#!/usr/bin/perl -w
use strict;
use File::Find;
use DB_File;
$|++;


my $WEB = "/home/user/public_html/";   
my $HOST = "http://www.easyya.com/"; # HOST and webdir must keep the same trailing

###############################################################
my %db;
my $fileno =0;

unlink("search_index.db");
dbmopen(%db,"search_index.db",0644) or die "dbmopen: $!";
finddepth(\&wanted,$WEB);

sub wanted {   
   return if ($File::Find::dir =~ /usage/i);
   return unless /\.(html|txt)$/;
   my $filename = $File::Find::name;   
   return if -l $filename;
   print "indexing No.$fileno $filename\n";

   my $last_mod = localtime( (stat)[9] );
   my ($t1,$m,$d,$t2,$y) = split /\s+/,$last_mod;
   $last_mod = $m." ".$d.", ".$y;

   my $parser = MyParser->new;
   $parser->parse_file($filename);
       
   # store link and title of the file into database
   my $title =$parser->{title}; 
   $title = $_ if ( !(defined $title) or $title=~ /^\s+$/);
   $filename =~ s/$WEB/$HOST/;
   $db{-$fileno} = "$filename<>$title<>$last_mod";

   # extract words from file & store in database    
   my $last = "";
   my @words = $parser->{TEXT}=~ /\w+/g;
   foreach my $word (sort @words) {       
     next if lc $word eq $last;
     $last = $word;
     $db{$word} .= "-$fileno";
   }     
   $fileno++;
}

BEGIN{
  package MyParser;
  require HTML::Parser;
  @MyParser::ISA = qw(HTML::Parser);  
  sub start {
     my ($self,$tag) = @_;
     $self->{TITLE} = "defined" if ($tag eq 'title');     
  }
  sub end {
     my ($self,$tag) = @_;
     undef $self->{TITLE} if ($tag eq 'title' && $self->{TITLE});
  }
  sub text {
     my ($self,$text) = @_;
     $self->{title} = $text if ($self->{TITLE});
     $self->{TEXT} .= $text;
  }
}
untie %db;
1;
__END__