# A little regexp help for a newbie **URL:** \ **Category:** Ruby\ **Created:** [August 4, 2006, 7:22am UTC](https://www.ruby-forum.com/t/a-little-regexp-help-for-a-newbie/101318 "2006-08-04T07:22:52Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![RJ\_Melton](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/f08c70/32.png) [@RJ\_Melton](https://www.ruby-forum.com/u/RJ_Melton)\ **Post date:** [August 4, 2006, 7:22am UTC](https://www.ruby-forum.com/t/a-little-regexp-help-for-a-newbie/101318/1 "2006-08-04T07:22:52Z")
Sorry this is such an absurdly simple question, but I can’t find the answer after some googling, probably just me not knowing how to look. Anyway, I have a simple string in the format 1a2+3 and want to basically take out the non-number bits and put the numbers into an array. It also has to match more than single-digit numbers. The rest is simple enough, but I’m just really new to regex and can’t make it work. Anyone feel like helping me out a little, or at least pointing me in the right direction? Thanks. ---
**Author:** ![RJ\_Melton](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/f08c70/32.png) [@RJ\_Melton](https://www.ruby-forum.com/u/RJ_Melton)\ **Post date:** [August 4, 2006, 7:34am UTC](https://www.ruby-forum.com/t/a-little-regexp-help-for-a-newbie/101318/2 "2006-08-04T07:34:52Z")
On 8/4/06, RJ Melton [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > Anyway, I have a simple string in the format 1a2+3 and want to basically > take out the non-number bits and put the numbers into an array. > It also has to match more than single-digit numbers. I’m pretty sure this is non-optimal, but you could use String.split: “1a23+4-123”.split(/[^\d]/) the argument to split is a regex that matches anything that’s not a digit ( \d is a digit; [^\d] is anything but that). split uses that argument to decide where to split the string up. ;D ---
**Author:** ![RJ\_Melton](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/f08c70/32.png) [@RJ\_Melton](https://www.ruby-forum.com/u/RJ_Melton)\ **Post date:** [August 4, 2006, 7:40am UTC](https://www.ruby-forum.com/t/a-little-regexp-help-for-a-newbie/101318/3 "2006-08-04T07:40:54Z")
I tried that and it works fine, optimal or not. Thanks. 🙂 ---
**Author:** ![RJ\_Melton](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/f08c70/32.png) [@RJ\_Melton](https://www.ruby-forum.com/u/RJ_Melton)\ **Post date:** [August 4, 2006, 7:46am UTC](https://www.ruby-forum.com/t/a-little-regexp-help-for-a-newbie/101318/4 "2006-08-04T07:46:54Z")
What about “1a2+b”.scan(/\d+) ? It returns the numbers as array. RJ Melton schrieb: ---
**Author:** ![RJ\_Melton](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/f08c70/32.png) [@RJ\_Melton](https://www.ruby-forum.com/u/RJ_Melton)\ **Post date:** [August 4, 2006, 7:52am UTC](https://www.ruby-forum.com/t/a-little-regexp-help-for-a-newbie/101318/5 "2006-08-04T07:52:55Z")
On 8/4/06, Robert R. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > What about “1a2+b”.scan(/\d+) ? > It returns the numbers as array. Muuuuch better! ---
**Author:** ![RJ\_Melton](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/f08c70/32.png) [@RJ\_Melton](https://www.ruby-forum.com/u/RJ_Melton)\ **Post date:** [August 4, 2006, 7:55am UTC](https://www.ruby-forum.com/t/a-little-regexp-help-for-a-newbie/101318/6 "2006-08-04T07:55:56Z")
On 8/4/06, Daniel B. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > [http://danielbaird.com](http://danielbaird.com) (TiddlyW;nks! :: Whiteboard Koala :: Blog :: > Things That Suck) Will this pick up negative numbers or floats? ---
**Author:** ![RJ\_Melton](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/f08c70/32.png) [@RJ\_Melton](https://www.ruby-forum.com/u/RJ_Melton)\ **Post date:** [August 4, 2006, 8:01am UTC](https://www.ruby-forum.com/t/a-little-regexp-help-for-a-newbie/101318/7 "2006-08-04T08:01:56Z")
> > On 8/4/06, Robert R. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > > > > > What about “1a2+b”.scan(/\d+) ? > > > It returns the numbers as array. > > Will this pick up negative numbers or floats? Nope, only plain digits. if you want other stuff, you need to add it to the regex. /\d+/ …is just digits /[±]?\d+/ …optional + or - /[±]?\d+.?\d\*/ …optional + or -, and optional . untested… (sorry, too lazy to test 🙂 ---
**Author:** ![RJ\_Melton](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/r/f08c70/32.png) [@RJ\_Melton](https://www.ruby-forum.com/u/RJ_Melton)\ **Post date:** [August 4, 2006, 4:15pm UTC](https://www.ruby-forum.com/t/a-little-regexp-help-for-a-newbie/101318/8 "2006-08-04T16:15:24Z")
p “-1.2 56 0.443 4.22E-12 000555”.scan( /(?:[±])?[0-9]+(?:[,.][0-9]+)?(?:E[±][1-9][0-9]\*)?/ ) #=\> ["-1.2", “56”, “0.443”, “4.22E-12”, “000555”] This looks okay for most purposes, in my opinion. 🙂 You can google for “real number regex” et cetera… Daniel B. schrieb: