# -----------------------------------------------------------------------------------------------
# Author: Dean Stringer
# Description/Purpose:
#
#	Takes a source XML file and reads into a DOM-like tree which
#	we can query using XPAth (or at least the functionality available
#	in the XML::Xpath.pm)
#
# Useful References/Resources:
#	Intro's to XPath (for anyone)...
#		http://html.about.com/library/weekly/aa110501a.htm
#		http://www.zvon.org/xxl/XPathTutorial/General/examples.html
#	XPath spec (for the brave)...
#		http://www.w3.org/TR/xpath
#
# -----------------------------------------------------------------------------------------------
# Dependencies:
#	Perl Modules:	XML::XPath (bundle)
# -----------------------------------------------------------------------------------------------
use strict;

my $syntax = "\nSyntax: $0 <file2parse> <xpathQuery>\n";
unless (@ARGV == 2) {
		print STDERR $syntax;
		exit;
}
my $file2Parse = $ARGV[0];
unless ((-e $file2Parse) && (-f $file2Parse)) {
		print STDERR "No such file $file2Parse $syntax";
		exit;
}

use XML::XPath;
use XML::XPath::XMLParser;
#use XML::XPath::Number;


my $xp = XML::XPath->new(filename => $file2Parse);
my $expression = $ARGV[1];

my $nodeset = $xp->find( $expression );
print "Found " . $nodeset->size . " nodes.\n";

foreach my $node ($nodeset->get_nodelist) {
	print XML::XPath::XMLParser::as_string($node);
}

exit;


# -----------------------------------------------------------------------------------------------
# sample XPath expressions, these are useful for querying HTML sources
#	see also: http://www.w3schools.com/xpath/xpath_examples.asp
# -----------------------------------------------------------------------------------------------
#	"//@id"				# all id values
#	"//link[@id=20]"		# links with id=20
#	"//link[@id=20]/text()"		# text contents of all links with id = '20'
#	"html/body/table",		# all tables in the body root
#	"/html/head/meta",		# all meta data
#	"/html/head/*",			# all head section
#	"//p",				# any <p> no matter where its contained
#	"//td",				# all <td>s
#	"//td/*",			# contents of all <td>s
#	"html/body/table/tr/td/table",	# 2nd level <table>s in
#	"/*/*/*/*/*/table",		# <table>s which have 5 ancestors
#	"/html/body/table/tr/td[2]/*",	# Contents of 2nd <td> of parent table
#	"/*/*/*/*/*/table[last()]",	# last <table> which has 5 ancestors - NO SUPPORT
#	"count(//p)",			# no. of <p>s
#	"//h1"				# all H1s