Quick RegEx question

[saywell]

Hi,

I'm trying to pick out 6-digit numbers from a string.

(\d{6,6}) works, but it includes numbers of more than 6 digits, and also 6-digit numbers with a prefix, neither of which do I want.

Could one of you RegEx gurus point me in the right direction to limit the start to the beginning of a 'word' and the whole thing to match 6 and 6 only consecutive digits?

I've tried playing with /b -- (\(\d{6}) but that gives a load of blank returns in between the correct ones, though it filters the numbers with a digit prefix OK.

Regards,

William

[Melba23]

saywell,

It is always easier if you can give a few samples with the required output.

Otherwise we risk producing an SRE which does not do exactly what you want and we have to start all over again. As I age considerably each time I try to write one I would prefer to do it as little as possible.

M23

[saywell]

OK, it's the agenda from a medical meeting, which starts as a word doc [table] and is copied to a string as per the attached text file [anonymised].

I need to get all the stand-alone 6 digit numbers - excluding components of the YH-prefixed 8-digit number, or indeed any other numbers with a letter prefix, and any that aren't exactly 6 digits long.

If I load the file into PCRE Toolkit and use (\(\d{6}) it gives all the correct matches, and no incorrect matches apart from null strings [which faut de mieux I could take out of the array using non regex techniques].

William

[Melba23]

saywell,

This works for me:

#include <Array.au3>

$sText = FileRead("test.txt")

$aArray = StringRegExp($sText, "\s(\d{6})\s", 3)

_ArrayDisplay($aArray)

Explanation:

\s       = any horizontal whitespace
(\d{6})  = match any six digit sequence
\s       = any horizontal whitespace

Does it work for you too?

M23

P.S. Not as much as an extra grey hair - I must be getting better at SREs!

[saywell]

Yup - works for me too!!

Thanks for your help - again!

William