The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.
# Copyright (c) 2002-2003 the World Wide Web Consortium :
#       Keio University,
#       European Research Consortium for Informatics and Mathematics
#       Massachusetts Institute of Technology.
# written by Olivier Thereaux <ot@w3.org> for W3C
#
# $Id: HTMLValidator.pm,v 1.17 2004/09/10 00:41:24 ot Exp $

package W3C::LogValidator::HTMLValidator;
use strict;


require Exporter;
our @ISA = qw(Exporter);
our %EXPORT_TAGS = ( 'all' => [ qw() ] );
our @EXPORT_OK = ( @{ $EXPORT_TAGS{'all'} } );
our @EXPORT = qw();
our $VERSION = sprintf "%d.%03d",q$Revision: 1.17 $ =~ /(\d+)\.(\d+)/;



###########################
# usual package interface #
###########################
our $verbose = 1;
our %config;

sub new
{
        my $self  = {};
        my $proto = shift;
        my $class = ref($proto) || $proto;
	# mandatory vars for the API
	$self->{RESULT}	= undef;
	# internal stuff
	$self->{VALID} = undef;
	$self->{VALID_ERR_NUM} = undef;
	$self->{VALID_SUCCESS} = undef;
	$self->{VALID_HEAD} = undef;
	# configuration for this module
	if (@_) {%config =  %{(shift)};}
	$config{ValidatorMethod} = "HEAD" ;
	$config{ValidatorHost} = "validator.w3.org" if (! exists $config{ValidatorHost});
	$config{ValidatorPort} = "80" if (!exists $config{ValidatorPort});
	$config{ValidatorString} = "/check\?uri=" if (!exists $config{ValidatorString});
	$config{ValidatorPostString} = "\;output=xml" if (!exists $config{ValidatorPostString});
        if (exists $config{AuthorizedExtensions})
        {
                $self->{AUTH_EXT} =  $config{AuthorizedExtensions};
        }
        else
        {
		$self->{AUTH_EXT} = ".html .xhtml .phtml .htm .shtml .php .svg .xml /";
	}
	if (exists $config{verbose}) {$verbose = $config{verbose}}
	@{$self->{URIs}} = undef;
        bless($self, $class);
        return $self;
}

sub uris
{
	my $self = shift;
	if (@_) { @{$self->{URIs}} = @_ }
	return @{$self->{URIs}};
}


sub valid
{
	my $self = shift;
	if (@_) { $self->{VALID} = shift }
		return $self->{VALID};
}

sub valid_err_num
{
	my $self = shift;
	if (@_) { $self->{VALID_ERR_NUM} = shift }
		return $self->{VALID_ERR_NUM};
}

sub valid_success
{
	my $self = shift;
	if (@_) { $self->{VALID_SUCCESS} = shift }
		return $self->{VALID_SUCCESS};
}

sub valid_head
{
        my $self = shift;
        if (@_) { $self->{VALID_HEAD} = shift }
                return $self->{VALID_HEAD};
}


sub auth_ext
{
	my $self=shift;
	if (@_) { $self->{AUTH_EXT} = shift}
	return $self->{AUTH_EXT};
}

sub new_doc{
	my $self=shift;
        $self->{VALID} = undef;
        $self->{VALID_ERR_NUM} = undef;
        $self->{VALID_SUCCESS} = undef;
        $self->{VALID_HEAD} = undef;
}

sub HEAD_check {
## Checking whether a document with no extension is actually an HTML/XML document
## causes a lot of requests, but internal - should be OK?
	my $self = shift;
	my $check_uri;
	use LWP::UserAgent;
	if (@_) { $check_uri = shift }
	my $ua = new LWP::UserAgent;
	my $method = "HEAD";
	my $request = new HTTP::Request("$method", "$check_uri");
	my $response = new HTTP::Response;
	$response = $ua->simple_request($request);
	my $is_html = 0;
	if ($response->is_success) # not an error, we could contact the server
	{
		my $type = $response->header('Content-Type');
		if ($type =~ /text\/html|application\/xhtml+xml|text\/xml/) #should be enough for a start
		{ 
			$is_html = 1;
			# print "URI with no extension $check_uri has content-type $type\n" if ($verbose > 2); # debug
		}
	}
	return $is_html;
}

sub trim_uris 
{
        my $self = shift;
        my @authorized_extensions = split(" ", $self->auth_ext);
        my @trimmed_uris;
        my $exclude_regexp = "";
	my @excluded_areas;
        $exclude_regexp = $config{ExcludeAreas};
	if ($exclude_regexp){
            $exclude_regexp =~ s/\//\\\//g ;
            @excluded_areas = split(" ", $exclude_regexp);
	}
	else { print "nothing to exclude" if ($verbose >2);}
	my $uri;
        while ($uri = shift)
        {
                my $uri_ext = "";
                my $match = 0;
                if ($uri =~ /(\.[0-9a-zA-Z]+)$/)
                {  
                   $uri_ext = $1;
                }
                elsif ($uri =~ /\/$/) { $uri_ext = "/";}
                elsif ( $self->HEAD_check($uri) ) { $match = 1; }
                foreach my $ext (@authorized_extensions)
                {   
                    if ($ext eq $uri_ext) { $match = 1; }
                }
		if ($match)
		{
            	  foreach my $area (@excluded_areas)
            	  {
                    if ($uri =~ /$area/)
                    {
			my $slasharea = $area;
                        $slasharea =~ s/\\\//\//g;
                        $slasharea =~ s/\\././g;
                        print "Ignoring $uri matching $slasharea \n" if ($verbose > 2) ;
                        $match = 0;
                    }

		  }
                }
		
                push @trimmed_uris,$uri if ($match);
        }
        return @trimmed_uris;
}

#########################################
# Actual subroutine to check the list of uris #
#########################################

sub process_list
{
	my $self = shift;
	print "Now using the HTML Validator module... " if $verbose;
	print "\n" if ($verbose > 1);
	my @uris = undef;
	my %hits;
	# Opening the file with the hits and URIs data
	if (defined ($config{tmpfile}))
	{
		use DB_File; 
		my $tmp_file = $config{tmpfile};
		tie (%hits, 'DB_File', "$tmp_file", O_RDONLY) || 
		    die ("Cannot create or open $tmp_file");
		@uris = sort { $hits{$b} <=> $hits{$a} } keys %hits;
	}
	elsif ($self->uris())
	{
		@uris = $self->uris();
		foreach my $uri (@uris) { $hits{$uri} = 0 }
	}
	print "\n (This may take a long time if you have many files to validate)\n" if ($verbose eq 1);
	print "\n" if ($verbose > 2); # trying to breathe in the debug volume...
	use LWP::UserAgent;
	use URI::Escape;
	my $max_invalid = undef;
	if (exists $config{MaxInvalid}) {$max_invalid = $config{MaxInvalid}}
	else {$max_invalid = 0}
	my $max_documents = undef;
	if (exists $config{MaxDocuments}) {$max_documents = $config{MaxDocuments}}
	else {$max_documents = 0}
	my $name = ""; 
	if (exists $config{ServerName}) {$name = $config{ServerName}}
	@uris = $self->trim_uris(@uris);
	my @result;
	my @result_head;
	my $intro="Here are the <census> most popular invalid document(s) that I could find in the 
logs for $name.";
	my $outro;
	push @result_head, "Rank";
	push @result_head, "Hits";
	push @result_head, "#Error(s)";
	push @result_head, "Address";
	my $invalid_census = 0; # number of invalid docs
	my $last_invalid_position = 0; # latest position at which we found an invalid doc
	my $total_census = 0; # number of documents checked
	my $ua = new LWP::UserAgent;
#	$ua->timeout([30]); # instead of 180. 3 minutes timeout is too long.
	my $uri = undef;
	while ( (@uris) and  (($invalid_census < $max_invalid) or (!$max_invalid)) and (($total_census < $max_documents) or (!$max_documents)) )
	{
		$uri = shift (@uris);
		$self->new_doc();
		my $uri_orig = $uri;
		$total_census++;
		print "	processing #$total_census $uri..." if ($verbose > 1);
		# escaping URI
		$uri = uri_escape($uri);
		# creating the HTTP query string with all parameters
		my $string=join ("", "http://",$config{ValidatorHost},":",$config{ValidatorPort},
		$config{ValidatorString},$uri,$config{ValidatorPostString});
		my $method = $config{ValidatorMethod};
		my $request = new HTTP::Request("$method", "$string");
		my $response = new HTTP::Response;
		$response = $ua->simple_request($request);
		if ($response->is_success) # not an error, we could contact the server
		{
			# set both valid and error number according to response
			$self->valid($response->header('X-W3C-Validator-Status'));
			$self->valid_err_num($response->header('X-W3C-Validator-Errors'));
			# we know the validator has been able to (in)validate if $self->valid is not NULL
			if ( ($self->valid) and ($self->valid_err_num) ) # invalid doc
#			if (1) # debug
			{
				my @result_tmp;
				push @result_tmp, $total_census;
				push @result_tmp, $hits{$uri_orig};
				push @result_tmp, $self->valid_err_num;
				push @result_tmp, $uri_orig;
				push @result, [@result_tmp];
				$invalid_census++;
				$last_invalid_position = $total_census;
			}
			printf (" %s!", $self->valid) if ( ($verbose > 1) and (defined ($self->valid)));
			print " Could not validate!" if (($verbose > 1) and(!defined ($self->valid)));

			if (($verbose > 1) and ($self->valid_err_num)) # verbose or debug
				{printf ", %s errors!",$self->valid_err_num}
		}
		else { print " Could not validate!" if ($verbose > 1) }
		print "\n" if ($verbose > 1);

		$self->valid_head($response->as_string); # for debug
		if ($verbose > 2) {printf "%s :\n%s", $string, $self->valid_head;} # debug
		sleep(1); # do not kill validator.w3.org

	}
	print "Done!\n" if $verbose;
	print "invalid_census $invalid_census \n" if ($verbose > 2 );
	if ($invalid_census) # we found invalid docs
	{
		if ($invalid_census eq 1)  # let's repect grammar here
		{
			$intro=~ s/are/is/;
			$intro=~ s/<census> //;
			$intro=~ s/document\(s\)/document/;
		}
		$intro =~s/<census>/$invalid_census/;
		my $ratio = 10000*$invalid_census/$total_census;
		$ratio = int($ratio)/100;
		if ($last_invalid_position eq $total_census )
		# usual case
		{
			$outro="Conclusion :
I had to check $last_invalid_position document(s) in order to find $invalid_census invalid HTML documents.
This means that about $ratio\% of your most popular documents were invalid.";
		}
		else
		# we didn't find as many invalid docs as requested
		{

			$outro="Conclusion :
You asked for $max_invalid invalid HTML document but I could only find $invalid_census 
by processing (all the) $total_census document(s) in your logs. 
This means that about $ratio\% of your most popular documents were invalid.";
		}
	}
	elsif (!$total_census)
	{
		$intro="There was nothing to validate in this log.";
		$outro="";
	}
	else # everything was actually valid!
	{
		$intro=~s/<census> //;
		$outro="I couldn't find any invalid document in this log. Congratulations!";
	}
	if (($total_census == $max_documents) and ($total_census)) # we stopped because of max_documents
	{
		$outro=$outro."\nNOTE: I stopped after processing $max_documents documents:\n     Maybe you could set MaxDocuments to a higher value?";
	}
	if (defined ($config{tmpfile}))
        {
		untie %hits;
	}
	my %returnhash;
        $returnhash{"name"}="HTMLValidator";
        $returnhash{"intro"}=$intro;
        $returnhash{"outro"}=$outro;
        @{$returnhash{"thead"}}=@result_head;
        @{$returnhash{"trows"}}=@result;
        return %returnhash;         
}

package W3C::LogValidator::HTMLValidator;

1;

__END__

=head1 NAME

W3C::LogValidator::HTMLValidator - check HTML validity vith validator.w3.org

=head1 SYNOPSIS

  use  W3C::LogValidator::HTMLValidator;
  my %config = ("verbose" => 2);
  my $validator = W3C::LogValidator::HTMLValidator->new(\%config);
  $validator->uris('http://www.w3.org/Overview.html', 'http://www.yahoo.com/index.html');
  my %results = $validator->process_list;

=head1 DESCRIPTION

This module is part of the W3C::LogValidator suite, and checks HTML validity
of a given document via the W3C HTML validator service.

=head1 API

=head2 Constructor

=over 2

=item $val = W3C::LogValidator::HTMLValidator->new

Constructs a new C<W3C::LogValidator:HTMLValidator> processor.  

You might pass it a configuration hash reference (see L<W3C::LogValidator/config_module> and L<W3C::LogValidator::Config>)

  $validator = W3C::LogValidator::HTMLValidator->new(\%config);  

=back

=head2 Main processing method

=over 4


=item $val->process_list

Processes a list of sorted URIs through the W3C Markup Validator.

The list can be set C<uris>. If the $val was given a config has when constructed, and if the has has a "tmpfile" key, C<process_list> will try to read this file as a hash of URIs and "hits" (popularity) with L<DB_File>.

Returns a result hash. Keys for this hash are: 


  name (string): the name of the module, i.e "HTMLValidator"
  intro (string): introduction to the processing results
  thead (array): headers of the results table
  trows (array of arrays): rows of the results table
  outro (string): conclusion of the processing results

=back

=head2 General methods

=over 4

=item $val->uris

Returns a  list of URIs to be processed (unless the configuration gives the location for the hash of URI/hits berkeley file, see C<process_list> 
If an array is given as a parameter, also sets the list of URIs and returns it.


=item $val->trim_uris 

Given a list of URIs of documents to process, returns a subset of this list containing the URIs of documents the module supposedly can handle.
The decision is made based on file extensions (see C<auth_ext>), content-type (see C<HEAD_check>) , and the setting for ExcludedAreas

=item $val->HEAD_check

Checks whether a document with no extension is actually an HTML/XML document through an HTTP HEAD request
returns 1 if the URI is of an expected content-type, 0 otherwise

=item $val->auth_ext

Returns the file extensions (space separated entries in a string) supported by the Module.
Public method accessing $self->{AUTH_EXT}, itself coming from either the AuthorizedExtensions configuration setting, or a default value

=item $val->valid

Sets / Returns whether the document being processed has been found to be valid or not.
If an argument is given, sets the variable, otherwise returns the current variable.

=item $val->valid_err_num

Sets / Returns the number of validation errors for the document being processed.
If an argument is given, sets the variable, otherwise returns the current variable.

=item $val->valid_success

Sets / Returns whether the module was able to process validation of the current document successfully (regardless of valid/invalid result)
If an argument is given, sets the variable, otherwise returns the current variable.

=item $val->valid_head

Sets / Returns all HTTP headers returned by the markup validator when attempting to validate the current document.
If an argument is given, sets the variable, otherwise returns the current variable.

=item $val->new_doc

Resets all validation variables to 'undef'. In effect, prepares the processing module to the handling of a new document.

=back

=head1 BUGS

Public bug-tracking interface at http://www.w3.org/Bugs/Public/




=head1 AUTHOR

Olivier Thereaux <ot@w3.org>

=head1 SEE ALSO

W3C::LogValidator, perl(1).
Up-to-date complete info at http://www.w3.org/QA/Tools/LogValidator/

=cut