Jason Orri

Does anyone know if there's anyway to filter, using regexp a url like 

 from a message/piece of text - maybe a way to recognise the "http" and cut it, and what follows, from the message. This is necessary as I am unsure what the URLs might be so stating a specific one to regexp isn't helpful here...

Does anyone know if there's anyway to filter, using regexp a url like http://t.co/tmkdWBcS from a message/piece of text - maybe a way to recognise the "http" and cut it, and what follows, from the message. This is necessary as I am unsure what the URLs might be so stating a specific one to regexp isn't helpful here...
Any help would be great
J


cutting-httpurl-links-from-text-messages

Assuming there is only one URL in the message/text it could be done like this.

Thanks Broc, that's really helpful but I should have said, I need to use everything else in the text except the link/URL so "dog cat 

 pig cat" would read "dog cat pig cat" .. hope that makes sense. It seems like it could be possible though.

Thanks Broc, that's really helpful but I should have said, I need to use everything else in the text except the link/URL so "dog cat http://t.co/tmkdWBcS pig cat" would read "dog cat pig cat" .. hope that makes sense. It seems like it could be possible though.


Yes, it can simply be inverted (with some cleaning from quotation marks).

This works great. Thanks broc, the only problem I'm having is that if the message does NOT contain a URL the message wont output anything.. there is a chance some incoming messages will have links/URLs and some wont.. thanks for your efforts

I think it can be solved with a simple trick by sending also the 4th outlet (unmatched) to output.

And cutting multiple URLs could be solved by iteration until the output doesn't change anymore.

Another solution, you can use %0 to remove elements:


cutting http/URL links from text (messages)