The London Perl and Raku Workshop takes place on 26th Oct 2024. If your company depends on Perl, please consider sponsoring and/or attending.




version 2.33


    use MP3::M3U::Parser;
    my $parser = MP3::M3U::Parser->new( %options );
    my $result = $parser->result;
    my %info   = $parser->info;
        -format   => 'xml',
        -file     => '/path/mp3.xml',
        -encoding => 'ISO-8859-9',
        -format   => 'html',
        -file     => '/path/mp3.html',
        -drives   => 'off',
    # convert all m3u files to individual html files.
    foreach ( <*.m3u> ) {
        $parser->parse( $_ )->export->reset;
    # convert all m3u files to one big html file.
    foreach ( <*.m3u> ) {
        $parser->parse( $_ );


MP3::M3U::Parser is a parser for M3U mp3 playlist files. It also parses the EXTINF lines (which contains id3 song name and time) if possible. You can get a parsed object or specify a format and export the parsed data to it. The format can be xml or html.



The object constructor. Takes several arguments like:


Format the seconds returned from parsed file? if you set this to the value format, it will convert the seconds to a format like MM:SS or H:MM:SS. Else: you get the time in seconds like; 256 (if formatted: 04:15).

If you don't want to get a list of every song in the m3u list, but want to get a specific group's/singer's songs from the list, set this to the string you want to search. Think this "search" as a parser filter.

Note that, the module will do a *very* basic case-insensitive search. It does dot accept multiple words (if you pass a string like "michael beat it", it will not search every word seperated by space, it will search the string "michael beat it" and probably does not return any results -- it will not match "michael jackson - beat it"), it does not have a boolean search support, etc. If you want to do something more complex, get the parsed tree and use it in your own search function, or subclass this module and write your own _search method (notice the underscore in the method name). See the tests for a subclassing example.


The module assumes that all of the songs in your M3U lists are (or were: the module does not check the existence of them) on the same drive. And it builds a seperate data table for drive names and removes that drive letter (if there is a drive letter) from the real file path. If there is no drive letter (eg: under linux there is no such thing, or you saved m3u file into the same volume as your mp3s), then the drive value is 'CDROM:'.

So, if you have a mixed list like:


set this parameter to 'asis' to not to remove the drive letter from the real path. Also, you "must" ignore the drive table contents which will still contain a possibly wrong value; export does take the drive letters from the drive tables. So, you can not use the drive area in the exported xml (for example).

Note: you probably want to set this parameter to 'asis' on a non-Windows machine.


Same as the -overwrite option in export but new sets this export option globally.


Same as the -encoding option in export but new sets this export option globally.


Same as the -format option in export but new sets this export option globally.


Same as the -drives option in export but new sets this export option globally.


It takes a list of arguments. The list can include file paths, scalar references or filehandle references. You can mix these types. Module interface can handle them correctly.

   open FILEHANDLE, ...

or with new versions of perl:

   open my $fh, ...

   my $scalar = "#EXTM3U\nFoo - bar.mp3";





Note that globs and scalars are passed as references.

Returns the object itself.


Must be called after parse. Returns the result set created from the parsed data(s). Returns the data as an array or arrayref.

   $result = $parser->result;
   @result = $parser->result;

Data structure is like this:

   $VAR1 = [
               'drive' => 'G:',
               'file' => '/path/to/mylist.m3u',
               'data' => [
                             'mp3\Singer - Song.mp3',
                             'Singer - Song',
                           # other songs in the list
               'total' => '3',
               'list' => 'mylist'
             # other m3u list

Each playlist is added as a hashref:

   $pls = {
           drive => "Drive letter if available",
           file  => "Path to the parsed m3u or generic name if GLOB/SCALAR",
           data  => "Songs in the playlist",
           total => "Total number of songs in the playlist",
           list  => "name of the list",

And the data key is an AoA:

   data => [
            ["MP3 PATH INFO", "ID3 INFO","TIME","ARTIST","SONG"],
            # other entries...

You can use the Data::Dumper module to see the structure yourself:

   use Data::Dumper;
   print Dumper $result;


You must call this after calling parse. It returns an info hash about the parsed data.

   my %info = $parser->info;

The keys of the %info hash are:

   songs   => Total number of songs
   files   => Total number of lists parsed
   ttime   => Total time of the songs 
   average => Average time of the songs
   drive   => Drive names for parsed lists

Note that the 'drive' key is an arrayref, while others are strings.

   printf "Drive letter for first list is %s\n", $info{drive}->[0];

But, maybe you do not want to use the $info{drive} table; see -parse_path option in new.


Exports the parsed data to a format. The format can be xml or html. The HTML File' s style is based on the popular mp3 player WinAmp' s HTML List file. Takes several arguments:


The full path to the file you want to write the resulting data. If you do not set this parameter, a generic name will be used.


Can be xml or html. Default is html.


The exported xml file's encoding. Default is ISO-8859-1. See for a list. If you don't define the correct encoding for xml, you can get "not well-formed" errors from the xml parsers. This value is also used in the meta tag section of the html file.


Only required for the html format. If set to off, you will not see the drive information in the resulting html file. Default is on. Also see -parse_path option in new.


If the file to export exists on the disk and you didn't set this parameter to a true value, export will die with an error.

If you set this parameter to a true value, the named file will be overwritten if already exists. Use carefully.

Has no effect if you use -toscalar option.


With the default configuration, export method will dump the exported data to a disk file, but you can alter this behaviour if you pass this parameter with a reference to a scalar.

   $parser->export(-toscalar => \$dumpvar);
   # then do something with $dumpvar

Returns the object itself.


Resets the parser object and returns the object itself. Can be usefull when exporting to html.

   $parser->parse($fh       )->export->reset;
   $parser->parse(\$scalar  )->export->reset;

Will create individual files while this code

   $parser->parse($fh       )->export;
   $parser->parse(\$scalar  )->export;

creates also individual files but, file2 content will include $fh + $scalar data and file3 will include $fh + $scalar + file.m3u data.


You may want to subclass the module to implement a more advanced search or to change the HTML template.

To override the default search method create a _search method in your class and to override the default template create a _template method in your class.

See the tests in the distribution for examples.

Error handling

Note that, if there is an error, the module will die with that error. So, using eval for all method calls can be helpful if you don't want to die:

    my $eval_ok = eval {
       $parser->parse( @list );
    die "Parser error: $@" if $@ || !$eval_ok;

As you can see, if there is an error, you can catch this with eval and access the error message with the special Perl variable $@.


MP3::M3U::Parser - MP3 playlist parser.


See the tests in the distribution for example codes. If you don't have the distro, you can download it from CPAN.



(For v2.80) If you don't see any EXTINF lines in your saved M3U lists, open preferences, go to "Options", set "Read titles on" to "Display", add songs to your playlist and scroll down/up in the playlist window until you see all songs' time infos. If you don't do this, you'll get only the file names or only the time infos for the songs you have played. Because, to get the time info, winamp must read/scan the file first.

Naming M3U Files

Give your M3U files unique names and put them into the same directory. This way, you can have an easy maintained archive.


HTML and XML escaping is limited to these characters: & " < > unless you have HTML::Entities installed.




Burak Gursoy <>


This software is copyright (c) 2003 by Burak Gursoy.

This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.