def filter(self):
try:
return es_regex_filter(
self._parsed_regex,
'content',
is_case_sensitive=self._term['case_sensitive'])
except NoTrigrams:
raise BadTerm('Regexes need at least 3 literal characters in a '
'row for speed.')
def filter(self):
glob = self._term['arg']
try:
return es_regex_filter(
regex_grammar.parse(glob_to_regex(glob)),
'path',
is_case_sensitive=self._term['case_sensitive'])
except NoTrigrams:
raise BadTerm('Path globs need at least 3 literal characters in a row '
'for speed.')
def _regex_filter(self, path_seg_property_name, no_trigrams_error_text):
"""Return an ES regex filter that matches this filter's glob against the
path segment at path_seg_property_name.
"""
glob = self._term['arg']
try:
return es_regex_filter(
regex_grammar.parse(glob_to_regex(glob)),
path_seg_property_name,
is_case_sensitive=self._term['case_sensitive'])
except NoTrigrams:
raise BadTerm(no_trigrams_error_text)
def __init__(self, term, enabled_plugins):
"""Compile the Python equivalent of the regex so we don't have to lean
on the regex cache during highlighting.
Python's regex cache is naive: after it hits 100, it just clears: no
LRU.
"""
super(RegexpFilter, self).__init__(term, enabled_plugins)
try:
self._parsed_regex = regex_grammar.parse(term['arg'])
except ParseError:
raise BadTerm('Invalid regex.')
self._compiled_regex = (
re.compile(PythonRegexVisitor().visit(self._parsed_regex),
flags=0 if self._term['case_sensitive'] else re.I))