checkin url_remove.py

Author: dave3d

comment_spell_check/utils/url_remove.py

@@ -0,0 +1,22 @@
+#! /usr/bin/env python
+
+"""Module to remove URLs from a string."""
+
+import re
+
+
+def remove_urls(text):
+    """
+    Removes URLs from a string using a regular expression.
+
+    Args:
+        text: The input string.
+
+    Returns:
+        The string with URLs removed.
+    """
+    url_pattern = re.compile(
+        r"(?:http(s)?:\/\/)?[\w.-]+(?:\.[\w\.-]+)+[\w\-\._~:/?#[\]@!\$&\'\(\)\*\+,;=.]+",
+        re.IGNORECASE,
+    )
+    return url_pattern.sub("", text)