man XML::SAX::Expat::Incremental () - XML::SAX::Expat subclass for non-blocking (incremental) parsing, with XML::Parser::ExpatNB.
NAME
XML::SAX::Expat::Incremental - XML::SAX::Expat subclass for non-blocking (incremental) parsing, with XML::Parser::ExpatNB.
SYNOPSIS
use XML::SAX::Expat::Incremental;
# don't do this, use XML::SAX::ParserFactory my $p = XML::SAX::Expat::Incremental->new( Handler => MyHandler->new );
$p->parse_start;
while (<DATA>){ $p->parse_more($_); # or $p->parse_string($_); }
$p->parse_done;
DESCRIPTION
Most XML parsers give a callback interface within an encapsulated loop. That is, you call
$p->parse_whatever($whatever);
And eventually, when CW$whatever is depleted by the parser, CW$p->parse will return.
Sometimes you don't want the parser to control the loop for you. For example, if you need to retrieve your XML in chunks in a funny way, you might need to do something like
my $doc = ''; while (defined(my $buffer = get_more_xml())) { $doc .= $buffer; }
$p->parse_string($doc);
which is not very convenient, or efficient. You could use perltie to tie a filehandle which does this for you, but that only works some of the time (for example, say you have two inputs coming in simultaneously).
XML::Parser::ExpatNB solves this by providing three methods:
- parse_start
- parse_more
- parse_done
This interface lets you move the loop to outside the parser, retaining control.
The callbacks are executed in the same manner, just that now, when there is no left to parse, instead of taking more data from a source on it's own, the parser returns control to you.
$p->parse_start; # you can omit this - parse_start will # be called automatically as needed
while(defined(my $buffer = get_more_xml())) { $p->parse_more($buffer); }
$p->parse_done;
This module is a subclass of XML::SAX::Expat which is to XML::Parser::ExpatXS as XML::SAX::Expat is to XML::Parser itself.
METHODS
- parse_string STRING
- parse_more STRING
- These have the same effect, except that parse_more actually calls parse_string with CW@_. You might want to use parse_string because in theory it's more efficient. This simply continues parsing with the new string, and sends SAX events for the data that is complete in the string.
- parse_start
- This calls parse_start on the underlying XML::Parser::ExpatNB object. It's called implicitly when you first call parse_string, though, so you don't have to worry about it.
- parse_done
- This calls parse_done on the underlying XML::Parser::ExpatNB object. You use it to tell the parser you have no more data to give it.
- parse
- This is used internally as a sort of parse-anything method. Don't use it, instead use CWparse_string, which invokes this method correctly, and takes simpler options.
SEE ALSO
XML::Parser, XML::SAX, XML::SAX::Expat, XML::SAX::ExpatNB
AUTHOR
Yuval Kogman <nothingmuch@woobling.org>
COPYRIGHT & LICENSE
Copyright (c) 2005 Yuval Kogman. All rights reserved This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself.