Fix article ID parsing bug

Since the article ID regular expression wasn't anchored to the end of
the URL, it would grab characters after a / or - that were hex
characters. For example /@user/bacon-123abc would just grab `bac`. Not
great.

This anchors the ID at the end of the string so that it will be more
likely to catch IDs.

CHANGELOG

@@ -2,6 +2,7 @@
 
 * Better article ID parsing
 * Link to full Medium URL on error page
+* Fix article ID parsing bug
 
 2022-02-12
 

spec/classes/article_id_parser_spec.cr

@@ -14,6 +14,14 @@ describe ArticleIdParser do
     result.should eq(Monads::Just.new("111111abcdef"))
   end
 
+  it "parses the post id for urls with hex characters after a /" do
+    request = resource_request("/@user/bacon-abcdef123456")
+
+    result = ArticleIdParser.parse(request)
+
+    result.should eq(Monads::Just.new("abcdef123456"))
+  end
+
   it "parses the post id for urls like /user/:post_slug" do
     request = resource_request("/user/my-post-222222abcdef")
 

src/classes/article_id_parser.cr

@@ -1,7 +1,7 @@
 class ArticleIdParser
   include Monads
 
-  ID_REGEX = /[\/\-]([0-9a-f]+)/i
+  ID_REGEX = /[\/\-]([0-9a-f]+)$/i
 
   def self.parse(request : HTTP::Request)
     new.parse(request)