From any URL I want to extract its path.
For example:
URL: https://stackoverflow.com/questions/ask Path: questions/ask
It shouldn't be difficult:
url[/(?:\w{2,}\/).+/]
But I think I use a wrong pattern for 'ignore this' ('?:' - doesn't work). What is the right way?
I would suggest you don't do this with a regular expression, and instead use the built in URI lib:
require 'uri'
uri = URI::parse('http://stackoverflow.com/questions/ask')
puts uri.path # results in: /questions/ask
It has a leading slash, but thats easy to deal with =)
You can use regex in this case, which is faster than URI.parse:
s = 'http://stackoverflow.com/questions/ask'
s[s[/.*?\/\/[^\/]*\//].size..-1]
# => "questions/ask" (6,8 times faster)
s[/\/(?!.*\.).*/]
# => "/questions/ask" (9,9 times faster, but with an extra slash)
But if you don't care with the speed, use uri, as ctcherry showed, is more readable.
s = 'http://stackoverflow.com/questions//ask/stuff'), use URI.parse. Don't worry about the speed difference until you have found URI parsing to be a real bottleneck in your code.
http://localhost/test. The negative lookahead expression will fail.
URI.parse regardless of such speed improvements of the regex version. URI.parse is much more solid and will work on any edges cases as @ClintPachl noted.