my $dist = $1 if ($str =~ /dist:.([0-9]+)/s );
[download]

Hi Everyone. I've tried your suggestions (and combos of) and the one that seems to work best is

(/dist:.(\d+)/s)

The only case I haven't gotten this one to work on is:

dist:
45.3 km;

Any suggestions?

I should have mentioned earlier that the number involved is a real number and that I'm reading these numbers from an input data file. Right now, only the integer part of the number is being read. I'm in the process of looking up how to handle that one right now. Thanks again!

You're really close. Try this:

( /dist:.*([\d.]+)/s )
[download]

Note that the period, when inside square brackets, is treated as a literal, not as a wildcard.

-- Dan

The /s modifier mainly allows the . char in the regex to also capture newlines whereas it otherwise would not.

-enlil

The latest version of coding I've been trying to use is:

(/dist:(.+)$/ms)

This works for all but the case where "dist:" and the number are on different lines.

This code seems to do what you're looking for:


#!/usr/bin/perl

use strict;
use warnings;

my @data = ( "dist: 45 km;", "dist: 45\nkm;", "dist:\n45 km;" );

foreach my $chunk( @data ){
    my( $distance ) = $chunk =~ /dist:.*?(\d+).*?km;/s;
    print STDOUT "$distance\n";
}
[download]


Here's a breakdown of the regexp:


    /dist:    # match 'dist:'
     .*?      # match anything (including newlines), but only until th
+e next part of the pattern starts to match
     (\d+)    # match 1 or more digits, and put these into $1
     .*?      # same as before
     km;      # match 'km;' (not needed unless you want to match a num
+ber of these in one string)
    /s        # treat the whole line as a single string
[download]



Wonko's code works pretty much the same way. I think part of your problem may be that you appended "/ms" to your pattern. 'm' makes the regexp treat the string as multiple lines, and 's' makes the regexp treat the string as a single line, which is what you want it to do in this case. I don't know how those opposing options interact, but that may have caused some of your difficulty.

One other option is to run $chunk =~ s/\n/ /; over each chunk. That will take all of the newlines out of each, so that all of the strings you gave as examples will evaluate to the same string. That would make a regexp much easier to write.
--
Love justice; desire mercy.

I believe your problem is the /m flag - you're telling perl to match '$' to end of lines in the middle of a string. Try removing it.

-- Dan