SplFileObject very high disk usage

[Red-Owl]

Description

// use fseek

<?php
set_time_limit(0);
![disk](https://github.com/user-attachments/assets/f02f3400-3645-4a8f-9be4-b9bf3c6dfa49)



echo "loading";
$filepath = "Video Music.txt";

$file = new SplFileObject($filepath);
$n = 0;
$lined = Array();
$data1 = Array();
$datar = "";



$file_sql = $filepath.'.sql';
// The new person to add to the file

// Write the contents to the file, 
// using the FILE_APPEND flag to append the content to the end of the file
// and the LOCK_EX flag to prevent anyone else writing to the file at the same time

$link = mysqli_connect('127.0.0.1', 'wlab', 'xxxxxx', 'test', '3307');
if (!$link) {
    die('Not connected : ' . mysql_error());
}

// make foo the current db
/* $db_selected = mysqi_select_db('torrent', $link);
if (!$db_selected) {
    die ('Can\'t use foo : ' . mysql_error());
}
*/


while (false !== ($char = $file->fgetc())) {
    //echo "$char\n";
    $n++;
    $lined[] = $char;
    if($char == "\n"){
        $datar = implode('', $lined);
        $data1 = explode("|", $datar);
        if(count($data1)==2){
        $sql = "INSERT INTO magnet SET mn_exid=NULL,mn_name='".addslashes($data1[1])."',mn_size=NULL,mn_seed=NULL,mn_leker=NULL,mn_magnet='".trim($data1[0])."', mn_categ='".basename($filepath)."', mn_origin='BitSnoop', mn_date='2012-02-12 00:00:00';\r\n";
        // echo "\n<br />"
        // file_put_contents($file_sql, $sql, FILE_APPEND | LOCK_EX);
        mysqli_query($link,$sql);
    }else{
            echo $datar; 
        }
        $datar = ""; 
        $lined = Array();
    }
    /* if($n==5000){
        exit;
    } */
 
}


echo $n;

?>

### PHP Version

 8.3.12 

### Operating System

win 11

[cmb69]

How did you conclude that the disk usage is caused by SplFileObject::fgetc(), and not by the database server?

[Red-Owl]

yes :)

[iluuu1994]

@Red-Owl In that case, can you simplify the reproducer? The queries shouldn't be needed. The file is read in 8192 byte chunks, so this shouldn't be incredibly inefficient. Of course, if you read a huge file, the process will be IO bound and then high disk usage is expected.