Fix hashtag matching pattern matching some URLs (mastodon#27584)

kmycode · Oct 27, 2023 · 93e4cdc · 93e4cdc
1 parent 08bdd57
commit 93e4cdc
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 1 deletion.
diff --git a/app/models/tag.rb b/app/models/tag.rb
@@ -35,7 +35,7 @@ class Tag < ApplicationRecord
   HASHTAG_LAST_SEQUENCE = '([[:word:]_]*[[:alpha:]][[:word:]_]*)'
   HASHTAG_NAME_PAT = "#{HASHTAG_FIRST_SEQUENCE}|#{HASHTAG_LAST_SEQUENCE}"
 
-  HASHTAG_RE = %r{(?<![=/)[:word]])#(#{HASHTAG_NAME_PAT})}i
+  HASHTAG_RE = %r{(?<![=/)\w])#(#{HASHTAG_NAME_PAT})}i
   HASHTAG_NAME_RE = /\A(#{HASHTAG_NAME_PAT})\z/i
   HASHTAG_INVALID_CHARS_RE = /[^[:alnum:]#{HASHTAG_SEPARATORS}]/
 

diff --git a/spec/models/tag_spec.rb b/spec/models/tag_spec.rb
@@ -32,6 +32,10 @@
       expect(subject.match('https://en.wikipedia.org/wiki/Ghostbusters_(song)#Lawsuit')).to be_nil
     end
 
+    it 'does not match URLs with hashtag-like anchors after a numeral' do
+      expect(subject.match('https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111895#c4')).to be_nil
+    end
+
     it 'does not match URLs with hashtag-like anchors after an empty query parameter' do
       expect(subject.match('https://en.wikipedia.org/wiki/Ghostbusters_(song)?foo=#Lawsuit')).to be_nil
     end