def _path_trigram_filter(path, is_case_sensitive):
"""Return an ES filter clause that returns docs whose paths match the
given path all the way to their ends.
If a given path starts with a /, the user is explicitly requesting a match
starting at the root level.
"""
if path.startswith('/'):
path = path[1:] # Leading slashes aren't stored in the index.
regex = '^{0}$' # Insist it start at the beginning.
else:
regex = '(/|^){0}$' # Start at any path segment.
return es_regex_filter(
regex_grammar.parse(
regex.format(
re.escape(
path.encode('ascii', 'backslashreplace')
)
)
),
'path',
is_case_sensitive
)
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 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):
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 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 _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 _path_trigram_filter(path, is_case_sensitive):
"""Return an ES filter clause that returns docs whose paths match the
given path all the way to their ends.
If a given path starts with a /, the user is explicitly requesting a match
starting at the root level.
"""
if path.startswith('/'):
path = path[1:] # Leading slashes aren't stored in the index.
regex = '^{0}$' # Insist it start at the beginning.
else:
regex = '(/|^){0}$' # Start at any path segment.
return es_regex_filter(regex_grammar.parse(regex.format(re.escape(path))),
'path', is_case_sensitive)
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.")
