MP3 and cue file joining

README.md

I wanted to combine some old fashioned MP3 files and I did not want to download a separate application. Only thing necessary when combining MP3 files is stripping the ID3 tags - if present.

php -f join_mp3.php {destination} [{source}]

Edit: Turned out I has some multidisc mixtapes with .cue files lying around as well. So I wrote a joiner script for that as well:

php -f join_cue.php {destination} [{source}:{duration_of_source}]

NB These scripts come without any warranty :)

NB2 Joining MP3 files can also be done via ffmpeg - https://bytefreaks.net/applications/how-we-concatenate-multiple-mp3-files-into-one-using-ffmpeg - but writing it yourself is way more fun.

join_cue.php

<?php

/*
 * https://www.gnu.org/software/ccd2cue/manual/html_node/INDEX-_0028CUE-Command_0029.html
 */

final class Time
{
    private $seconds;
    private $frames;
    public function __construct(int $seconds, int $frames = 0)
    {
        $this->seconds = $seconds;
        $this->frames = $frames;
    }

    public function seconds(): int { return $this->seconds; }
    public function frames(): int { return $this->frames; }

    public function add(Time $time): Time
    {
        $frames = $this->frames() + $time->frames();

        return new Time(
            $this->seconds() + $time->seconds() + floor($frames / 75),
            $frames % 75
        );
    }

    public function toCueNotation(): string
    {
        return sprintf(
            '%02d:%02d:%02d', 
            floor($this->seconds / 60),
            $this->seconds % 60,
            $this->frames
        );
    }

    public static function fromTimeNotation(string $duration): Time
    {
        $parts = array_reverse(explode(':', $duration));
        $seconds = 
            $parts[0] +
            intval($parts[1] ?? '0') * 60 +
            intval($parts[2] ?? '0') * 3600;

        return new Time($seconds);
    }

    public static function fromCueNotation(string $duration): Time
    {
        $parts = array_reverse(explode(':', $duration));

        return new Time(
            intval($parts[1] ?? '0') + intval($parts[2] ?? '0') * 60,
            intval($parts[0] ?? '0')
        );
    }
}

final class Track
{
    private $raw;
    private $number;
    private $indices;

    public function __construct(string $raw, int $number, string ...$indices)
    {   
        $this->raw = $raw;
        $this->number = $number;
        $this->indices = $indices;
    }

    public function raw(): string { return $this->raw; }
    public function number(): int { return $this->number; }
    public function indices(): array { return $this->indices; }
}

function extract_tracks(string $contents): array
{
    preg_match_all('/(?P<track>\s*TRACK (?P<number>\d+) AUDIO.+?(\s*INDEX \d{2} \d+(:\d+)*)+)/s', $contents, $matches);

    $tracks = [];
    for($i = 0; $i < count($matches['track']); $i++) {
        preg_match_all('/INDEX \d+ (?P<time>\d+(:\d+)*)/s', $matches['track'][$i], $indexMatches);

        $tracks[] = new Track(
            $matches['track'][$i],
            intval($matches['number'][$i]),
            ...$indexMatches['time']
        );
    }

    return $tracks;
}

function debug(string $line): void
{
    print $line . PHP_EOL;
}

if (count($argv) < 3) {
    $filename = basename(__FILE__);
    print "Usage php -f ${filename} <destination> [<source>:<duration>]" . PHP_EOL;
    exit(0);
}

$destination = $argv[1];
$sources = array_slice($argv, 2);

$trackNumber = 0;
$totalDuration = new Time(0);

foreach ($sources as $index => $source) {
    if (preg_match('/^(?P<filename>.+?):(?P<duration>(\d+:)*\d+)$/si', $source, $matches) !== 1) {
        print "Unable to split filename and duration for ${source}" . PHP_EOL;
        exit(1);
    }

    $filename = $matches['filename'];
    $time = Time::fromTimeNotation($matches['duration']);

    $contents = utf8_encode(file_get_contents($filename));
    $tracks = extract_tracks($contents);

    if ($index === 0) {
        // Copy first file as is
        file_put_contents($destination, trim($contents));
    } else {
        foreach ($tracks as $track) {
            $raw = str_replace( // update track number
                [sprintf('TRACK %d', $track->number()), sprintf('TRACK %02d', $track->number())], 
                sprintf('TRACK %02d', $trackNumber + $track->number()), 
                $track->raw()
            );

            foreach ($track->indices() as $index) {
                $newIndex = Time::fromCueNotation($index)->add($totalDuration);

                $raw = str_replace(
                    $index,
                    $newIndex->toCueNotation(),
                    $raw
                );
            }

            file_put_contents($destination, $raw, FILE_APPEND);
        }
    }

    $trackNumber += count($tracks); // let's assume sequential track numbers
    $totalDuration = $totalDuration->add($time);
}

// Leave a nice line break at the end of the file
file_put_contents($destination, PHP_EOL, FILE_APPEND); 

join_mp3.php

<?php

/*
 * SETTINGS
 */
$CHUNK_SIZE = 8096;
$KEEP_ID3_TAGS_OF_FIRST_FILE = true;

function skip_id3_v1 ($handle): void 
{
    $position = ftell($handle);

    $id = fread($handle, 3);
    if ($id === 'TAG') {
        // ID3 v1 is always 128 byte long
        fseek($handle, 128 - 3, SEEK_CUR);
    } else {
        // reset
        fseek($handle, $position, SEEK_SET);
    }

}

function skip_id3_v2($handle): void
{
    /*
    * Inspiration for the ID3v2 code is drawn from http://www.zedwood.com/article/php-calculate-duration-of-mp3
    */

    $position = ftell($handle);

    // pre-emptive read 10 bytes
    $block = fread($handle, 10);

    if (substr($block, 0, 3) === 'ID3') {
        $flagIsFooterPresent = (bool)ord($block[5]) & 0x10;
        $z0 = ord($block[6]);
        $z1 = ord($block[7]);
        $z2 = ord($block[8]);
        $z3 = ord($block[9]);
        if ( (($z0&0x80)==0) && (($z1&0x80)==0) && (($z2&0x80)==0) && (($z3&0x80)==0) ) {
            $tagSize = (($z0&0x7f) * 2097152) + (($z1&0x7f) * 16384) + (($z2&0x7f) * 128) + ($z3&0x7f);

            fseek($handle, $tagSize + ($flagIsFooterPresent ? 10 : 0), SEEK_CUR);
        } else {
            // reset
            fseek($handle, $position, SEEK_SET);    
        }
    } else {
        // reset
        fseek($handle, $position, SEEK_SET);
    }
}

function debug(string $line): void 
{
    print $line . PHP_EOL;
}

if (count($argv) < 3) {
    $filename = basename(__FILE__);
    print "Usage php -f ${filename} <destination> [<source>]" . PHP_EOL;
    exit(0);
}

$destination = $argv[1];
$sources = array_slice($argv, 2);

$writer = fopen($destination, 'wb+');

foreach ($sources as $index => $source) {
    debug("Adding ${source}");

    $reader = fopen($source, 'rb+');

    if ($index !== 0 || $KEEP_ID3_TAGS_OF_FIRST_FILE === false) {
        skip_id3_v2($reader);
        skip_id3_v1($reader);
    }    

    while (!feof($reader)) {
        fwrite($writer, fread($reader, $CHUNK_SIZE));
    }

    fclose($reader);
}

fclose($writer);
exit(0);