plugins.rules: URLCallback shall handle invalid URLs by ignoring them

sopel/plugins/rules.py

@@ -1678,14 +1678,16 @@ def match(self, bot, pretrigger):
         if not self.match_preconditions(bot, pretrigger):
             return
 
-        urls = (
-            url
-            for url in pretrigger.urls
-            if urlparse(url).scheme in self._schemes
-        )
+        # Parse only valid URLs with wanted schemes
+        for url in pretrigger.urls:
+            try:
+                if urlparse(url).scheme not in self._schemes:
+                    # skip URLs with unwanted scheme
+                    continue
+            except ValueError:
+                # skip invalid URLs
+                continue
 
-        # Parse URL for each found
-        for url in urls:
             # TODO: convert to 'yield from' when dropping Python 2.7
             for result in self.parse(url):
                 yield result

test/plugins/test_plugins_rules.py

@@ -2609,13 +2609,27 @@ def test_url_callback_parse():
         re.escape('https://wikipedia.com/') + r'(\w+)'
     )
 
-    rule = rules.SearchRule([regex])
+    rule = rules.URLCallback([regex])
     results = list(rule.parse('https://wikipedia.com/something'))
     assert len(results) == 1, 'URLCallback on word must match only once'
     assert results[0].group(0) == 'https://wikipedia.com/something'
     assert results[0].group(1) == 'something'
 
 
+def test_url_callback_match(mockbot):
+    regex = re.compile(r'.*')
+    rule = rules.URLCallback([regex])
+
+    line = (
+        ':Foo!foo@example.com PRIVMSG #sopel :'
+        'two links http://example.com one invalid https://[dfdsdfsdf'
+    )
+    pretrigger = trigger.PreTrigger(mockbot.nick, line)
+    matches = list(rule.match(mockbot, pretrigger))
+    assert len(matches) == 1, 'URLCallback must ignore invalid URLs'
+    assert matches[0].group(0) == 'http://example.com'
+
+
 def test_url_callback_execute(mockbot):
     regex = re.compile(r'.*')
     rule = rules.URLCallback([regex])