PERLRE RegExp help

[iCode]

hi

i'm trying to match phone numbers (not that that matters) with a RegEx and having trouble trying to figure this out, even after reading several documents and posts here.

this is the closest i have gotten:

(?:^|\s)((?:1 |1-|\+1 )?\(?\d{3}\)?[ -]\d{3}-\d{4})(?:$|\s)

here's a sample test string:

555-5642-1212 787-888-7788 799-000-0000 344-343-4433

this matches...

787-888-7788

344-343-4433

the problem is i'm not matching the middle number and i know it's because of the outside non-capturing groups

(?:^|\s)and (?:$|\s), specifically the \s. i can understand why this is not working, but i don't know how to get around it since i need the space before and after each phone num.

i'm stuck

Edited July 7, 2011 by iCode

[enaiman]

IMO it can be done much easier with a simpler expression:

#include<array.au3>
$str = "555-5642-1212 787-888-7788 799-000-0000 344-343-4433"
$arr = StringRegExp($str, "\d{3}-\d{3}-\d{4}", 3)
_ArrayDisplay($arr)

[iCode]

sorry, i neglected to mention that i need to pull the numbers from a file where they could be mixed with whatever else, such as a web page for example. so in order to avoid what may be non-phone numbers ('abc222-333-4444' or '333-444-5555-333-444-5555'), i need to look for a space both before and after the number.

the regex i'm using will match stuff like...

+1 222-333-4444

1 222-333-4444

1-222-333-4444

which works fine until you run multiple numbers together on the same line...

222-333-4444 333-444-5555 444-555-6666

in which case it won't match the middle num

[enaiman]

Why don't you StringSplit for " " your content and search every member after that for matches?

You'll need some good logic involved to detect such cases (+1 222-333-4444, 1 222-333-4444 ...) but it can be done.

[jchd]

i need to look for a space both before and after the number.

...

222-333-4444 333-444-5555 444-555-6666

in which case it won't match the middle num

If you use a non-capturing group to match a space after a phone number, then this space is no more "available" for the space you require ahead of a patternin themiddle of a string.

You need to use look-behind and look-ahead assertions resp. which do not bump the index in the subject line and don't capture either.

(?<=^|\s)((?:1 |1-|\+1 )?\(?\d{3}\)?[ -]\d{3}-\d{4})(?=$|\s)

[iCode]

thanks very much jchd - that is exactly what i needed to know

i knew i was "using" the space, but wasn't sure why or how to get around it

brilliant!