Accepting only ASCII characters left from TLD.

lipoja · Feb 26, 2024 · 562bf74 · 562bf74
1 parent 1cda63c
commit 562bf74
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 2 deletions.
diff --git a/tests/unit/test_find_urls.py b/tests/unit/test_find_urls.py
@@ -12,6 +12,8 @@
 @pytest.mark.parametrize(
     "text, expected",
     [
+        ("%sexample.com" % chr(8231), ["example.com"]),
+        ("some%sdomain.example.com" % chr(8231), ["domain.example.com"]),
         ("Let's have URL http://janlipovsky.cz", ["http://janlipovsky.cz"]),
         ("Let's have text without URLs.", []),
         ("Dot after TLD: http://janlipovsky.cz.", ["http://janlipovsky.cz"]),
@@ -57,7 +59,7 @@
             "<script src='//www.example.com/somejsfile.js'>",
             ["www.example.com/somejsfile.js"],
         ),
-        ("bad.email @address.net>", ['bad.email']),
+        ("bad.email @address.net>", ["bad.email"]),
         ('[[ "$(giturl)" =~ ^https://gitlab.com ]] echo "found" || echo "didnt', []),
     ],
 )

diff --git a/urlextract/urlextract_core.py b/urlextract/urlextract_core.py
@@ -484,7 +484,12 @@ def _complete_url(
                         and text[start_pos - 1] in self._stop_chars_left_from_schema
                     ):
                         left_ok = False
-                    if left_ok and text[start_pos - 1] not in self._stop_chars_left:
+                    if (
+                        left_ok
+                        and text[start_pos - 1] not in self._stop_chars_left
+                        # Allow only ASCII characters in authority and schema
+                        and ord(text[start_pos - 1]) <= 127
+                    ):
                         start_pos -= 1
                     else:
                         left_ok = False