<?php
require_once 'libs/lib.php';
//global used fonctions
@(include 'config.php');
//configuration file generated by install.php
if (!defined('CONFIG')) {
exit(setup());
}
clean_token();
//Destroy tokens for more security
clean_add_token();
ob_start();
?>
<!DOCTYPE HTML PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1" />
<link rel="stylesheet" type="text/css" href="design/default.css" media="screen,projection" />
<link rel="alternate" type="application/rss+xml" title="<?php
echo $config['team'];
?>
Releases RSS Feed" href="rss.xml" />
<link rel="shortcut icon" href="design/favicon.ico" />
<title><?php
echo $config['team'];
?>
Release Portal</title>
</head>
<body>
<div id="contentwrapper">
/**
* @brief Align words in two strings using Smith-Waterman algorithm
*
* Strings are split into words, and the resulting arrays are aligned using Smith-Waterman algorithm
* which finds a local alignment of the two strings. Aligning words rather than characters saves
* memory
*
* @param str1 First string (haystack)
* @param str2 First string (needle)
* @param html Will contain the alignment between str1 and str2 in HTML format
*
* @return The score (0-1) of the alignment, where 1 is a perfect match between str2 and a subsequence of str1
*/
function smith_waterman($str1, $str2, &$html)
{
$score = 0.0;
// Weights
$match = 2;
$mismatch = -1;
$deletion = -1;
$insertion = -1;
// Tokenise input strings, and convert to lower case
$X = tokenise_string($str1);
$Y = tokenise_string($str2);
// Lengths of strings
$m = count($X);
$n = count($Y);
// Create and initialise matrix for dynamic programming
$H = array();
for ($i = 0; $i <= $m; $i++) {
$H[$i][0] = 0;
}
for ($j = 0; $j <= $m; $j++) {
$H[0][$j] = 0;
}
$max_i = 0;
$max_j = 0;
$max_H = 0;
for ($i = 1; $i <= $m; $i++) {
for ($j = 1; $j <= $n; $j++) {
$a = $H[$i - 1][$j - 1];
$s1 = clean_token($X[$i - 1]);
$s2 = clean_token($Y[$j - 1]);
// Compute score of four possible situations (match, mismatch, deletion, insertion
if (strcasecmp($s1, $s2) == 0) {
// Strings are identical
$a += $match;
} else {
// Strings are different
//$a -= levenshtein($X[$i-1], $Y[$i-1]); // allow approximate string match
$a += $mismatch;
// you're either the same or you're not
}
$b = $H[$i - 1][$j] + $deletion;
$c = $H[$i][$j - 1] + $insertion;
$H[$i][$j] = max(max($a, $b), $c);
if ($H[$i][$j] > $max_H) {
$max_H = $H[$i][$j];
$max_i = $i;
$max_j = $j;
}
}
}
// Best possible score is perfect alignment with no mismatches or gaps
$maximum_possible_score = count($Y) * $match;
$score = $max_H / $maximum_possible_score;
//echo "<p>Score=$score</p>";
// Traceback to recover alignment
$alignment = array();
$value = $H[$max_i][$max_j];
$i = $max_i - 1;
$j = $max_j - 1;
while ($value != 0 && ($i != 0 && $j != 0)) {
//echo $H[$i][$j] . "\n";
//echo $i . ',' . $j . "\n";
//echo $X[$i] . '-' . $Y[$j] . "\n";
//print_r($X);
//print_r($Y);
$s1 = clean_token($X[$i]);
$s2 = clean_token($Y[$j]);
if ($s2 != '') {
array_unshift($alignment, array('pos' => $i, 'match' => strcasecmp($s1, $s2) == 0 ? 1 : 0, 'token' => $X[$i]));
}
$up = $H[$i - 1][$j];
$left = $H[$i][$j - 1];
$diag = $H[$i - 1][$j - 1];
if ($up > $left) {
if ($up > $diag) {
$i -= 1;
} else {
$i -= 1;
$j -= 1;
}
} else {
if ($left > $diag) {
$j -= 1;
} else {
$i -= 1;
$j -= 1;
}
}
}
//echo $i . ',' . $j . "\n";
//echo $X[$i] . '-' . $Y[$j] . "\n";
// Store last token in alignment
$s1 = clean_token($X[$i]);
$s2 = clean_token($Y[$j]);
array_unshift($alignment, array('pos' => $i, 'match' => strcasecmp($s1, $s2) == 0 ? 1 : 0, 'token' => $X[$i]));
// HTML snippet showing alignment
// Local alignment
$snippet = '';
$last_pos = -1;
foreach ($alignment as $a) {
if ($a['pos'] != $last_pos) {
if ($a['match'] == 1) {
$snippet .= '<span style="color:black;font-weight:bold;background-color:yellow;">';
} else {
$snippet .= '<span style="color:rgb(128,128,128);font-weight:bold;background-color:yellow;">';
}
$snippet .= $a['token'] . ' ';
//$Z[$a['pos']] . ' ';
$snippet .= '</span>';
}
$last_pos = $a['pos'];
}
// Embed this in haystack string
// Before alignment
$start_pos = $alignment[0]['pos'] - 1;
$prefix_start = max(0, $start_pos - 10);
$prefix = '';
while ($start_pos > $prefix_start) {
$prefix = $X[$start_pos] . ' ' . $prefix;
$start_pos--;
}
if ($start_pos > 0) {
$prefix = '…' . $prefix;
}
// After alignment
$end_pos = $alignment[count($alignment) - 1]['pos'] + 1;
$suffix_end = min(count($X), $end_pos + 10);
$suffix = '';
while ($end_pos < $suffix_end) {
$suffix .= ' ' . $X[$end_pos];
$end_pos++;
}
if ($end_pos < count($X)) {
$suffix .= '…';
}
$html = $prefix . $snippet . $suffix;
return $score;
}
