Use regex /r\d+=\d+/: “http://test.com?t&r12=1&r122=1&r1=1&r124=1”.scan(/r\d+=\d+/) # => [“r12=1”, “r122=1”, “r1=1”, “r124=1”] “http://test.com?t&r12=1&r124=1”.scan(/r\d+=\d+/) # => [“r12=1”, “r124=1”] You can use join to get a string output. Here: “http://test.com?t&r12=1&r122=1&r1=1&r124=1”.scan(/r\d+=\d+/).join(‘,’) # => “r12=1,r122=1,r1=1,r124=1” Update If the URL contains other parameters that may include r in end, the regex can be made stricter: a = [] “http://test.com?r1=2&r12=1&r122=1&r1=1&r124=1&ar1=2&tr2=3&xy4=5”.scan(/(&|\?)(r+\d+=\d+)/) {|x,y| a << … Read more