Regex keyword filter spec

gistfile1.txt

Keywords here are "SVG" and "CSS3"

The regex should find only words found in the text or in individual keywords in a URL seperated by dashes. It should NOT match any keywords located in shortned URL hashes.

Should match:

"I love this thing about SVG. http://foo.com/article"
"This is awesome http://super-awesome-svg.com"
"I <3 this css3 button thing http://codepen.com/abcdefg"

Should NOT match:

"This is a thing we are testing. It should not match this http://t.co/ifesVgief"
"Boogy buggy booger. It should not match this http://t.co/ewfijCSs3few"

// match any string that contains either of the whole words svg or css (\b means word boundary, which works for "-"
var r = /.*\b(svg|css3)\b.*/i;
r.exec("I love this thing about SVG. http://foo.com/article");
["I love this thing about SVG. http://foo.com/article", "SVG"]
r.exec("This is awesome http://super-awesome-svg.com");
["This is awesome http://super-awesome-svg.com", "svg"]
r.exec("I <3 this css3 button thing http://codepen.com/abcdefg")
["I <3 this css3 button thing http://codepen.com/abcdefg", "css3"]
r.exec("This is a thing we are testing. It should not match this http://t.co/ifesVgief")
null
r.exec("Boogy buggy booger. It should not match this http://t.co/ewfijCSs3few")
null

Ted -- that looks great!

Looks like @jonfriskics got it working as well --> https://gist.github.com/anonymous/11278042

var reg = new RegExp(/(\bcss3|svg\b)+|https?://(www.)?\w+.\w+\/(-css3|svg-)/ig);