Match links with no http prefix. Remove trailing garbage from tags.

federation/entities/activitypub/models.py

@@ -849,11 +849,13 @@ class Note(Object, RawContentMixin):
 
         for link in self._soup.find_all('a', href=True):
             parsed = urlparse(link['href'].lower())
-            # remove the query part, if any
-            url = f'{parsed.scheme}://{parsed.netloc}{parsed.path}'
+            # remove the query part and trailing garbage, if any
+            path = re.match(r'(/[\w/]+)', parsed.path).group()
+            url = f'{parsed.scheme}://{parsed.netloc}{path}'
             links = {link['href'].lower(), url}
             if links.intersection(hrefs):
-                link['data-hashtag'] = link.text.lstrip('#').lower()
+                tag = re.match(r'#?([\w]+)', link.text).group(1).lower()
+                link['data-hashtag'] = tag
 
     def _find_and_mark_mentions(self):
         mentions = [mention for mention in self.tag_objects if isinstance(mention, Mention)]

federation/utils/text.py

@@ -10,7 +10,7 @@ ILLEGAL_TAG_CHARS = "!#$%^&*+.,@£/()=?`'\\{[]}~;:\"’”—\xa0"
 TAG_PATTERN = re.compile(r'(#[\w]+)', re.UNICODE)
 # This will match non matching braces. I don't think it's an issue.
 MENTION_PATTERN = re.compile(r'(@\{?(?:[\w\-. \u263a-\U0001f645]*; *)?[\w]+@[\w\-.]+\.[\w]+}?)', re.UNICODE)
-URL_PATTERN = re.compile(r'(https?://[\w_\-.#?&/]+)', re.UNICODE)
+URL_PATTERN = re.compile(r'((?:https?://)?[\w_\-.#?&/~@!$()*,;%=+]+)', re.UNICODE)
 
 def decode_if_bytes(text):
     try:
@@ -52,7 +52,7 @@ def find_elements(soup: BeautifulSoup, pattern: re.Pattern) -> List[NavigableStr
     for candidate in soup.find_all(string=True):
         if candidate.parent.name == 'code': continue
         ns = [NavigableString(r) for r in re.split(pattern, candidate.text)]
-        candidate.replace_with(*ns)
+        if ns: candidate.replace_with(*ns)
     return list(soup.find_all(string=re.compile(r'\A'+pattern.pattern+r'\Z')))