regex {min, max} range

[spaghettio05]

I'm new to regex and having some difficulty. Here is some sample code:

 

<?php

 

$t = "Apple: blah blah blah 12 Mar 2003 http";

$pat = '/(Apple|Orange): (?P<extract>.+)(\d{1,2} \w{3} \d{4})? http/';

preg_match($pat, $t, $matches);

print trim($matches["extract"]);

 

?>

 

I have 2 questions:

 

1) I want it such that the day in the date can be either 1 or 2 digits, so I would like it to return "blah blah blah" instead of "blah blah blah 1".

 

2) I would also like the date to be optional (it may be there sometimes, not other times), but can't get it to work by adding a ? after the date grouping:

$pat = '/(Apple|Orange:): (?P<extract>.+)(\d{1,2} \w{3} \d{4})? http/';

In this case it returns: "blah blah blah 12 Mar 2003"

 

This seems simple but I can't figure it out. Help would be greatly appreciated!

 

 

[btherl]

The problem is that ".+" will match anything, including the date string.  If you make the date string optional, the ".+" will eat up the date string, because it knows it doesn't need the date string in order to match the expression.

 

One possible solution is to use two regexp, one with a compulsory date and a second one with no date.  First you try to match the full expression with compulsory date string.  If that fails, then try matching without the date string.

[requinix]

Stick a space in there.

(?P.+) (\d{1,2}

[JAY6390]

Couldn't you just add ? after the .+

/(Apple|Orange): (?P<extract>.+?)(\d{1,2} \w{3} \d{4})? http/

 

so that it's not greedy

[spaghettio05]

Great, thanks for the responses - it worked!