Add automatic hashtag parsing from note-like posts

redwind/admin.py

@@ -275,6 +275,10 @@ def save_post(post):
         post.audience = util.multiline_string_to_list(audience)
 
     tags = request.form.getlist('tags')
+    if post.post_type != "article" and post.content:
+        # parse out hashtags as tag links from note-like posts
+        post.content, htags = util.parse_hashtags(post.content)
+        tags += htags
     tags = list(filter(None, map(util.normalize_tag, tags)))
     post.tags = [Tag.query.filter_by(name=tag).first() or Tag(tag)
                  for tag in tags]

redwind/templates/admin/venues.jinja2

@@ -24,8 +24,8 @@
                     {{venue.name}}
                   </a>
                   {% if current_user.is_authenticated() %}
-                    <a href="{{url_for('edit_venue', id=venue.id)}}"><i class="glyphicon glyphicon-edit"></i> Edit</a>
-                    <a href="{{url_for('delete_venue', id=venue.id)}}"><i class="glyphicon glyphicon-trash"></i> Delete</a>
+                    <a href="{{url_for('admin.edit_venue', id=venue.id)}}"><i class="glyphicon glyphicon-edit"></i> Edit</a>
+                    <a href="{{url_for('admin.delete_venue', id=venue.id)}}"><i class="glyphicon glyphicon-trash"></i> Delete</a>
                   {% endif %}
                 </li>
               {% endfor %}

redwind/util.py

@@ -36,6 +36,7 @@
 INSTAGRAM_RE = re.compile(r'https?://(?:www\.|mobile\.)?instagram\.com/p/([a-zA-Z0-9_\-]+)/?')
 PEOPLE_RE = re.compile(r"\[\[([\w ]+)(?:\|([\w\-'. ]+))?\]\]")
 RELATIVE_PATH_RE = re.compile('\[([^\]]*)\]\(([^/)]+)\)')
+HASHTAG_RE = re.compile('(?<!\w)#(\w\w+)')
 
 AT_USERNAME_RE = re.compile(r"""(?<!\w)@([a-zA-Z0-9_]+)(?=($|[\s,:;.?'")]))""")
 LINK_RE = re.compile(
@@ -160,15 +161,7 @@ def person_to_microcard(contact, nick, soup):
     return a_tag
 
 
-def autolink(plain, url_processor=url_to_link,
-             person_processor=person_to_microcard):
-    """Replace bare URLs in a document with an HTML <a> representation
-    """
-    blacklist = ('a', 'script', 'pre', 'code', 'embed', 'object',
-                      'audio', 'video')
-    soup = bs4.BeautifulSoup(plain)
-
-    def bs4_sub(regex, repl):
+def bs4_sub(soup, regex, repl, blacklist):
         """Process text elements in a BeautifulSoup document with a regex and
         replacement string.
 
@@ -199,6 +192,15 @@ def bs4_sub(regex, repl):
             for offset, node in enumerate(nodes):
                 parent.insert(ii + offset, node)
 
+
+def autolink(plain, url_processor=url_to_link,
+             person_processor=person_to_microcard):
+    """Replace bare URLs in a document with an HTML <a> representation
+    """
+    blacklist = ('a', 'script', 'pre', 'code', 'embed', 'object',
+                      'audio', 'video')
+    soup = bs4.BeautifulSoup(plain)
+
     def link_repl(m):
         url = (m.group(1) or 'http://') + m.group(2)
         return url_processor(url, soup)
@@ -216,10 +218,10 @@ def process_nick(m):
         return m.group(0)
 
     if url_processor:
-        bs4_sub(LINK_RE, link_repl)
+        bs4_sub(soup, LINK_RE, link_repl, blacklist)
 
     if person_processor:
-        bs4_sub(AT_USERNAME_RE, process_nick)
+        bs4_sub(soup, AT_USERNAME_RE, process_nick, blacklist)
 
     return ''.join(str(t) for t in soup.body.contents) if soup.body else ''
 
@@ -359,6 +361,8 @@ def format_as_text(html, link_fn=None):
     for a in soup.find_all('a'):
         if link_fn:
             link_fn(a)
+        elif a.text[0] == '#':
+            a.replace_with(a.text)
         else:
             a.replace_with(a.get('href') or '[link]')
 
@@ -471,3 +475,29 @@ def find_first_syndicated(originals):
         find_first_syndicated(post.repost_of),
         find_first_syndicated(post.like_of),
     )
+
+
+def parse_hashtags(s):
+    """ Parses out hashtags from a string and replaces them with links, then
+        returns the new string and a list of tags encountered
+    """
+    blacklist = ('a', 'script', 'pre', 'code', 'embed', 'object',
+                      'audio', 'video')
+    soup = bs4.BeautifulSoup(s)
+    tags = []
+
+    def process_hashtag(m):
+        url = '/tags/' + m.group(1).lower()
+        a = soup.new_tag('a', href=url)
+        a.string = m.group(0)
+
+        tags.append(m.group(1).lower())
+
+        return a
+
+    try:
+        bs4_sub(soup, HASHTAG_RE, process_hashtag, blacklist)
+        s = ''.join(str(t) for t in soup.body.contents) if soup.body else ''
+        return s, tags
+    except TypeError:
+        return s, []

tests/util_test.py

@@ -141,3 +141,55 @@ def simple_name_marker(contact, name, soup):
         assert out == util.autolink(
             inp, person_processor=simple_name_marker,
             url_processor=None)
+
+
+def test_parsing_hashtags():
+    """Exercise the #-tag matching regex
+    """
+
+    test_cases = [
+        ('#hashtag should be linked', 
+         '<a href="/tags/hashtag">#hashtag</a> should be linked',
+         ['hashtag']),
+        ('hashtag should not be linked',
+         'hashtag should not be linked',
+         []),
+        ('match #hashtags in the middle',
+         'match <a href="/tags/hashtags">#hashtags</a> in the middle',
+         ['hashtags']),
+        ('match a tag at the #end',
+         'match a tag at the <a href="/tags/end">#end</a>',
+         ['end']),
+        ('#1 should not be linked',
+         '#1 should not be linked',
+         []),
+        ('#12345 should be linked',
+         '<a href="/tags/12345">#12345</a> should be linked',
+         ['12345']),
+        ('#.foobar should not be linked',
+         '#.foobar should not be linked',
+         []),
+        ('#foo.bar should be partially linked',
+         '<a href="/tags/foo">#foo</a>.bar should be partially linked',
+         ['foo']),
+        ('capital letters in #HashTags will be lowercased',
+         'capital letters in <a href="/tags/hashtags">#HashTags</a> will be lowercased',
+         ['hashtags']),
+        ('duplicate #hashtags should parse both #hashtags fine',
+         'duplicate <a href="/tags/hashtags">#hashtags</a> should parse both <a href="/tags/hashtags">#hashtags</a> fine',
+         ['hashtags','hashtags']),
+        ('Lorem ipsum dolor sit amet, consectetur adipiscing elit. Praesent tincidunt aliquam sem, in tempus elit lacinia vel. Integer accumsan cursus purus et euismod. Nullam ultricies nunc sit amet ante consequat porta. Pellentesque et porta odio. Sed et neque cursus, iaculis lorem nec, laoreet odio. Donec molestie volutpat vestibulum. Curabitur rhoncus elit ut massa pretium luctus. #Nullam sollicitudin ligula vitae tincidunt suscipit. Maecenas in neque porta, scelerisque metus at, mollis nunc. Fusce accumsan imperdiet velit, in tincidunt tellus aliquam ac. Nullam iaculis vel urna sed vulputate. Aliquam erat volutpat. Etiam et tortor turpis. Vivamus mattis enim lacus, in aliquet nulla blandit.',
+         'Lorem ipsum dolor sit amet, consectetur adipiscing elit. Praesent tincidunt aliquam sem, in tempus elit lacinia vel. Integer accumsan cursus purus et euismod. Nullam ultricies nunc sit amet ante consequat porta. Pellentesque et porta odio. Sed et neque cursus, iaculis lorem nec, laoreet odio. Donec molestie volutpat vestibulum. Curabitur rhoncus elit ut massa pretium luctus. <a href="/tags/nullam">#Nullam</a> sollicitudin ligula vitae tincidunt suscipit. Maecenas in neque porta, scelerisque metus at, mollis nunc. Fusce accumsan imperdiet velit, in tincidunt tellus aliquam ac. Nullam iaculis vel urna sed vulputate. Aliquam erat volutpat. Etiam et tortor turpis. Vivamus mattis enim lacus, in aliquet nulla blandit.',
+         ['nullam']),
+        ('this hash#tag will not be parsed',
+         'this hash#tag will not be parsed',
+         []),
+        ('http://example.com/path#fragment',
+         'http://example.com/path#fragment',
+         []),
+    ]
+
+    for inp, out, tags in test_cases:
+        res, ts = util.parse_hashtags(inp)
+        assert out == res
+        assert tags == ts