Index
NAME
MusicRoom::CoverArt - Identify suitable cover images for tracks
DESCRIPTION
This package scans a directory full of images to identify a suitable image for a nominated music track. The software is configured by a specification file that tells it where to find images. Here is an example specification file:
# Look for the ID, then the artist and finally use a default
by_id/<id>.jpg
by_artist/<artist>.jpg
default.jpg
In the function calls the $track argument is assumed to contain a reference to a hash with values for all the attributes that are mentioned in the specification file.
If the package is called like:
use MusicRoom::CoverArt;
my $t1 = {id => "tf6ty2", artist => "Paul Simon",
song => "That Was Your Mother",
album => "Graceland"};
my $jpg1_file = MusicRoom::CoverArt::locate($t1);
The specification file above causes the software to look for the files:
/data/music/meta/art/by_id/tf6ty2.jpg
/data/music/meta/art/by_artist/Paul Simon.jpg
/data/music/meta/art/default.jpg
returning the name of the first one that is found or undefined if there are none of them.
A more reasonable specification file might look like:
by_id/<id>.jpg
by_dirartist/<dir_artist>/<artist> - <name>.jpg
by_dir/<dir_name>.jpg
by_artist/<artist>.jpg
default.jpg
This tells the software to look for:
1. The exact ID
2. A combination of the artist associated with the directory,
the real artist and the track name. This lets us specify
an image like "Various Artist/The Byrds - Mr Tambourine Man.jpg"
which is used whenever that song is found on a Various Artist
collection
3. An image that is named the same as the album directory
4. An image of the artist
5. A default image if none of the above are available
my $t1 = {id => "tf6ty2", track => "10", length => "02:51",
artist => "Paul Simon", song => "That Was Your Mother",
album => "Graceland", dir_artist => "Paul Simon",
dir_album => "Graceland", dir_name => "Paul Simon - Graceland",
size => "1371741", quality => "7", year => "1986"};
my $t2 = {id => "8nj3vp", track => "06", length => "03:16",
artist => "The Clash", song => "Tommy Gun",
album => "The Singles", dir_artist => "The Clash",
dir_album => "The Singles", dir_name => "The Clash - The Singles",
size => "1575705",quality => "7",year => "1978"};
=cut
use strict; use warnings; use Carp;
use MP3::Tag; use MusicRoom::File; use MusicRoom::Track; use MusicRoom::Locate;
use constant COVERART_LOCATOR => "coverart"; use constant PICTURE_TYPE => "Cover (front)"; use constant PICTURE_COMMENT => "Cover Image"; use constant APIC => "APIC";
my $configured;
sub init { # If we already have loaded the directory details then we can # just move on
return if(defined $configured);
MusicRoom::Locate::scan_dir(COVERART_LOCATOR);
$configured = 1;
}
sub locate { my($track) = @_;
init();
return MusicRoom::Locate::locate($track,COVERART_LOCATOR);
}
sub search_for { my($track) = @_;
init();
return MusicRoom::Locate::search_for($track,COVERART_LOCATOR);
}
sub attach { # Find a suitable image and attach it to the suggested mp3 file my($track,$mp3_file) = @_;
my $image_file = locate($track);
return undef
if(!defined $image_file);
if(!-w $mp3_file)
{
# This shouldn't happen but...
chmod 0755,$mp3_file;
}
if(!-w $mp3_file)
{
carp("Cannot write to $mp3_file");
return undef;
}
my $mp3 = MP3::Tag->new($mp3_file);
if(!defined $mp3)
{
carp("Cannot read tags from $mp3_file");
return undef;
}
# Attempt to read the tags
$mp3->get_tags();
if(!defined $mp3->{ID3v2})
{
# Need to create a new set of tags
$mp3->new_tag("ID3v2");
}
if(!defined $mp3->{ID3v2})
{
carp("Cannot create ID3v2 tags in $mp3_file");
return undef;
}
my($mime_type,$image_data) = read_image($image_file);
return undef if(!defined $mime_type);
my $encoding = 0;
my @apic_parts = ($encoding, $mime_type,
picture_type_idx(PICTURE_TYPE),
PICTURE_COMMENT, $image_data);
if(defined $mp3->{ID3v2}->get_frame(APIC))
{
# Modifying an existing image
$mp3->{ID3v2}->change_frame(APIC,@apic_parts);
}
else
{
# Create a new frame
$mp3->{ID3v2}->add_frame(APIC,@apic_parts);
}
$mp3->{ID3v2}->write_tag();
return $image_file;
}
sub read_image { # Read the image file my($file_name) = @_;
my $image_type;
my $image_data;
if(!-f $file_name)
{
error("Cannot read file \"$file_name\"");
return;
}
if($file_name =~ /\.jpg$/i)
{
$image_type = "jpg";
my $ifh = IO::File->new($file_name);
if(!defined $ifh)
{
error("Failed to open \"$file_name\"");
return;
}
binmode $ifh;
$image_data = "";
# This reads the data in 16k chunks, but the images should be small anyway
while(!$ifh->eof())
{
my $c = $ifh->read($image_data,1024*16,length($image_data));
}
$ifh = undef;
}
if(!defined $image_type)
{
error("Does not yet support file type for \"$file_name\"");
return;
}
if(!defined $image_data)
{
error("Cannot extract $image_type data from \"$file_name\"");
return;
}
return("image/$image_type",$image_data);
}
sub picture_type_idx { # Given a picture type string convert it into a number suitable # for MP3::Tag my($picture_type) = @_;
# The picture types that are currently understood (from MP3::Tag::ID3v2):
my @picture_types =
("Other", "32x32 pixels 'file icon' (PNG only)", "Other file icon",
"Cover (front)", "Cover (back)", "Leaflet page",
"Media (e.g. lable side of CD)", "Lead artist/lead performer/soloist"
,
"Artist/performer", "Conductor", "Band/Orchestra", "Composer",
"Lyricist/text writer", "Recording Location", "During recording",
"During performance", "Movie/video screen capture",
"A bright coloured fish", "Illustration", "Band/artist logotype",
"Publisher/Studio logotype");
# This approach is easy to understand
for(my $i=0;$i<=$#picture_types;$i++)
{
if(lc($picture_type) eq lc($picture_types[$i]))
{
return chr($i);
}
}
error("The picture type \"$picture_type\" is not valid");
return chr(3);
}
1;