johngg has asked for the wisdom of the Perl Monks concerning the following question:
Prompted by the responses in Newbie Q:How do I compare items within a string?, in particular those of Zaxo and TedPride I had a go at expanding my attempt with some of their ideas. This lead to some confusion about the use of pos() as shown here. In order to explore this further I have written this script which attempts to find occurrences of one string within another and to annotate each occurrence with the offset (zero-based) within the string
use strict; use warnings; # 1 2 # 012345678901234567890123456789 my $string = q(The cat scattered caterpillars\n); # ^ ^ ^ # 4 9 18 print "\n Match against string - $string\n"; print "Start matching\n\n"; my $match = 1; while($string =~ /(cat)/g) { print "Match @{[$match ++]}\n"; print " found - $1\n"; print " Value of \$-[1] - $-[1]\n"; print " Value of \$+[1] - $+[1]\n"; print " Value of \$+[0] - $+[0]\n"; print " Value of \$#+ - $#+\n"; print "Value of pos(\$string) - @{[pos($string)]}\n"; } print "\nStart annotation\n\n"; $string =~ s { (cat)(?{print "Value of pos(\$string) - @{[pos($string)]}\n"}) } { $1 . "[@{[pos($string)]}]" }xeg; print "\n Annotated string - $string\n";
which when run produces
Match against string - The cat scattered caterpillars Start matching Match 1 found - cat Value of $-[1] - 4 Value of $+[1] - 7 Value of $+[0] - 7 Value of $#+ - 1 Value of pos($string) - 7 Match 2 found - cat Value of $-[1] - 9 Value of $+[1] - 12 Value of $+[0] - 12 Value of $#+ - 1 Value of pos($string) - 12 Match 3 found - cat Value of $-[1] - 18 Value of $+[1] - 21 Value of $+[0] - 21 Value of $#+ - 1 Value of pos($string) - 21 Start annotation Value of pos($string) - 7 Value of pos($string) - 12 Value of pos($string) - 21 Annotated string - The cat[4] scat[9]tered cat[18]erpillars
As per the documentation, during the matching phase the value returned by pos() corresponds with the value of $+[1], pointing to just after the match. As Zaxo pointed out, $-[1] points to the start of the match.
In the annotation stage pos() again points to just after the match in the (?{ ... }) block, where the last / .../g match left off, I think the documentation says. However, when I use pos() as part of the substitution it seems to return values corresponding to $-[1] and not $+[1].
My question is, what could be causing this apparent change in behaviour?
Cheers,
JohnGG
|
---|
Replies are listed 'Best First'. | |
---|---|
Re: Inconsistent behaviour of pos() in s/ ... / ... /
by Errto (Vicar) on May 10, 2006 at 23:10 UTC | |
by johngg (Canon) on May 11, 2006 at 08:36 UTC |
Back to
Seekers of Perl Wisdom