<?php  

$fd = fopen( "text.txt", "r" ); // open file for reading
$dict = array();
$text = array();

if( $fd ) {
    //
    // read contents of text file
    //
    $c = 0;
    while( ($line = fgets($fd)) != False ) { // keep reading until end of file is reached
	    // build dictionary for each word
        $word = strtok( $line, ' ' ); // extract first word
        $word = stripPunctuation( $word );
    	$text[$c] = $text[$c] . " " . $word; // store parsed words for later
    	while( $word != False ) { // extract one word at a time
	       $word = strtok( " " );
               $word = stripPunctuation( $word );
               $text[$c] = $text[$c] . " " . $word; // store parsed words for later
	           if( strlen($word) > 0 ) {
    	       	   if( array_key_exists( $word, $dict ) ) {
        		        $dict[$word] = $dict[$word] + 1; // increment existing entry
        		   } else {
        		        $dict[$word] = 1; // add new entry to dictionary
    		   }
	       }
	}
        $c = $c + 1;
    }
    fclose( $fd ); // close file

    // build word frequency based on each word in dictionary
    $features = Array();
    $keys = array_keys( $dict );
    $i = 0;
    foreach( $text as $t ) { 
    	 $j = 0;
         foreach( $keys as $k ) {
	        $k2 = " " . $k . " ";
            $c = substr_count( $t, $k2 );
    	    $features[$i][$j] = $features[$i][$j] + $c;
	        $j = $j + 1;
         }
         $i = $i + 1;
    }

    // K-means clustering
    // your code here...

} else {
    print( "Could not open file... aborting\n" );
} 


//
// point-wise add two arrays
//
function addArrays( $A, $B ) {
    $C = Array();
    $n = sizeof( $A );
    for( $i=0 ; $i<$n ; $i++ ) {
         $C[$i] = $A[$i] + $B[$i];
    }
    return( $C );
}

//
// compute Euclidean distance between two vectors (cluster center and feature)
//
function computeDistance( $x, $y ) {
    $n = sizeof( $x );
    $dist = 0;
    for( $i=0 ; $i<$n ; $i++ ) {
         $dist = $dist + pow( $x[$i]-$y[$i], 2 );
    }
    $dist = sqrt( $dist );
    return( $dist );
}

//
// strip punctuation from a string
//
function stripPunctuation( $s ) {
     $s = str_replace( ",", "", $s );
     $s = str_replace( ".", "", $s );
     $s = str_replace( ";", "", $s );
     $s = str_replace( ":", "", $s );
     $s = str_replace( "-", "", $s );
     $s = str_replace( "(", "", $s );
     $s = str_replace( ")", "", $s );
     $s = str_replace( "?", "", $s );
     $s = str_replace( "\n", "", $s );
     return( $s );
}

?>