What's wrong with this code ?

[codefixer]

Hi,

I am trying to do substitution for XML characters. I am getting the
following error for this piece of code.

Thanks.

sub parseXML
{
my $ret= $_;
switch ($_)
{
case "&" { print "&"; }
case "<" { print "<"; }
case ">" { print ">"; }
case "\'" { print "'"; }
case "\"" { print """; }


}


return $ret;

}
String found where operator expected at generateXML.pl line 107, near
"case "&""

(Do you need to predeclare case?)
String found where operator expected at generateXML.pl line 108, near
"case "<""

(Do you need to predeclare case?)
syntax error at generateXML.pl line 106, near ")
{"
syntax error at generateXML.pl line 108, near "case "<""
generateXML.pl had compilation errors.

 

[Tony Curtis]

On 9 Apr 2005 12:19:10 -0700,

Hi, I am trying to do substitution for XML characters. I am
getting the following error for this piece of code.

switch ($_)

Have you

use Switch;

'ed? There's no built-in switch statement.

There are also modules to do entity encoding, e.g.

HTML::Entities

but I'll let someone else narrow in on that for XML
specifically.

hth
t

 

[codefixer]

Have you

use Switch;

No when I copied the sub I forgot to add that. That was the
problem.
Thanks.

 

[Anno Siegel]

Hi,

I am trying to do substitution for XML characters. I am getting the
following error for this piece of code.

Thanks.

sub parseXML
{
my $ret= $_;
switch ($_)
{
case "&" { print "&"; }
case "<" { print "<"; }
case ">" { print ">"; }
case "\'" { print "'"; }
case "\"" { print """; }


}


return $ret;

}
String found where operator expected at generateXML.pl line 107, near
"case "&""

Tony Curtis has explained how to make the case-statement work, but in
Perl "case" is rarely needed (that's why it doesn't have it in the first
place). The routine parseXML (a rather pretentious name for a modest
routine, I'd say) would be better written like this (untested):

{
my %trans = map { $_ => sprintf '&#%02d;', ord $_ } qw( & < > ' ");
sub parseXML {
print $trans{ $_} if $trans{ $_};
$_;
}
}

This uses a hash (%trans) to distinguish the cases. A bare block is used
to give %trans a small lexical scope. It doesn't hard-code the ASCII
equivalents but lets Perl do the work, though the wisdom of this could be
debated. It should work like the original, but is much more compact.

Anno