Fix some characters stopping tags being identified

Closes socialhome/socialhome#222

CHANGELOG.md

@@ -63,6 +63,8 @@
 
 * Don't try to relay AP payloads to Diaspora receivers and vice versa, for now, until cross-protocol
   relaying is supported.
+  
+* Fix some characters stopping tags being identified ([related issue](https://git.feneas.org/socialhome/socialhome/-/issues/222))
 
 ## [0.19.0] - 2019-12-15
 

federation/tests/utils/test_text.py

@@ -34,12 +34,12 @@ class TestFindTags:
         assert text == "foo\n```\n#code\n```\n#notcode/notcode\n\n    #alsocode\n"
 
     def test_endings_are_filtered_out(self):
-        source = "#parenthesis) #exp! #list]"
+        source = "#parenthesis) #exp! #list] *#doh* _#bah_ #gah%"
         tags, text = find_tags(source)
-        assert tags == {"parenthesis", "exp", "list"}
+        assert tags == {"parenthesis", "exp", "list", "doh", "bah", "gah"}
         assert text == source
         tags, text = find_tags(source, replacer=self._replacer)
-        assert text == "#parenthesis/parenthesis) #exp/exp! #list/list]"
+        assert text == "#parenthesis/parenthesis) #exp/exp! #list/list] *#doh/doh* _#bah/bah_ #gah/gah%"
 
     def test_finds_tags(self):
         source = "#post **Foobar** #tag #OtherTag #third\n#fourth"

federation/utils/text.py

@@ -50,7 +50,7 @@ def find_tags(text: str, replacer: callable = None) -> Tuple[Set, str]:
         words = line.split(" ")
         for word in words:
             if word.find('#') > -1:
-                candidate = word.strip().strip("([]),.!?:")
+                candidate = word.strip().strip("([]),.!?:*_%")
                 if candidate.find('<') > -1 or candidate.find('>') > -1:
                     # Strip html
                     candidate = bleach.clean(word, strip=True)