README

pyppd
=====

``pyppd`` is a CUPS PPD generator. It holds an compressed archive of PPDs, which
can be listed and retrieved only when needed by CUPS, saving disk space.

Installation
------------

To install ``pyppd``, you can use:

    # pip install pyppd

Or download the source package, uncompress, and run as root:

    # python3 setup.py install

It depends on Python 2.x or 3.x (http://www.python.org) and XZ Utils
(http://tukaani.org/xz/).

Usage
-----

At first, you have to create a PPD archive. For such, put all PPDs (they might
be gzipped) you want to add in the archive inside a single folder (which can
have subfolders), then run:

    $ pyppd /path/to/your/ppd/folder

It'll create ``pyppd-ppdfile`` in your current folder. This executable only
works with the same Python version that you used to generate it. You can test
it by running:

    $ ./pyppd-ppdfile list

And, for reading a PPD from the archive, simply do:

    $ ./pyppd-ppdfile cat pyppd-ppdfile:MY-PPD-FILE.PPD

For CUPS to be able to use your newly-created archive, copy ``pyppd-ppdfile``
to ``/usr/lib/cups/driver/`` and you're done.

The generated ``pyppd-ppdfile`` can be arbitrarily renamed, so that more than
one packed repository can be installed on one system. This can be useful if
you need a better performance, be it in time or memory usage. Note that also
the PPD URIs will follow the new name:

    $ ./pyppd-ppdfile list
    pyppd-ppdfile:LasterStar/LaserStar-XX100.ppd
    $ mv pyppd-ppdfile laserstar
    $ ./laserstar list
    laserstar:LaserStar/LaserStar-XX100.ppd

Contributors
------------

* **Till Kamppeter** - Original idea, mentoring and feedback. User #0.

* **Hin-Tak Leung** - Lots of technical suggestions.

* **Martin Pitt** - Python 3 port.

* **Flávio Ribeiro** and **Diógenes Fernandes** - Refactorings and general Python's best practices tips.

* **Didier Raboud** - Make archives reproducible, by sorting the list of found PPDs and using JSON dumps instead of Pickle dumps.

* **Sambhav Dusad** - Streaming decompression, to not need to hold the whole decompressed archive in memory.

* **Google's OSPO** - Initial funding at GSoC 2010.