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

I'm trying to use the following code to change some xml:

$xml = '<funcCall>
<funcName>write</funcName>
<rhsValue>crlf</rhsValue>
<symConstant>loading</symConstant>
<symConstant>| |</symConstant>
<symConstant>lemmas</symConstant>
<symConstant>| |</symConstant>
<symConstant>for</symConstant>
<symConstant>| |</symConstant>
<variable>&lt;lx&gt;</variable>
<symConstant>| |</symConstant>
<symConstant>onto</symConstant>
<symConstant>| |</symConstant>
<variable>&lt;lms&gt;</variable>
</funcCall>';
$xml =~ 
    s&
        <funcCall>.
        <funcName>write</funcName>.
        <rhsValue>crlf</rhsValue>
        (.<(?:symConstant|variable)>[^<]*</(?:symConstant|variable)>)+
+.
        </funcCall>
    &
        my $text;
        foreach my $expr (1..$#-) {
            $text .= ${$expr};
        }
        $text =~ s#<rhsValue>crlf</rhsValue>#\\n#;
        $text =~ s#</?symConstant>##sg;
        $text =~ s#</?variable>##sg;
        $text =~ s#.\| \|.# #sg;
        "<write string=\"$text\"/>";
    &xse;
    
print $xml;

[download]

I want "<write string="loading lemmas for <lx> onto <lms>"/>" but am getting "<write string=" <lms>"/>".

I don't understand why, but the @- only has one element. I know the match is fine because when I print $& I can see that the ()+ matched several items. Any suggestions?

Replies are listed 'Best First'.
Re: regex replace using position loop
by wind (Priest) on Mar 25, 2011 at 20:01 UTC

    If you're working with XML, then you should consider using a cpan module like XML::Simple instead of hacking together a regex.

    use Data::Dumper;
use XML::Simple;

use strict;

my $xml = '<funcCall>
<funcName>write</funcName>
<rhsValue>crlf</rhsValue>
<symConstant>loading</symConstant>
<symConstant>| |</symConstant>
<symConstant>lemmas</symConstant>
<symConstant>| |</symConstant>
<symConstant>for</symConstant>
<symConstant>| |</symConstant>
<variable>&lt;lx&gt;</variable>
<symConstant>| |</symConstant>
<symConstant>onto</symConstant>
<symConstant>| |</symConstant>
<variable>&lt;lms&gt;</variable>
</funcCall>';

my $ref = XMLin($xml);

print Dumper($ref);

    [download]
[reply]
[d/l]
Re: regex replace using position loop
by Eliya (Vicar) on Mar 25, 2011 at 19:49 UTC
    I don't understand why, but the @- only has one element

    That's the way capturing works if you have the "+" outside of the capturing parens (i.e. only the last match is captured).  Try putting the parens around the entire subpattern: 

        ((?:.<(?:symConstant|variable)>[^<]*</(?:symConstant|variable)>)+)

    [download]

    This way you also don't need the foreach loop assembling $text — just set $text = $1

[reply]
[d/l]
[select]
Re: regex replace using position loop
by choroba (Cardinal) on Mar 25, 2011 at 20:03 UTC
    Using XML::XSH2:
    open 895570.xml ;
my $fullout ;
for funcCall/*[name()!='funcName'] {
    if (name()='rhsValue' and .='crlf') {
        $out = "" ;
    } elsif (name()='symConstant' and .='| |') {
        $out = " " ;
    } else {
        $out = (.) ;
    } ;
    $fullout .= $out ;
}
echo :s '<' (funcCall/funcName) ' string="' $fullout '"/>' ;

    [download]
[reply]
[d/l]

Back to Seekers of Perl Wisdom