Mirrors
/
synapse
mirror of https://github.com/element-hq/synapse.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Use better method for word boundary searching 

From ebc95667b8
This commit is contained in:
David Baker 2017-10-05 11:57:43 +01:00
parent cbe3c3fdd4
commit 0c8da8b519
1 changed files with 3 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
synapse/push/push_rule_evaluator.py
View File

@ -26,8 +26,6 @@ logger = logging.getLogger(__name__)
GLOB_REGEX = re.compile(r'\\\[(\\\!|)(.*)\\\]') GLOB_REGEX = re.compile(r'\\\[(\\\!|)(.*)\\\]')
IS_GLOB = re.compile(r'[\?\*\[\]]') IS_GLOB = re.compile(r'[\?\*\[\]]')
INEQUALITY_EXPR = re.compile("^([=<>]*)([0-9]*)$") INEQUALITY_EXPR = re.compile("^([=<>]*)([0-9]*)$")
STARTS_WITH_WORD_CHAR_REGEX = re.compile(r"^\w")
ENDS_WITH_WORD_CHAR_REGEX = re.compile(r"\w$")
def _room_member_count(ev, condition, room_member_count): def _room_member_count(ev, condition, room_member_count):
@ -209,15 +207,9 @@ def _re_word_boundary(r):
but do so respecting the fact that strings starting or ending but do so respecting the fact that strings starting or ending
with non-word characters will change word boundaries. with non-word characters will change word boundaries.
""" """
# Matching a regex string aginst a regex, since by definition # we can't use \b as it chokes on unicode. however \W seems to be okay
# \b is the boundary between a \w and a \W, so match \w at the # as shorthand for [^0-9A-Za-z_].
# start or end of the expression (although this will miss, eg. return r"(^|\W)%s(\W|$)" % (r,)
# "[dl]og")
if STARTS_WITH_WORD_CHAR_REGEX.search(r):
r = r"\b%s" % (r,)
if ENDS_WITH_WORD_CHAR_REGEX.search(r):
r = r"%s\b" % (r,)
return r
def _flatten_dict(d, prefix=[], result=None): def _flatten_dict(d, prefix=[], result=None):