Parsing tweets with PHP
A port of Remy Sharp’s excellent ‘ify’ code.
Just a quick post to announce my release of a PHP port of Remy Sharp’s ‘ify’ script, a very simple chunk of code to auto-link the following parts of a tweet:
- URLs (into direct links)
- hashtags (link to a twitter search for that hashtag)
- usernames (link to that user’s profile page)
such that this text:
I wrote an open letter to @taylorswift13 http://stuffandnonsense.co.uk/s/1239 #sn_1239 (RT please)
becomes this:
I wrote an open letter to @taylorswift13 http://stuffandnonsense.co.uk/s/1239 #sn_1239 (RT please)
Remy’s code is nothing more than 3 regular expressions, so is highly portable. There are many other very similar solutions out there, but Remy’s is the best at handling edge-cases that I’ve come across. I’ll endeavour to update this port whenever the original is updated. All feedback is welcome!
Tweet
Comments
Five Minute Argument said:
Thanks, Russell. All the work is in Remy’s regexps which are much better than anything else I've seen; I might write a detailed analysis of them if I think it’s worth it!
Ante Vukorepa said:
I've improved the URL regexp a bit, it was pretty limited:
([a-z]*(?:\:\/\/)*[a-z0–9-_]+(?:\.[a-zA-Z0–9-_:~%&\?\+#\=\/]+)+[^:\.,\)\s*$])
It now handles URLs not prefixed by "http://", but eliminates spurious ones ("foo..something" or "foo…something" or… you get the idea). I ran across a bug or two with it and fixed those as well.
Hope they work out for you.
Mon 15 Mar 2010 08:07
Russell Heimlich said:
Awesome stuff! I used to write my own but these look a lot more accomodating.