diff options
Diffstat (limited to 'synapse')
| -rw-r--r-- | synapse/config/room_directory.py | 3 | ||||
| -rw-r--r-- | synapse/config/tls.py | 3 | ||||
| -rw-r--r-- | synapse/federation/federation_server.py | 3 | ||||
| -rw-r--r-- | synapse/push/push_rule_evaluator.py | 7 | ||||
| -rw-r--r-- | synapse/python_dependencies.py | 1 | ||||
| -rw-r--r-- | synapse/util/__init__.py | 59 |
6 files changed, 12 insertions, 64 deletions
diff --git a/synapse/config/room_directory.py b/synapse/config/room_directory.py index 57316c59b6..3c5e0f7ce7 100644 --- a/synapse/config/room_directory.py +++ b/synapse/config/room_directory.py @@ -15,8 +15,9 @@ from typing import List +from matrix_common.regex import glob_to_regex + from synapse.types import JsonDict -from synapse.util import glob_to_regex from ._base import Config, ConfigError diff --git a/synapse/config/tls.py b/synapse/config/tls.py index ffb316e4c0..6e673d65a7 100644 --- a/synapse/config/tls.py +++ b/synapse/config/tls.py @@ -16,11 +16,12 @@ import logging import os from typing import List, Optional, Pattern +from matrix_common.regex import glob_to_regex + from OpenSSL import SSL, crypto from twisted.internet._sslverify import Certificate, trustRootFromCertificates from synapse.config._base import Config, ConfigError -from synapse.util import glob_to_regex logger = logging.getLogger(__name__) diff --git a/synapse/federation/federation_server.py b/synapse/federation/federation_server.py index cf067b56c6..ee71f289c8 100644 --- a/synapse/federation/federation_server.py +++ b/synapse/federation/federation_server.py @@ -28,6 +28,7 @@ from typing import ( Union, ) +from matrix_common.regex import glob_to_regex from prometheus_client import Counter, Gauge, Histogram from twisted.internet.abstract import isIPAddress @@ -65,7 +66,7 @@ from synapse.replication.http.federation import ( ) from synapse.storage.databases.main.lock import Lock from synapse.types import JsonDict, get_domain_from_id -from synapse.util import glob_to_regex, json_decoder, unwrapFirstError +from synapse.util import json_decoder, unwrapFirstError from synapse.util.async_helpers import Linearizer, concurrently_execute, gather_results from synapse.util.caches.response_cache import ResponseCache from synapse.util.stringutils import parse_server_name diff --git a/synapse/push/push_rule_evaluator.py b/synapse/push/push_rule_evaluator.py index 7f68092ec5..659a53805d 100644 --- a/synapse/push/push_rule_evaluator.py +++ b/synapse/push/push_rule_evaluator.py @@ -17,9 +17,10 @@ import logging import re from typing import Any, Dict, List, Optional, Pattern, Tuple, Union +from matrix_common.regex import glob_to_regex, to_word_pattern + from synapse.events import EventBase from synapse.types import JsonDict, UserID -from synapse.util import glob_to_regex, re_word_boundary from synapse.util.caches.lrucache import LruCache logger = logging.getLogger(__name__) @@ -184,7 +185,7 @@ class PushRuleEvaluatorForEvent: r = regex_cache.get((display_name, False, True), None) if not r: r1 = re.escape(display_name) - r1 = re_word_boundary(r1) + r1 = to_word_pattern(r1) r = re.compile(r1, flags=re.IGNORECASE) regex_cache[(display_name, False, True)] = r @@ -213,7 +214,7 @@ def _glob_matches(glob: str, value: str, word_boundary: bool = False) -> bool: try: r = regex_cache.get((glob, True, word_boundary), None) if not r: - r = glob_to_regex(glob, word_boundary) + r = glob_to_regex(glob, word_boundary=word_boundary) regex_cache[(glob, True, word_boundary)] = r return bool(r.search(value)) except re.error: diff --git a/synapse/python_dependencies.py b/synapse/python_dependencies.py index 13fb69460e..d844fbb3b3 100644 --- a/synapse/python_dependencies.py +++ b/synapse/python_dependencies.py @@ -88,6 +88,7 @@ REQUIREMENTS = [ # with the latest security patches. "cryptography>=3.4.7", "ijson>=3.1", + "matrix-common==1.0.0", ] CONDITIONAL_REQUIREMENTS = { diff --git a/synapse/util/__init__.py b/synapse/util/__init__.py index 95f23e27b6..f157132210 100644 --- a/synapse/util/__init__.py +++ b/synapse/util/__init__.py @@ -14,9 +14,8 @@ import json import logging -import re import typing -from typing import Any, Callable, Dict, Generator, Optional, Pattern +from typing import Any, Callable, Dict, Generator, Optional import attr from frozendict import frozendict @@ -35,9 +34,6 @@ if typing.TYPE_CHECKING: logger = logging.getLogger(__name__) -_WILDCARD_RUN = re.compile(r"([\?\*]+)") - - def _reject_invalid_json(val: Any) -> None: """Do not allow Infinity, -Infinity, or NaN values in JSON.""" raise ValueError("Invalid JSON value: '%s'" % val) @@ -185,56 +181,3 @@ def log_failure( if not consumeErrors: return failure return None - - -def glob_to_regex(glob: str, word_boundary: bool = False) -> Pattern: - """Converts a glob to a compiled regex object. - - Args: - glob: pattern to match - word_boundary: If True, the pattern will be allowed to match at word boundaries - anywhere in the string. Otherwise, the pattern is anchored at the start and - end of the string. - - Returns: - compiled regex pattern - """ - - # Patterns with wildcards must be simplified to avoid performance cliffs - # - The glob `?**?**?` is equivalent to the glob `???*` - # - The glob `???*` is equivalent to the regex `.{3,}` - chunks = [] - for chunk in _WILDCARD_RUN.split(glob): - # No wildcards? re.escape() - if not _WILDCARD_RUN.match(chunk): - chunks.append(re.escape(chunk)) - continue - - # Wildcards? Simplify. - qmarks = chunk.count("?") - if "*" in chunk: - chunks.append(".{%d,}" % qmarks) - else: - chunks.append(".{%d}" % qmarks) - - res = "".join(chunks) - - if word_boundary: - res = re_word_boundary(res) - else: - # \A anchors at start of string, \Z at end of string - res = r"\A" + res + r"\Z" - - return re.compile(res, re.IGNORECASE) - - -def re_word_boundary(r: str) -> str: - """ - Adds word boundary characters to the start and end of an - expression to require that the match occur as a whole word, - but do so respecting the fact that strings starting or ending - with non-word characters will change word boundaries. - """ - # we can't use \b as it chokes on unicode. however \W seems to be okay - # as shorthand for [^0-9A-Za-z_]. - return r"(^|\W)%s(\W|$)" % (r,) |