validating URL with regular expression

rubyonrails-talk
1

Hi All,

I am new to regular expressions.Even if I try it will take sometime for me to get one regular expression for URL validation. Can anyone help me by providing a regular expression for validating URL http,https to be used with validates_format_of ?

I found this https?://([-\w.]+)+(:\d+)?(/([\w/_.]*(?\S+)?)?)? but its not working with validates_format_of.

I’ll appreciate if anyone helps me with this.

regards gaurav

2

Hi!

I think you’ll need to escape / by putting \ before, otherwise you’ll close the regexp too soon. I came out with something like this:

irb(main):001:0> r=/^(https?://)?[a-z0-9]+([.-_=&+/?]?[a-z0-9]+)+$/i

=> /^(https?://)?[a-z0-9]+([.-_=&+/?]?[a-z0-9]+)+$/i irb(main):002:0> “www.google.com”=~r => 0 irb(main):003:0> " http://www.google.fr"=~r => 0 irb(main):004:0> “https://www.google.de”=~r => 0 irb(main):005:0> “http://.google.de”=~r => nil irb(main):006:0> “http://www.ruby-doc.org/core/classes/Regexp.html”=~r => 0 irb(main):007:0> " http://www.google.fr/search?hl=fr&q=ruby&btnG=Recherche+Google&meta=test"=~r => 0 irb(main):008:0> “http://gogo…fr”=~r => nil irb(main):009:0> " http://gogol.de."=~r => nil

Have a good day, Eric

3

hi eric, thanks for help regards gaurav

4

hi eric,

  one issue   when i try   irb(main):004:0> "http://groups-beta.google.com/group/rubyonrails-talk/browse_th read/thread/8f085b191387d799/e78a71cbd7354c0c#e78a71cbd7354c0c"=~r

irb just hangs and cpu usage on windows xp goes 100%

when i removed that # in the last part of url it gives the result

wonder why it hangs if not of proper format as given it should just fail.

regards gaurav

5

yes, indeed there are problems with Regexp engine. In meanwhile, you can try this regular expression:

a = "http://groups-beta.google.com/group/rubyonrails-talk/browse_thread/thread/8f085b191387d799/e78a71cbd7354c0c#e78a71cbd7354c0c"

p $& if a=~ /(^http?:\/{2})\S+\.(\w+)(\S+)$/

#=> "http://groups-beta.google.com/group/rubyonrails-talk/browse_thread/thread/8f085b191387d799/e78a71cbd7354c0c#e78a71cbd7354c0c"

p $1 if a=~ /(^http?:\/{2})\S+\.(\w+)(\S+)$/

#=> "http://"

p $3 if a=~ /(^http?:\/{2})\S+\.(\w+)(\S+)$/

#=>"/group/rubyonrails-talk/browse_thread/thread/8f085b191387d799/e78a71cbd7354c0c#e78a71cbd7354c0c"

Yes, in above regular expression doesn't serve your purpose reply back with what exactly you what?

6

hi, thanks for help that will do ~gaurav

7

You’re welcome!

Sorry for my RegExp, there might be too much recursion for the Regexp engine… Take care with this one (^http?:/{2})\S+.(\w+)(\S+)$ though :

r=/(^http?:/{2})\S+.(\w+)(\S+)$/ ‘htt://==.f-é#-[==/’=~r

=>0