
unescaped = unescape(file_id)
2142
unescape_fileid_cache[file_id] = unescaped
2143
file_id = unescaped
2144
2145
key = (file_id, revision_id)
2146
setdefault(key, False)
2147
if revision_id == line_key[-1]:
2148
result[key] = True
2149
return result
2150
2151
2064
def _inventory_xml_lines_for_keys(self, keys):
2152
2065
"""Get a line iterator of the sort needed for findind references.
2153
2066
2183
2096
revision_ids. Each altered file-ids has the exact revision_ids that
2184
2097
altered it listed explicitly.
2185
2098
"""
2186
seen = set(self._find_text_key_references_from_xml_inventory_lines(
2099
seen = set(self._serializer._find_text_key_references(
2187
2100
line_iterator).iterkeys())
2188
2101
parent_keys = self._find_parent_keys_of_revisions(revision_keys)
2189
parent_seen = set(self._find_text_key_references_from_xml_inventory_lines(
2102
parent_seen = set(self._serializer._find_text_key_references(
2190
2103
self._inventory_xml_lines_for_keys(parent_keys)))
2191
2104
new_keys = seen - parent_seen
2192
2105
result = {}
2820
2733
return result
2821
2734
2822
2735
def _warn_if_deprecated(self, branch=None):
2736
if not self._format.is_deprecated():
2737
return
2823
2738
global _deprecation_warning_done
2824
2739
if _deprecation_warning_done:
2825
2740
return
3252
3167
"""
3253
3168
return True
3254
3169
3170
def is_deprecated(self):
3171
"""Is this format deprecated?
3172
3173
Deprecated formats may trigger a user-visible warning recommending
3174
the user to upgrade. They are still fully supported.
3175
"""
3176
return False
3177
3255
3178
def network_name(self):
3256
3179
"""A simple byte string uniquely identifying this format for RPC calls.
3257
3180
4051
3974
pb.finished()
4052
3975
4053
3976
4054
_unescape_map = {
4055
'apos':"'",
4056
'quot':'"',
4057
'amp':'&',
4058
'lt':'<',
4059
'gt':'>'
4060
}
4061
4062
4063
def _unescaper(match, _map=_unescape_map):
4064
code = match.group(1)
4065
try:
4066
return _map[code]
4067
except KeyError:
4068
if not code.startswith('#'):
4069
raise
4070
return unichr(int(code[1:])).encode('utf8')
4071
4072
4073
_unescape_re = None
4074
4075
4076
def _unescape_xml(data):
4077
"""Unescape predefined XML entities in a string of data."""
4078
global _unescape_re
4079
if _unescape_re is None:
4080
_unescape_re = re.compile('\&([^;]*);')
4081
return _unescape_re.sub(_unescaper, data)
4082
4083
4084
3977
class _VersionedFileChecker(object):
4085
3978
4086
3979
def __init__(self, repository, text_key_references=None, ancestors=None):
Loggerhead is a web-based interface for Breezy
Version: 2.0.1