Using regex to find a pattern which does not start with a certain String

Question

I need to regex-match numbers in a certain pattern which works already, but only if there is not (+ right in front of it.

Example Strings I want to have a valid match within: 12, 12.5, 200/300, 200/300/400%, 1/2/3/4/5/6/7

Example Strings I want to have no valid match within: (+10% juice), (+4)

I can already get all the valid matches with (\d+[/%.]?)+, but I need help to exclude the example Strings I want to have no valid match in (which means, only match if there is NOT the String (+ right in front of the mentioned pattern).

Can someone help me out? I have already experimented with the ! (like ?!(\(\+)(\d+[/%.]?)+) but for some reason I can't it to work the way I need it.

(You can use http://gskinner.com/RegExr/ to test regex live)

EDIT: I did maybe use wrong words. I don't want to check if the searchstring does start with (+ but I want to make sure that there is no (+ right in front of my String.

Try regexr with the following inputs:

Match: (\d+[/%.]?)+

Check the checkbox for global (to search for more than one match within the text)

Text:

this should find a match: 200/300/400
this shouldnt find any match at all: (+100%) 
this should find a match: 40/50/60%
this should find a match: 175

Currently it will find a match in all 4 lines. I want a regex that does no longer find a match in line 2.

Try ^(\d[/%.]?)+, i.e. "match only strings that start with a number", so it will work for the examples you provided, but will not be able to extract those patterns from a longer string (which you didn't ask for anyway). — n.st
– n.st, Commented Oct 11, 2013 at 8:34
Are you interested in other possibilities, like +100 returning 100 as a match? Every answer so far seems to think that you only want to disregard numbers occurring after +. — Nightfirecat
– Nightfirecat, Commented Oct 11, 2013 at 8:36

James S · Accepted Answer · 2013-10-11 09:22:03Z

1

The regex construct you are wanting is "Negative Lookbehind" - See http://www.regular-expressions.info/lookaround.html. A negative lookbehind is defined like (?<!DONTMATCHME) where DONTMATCHME is the expression you don't want to find just before the next bit of the expression. Crutially, the lookbehind bit is not considered part of the match itself

Your expression should be:

(?<![+\d/\.])(\d+[/%.]?)+

Edit - changed negative lookbehind to any character that is not a + or another digit! Edit #2 - moved the lookbehind outside the main capture brackets. Expanded the range of not acceptable characters before the match to include / & .

edited Oct 11, 2013 at 9:22

answered Oct 11, 2013 at 8:30

James S

3,59818 silver badges26 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

pimeys Over a year ago

This would still match 00% within (+100%)

James S Over a year ago

Also Note - as Sniffer rightly points out, this answer doesn't work in engines that don't support lookbehind - i.e. javascript

pimeys Over a year ago

Hm, this would still match 5% within (+6.5%)

pimeys Over a year ago

This seems to be flawless now, thanks alot for your effort :)

James S Over a year ago

TBH I wonder if a positive lookbehind might be even better: Try (?:^|(?<=\s))(\d+[/%.]?)+ if theres any other issues - which matches either the start of the string (with ^) or uses a positive lookbehind to say there must be a whitespace character before the expression.

Collectives™ on Stack Overflow

Using regex to find a pattern which does not start with a certain String

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related