XML-Twig

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

Oh great monks, I find I am needful your wisdom. I am trying to use Twig to process an ungainly XML file as a stream of descrete parts. I am using ActivePerl 5.6.1.631 on Win32. I have used variations of the XML tools provided for Perl before for other tasks, but these files are too big for things like XPath, etc. I have loaded the Twig module and am trying to use the twig_roots function, since all I need to do is read data from the tags of interest and then purge the rest of the tree. When I run the code below I get this error message: "Undefined subroutine &main:: called at C:/Perl/site/lib/XML/Parser/Expat.pm line 439." In Expat.pm I see the following (lines 412 through 448):

sub parse {
  my $self = shift;
  my $arg = shift;
  croak "Parse already in progress (Expat)" if $self->{Used};
  $self->{Used} = 1;
  my $parser = $self->{Parser};
  my $ioref;
  my $result = 0;
  
  if (defined $arg) {
    if (ref($arg) and UNIVERSAL::isa($arg, 'IO::Handler')) {
      $ioref = $arg;
    } else {
      eval {
        no strict 'refs';
    $ioref = *{$arg}{IO};
      };
    }
  }
  
  if (defined($ioref)) {
    my $delim = $self->{Stream_Delimiter};
    my $prev_rs;
    
    $prev_rs = ref($ioref)->input_record_separator("\n$delim\n")
      if defined($delim);
    
    $result = ParseStream($parser, $ioref, $delim);  ### Line 439
        
    ref($ioref)->input_record_separator($prev_rs)
      if defined($delim);
  } else {
    $result = ParseString($parser, $arg);
  }
  
  $result or croak $self->{ErrorMessage};
}
[download]

The only thing I can think is that ParseStream is undefined. A quick search through the Expat, Parser, and Twig modules reveals no other instance of "ParseStream." What could be going on? I have used PPM to "verify --force update" all relevant modules, but to no avail. Here is a snippet of my code:

#!C:/Perl

use XML::Parser;
use XML::Parser::Expat;
use XML::Twig;

my $t = XML::Twig->new( 
   twig_roots => { 
   '/invoice/header/accountnumber' => \&print_elt_text 
   } );
   
$t->parsefile('MyHugeXML.xml');

sub print_elt_text { 
     my( $t, $elt ) = @_;
     print $elt->text;
     $t->purge;
   }
[download]

Thank you wise ones. Your wisdom is like ambrosia.

Comment on XML-Twig Select or Download Code

Replies are listed 'Best First'.
Re: XML-Twig by bart (Canon) on Aug 27, 2002 at 00:49 UTC
The only thing I can think is that ParseStream is undefined. Urm... I don't think so. Let me tell you that I'm just guessing... It looks to me like ParseStream is defined in XML/Parser/Expat.xs, the C bind layer between plain Perl and the C library. At least I can see this in there: `int XML_ParseStream(parser, ioref, delim) XML_Parser parser SV * ioref SV * delim CODE: ...` [download] What I think, is that ParseStream() needs a callback for each "event". That means that this function calls some plain Perl subs. And I think that there is where something goes wrong. Perhaps it simply misses a proper callback sub. Could you make a small XML file available, that is compatible with your program, and which exhibits the problem? That way, I, and maybe some other people here as well, could play a little with your program.	[reply] [d/l]
Re: XML-Twig by PodMaster (Abbot) on Aug 27, 2002 at 07:49 UTC
Works perfectly well for me. This is what I get after I run my example (yours modified) DON't LEAVE OUT RELEVANT INFORMATION XML::Twig::VERSION 3.04 XML::Parser::VERSION 2.31 XML::Parser::Expat::VERSION 2.31 ... EEEEEEEEEEEEEEEEEEEEEEEKKKKK ... use XML::Parser; use XML::Parser::Expat; use XML::Twig; $\="\n"; print "DON't LEAVE OUT RELEVANT INFORMATION\n"; print "XML::Twig::VERSION $XML::Twig::VERSION"; print "XML::Parser::VERSION $XML::Parser::VERSION"; print "XML::Parser::Expat::VERSION $XML::Parser::Expat::VERSION\n"; my $t = XML::Twig->new( twig_roots => { '/node/data/field' => \&print_elt_text } ); $t->parse( \DATA ); sub print_elt_text { my( $t, $elt ) = @_; print $elt->text; $t->purge; } __DATA__ <?xml version="1.0" encoding="ISO-8859-1"?> <node id="193032" title="XML-Twig" > <type id="115"> perlquestion</type> <author id="961"> Anonymous Monk</author> <data> <field name="doctext"> ... EEEEEEEEEEEEEEEEEEEEEEEKKKKK ... </field> </data> </node> [download] I suggest removing XML::Parser and XML::Twig completely and reinstalling ( something gone bad). `____________________________________________________` * The Third rule of perl club is a statement of fact: pod is sexy.	[reply] [d/l]