#!/usr/bin/perl
#-------------------------------------------
# MiniSearch v0.2
# Personal Website Search Engine
# http://www.dansteinman.com/minisearch/
#
# Copyright (C) 1999 Dan Steinman
# Distributed under the terms of the of the GNU General Public License
# You may modify this progam, redistribute it in it's entirety,
# but any significant improvements must be made public
# Read the LICENSE file for more details
#-------------------------------------------
# search
# gets and displays the results of a query
#-------------------------------------------

#-------------------------------------------

if (-e "searchdata/config") {
	&getConfig;
	&ReadParse;
	
	print "Content-type: text/html\n\n";
	
	$root = $config{'root'};
	$datadir = $config{'datadir'};
	$webaddress = $config{'webaddress'};

	$view = $in{'view'} || 'default';
	@parts = split(/:/,$templates{$view});
	$promptpage = "$datadir$parts[1]";
	$resultspage = "$datadir$parts[2]";

	$qstr = $ENV{'QUERY_STRING'};

	if ($qstr eq "") {
		&displayPrompt;
	}
	else {
		&displayResults;
	}
}
else {
	# if no config go to searchindex to create one
	print "Location: searchindex.pl\n\n";
	exit(0);
}

#-------------------------------------------
# write the search prompt page
sub displayPrompt {
	open(FILE,"$promptpage");
	$output = join('',<FILE>);
	close(FILE);
	print $output;
}

#-------------------------------------------
# show the results of the search
sub displayResults {

if ($in{'query'} eq "") {
	&displayPrompt;
	exit(0);
}

$in{'query'} =~ s/'//g;
$in{'query'} =~ tr/A-Z/a-z/;
@queries = split(/ /,$in{'query'});  # the words to search for
$totalqueries = @queries;
$match = ($in{'match'} || "all");
$set = ($in{'set'} || 1);
$max = ($in{'max'} || 10);

$where = ($in{'where'} || 'all');
# convert if the where directory was a forum
if ($where ne 'all') {
	@wheredirs = split(/:/,$where); # split in case there were multiple directories to filter
	for ($i=0;$i<@wheredirs;$i++) {
		foreach $forumdir (keys(%forums)) {
			if ($wheredirs[$i] eq $forums{$forumdir}) {
				$wheredirs[$i] = $forumdir;
			}
		}
	}
}

# retrieve the list of files and get the size/date/titles
open(FILE,"$datadir/files");
while ($line = <FILE>) {
	@parts = split(/,/,$line);
	push(@files,$parts[1]);
	$file = $parts[1];
	$matches{$file} = 0;
	$sizes{$file} = $parts[2];
	$dates{$file} = $parts[3];
	$titles{$file} = $parts[4];
}
close(FILE);

# split words
open(FILE,"$datadir/words");
while (<FILE>) {
	next unless s/^(.*?) = \s*//;
	$words{$1} = [ split ];
}
close(FILE);

# find the potentially matched filenames
foreach $query (@queries) {  # loop through each of the query words
	if (@{$words{$query}}) {  # if the word has an associated file
		foreach $index (@{$words{$query}}) {
			$potmatches{$files[$index]} += 1;  # mark the file as a potential match by uping the potmatches count
		}
	}
}

# loop through all the potentially matched files and determine if it is a true match
foreach $file (keys(%potmatches)) {
	# skip if 
	next if ($match eq "any" && $potmatches{$file} == 0);

	# skip a potential match if the number of queries found was less than the number of queries searched for
	next if ($match eq "all" && $potmatches{$file} < $totalqueries);
	
	if ($where ne "all") {  # if all queries are needed to be matched
		foreach $wheredir (@wheredirs) {  # if where was given only make those filenames final matches
			if ($file =~ /^$wheredir/) {
				push(@matches,$file);
			}
		}
	}
	else {
		push(@matches,$file);  # if any queries are needed add to final matches
	}
}

$nummatches = @matches;
if ($in{'max'} eq 'unlimited') {$max = $nummatches;}

# display the results page

open(FILE,"$resultspage");
$output = join('',<FILE>);
close(FILE);

$output =~ s/##QUERY##/$in{'query'}/g;
$output =~ s/##NUMMATCHES##/$nummatches/g;
$output =~ s/##SCRIPTSEARCH##/$config{'scriptsearch'}/g;

if ($nummatches > 0) {
	$lastmatch = $max+($set-1)*$max;
	$lastmatch = $nummatches if ($lastmatch>$nummatches);
	for ($i=($set-1)*$max;$i<$lastmatch;$i++) {
		$file = $matches[$i];
		$ip1 = $i+1;
		my($ksize) = ($sizes{$file}/1000);
		$size = &round($ksize,1);
		$date = &formattedDate($dates{$file});
		# convert the url for the forum messages
		$url = $file;
		foreach $dir (keys(%forums)) {
			if ($file =~ /$dir/ && $file =~ /(.*)\/(.*)\.txt/) {
				$url = "$forums{$dir}\?msg=$2";
			}
		}		
		$resultstr .= "<p><b>$ip1.</b> <a href=\"$url\">$titles{$file}</a>";
		$resultstr .= "<br><b>URL:</b> $config{'webaddress'}$url";
		$resultstr .= "<br><b>Size:</b> ${size}KB, <b>Date:</b> $date";

	}
	$numsets = int($nummatches/$max)+1;
	$next = $set+1;
	$previous = $set-1;
	$qstr =~ s/\&set=(.*[0-9])//;
	$qstr = "$config{'scriptsearch'}?$qstr&set=";
}

$resultstr .= "<form action=\"search.pl\" method=\"get\">";
$resultstr .= "<td><input name=\"query\" type=\"text\" size=20 value=\"$in{'query'}\"></td>";
$resultstr .= "<input type=\"hidden\" name=\"match\" value=\"$in{'match'}\">";
$resultstr .= "<input type=\"hidden\" name=\"max\" value=\"$in{'max'}\">";
$resultstr .= "<input type=\"hidden\" name=\"where\" value=\"$in{'where'}\">";
$resultstr .= "<input type=\"hidden\" name=\"view\" value=\"$in{'view'}\">";
$resultstr .= "<input type=\"submit\" value=\"Search\">";

# the number link bar "more matches"
if ($numsets>1) {
	$resultstr .= "<p>More Matches:<br>";
	if ($set!=1) {
		$resultstr .= "<a href=\"${qstr}${previous}\">Previous</a> | ";
	}
	if ($set>20 && $numsets>20) {$resultstr .= " <a href=\"${qstr}20\">&lt;</a> ";}
	if ($numsets>20) {
		if ($set>20) {
			$startset = 21;
			$endset = $numsets;
		}
		else {
			$startset = 1;
			$endset = 20;
		}
	}
	else {
		$startset = 1;
		$endset = $numsets;
	}
	
	for ($i=$startset;$i<=$endset;$i++) {
		if ($i == $set) {
			$resultstr .= "$i ";
		}
		else {
			$resultstr .= "<a href=\"$qstr$i\">$i</a> ";
		}
	}
	if ($set<=20 && $numsets>20) {$resultstr .= " <a href=\"${qstr}21\">&gt;</a> ";}
	if ($set<$numsets) {
		$resultstr .= "| <a href=\"${qstr}${next}\">Next</a>";
	}
}

$output =~ s/##RESULTS##/$resultstr/;

print $output;

}

#-------------------------------------------
# some helper routines
#-------------------------------------------

# get configuration variables
sub getConfig {
	if (open(FILE,"searchdata/config")) {
		@configfile = <FILE>;
		close(FILE);	
		foreach $line (@configfile) {
			if ($line =~ /template = (.*)/) {
				@parts = split(/:/, $1);
				$templates{$parts[0]} = $1;
			}
			if ($line =~ /forum = (.*)/) {
				@parts = split(/:/, $1);
				$forums{$parts[0]} = $parts[1];
			}
			elsif ($line =~ /(.*) = (.*)/) {
				$config{$1} = $2;
			}
		}
	}
	else {
		&setDefaultConfig;
		&generateConfig(1);
	}
}

# get form input
sub ReadParse {
	local (*in) = @_ if @_;
	local ($i, $loc, $key, $val);
	if ($ENV{'REQUEST_METHOD'} eq "GET") {
		$in = $ENV{'QUERY_STRING'};
	} elsif ($ENV{'REQUEST_METHOD'} eq "POST") {
		read(STDIN,$in,$ENV{'CONTENT_LENGTH'});
	}
	@in = split(/&/,$in);
	foreach $i (0 .. $#in) {
		$in[$i] =~ s/\+/ /g;
		($key, $val) = split(/=/,$in[$i],2);
		$key =~ s/%(..)/pack("c",hex($1))/ge;
		$val =~ s/%(..)/pack("c",hex($1))/ge;
		$in{$key} .= "\0" if (defined($in{$key}));
		$in{$key} .= $val;
	}
	return 1;
}

# round $num to $dec decimal places
sub round {
	my($num,$dec) = @_;
	my($exp) = 10**$dec;
	return(int($num*$exp+$exp/20)/$exp);
}

# make it look nicer
sub formattedDate {
	my($date) = @_;
	($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,$isdst) = localtime($date);
	@months = ('Jan','Feb','Mar','Apr','May','Jun','Jul','Aug','Sep','Oct','Nov','Dec');
	if ($year<100) {$year += 1900;}
	$fdate = "$months[$mon] $mday, $year";
	return $fdate;
}

#--end-----------------------------------------