NAME

App::MathImage::Image::Base::Imlib2 -- draw images using Imlib2

SYNOPSIS

use App::MathImage::Image::Base::Imlib2;
my $image = App::MathImage::Image::Base::Imlib2->new (-width => 100,
                                      -height => 100);
$image->rectangle (0,0, 99,99, 'white');
$image->xy (20,20, 'black');
$image->line (50,50, 70,70, '#FF00FF');
$image->line (50,50, 70,70, '#0000AAAA9999');
$image->save ('/some/filename.png');

CLASS HIERARCHY

App::MathImage::Image::Base::Imlib2 is a subclass of Image::Base,

Image::Base
  App::MathImage::Image::Base::Imlib2

DESCRIPTION

App::MathImage::Image::Base::Imlib2 extends Image::Base to create or update image files using the Image::Imlib2 module.

As of Imlib2 1.4.4 the following file formats are available, provided for some of them that Imlib2 was built with supporting libraries (such as libpng).

PNG, JPEG, TIFF, PNM, BMP
GIF    read-only
XPM    read-only
TGA    Targa
LBM    Amiga Paint, read-only
ARGB   raw format from Imlib (?)

Colour Names

There's no named colours as such, only hex and a special "None" for transparent

#RGB
#RRGGBB
#RRRGGGBBB
#RRRRGGGGBBBB
None             transparent

When loading XPM files /usr/share/X11/rgb.txt or similar is consulted for named colours in the file, but that's not made available to later drawing operations.

FUNCTIONS

$image = App::MathImage::Image::Base::Imlib2->new (key=>value,...)

Create and return a new image object. A new image can be started with -width and -height,

$image = App::MathImage::Image::Base::Imlib2->new (-width => 200, -height => 100);

Or an existing file can be read,

$image = App::MathImage::Image::Base::Imlib2->new (-file => '/some/filename.png');

Or an Image::Imlib2 object can be given,

my $imlibobj = Image::Imlib2->new (20, 10);
$image = App::MathImage::Image::Base::Imlib2->new (-imlib => $imlibobj);

$new_image = $image->new (key=>value,...)

Create and return a copy of $image. The Imlib object within $image is cloned per $imlib->clone(). Optional key/value parameters are applied to the new image as per set().

# copy image, new compression level
my $new_image = $image->new (-zlib_compression => 9);

$image->load

$image->load ($filename)

Read the -file, or set -file to $filename and then read.

This creates and sets a new underlying -imlib because it's not possible to read into an existing Imlib2 object, only create a new one from the load.

$image->save ()

$image->save ($filename)

Save to -file, or with a $filename argument set -file then save to that.

The file format is taken from the -file_format (see below) if that was set by a load or explicit set, otherwise Imlib2 follows the filename extension. In both cases if format or extension is unrecognised then save() croaks.

ATTRIBUTES

-width (integer, create-only)

-height (integer, create-only)

The size of the underlying Image::Imlib2 is set when created and can't be changed after that. (Is that right?)

-imlib

The underlying Image::Imlib2 object.

-file_format (string or undef, write-only)

The file format as a string like "png" or "jpeg".

After load() the -file_format is the format read. Setting -file_format can change the format for a subsequent save().

This is applied with $imlib->set_file_format() and is currently write-only.

-quality_percent (0 to 100, write-only)

The image quality for saving to JPEG format. JPEG compresses by reducing colours and resolution in ways that are not too noticeable to the human eye. 100 means full quality, no such reductions.

BUGS

Imlib2 interprets filenames with a colon like "name:foo.bar" as meaning a sub-part of certain file formats. If the full "name:foo.bar" exists and is a ordinary file (not a char special) then it loads that, otherwise it tries "name" alone. This is unlike other Image::Base modules and the intention would be to avoid it in load() and save() in the interests of consistency, if possible.

In the current implementation ellipse() is done with the Image::Base code, since in Imlib2 1.4.4 the native ellipse drawing seems a bit dubious, going outside the requested size. In any case they would only be applicable for odd sides, meaning $x2-$x1+1 and $y2-$y1+1 both odd numbers, since they take an x,y centre and a,b semi-radii must be integers.

SEE ALSO

Image::Base, Image::Imlib2

file://usr/share/doc/libimlib2-dev/html/index.html