Regex Help

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

Hi all,

I am trying to parse the following data for gap(-) counts and their positions:

KPHVNVGTIGHVDHGKTTL-------AISHVLAKTYGGEAKDFSQ--------IDNAP  
KPVLNVAFIGHVDAGKSTTVGRLLLDSGTIDPQIIERLKREASEKGKAGFEFAYVMDGLK
[download]

I am using

$qgapstart= $-[0]; 

$qgapend= $+[0];
[download]

to catch the positions of the /-+/.

My Question is :it reads only TL-------AAI and skips FSQ--------IDN occuring in the same line.

How do i do recursive search for a pattern withinn the same line.

Please help me.
Thanks in advance,

Comment on Regex Help Select or Download Code

Replies are listed 'Best First'.
Re: Regex Help by pc88mxer (Vicar) on Jul 10, 2008 at 15:27 UTC
Here's one way to do it: `$x = "KPHVNVGTIGHVDHGKTTL-------AISHVLAKTYGGEAKDFSQ--------IDNAPKPVLNV +AFIGHVDAGKSTTVGRLLLDSGTIDPQIIERLKREASEKGKAGFEFAYVMDGLK"; while ($x =~ m/(-+)/g) { print pos($x), " ", length($1), "\n"; } __END__ 26 7 53 8` [download]	[reply] [d/l]
Re^2: Regex Help by Anonymous Monk on Jul 10, 2008 at 15:35 UTC
Got it Buddy, Thanks a lot. Allright,I am sorry -I would use editors in future.	[reply]
Re: Regex Help by moritz (Cardinal) on Jul 10, 2008 at 15:30 UTC
It's nearly impossible to tell what you want the result to be, what your regex is, and in which way it fails for you. Try providing us with some sample data and a runnable script along with what you want it to produce. See also How (Not) To Ask A Question. That being said, perhaps you want to use a while-loop and the `/g` modifier. But it's just guesswork from my part.	[reply] [d/l]
Re: Regex Help by linuxer (Curate) on Jul 10, 2008 at 15:29 UTC
You could have read How do I post an effective node title? to format your question... You could search globally (modifier /g) with your regex and use `pos()` to identify the position of your match... `perl -e ' $str = "a---bc---d"; while ( $str =~ m/-+/g ) { print pos( $str ), $/; } '` perldoc -f pos	[reply] [d/l] [select]