From 6cb461fed424f07bf50a1fdc0693d40ed8bbee12 Mon Sep 17 00:00:00 2001
From: David Teller <davidt@element.io>
Date: Wed, 9 Feb 2022 13:07:23 +0100
Subject: [PATCH] Move glob_to_regex into the source - resolves #173 (#218)

---
 synapse_antispam/mjolnir/list_rule.py |  5 +-
 synapse_antispam/mjolnir/matching.py  | 73 +++++++++++++++++++++++++++
 2 files changed, 75 insertions(+), 3 deletions(-)
 create mode 100644 synapse_antispam/mjolnir/matching.py

diff --git a/synapse_antispam/mjolnir/list_rule.py b/synapse_antispam/mjolnir/list_rule.py
index b21d1e3..ad0ca3f 100644
--- a/synapse_antispam/mjolnir/list_rule.py
+++ b/synapse_antispam/mjolnir/list_rule.py
@@ -1,5 +1,5 @@
 # -*- coding: utf-8 -*-
-# Copyright 2019 The Matrix.org Foundation C.I.C.
+# Copyright 2022 The Matrix.org Foundation C.I.C.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -13,7 +13,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from synapse.util import glob_to_regex
+from .matching import glob_to_regex
 
 RECOMMENDATION_BAN = "m.ban"
 RECOMMENDATION_BAN_TYPES = [RECOMMENDATION_BAN, "org.matrix.mjolnir.ban"]
@@ -26,7 +26,6 @@ ROOM_RULE_TYPES = [RULE_ROOM, "m.room.rule.room", "org.matrix.mjolnir.rule.room"
 SERVER_RULE_TYPES = [RULE_SERVER, "m.room.rule.server", "org.matrix.mjolnir.rule.server"]
 ALL_RULE_TYPES = [*USER_RULE_TYPES, *ROOM_RULE_TYPES, *SERVER_RULE_TYPES]
 
-
 def recommendation_to_stable(recommendation):
     if recommendation in RECOMMENDATION_BAN_TYPES:
         return RECOMMENDATION_BAN
diff --git a/synapse_antispam/mjolnir/matching.py b/synapse_antispam/mjolnir/matching.py
new file mode 100644
index 0000000..4be8bed
--- /dev/null
+++ b/synapse_antispam/mjolnir/matching.py
@@ -0,0 +1,73 @@
+# -*- coding: utf-8 -*-
+# Copyright 2022 The Matrix.org Foundation C.I.C.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# Tools in this file were copied from Synapse as these functions will not
+# remain publicly accessible in the module, see https://github.com/matrix-org/mjolnir/pull/174
+
+import re
+from typing import Pattern
+
+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,)
+
+_WILDCARD_RUN = re.compile(r"([\?\*]+)")
+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)
+