Allow '-' in tags. Make AP tag discovery more robust.

federation/entities/activitypub/models.py

@@ -856,7 +856,7 @@ class Note(Object, RawContentMixin):
             parsed = urlparse(unquote(link['href']).lower())
             # remove the query part and trailing garbage, if any
             path = parsed.path
-            trunc = re.match(r'(/[\w/]+)', parsed.path)
+            trunc = re.match(r'(/[\w/\-]+)', parsed.path)
             if trunc:
                 path = trunc.group()
             url = f'{parsed.scheme}://{parsed.netloc}{path}'
@@ -865,8 +865,9 @@ class Note(Object, RawContentMixin):
             normalized_url = f'{parsed.scheme}://{parsed.netloc}{normalized_path.decode()}'
             links = {link['href'].lower(), unquote(link['href']).lower(), url, normalized_url}
             if links.intersection(hrefs):
-                tag = re.match(r'#?([\w]+)', link.text).group(1).lower()
-                link['data-hashtag'] = tag
+                tag = re.match(r'^#?([\w\-]+$)', link.text)
+                if tag:
+                    link['data-hashtag'] = tag.group(1).lower()
 
     def _find_and_mark_mentions(self):
         mentions = [mention for mention in self.tag_objects if isinstance(mention, Mention)]

federation/utils/text.py

@@ -7,7 +7,7 @@ from bs4.element import NavigableString
 from commonmark import commonmark
 
 ILLEGAL_TAG_CHARS = "!#$%^&*+.,@£/()=?`'\\{[]}~;:\"’”—\xa0"
-TAG_PATTERN = re.compile(r'(#[\w]+)', re.UNICODE)
+TAG_PATTERN = re.compile(r'(#[\w\-]+)([)\]_!?*%/.,;\s]+\s*|\Z)', 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'(^|[#*_\s])((?:https?://)?[\w\-.]+\.[\w]{1}[\w_\-.#?&/~@!$()*,;%=+]*)', re.UNICODE)
@@ -56,7 +56,8 @@ def find_elements(soup: BeautifulSoup, pattern: re.Pattern) -> List[NavigableStr
         ns = [NavigableString(r) for r in re.split(pattern, candidate.text)]
         if ns:
             candidate.replace_with(*ns)
-            found.extend([child for child in parent.find_all(string=pattern) if child in ns])
+            found.extend([child for child in parent.find_all(
+                string=re.compile(r'\A'+pattern.pattern+r'\Z')) if child in ns])
     return found