Problems with ereg

knight

I'm trying to use ereg to just get the version number of out the command perl -v but nothing outputs what am I doing wrong?

<?
$doh = ( perl -v);
$regex = (ereg ("v\d.\d.\d", $doh) );
echo "$regex";
?>

Anon

What's happening here is that you're feeding the PHP regex engine with the letter d, rather than a digit character class. It won't work the same way as you expect with perl.

Instead, just try using the expression "v[0-9].[0-9].[0-9]"

knight

alright I did what you said but instead of getting the out put v5.6.0 which is the version I know that is on the server I just get the number 1.

Anon

Try reading up on posix regular expressions. On most Unix / Linux / BSD boxen, a 'man 7 regex' will get you a nice little page on how they work. They're so simple it's not even humorous (WAYYY simpler than perl regexes are)

Also, ereg returns an int as a true/false indicator, so the 1 meant it found a match, but it returns matches in an array defined when you call it. (i.e. the third arg to ereg)

Try this:

$doh = perl -v;
$regex = ereg ("[0-9].[0-9].[0-9]", $doh, $args);
echo "$args[0]";

I got rid of those annoying extra parens for you, they were making my eyes cross :⁾

knight

hehe thank you this thing has been giving me a headache since I'm converting a system info utility done in perl to php

reagent

It sounds as if you are already familiar with regular expressions in perl, so i suggest reading up on the PCRE functions in php (i.e. preg_match, preg_replace,...)

I don't know why people bother with ereg anymore and the [:digit:] [:space:] character classes, when you have \d and \s

smarlowe

Because posix regexes are more transportable, easier to read, and, in my testing, generally quicker. Plus there's no accidental back references to make your life hell when Perl gives you 20 things instead of the one you expected.

They both have their uses, but for beginners to regex, posix is MUCH more simple than perl, and much less error pront.

What is the PERL equivalent for what I posted?

reagent

They both have their uses, but for beginners to regex, posix is MUCH more simple than perl, and much less error pront.

True to some extent. I did find that posix was easier to learn and understand special character classes, but when my regex's got to be too long, i decided to suck it up and learn PCRE 😉 I can't imagine going back now, but it's all what you're comfortable with.

What is the PERL equivalent for what I posted?

$regex = ereg ("[0-9].[0-9].[0-9]", $doh, $args);

did you mean:

$regex = ereg ("[0-9].[0-9].[0-9]", $doh, $args);

escaping the special '.' char that matches 'any' character

i would have written it
$regex = ereg ("[[:digit:]].[[:digit:]].[[:digit:]]", $doh, $args);

but that's just me.
for PCRE

$regex = preg_match('/\d.\d.\d/', $doh, $args)

sorry for the long-winded response 🙂
p.