Some URL's cannot be parsed

We came across an issue with the regular expression to detect valid URL's.

Some example URL's from mailings we did with CiviCRM that wouldn't get parsed:

• http://www.vib.be/en/training/research-training/courses/Pages/Scientific-Writing-and-Oral-Presentations--(Leuven).aspx
• http://www.ucsia.org/main.aspx?c=*UCSIAENG2&n=121273&ct=121273
• One more which I can't disclose (it's an internal company one), but it had "|" symbols in a parameter.