Long regexp un multiple lines

Lorand has asked for the wisdom of the Perl Monks concerning the following question:

Replies are listed 'Best First'.
Re: Long regexp un multiple lines by davido (Cardinal) on Apr 22, 2004 at 17:41 UTC
`/^[\da-f][\da-f]: [\da-f][\da-f]: [\da-f][\da-f]: [\da-f][\da-f]: [\da-f][\da-f]: [\da-f][\da-f]$/ix` [download] Or with less typing: `/^ (?: [\da-f]{2}: ){5} [\da-f]{2} $/ix;` [download] I used the /x modifier to allow for freeform whitespace within the RE. I used the /i modifier so that hex digits can be represented in either lower or upper case. And in the second example, quantifiers were used to factor out the unnecessary repetition. Update: Fixed quantifier on last term. Dave	[reply] [d/l] [select]
Re: Long regexp un multiple lines by Enlil (Parson) on Apr 22, 2004 at 17:49 UTC
Another good place to look for use of the /x modifier is in perlretut. That said the same functionality as in the above stated regular expression could be written more tersely as: `/^(?:[\da-f]{2}:){5}[/da-f]{2}$/` [download] or with /x `/^(?: [/da-f]{2}: ){5} [/da-f]{2} $ /x` [download] or even shorter if you use qr. `$x = qr/[\da-f]{2}/;` [download] (and later on) `/^(?:$x:){5}$x$/;` [download] -enlil	[reply] [d/l] [select]
Re: Re: Long regexp un multiple lines by japhy (Canon) on Apr 22, 2004 at 18:19 UTC
Enlil (et. al.), one of the convenient things about `qr//` is that it returns a self-contained regex. That means your "later on" code can be `/^$x{5}$x$/` It seems weird to me that there's no simpler way to say "pattern X, N times, with pattern Y in between"... you have to write `/^(?:XY){N-1}X$/`. I wonder if there's something in Regexp::Common for that. _____________________________________________________ Jeff`[japhy]`Pinyan: Perl, regex, and perl hacker, who'd like a job (NYC-area) `s++=END;++y(;-P)}y js++=;shajsj<++y(p-q)}?print:??;`	[reply] [d/l] [select]
Re: Re: Re: Long regexp un multiple lines by davido (Cardinal) on Apr 22, 2004 at 20:05 UTC
Actually the reason he had to use the `(?:$x:){5}` approach is because of the colon character that's supposed to delimit each grouping. Your solution leaves that out entirely. Yours will expand to: `/^(?:[\da-f]{2}){5}(?:[\da-f]{2})$/` [download] ...note the missing : colons. Dave	[reply] [d/l] [select]
Re: Re: Re: Re: Long regexp un multiple lines by japhy (Canon) on Apr 22, 2004 at 20:46 UTC
Re: Long regexp un multiple lines by Fletch (Bishop) on Apr 22, 2004 at 17:29 UTC
`perldoc perlre`, specifically for the `/x` modifier (and you might check out the parts on quantifiers and on character classes such as `[:xdigit:]` while you're there).	[reply] [d/l] [select]
Re: Long regexp un multiple lines by ybiC (Prior) on Apr 22, 2004 at 18:29 UTC
While not directly addressing your explicit question... If your underlying problem is validating MAC addresses, then Net::Address::Ethernet might be of some interest. Particularly the `is_address` function. cheers, ybiC striving toward Perl Adept (it's pronounced "why-bick") Update: I looked through Regexp::Common, but found nothing I recognized as parsing MAC address. 8^P	[reply] [d/l]
Re: Re: Long regexp un multiple lines by Lorand (Beadle) on Apr 23, 2004 at 16:29 UTC
Yes, when I searched on the net for the regexp, I've seen that module, but why use an extra module, if it is possible without and quite simply? If I needed to check more address types, maybe...	[reply]
Re: Long regexp un multiple lines by asdfgroup (Beadle) on Apr 23, 2004 at 00:20 UTC
`6==grep {/^[\da-f]{2}$/} split ':'`	[reply] [d/l]
Re: Long regexp un multiple lines by tkil (Monk) on Apr 24, 2004 at 00:08 UTC
Funny, I'm finding all sorts of great uses for my qr//e idea today... `my $mac_re = do { my $tmp = join ':', ('[0-9a-f]' x 2) x 6; qr/$tmp/i; };` [download] This is left intentionally non-anchored, so that you can use it even when searching inside a line: `my @macs = map { m/($mac_re)/g } qx( /sbin/ifconfig -a );` [download]	[reply] [d/l] [select]