pyudev – pure Python libudev binding

pyudev 0.21.0 (Changelog, installation)

pyudev is a LGPL licenced, pure Python 2/3 binding to libudev, the device and hardware management and information library of Linux.

Almost the complete libudev functionality is exposed. You can:

  • Enumerate devices, filtered by specific criteria (pyudev.Context)
  • Query device information, properties and attributes,
  • Monitor devices, both synchronously and asynchronously with background threads, or within the event loops of Qt (pyudev.pyqt4, pyudev.pyside), glib (pyudev.glib) and wxPython (pyudev.wx).

Documentation

Thanks to the power of libudev, usage of pyudev is very simple. Getting the labels of all partitions just takes a few lines:

>>> import pyudev
>>> context = pyudev.Context()
>>> for device in context.list_devices(subsystem='block', DEVTYPE='partition'):
...     print(device.get('ID_FS_LABEL', 'unlabeled partition'))
...
boot
swap
system

A user guide gives an introduction into common operations and concepts of pyudev, the API documentation provides a detailed reference:

Installation

Python versions and implementations

pyudev supports CPython from 2.6 up to the latest Python 3 release, and PyPy 1.5. Jython may work, too, but is not tested. Generally any Python implementation compatible with CPython 2.6 should work.

Dependencies

pyudev needs libudev 151 and newer, earlier versions of libudev as found on dated Linux systems may work, but are not tested and not officially supported. It is written in pure Python based on ctypes, so no compilers or headers are required for installation.

To use any of the toolkit integration modules. the corresponding toolkit must be available, but no toolkit is required during installation.

Installation from Cheeseshop

Install pyudev from the Cheeseshop with pip:

pip install pyudev

Installation from source code

Close the public repository:

git clone https://github.com/lunaryorn/pyudev.git

Or download tarball:

curl -OL https://github.com/lunaryorn/pyudev/tarball/master

Then install pyudev from the source code tree:

python setup.py install

User guide

This guide gives an introduction in how to use pyudev for common operations like device enumeration or monitoring:

A detailed reference is provided in the API documentation.

Getting started

Import pyudev and verify that you’re using the latest version:

>>> import pyudev
>>> pyudev.__version__
u'0.16'
>>> pyudev.udev_version()
181

This prints the version of pyudev itself and of the underlying libudev.

A note on versioning

pyudev supports libudev 151 or newer, but still tries to cover the most recent libudev API completely. If you are using older libudev releases, some functionality of pyudev may be unavailable, simply because libudev is too old to support a specific feature. Whenever this is the case, the minimum required version of udev is noted in the documentation (see Device.is_initialized for an example). If no version is specified for an attribute or a method, it is available on all supported libudev versions. You can check the version of the underlying libudev with pyudev.udev_version().

Enumerating devices

A common use case is to enumerate available devices, or a subset thereof. But before you can do anything with pyudev, you need to establish a “connection” to the udev device database first. This connection is represented by a library Context:

>>> context = pyudev.Context()

The Context is the central object of pyudev and libudev. You will need a Context object for almost anything in pyudev. With the context you can now enumerate the available devices:

>>> for device in context.list_devices(): 
...     device
...
Device(u'/sys/devices/LNXSYSTM:00')
Device(u'/sys/devices/LNXSYSTM:00/LNXCPU:00')
Device(u'/sys/devices/LNXSYSTM:00/LNXCPU:01')
...

By default, list_devices() yields all devices available on the system as Device objects, but you can filter the list of devices with keyword arguments to enumerate all available partitions for example:

>>> for device in context.list_devices(subsystem='block', DEVTYPE='partition'):
...    print(device)
...
Device(u'/sys/devices/pci0000:00/0000:00:0d.0/host2/target2:0:0/2:0:0:0/block/sda/sda1')
Device(u'/sys/devices/pci0000:00/0000:00:0d.0/host2/target2:0:0/2:0:0:0/block/sda/sda2')
Device(u'/sys/devices/pci0000:00/0000:00:0d.0/host2/target2:0:0/2:0:0:0/block/sda/sda3')

The choice of the right filters depends on the use case and generally requires some knowledge about how udev classifies and categorizes devices. This is out of the scope of this guide. Poke around in /sys/ to get a feeling for the udev-way of device handling, read the udev documentation or one of the tutorials in the net.

The keyword arguments of list_devices() provide the most common filter operations. You can apply other, less common filters by calling one of the match_* methods on the Enumerator returned by of list_devices().

Accessing individual devices directly

If you just need a single specific Device, you don’t need to enumerate all devices with a specific filter criterion. Instead, you can directly create Device objects from a device path (Devices.from_path()), by from a subsystem and device name (Devices.from_name()) or from a device file (Devices.from_device_file()). The following code gets the Device object for the first hard disc in three different ways:

>>> pyudev.Devices.from_path(context, '/sys/block/sda')
Device(u'/sys/devices/pci0000:00/0000:00:0d.0/host2/target2:0:0/2:0:0:0/block/sda')
>>> pyudev.Devices.from_name(context, 'block', 'sda')
Device(u'/sys/devices/pci0000:00/0000:00:0d.0/host2/target2:0:0/2:0:0:0/block/sda')
>>> pyudev.Devices.from_device_file(context, '/dev/sda')
Device(u'/sys/devices/pci0000:00/0000:00:0d.0/host2/target2:0:0/2:0:0:0/block/sda')

As you can see, you need to pass a Context to both methods as reference to the udev database from which to retrieve information about the device.

Note

The Device objects created in the above example refer to the same device. Consequently, they are considered equal:

>>> pyudev.Devices.from_path(context, '/sys/block/sda') == pyudev.Devices.from_name(context, 'block', 'sda')
True

Whereas Device objects referring to different devices are unequal:

>>> pyudev.Devices.from_name(context, 'block', 'sda') == pyudev.Devices.from_name(context, 'block', 'sda1')
False

Querying device information

As you’ve seen, Device represents a device in the udev database. Each such device has a set of “device properties” (not to be confused with Python properties as created by property()!) that describe the capabilities and features of this device as well as its relationship to other devices.

Common device properties are also available as properties of a Device object. For instance, you can directly query the device_node and the device_type of block devices:

>>> for device in context.list_devices(subsystem='block'):
...     print('{0} ({1})'.format(device.device_node, device.device_type))
...
/dev/sr0 (disk)
/dev/sda (disk)
/dev/sda1 (partition)
/dev/sda2 (partition)
/dev/sda3 (partition)

For all other properties, Device provides a dictionary-like interface to directly access the device properties. You’ll get the same information as with the generic properties:

>>> for device in context.list_devices(subsystem='block'):
...    print('{0} ({1})'.format(device['DEVNAME'], device['DEVTYPE']))
...
/dev/sr0 (disk)
/dev/sda (disk)
/dev/sda1 (partition)
/dev/sda2 (partition)
/dev/sda3 (partition)

Warning

When filtering devices, you have to use the device property names. The names of corresponding properties of Device will generally not work. Compare the following two statements:

>>> [device.device_node for device in context.list_devices(subsystem='block', DEVTYPE='partition')]
[u'/dev/sda1', u'/dev/sda2', u'/dev/sda3']
>>> [device.device_node for device in context.list_devices(subsystem='block', device_type='partition')]
[]

But you can also query many device properties that are not available as Python properties on the Device object with a convenient mapping interface, like the filesystem type. Device provides a convenient mapping interface for this purpose:

>>> for device in context.list_devices(subsystem='block', DEVTYPE='partition'):
...     print('{0} ({1})'.format(device.device_node, device.get('ID_FS_TYPE')))
...
/dev/sda1 (ext3)
/dev/sda2 (swap)
/dev/sda3 (ext4)

Note

Such device specific properties may not be available on devices. Either use get() to specify default values for missing properties, or be prepared to catch KeyError.

Most device properties are computed by udev rules from the driver- and device-specific “device attributes”. The Device.attributes mapping gives you access to these attributes, but generally you should not need these. Use the device properties whenever possible.

Examing the device hierarchy

A Device is part of a device hierarchy, and can have a parent device that more or less resembles the physical relationship between devices. For instance, the parent of partition devices is a Device object that represents the disc the partition is located on:

>>> for device in context.list_devices(subsystem='block', DEVTYPE='partition'):
...    print('{0} is located on {1}'.format(device.device_node, device.parent.device_node))
...
/dev/sda1 is located on /dev/sda
/dev/sda2 is located on /dev/sda
/dev/sda3 is located on /dev/sda

Generally, you should not rely on the direct parent-child relationship between two devices. Instead of accessing the parent directly, search for a parent within a specific subsystem, e.g. for the parent block device, with find_parent():

>>> for device in context.list_devices(subsystem='block', DEVTYPE='partition'):
...    print('{0} is located on {1}'.format(device.device_node, device.find_parent('block').device_node))
...
/dev/sda1 is located on /dev/sda
/dev/sda2 is located on /dev/sda
/dev/sda3 is located on /dev/sda

This also save you the tedious work of traversing the device tree manually, if you are interested in grand parents, like the name of the PCI slot of the SCSI or IDE controller of the disc that contains a partition:

>>> for device in context.list_devices(subsystem='block', DEVTYPE='partition'):
...    print('{0} attached to PCI slot {1}'.format(device.device_node, device.find_parent('pci')['PCI_SLOT_NAME']))
...
/dev/sda1 attached to PCI slot 0000:00:0d.0
/dev/sda2 attached to PCI slot 0000:00:0d.0
/dev/sda3 attached to PCI slot 0000:00:0d.0

Monitoring devices

Synchronous monitoring

The Linux kernel emits events whenever devices are added, removed (e.g. a USB stick was plugged or unplugged) or have their attributes changed (e.g. the charge level of the battery changed). With pyudev.Monitor you can react on such events, for example to react on added or removed mountable filesystems:

>>> monitor = pyudev.Monitor.from_netlink(context)
>>> monitor.filter_by('block')
>>> for device in iter(monitor.poll, None):
...     if 'ID_FS_TYPE' in device:
...         print('{0} partition {1}'.format(device.action, device.get('ID_FS_LABEL')))
...
add partition MULTIBOOT
remove partition MULTIBOOT

After construction of a monitor, you can install an event filter on the monitor using filter_by(). In the above example only events from the block subsystem are handled.

Note

Always prefer filter_by() and filter_by_tag() over manually filtering devices (e.g. by device.subsystem == 'block' or tag in device.tags). These methods install the filter on the kernel side. A process waiting for events is thus only woken up for events that match these filters. This is much nicer in terms of power consumption and system load than executing filters in the process itself.

Eventually, you can receive events from the monitor. As you can see, a Monitor is iterable and synchronously yields occurred events. If you iterate over a Monitor, you will synchronously receive events in an endless loop, until you raise an exception, or break the loop.

This is the quick and dirty way of monitoring, suitable for small scripts or quick experiments. In most cases however, simply iterating over the monitor is not sufficient, because it blocks the main thread, and can only be stopped if an event occurs (otherwise the loop is not entered and you have no chance to break it).

Asynchronous monitoring

For such use cases, pyudev provides asynchronous monitoring with MonitorObserver. You can use it to log added and removed mountable filesystems to a file, for example:

>>> monitor = pyudev.Monitor.from_netlink(context)
>>> monitor.filter_by('block')
>>> def log_event(action, device):
...    if 'ID_FS_TYPE' in device:
...        with open('filesystems.log', 'a+') as stream:
...            print('{0} - {1}'.format(action, device.get('ID_FS_LABEL')), file=stream)
...
>>> observer = pyudev.MonitorObserver(monitor, log_event)
>>> observer.start()

The observer gets an event handler (log_event() in this case) which is asynchronously invoked on every event emitted by the underlying monitor after the observer has been started using start().

Warning

The callback is invoked from a different thread than the one in which the observer was created. Be sure to protect access to shared resource properly when you access them from the callback (e.g. by locking).

The observer can be stopped at any moment using stop()`():

>>> observer.stop()

Warning

Do not call stop() from the event handler, neither directly nor indirectly. Use send_stop() if you need to stop monitoring from inside the event handler.

GUI toolkit integration

If you’re using a GUI toolkit, you already have the event system of the GUI toolkit at hand. pyudev provides observer classes that seamlessly integration in the event system of the GUI toolkit and relieve you from caring with synchronisation issues that would occur with thread-based monitoring as implemented by MonitorObserver.

pyudev supports all major GUI toolkits available for Python:

Each of these modules provides an observer class that observers the monitor asynchronously and emits proper signals upon device events.

For instance, the above example would look like this in a PySide application:

>>> from pyudev.pyside import QUDevMonitorObserver
>>> monitor = pyudev.Monitor.from_netlink(context)
>>> observer = QUDevMonitorObserver(monitor)
>>> observer.deviceEvent.connect(log_event)
>>> monitor.start()
Device objects as booleans

The use of a Device object in a boolean context as a shorthand for a comparison with None is an error.

The Device class inherits from the abstract Mapping class, as it maps udev property names to their values. Consequently, if a Device object has no udev properties, an unusual but not impossible occurance, the object is interpreted as False in a boolean context.

API documentation

This document provides API reference documentation for pyudev. Refer to the User guide for an introduction into pyudev.

pyudev The Context provides the connection to the udev device database and enumerates devices.
pyudev.pyqt4
pyudev.pyqt5
pyudev.pyside
pyudev.glib MonitorObserver integrates device monitoring into the Glib
pyudev.wx MonitorObserver integrates device monitoring into the wxPython_

Support

Please report issues, bugs and questions to the issue tracker, but respect the following guidelines:

  • Check that the issue has not already been reported.
  • Check that the issue is not already fixed in the master branch.
  • Open issues with clear title and a detailed description in grammatically correct, complete sentences.
  • Include the Python version and the udev version (see udevadm --version) in the description of your issue.

Development

The source code is hosted on GitHub:

git clone https://github.com/lunaryorn/pyudev.git

If you want to contribute to pyudev, please read the guidelines for contributions and the testsuite documentation.

Contribute

Please fork the repository, and send pull requests with new features or bug fixes, but respect the following guidelines:

Complying to these guidelines greatly increase the change of getting your pull request merged. You will be asked to improve your changeset if your pull request breaks any of the above guidelines.

If you intend to make larger changes, especially if these changes break the ABI, please ask on the mailing list first.

Testsuite documentation

This document explains the pyudev test suite and how to add new tests to this suite.

The pyudev testsuite uses the powerful pytest unittest framework, accompied by the nice mock library for mocking native functions and heavily extended with plugins to support the tests.

Test running

Direct testing using tox_

If you are on a Linux system run all tests with tox_. This tool automatically creates virtualenvs (see virtualenv_), installs all packages required by the test suite, and runs the tests.

Run all pyudev tests against Python 2.7, Python 3.2 and PyPy:

tox -e py27,py32,pypy

Pass any arguments you want to py.test after two dashes --:

tox -e py27,py32,pypy -- --enable-privileged
Notes
Device samples

Many pyudev tests run against the real device database of the system the tests are executed on. As testing against the whole database takes a long time, tests are run against a random sample by default. With the command line options provided by udev_database you can configure the size of this sample, or run the tests against a single device or the whole database.

Privileged tests

Some tests need to execute privileged operations like loading or unloading of kernel modules to trigger real udev events. These tests are disabled by default. Refer to privileged for more information on how to enable these tests and configure them properly.

plugins – Testsuite plugins

Plugins to support the pyudev testsuite.

The following plugins are provided and enabled:

privileged – Privileged operations

Support privileged operations to trigger real udev events.

This plugin adds load_dummy() and unload_dummy() to the pytest namespace.

Command line options

The plugin adds the following command line options to py.test:

--enable-privileged

Enable privileged tests. You’ll need to have sudo configured correctly in order to run tests with this option.

Configuration

In order to execute these tests without failure, you need to configure sudo to allow the user that executes the test to run the following commands:

  • modprobe dummy
  • modprobe -r dummy

To do so, create a file /etc/sudoers.d/20pyudev-tests with the following content:

me ALL = (root) NOPASSWD: /sbin/modprobe dummy, /sbin/modprobe -r dummy

Replace me with your actual user name. NOPASSWD: tells sudo not to ask for a password when executing these commands. This is simply for the sake of convenience and to allow unattended test execution. Remove this word if you want to be asked for a password.

Make sure to change the owner and group to root:root and the permissions of this file to 440 afterwards, other sudo will refuse to load the file. Also check the file with visudo to prevent syntactic errors:

$ chown root:root /etc/sudoers.d/20pyudev-tests
$ chmod 440 /etc/sudoers.d/20pyudev-tests
$ visudo -c -f /etc/sudoers.d/20pyudev-tests
pytest namespace

The plugin adds the following functions to the pytest namespace:

plugins.privileged.load_dummy()

Load the dummy module.

If privileged tests are disabled, the current test is skipped.

plugins.privileged.unload_dummy()

Unload the dummy module.

If privileged tests are disabled, the current test is skipped.

fake_monitor – A fake Monitor

Provide a fake Monitor.

This fake monitor allows to trigger arbitrary events. Use this class to test class building upon monitor without the need to rely on real events generated by privileged operations as provided by the privileged plugin.

class plugins.fake_monitor.FakeMonitor(device_to_emit)

A fake Monitor which allows you to trigger arbitrary events.

This fake monitor implements the complete Monitor interface and works on real file descriptors so that you can select() the monitor.

close()

Close sockets acquired by this monitor.

trigger_event()

Trigger an event on clients of this monitor.

Funcargs

The plugin provides the following funcargs:

plugins.fake_monitor.fake_monitor(request)

Return a FakeMonitor, which emits the platform device as returned by the fake_monitor_device funcarg on all triggered actions.

Warning

To use this funcarg, you have to provide the fake_monitor_device funcarg!

mock_libudev – Mock calls to libudev

Plugin to mock calls to libudev.

This plugin adds libudev_list() to the pytest namespace.

plugins.mock_libudev.libudev_list(function, items)

Mock a libudev linked list:

with pytest.libudev_list(device._libudev, 'udev_device_get_tag_list_entry', ['foo', 'bar']):
    assert list(device.tags) == ['foo', 'bar']

function is a string containing the name of the libudev function that returns the list. items is an iterable yielding items which shall be returned by the mocked list function. An item in items can either be a tuple with two components, where the first component is the item name, and the second the item value, or a single element, which is the item name. The item value is None in this case.

travis – Support for Travis CI

Support for Travis CI.

Test markers
pytest.mark.not_on_travis

Do not run the decorated test on Travis CI:

@pytest.mark.not_on_travis
def test_foo():
    assert True

test_foo will not be run on Travis CI.

pytest namespace

The plugin adds the following functions to the pytest namespace:

plugins.travis.is_on_travis_ci()

Determine whether the tests run on Travis CI.

Return True, if so, or False otherwise.

Endorsements

If you’re using pyudev and want to say something about it please add yourself to the endorsements page.

pyudev Users

If you are using pyudev and would like the world to know how and why, here is the place. Just submit a PR with an addition to the documentation, something like:

Choice of information about yourself.

What you are doing with pyudev and why it beats the alternatives.

Other reading

Changelog

0.21.0 (July 20, 2016)

  • Deprecate use of Device object as mapping from udev property names to values.
  • Add a Properties class and a Device.properties() method for udev properties.
  • Fix places where Device object was incorrectly used in a boolean context.
  • Return an empty string, not None, if the property value is an empty string.
  • Exceptions re-raised from libudev calls now have a message string.
  • Insert a warning about using a Device in a boolean context in docs.
  • Infrastructure for vagrant tests is removed.
  • Various internal refactorings.
  • Extensive test improvements.
  • Numerous documentation fixes.

0.20.0 (April 29, 2016)

  • Remove parsing code added in previous release.
  • No longer do CI for Python 2.6.
  • Eliminate all wildcard imports and __all__ statements.
  • No longer use deprecated Device.from_sys_path() method.
  • Minor pylint induced changes.
  • Documentation fixes.

0.19.0 (Feb 3, 2016)

  • Restore raising KeyError by Attributes.as* methods when attribute not found.
  • Explicitly require six module.
  • Never raise a DeviceNotFoundError when iterating over a device enumeration.
  • Device.subsystem() now returns None if device has no subsystem.
  • Add DeprecationWarnings to deprecated Device methods.
  • Replace “/” with ”!” in Device.from_name() sys_name parameter.
  • Add some unstable classes for parsing some kinds of values.
  • Make version info more like Python’s including micro numbers and levels.
  • Refactor some internal modules into subdirectories.
  • Work on tests and reproducers.

0.18 (Dec 1, 2015)

  • DeviceNotFoundError is no longer a subtype of LookupError
  • Added support for pyqt5 monitor observer
  • Added discover module, which looks up a device on limited information
  • Attributes class no longer extends Mapping, extends object instead
  • Attributes class no longer inherits [] operator, Mapping methods
  • Attributes class objects are no longer iterable
  • Attributes.available_attributes property added
  • Attributes.get() method, with usual semantics, defined
  • Device.from_* methods are deprecated, uses Devices.from_* methods instead
  • Device.from_device_file() now raises DeviceNotFoundByFileError
  • Device.from_device_number() now raises DeviceNotFoundByNumberError
  • Devices.from_interface_index() method added
  • Devices.from_kernel_device() method added
  • Numerous testing infrastructure changes

0.17 (Aug 26, 2015)

0.16.1 (Aug 02, 2012)

  • #53: Fix source distribution
  • #54: Fix typo in test

0.16 (Jul 25, 2012)

0.15 (Mar 1, 2012)

0.14 (Feb 10, 2012)

0.13 (Nov 4, 2011)

0.12 (Aug 31, 2011)

  • #32: Fix memory leak.
  • #33: Fix Python 3 support for pyudev.glib.
  • Fix license header in pyudev._compat.

0.11 (Jun 26, 2011)

0.10 (Apr 20, 2011)

0.9 (Mar 09, 2011)

0.8 (Jan 08, 2011)

0.7 (Nov 15, 2010)

0.6 (Oct 03, 2010)

0.5 (Sep 06, 2010)

0.4 (Aug 23, 2010)

API changes

0.3 (Jul 28, 2010)

  • #1: Fix documentation to reflect the actual behaviour of the underlying API
  • Raise TypeError if udev.Device are compared with >, >=, < or <=.
  • Add udev.Enumerator.match_children().
  • Add udev.Device.children.
  • Add qudev.QUDevMonitorObserver.deviceChanged().
  • Add qudev.QUDevMonitorObserver.deviceMoved().

0.2 (Jun 28, 2010)

  • Add udev.Monitor.
  • Add udev.Device.asbool().
  • Add udev.Device.asint().
  • Remove type magic in udev.Device.__getitem__().
  • Add qudev.

0.1 (May 03, 2010)

  • Initial release.
  • Add udev.Context.
  • Add udev.Device.
  • Add udev.Enumerator.

Licencing

		  GNU LESSER GENERAL PUBLIC LICENSE
		       Version 2.1, February 1999

 Copyright (C) 1991, 1999 Free Software Foundation, Inc.
     51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 Everyone is permitted to copy and distribute verbatim copies
 of this license document, but changing it is not allowed.

[This is the first released version of the Lesser GPL.  It also counts
 as the successor of the GNU Library Public License, version 2, hence
 the version number 2.1.]

			    Preamble

  The licenses for most software are designed to take away your
freedom to share and change it.  By contrast, the GNU General Public
Licenses are intended to guarantee your freedom to share and change
free software--to make sure the software is free for all its users.

  This license, the Lesser General Public License, applies to some
specially designated software packages--typically libraries--of the
Free Software Foundation and other authors who decide to use it.  You
can use it too, but we suggest you first think carefully about whether
this license or the ordinary General Public License is the better
strategy to use in any particular case, based on the explanations below.

  When we speak of free software, we are referring to freedom of use,
not price.  Our General Public Licenses are designed to make sure that
you have the freedom to distribute copies of free software (and charge
for this service if you wish); that you receive source code or can get
it if you want it; that you can change the software and use pieces of
it in new free programs; and that you are informed that you can do
these things.

  To protect your rights, we need to make restrictions that forbid
distributors to deny you these rights or to ask you to surrender these
rights.  These restrictions translate to certain responsibilities for
you if you distribute copies of the library or if you modify it.

  For example, if you distribute copies of the library, whether gratis
or for a fee, you must give the recipients all the rights that we gave
you.  You must make sure that they, too, receive or can get the source
code.  If you link other code with the library, you must provide
complete object files to the recipients, so that they can relink them
with the library after making changes to the library and recompiling
it.  And you must show them these terms so they know their rights.

  We protect your rights with a two-step method: (1) we copyright the
library, and (2) we offer you this license, which gives you legal
permission to copy, distribute and/or modify the library.

  To protect each distributor, we want to make it very clear that
there is no warranty for the free library.  Also, if the library is
modified by someone else and passed on, the recipients should know
that what they have is not the original version, so that the original
author's reputation will not be affected by problems that might be
introduced by others.

  Finally, software patents pose a constant threat to the existence of
any free program.  We wish to make sure that a company cannot
effectively restrict the users of a free program by obtaining a
restrictive license from a patent holder.  Therefore, we insist that
any patent license obtained for a version of the library must be
consistent with the full freedom of use specified in this license.

  Most GNU software, including some libraries, is covered by the
ordinary GNU General Public License.  This license, the GNU Lesser
General Public License, applies to certain designated libraries, and
is quite different from the ordinary General Public License.  We use
this license for certain libraries in order to permit linking those
libraries into non-free programs.

  When a program is linked with a library, whether statically or using
a shared library, the combination of the two is legally speaking a
combined work, a derivative of the original library.  The ordinary
General Public License therefore permits such linking only if the
entire combination fits its criteria of freedom.  The Lesser General
Public License permits more lax criteria for linking other code with
the library.

  We call this license the "Lesser" General Public License because it
does Less to protect the user's freedom than the ordinary General
Public License.  It also provides other free software developers Less
of an advantage over competing non-free programs.  These disadvantages
are the reason we use the ordinary General Public License for many
libraries.  However, the Lesser license provides advantages in certain
special circumstances.

  For example, on rare occasions, there may be a special need to
encourage the widest possible use of a certain library, so that it becomes
a de-facto standard.  To achieve this, non-free programs must be
allowed to use the library.  A more frequent case is that a free
library does the same job as widely used non-free libraries.  In this
case, there is little to gain by limiting the free library to free
software only, so we use the Lesser General Public License.

  In other cases, permission to use a particular library in non-free
programs enables a greater number of people to use a large body of
free software.  For example, permission to use the GNU C Library in
non-free programs enables many more people to use the whole GNU
operating system, as well as its variant, the GNU/Linux operating
system.

  Although the Lesser General Public License is Less protective of the
users' freedom, it does ensure that the user of a program that is
linked with the Library has the freedom and the wherewithal to run
that program using a modified version of the Library.

  The precise terms and conditions for copying, distribution and
modification follow.  Pay close attention to the difference between a
"work based on the library" and a "work that uses the library".  The
former contains code derived from the library, whereas the latter must
be combined with the library in order to run.

		  GNU LESSER GENERAL PUBLIC LICENSE
   TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION

  0. This License Agreement applies to any software library or other
program which contains a notice placed by the copyright holder or
other authorized party saying it may be distributed under the terms of
this Lesser General Public License (also called "this License").
Each licensee is addressed as "you".

  A "library" means a collection of software functions and/or data
prepared so as to be conveniently linked with application programs
(which use some of those functions and data) to form executables.

  The "Library", below, refers to any such software library or work
which has been distributed under these terms.  A "work based on the
Library" means either the Library or any derivative work under
copyright law: that is to say, a work containing the Library or a
portion of it, either verbatim or with modifications and/or translated
straightforwardly into another language.  (Hereinafter, translation is
included without limitation in the term "modification".)

  "Source code" for a work means the preferred form of the work for
making modifications to it.  For a library, complete source code means
all the source code for all modules it contains, plus any associated
interface definition files, plus the scripts used to control compilation
and installation of the library.

  Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope.  The act of
running a program using the Library is not restricted, and output from
such a program is covered only if its contents constitute a work based
on the Library (independent of the use of the Library in a tool for
writing it).  Whether that is true depends on what the Library does
and what the program that uses the Library does.
  
  1. You may copy and distribute verbatim copies of the Library's
complete source code as you receive it, in any medium, provided that
you conspicuously and appropriately publish on each copy an
appropriate copyright notice and disclaimer of warranty; keep intact
all the notices that refer to this License and to the absence of any
warranty; and distribute a copy of this License along with the
Library.

  You may charge a fee for the physical act of transferring a copy,
and you may at your option offer warranty protection in exchange for a
fee.

  2. You may modify your copy or copies of the Library or any portion
of it, thus forming a work based on the Library, and copy and
distribute such modifications or work under the terms of Section 1
above, provided that you also meet all of these conditions:

    a) The modified work must itself be a software library.

    b) You must cause the files modified to carry prominent notices
    stating that you changed the files and the date of any change.

    c) You must cause the whole of the work to be licensed at no
    charge to all third parties under the terms of this License.

    d) If a facility in the modified Library refers to a function or a
    table of data to be supplied by an application program that uses
    the facility, other than as an argument passed when the facility
    is invoked, then you must make a good faith effort to ensure that,
    in the event an application does not supply such function or
    table, the facility still operates, and performs whatever part of
    its purpose remains meaningful.

    (For example, a function in a library to compute square roots has
    a purpose that is entirely well-defined independent of the
    application.  Therefore, Subsection 2d requires that any
    application-supplied function or table used by this function must
    be optional: if the application does not supply it, the square
    root function must still compute square roots.)

These requirements apply to the modified work as a whole.  If
identifiable sections of that work are not derived from the Library,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works.  But when you
distribute the same sections as part of a whole which is a work based
on the Library, the distribution of the whole must be on the terms of
this License, whose permissions for other licensees extend to the
entire whole, and thus to each and every part regardless of who wrote
it.

Thus, it is not the intent of this section to claim rights or contest
your rights to work written entirely by you; rather, the intent is to
exercise the right to control the distribution of derivative or
collective works based on the Library.

In addition, mere aggregation of another work not based on the Library
with the Library (or with a work based on the Library) on a volume of
a storage or distribution medium does not bring the other work under
the scope of this License.

  3. You may opt to apply the terms of the ordinary GNU General Public
License instead of this License to a given copy of the Library.  To do
this, you must alter all the notices that refer to this License, so
that they refer to the ordinary GNU General Public License, version 2,
instead of to this License.  (If a newer version than version 2 of the
ordinary GNU General Public License has appeared, then you can specify
that version instead if you wish.)  Do not make any other change in
these notices.

  Once this change is made in a given copy, it is irreversible for
that copy, so the ordinary GNU General Public License applies to all
subsequent copies and derivative works made from that copy.

  This option is useful when you wish to copy part of the code of
the Library into a program that is not a library.

  4. You may copy and distribute the Library (or a portion or
derivative of it, under Section 2) in object code or executable form
under the terms of Sections 1 and 2 above provided that you accompany
it with the complete corresponding machine-readable source code, which
must be distributed under the terms of Sections 1 and 2 above on a
medium customarily used for software interchange.

  If distribution of object code is made by offering access to copy
from a designated place, then offering equivalent access to copy the
source code from the same place satisfies the requirement to
distribute the source code, even though third parties are not
compelled to copy the source along with the object code.

  5. A program that contains no derivative of any portion of the
Library, but is designed to work with the Library by being compiled or
linked with it, is called a "work that uses the Library".  Such a
work, in isolation, is not a derivative work of the Library, and
therefore falls outside the scope of this License.

  However, linking a "work that uses the Library" with the Library
creates an executable that is a derivative of the Library (because it
contains portions of the Library), rather than a "work that uses the
library".  The executable is therefore covered by this License.
Section 6 states terms for distribution of such executables.

  When a "work that uses the Library" uses material from a header file
that is part of the Library, the object code for the work may be a
derivative work of the Library even though the source code is not.
Whether this is true is especially significant if the work can be
linked without the Library, or if the work is itself a library.  The
threshold for this to be true is not precisely defined by law.

  If such an object file uses only numerical parameters, data
structure layouts and accessors, and small macros and small inline
functions (ten lines or less in length), then the use of the object
file is unrestricted, regardless of whether it is legally a derivative
work.  (Executables containing this object code plus portions of the
Library will still fall under Section 6.)

  Otherwise, if the work is a derivative of the Library, you may
distribute the object code for the work under the terms of Section 6.
Any executables containing that work also fall under Section 6,
whether or not they are linked directly with the Library itself.

  6. As an exception to the Sections above, you may also combine or
link a "work that uses the Library" with the Library to produce a
work containing portions of the Library, and distribute that work
under terms of your choice, provided that the terms permit
modification of the work for the customer's own use and reverse
engineering for debugging such modifications.

  You must give prominent notice with each copy of the work that the
Library is used in it and that the Library and its use are covered by
this License.  You must supply a copy of this License.  If the work
during execution displays copyright notices, you must include the
copyright notice for the Library among them, as well as a reference
directing the user to the copy of this License.  Also, you must do one
of these things:

    a) Accompany the work with the complete corresponding
    machine-readable source code for the Library including whatever
    changes were used in the work (which must be distributed under
    Sections 1 and 2 above); and, if the work is an executable linked
    with the Library, with the complete machine-readable "work that
    uses the Library", as object code and/or source code, so that the
    user can modify the Library and then relink to produce a modified
    executable containing the modified Library.  (It is understood
    that the user who changes the contents of definitions files in the
    Library will not necessarily be able to recompile the application
    to use the modified definitions.)

    b) Use a suitable shared library mechanism for linking with the
    Library.  A suitable mechanism is one that (1) uses at run time a
    copy of the library already present on the user's computer system,
    rather than copying library functions into the executable, and (2)
    will operate properly with a modified version of the library, if
    the user installs one, as long as the modified version is
    interface-compatible with the version that the work was made with.

    c) Accompany the work with a written offer, valid for at
    least three years, to give the same user the materials
    specified in Subsection 6a, above, for a charge no more
    than the cost of performing this distribution.

    d) If distribution of the work is made by offering access to copy
    from a designated place, offer equivalent access to copy the above
    specified materials from the same place.

    e) Verify that the user has already received a copy of these
    materials or that you have already sent this user a copy.

  For an executable, the required form of the "work that uses the
Library" must include any data and utility programs needed for
reproducing the executable from it.  However, as a special exception,
the materials to be distributed need not include anything that is
normally distributed (in either source or binary form) with the major
components (compiler, kernel, and so on) of the operating system on
which the executable runs, unless that component itself accompanies
the executable.

  It may happen that this requirement contradicts the license
restrictions of other proprietary libraries that do not normally
accompany the operating system.  Such a contradiction means you cannot
use both them and the Library together in an executable that you
distribute.

  7. You may place library facilities that are a work based on the
Library side-by-side in a single library together with other library
facilities not covered by this License, and distribute such a combined
library, provided that the separate distribution of the work based on
the Library and of the other library facilities is otherwise
permitted, and provided that you do these two things:

    a) Accompany the combined library with a copy of the same work
    based on the Library, uncombined with any other library
    facilities.  This must be distributed under the terms of the
    Sections above.

    b) Give prominent notice with the combined library of the fact
    that part of it is a work based on the Library, and explaining
    where to find the accompanying uncombined form of the same work.

  8. You may not copy, modify, sublicense, link with, or distribute
the Library except as expressly provided under this License.  Any
attempt otherwise to copy, modify, sublicense, link with, or
distribute the Library is void, and will automatically terminate your
rights under this License.  However, parties who have received copies,
or rights, from you under this License will not have their licenses
terminated so long as such parties remain in full compliance.

  9. You are not required to accept this License, since you have not
signed it.  However, nothing else grants you permission to modify or
distribute the Library or its derivative works.  These actions are
prohibited by law if you do not accept this License.  Therefore, by
modifying or distributing the Library (or any work based on the
Library), you indicate your acceptance of this License to do so, and
all its terms and conditions for copying, distributing or modifying
the Library or works based on it.

  10. Each time you redistribute the Library (or any work based on the
Library), the recipient automatically receives a license from the
original licensor to copy, distribute, link with or modify the Library
subject to these terms and conditions.  You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties with
this License.

  11. If, as a consequence of a court judgment or allegation of patent
infringement or for any other reason (not limited to patent issues),
conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License.  If you cannot
distribute so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you
may not distribute the Library at all.  For example, if a patent
license would not permit royalty-free redistribution of the Library by
all those who receive copies directly or indirectly through you, then
the only way you could satisfy both it and this License would be to
refrain entirely from distribution of the Library.

If any portion of this section is held invalid or unenforceable under any
particular circumstance, the balance of the section is intended to apply,
and the section as a whole is intended to apply in other circumstances.

It is not the purpose of this section to induce you to infringe any
patents or other property right claims or to contest validity of any
such claims; this section has the sole purpose of protecting the
integrity of the free software distribution system which is
implemented by public license practices.  Many people have made
generous contributions to the wide range of software distributed
through that system in reliance on consistent application of that
system; it is up to the author/donor to decide if he or she is willing
to distribute software through any other system and a licensee cannot
impose that choice.

This section is intended to make thoroughly clear what is believed to
be a consequence of the rest of this License.

  12. If the distribution and/or use of the Library is restricted in
certain countries either by patents or by copyrighted interfaces, the
original copyright holder who places the Library under this License may add
an explicit geographical distribution limitation excluding those countries,
so that distribution is permitted only in or among countries not thus
excluded.  In such case, this License incorporates the limitation as if
written in the body of this License.

  13. The Free Software Foundation may publish revised and/or new
versions of the Lesser General Public License from time to time.
Such new versions will be similar in spirit to the present version,
but may differ in detail to address new problems or concerns.

Each version is given a distinguishing version number.  If the Library
specifies a version number of this License which applies to it and
"any later version", you have the option of following the terms and
conditions either of that version or of any later version published by
the Free Software Foundation.  If the Library does not specify a
license version number, you may choose any version ever published by
the Free Software Foundation.

  14. If you wish to incorporate parts of the Library into other free
programs whose distribution conditions are incompatible with these,
write to the author to ask for permission.  For software which is
copyrighted by the Free Software Foundation, write to the Free
Software Foundation; we sometimes make exceptions for this.  Our
decision will be guided by the two goals of preserving the free status
of all derivatives of our free software and of promoting the sharing
and reuse of software generally.

			    NO WARRANTY

  15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR
OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT WARRANTY OF ANY
KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE.  THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE
LIBRARY IS WITH YOU.  SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME
THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.

  16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY
AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU
FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE
LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING
RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A
FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF
SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
DAMAGES.

		     END OF TERMS AND CONDITIONS

           How to Apply These Terms to Your New Libraries

  If you develop a new library, and you want it to be of the greatest
possible use to the public, we recommend making it free software that
everyone can redistribute and change.  You can do so by permitting
redistribution under these terms (or, alternatively, under the terms of the
ordinary General Public License).

  To apply these terms, attach the following notices to the library.  It is
safest to attach them to the start of each source file to most effectively
convey the exclusion of warranty; and each file should have at least the
"copyright" line and a pointer to where the full notice is found.

    <one line to give the library's name and a brief idea of what it does.>
    Copyright (C) <year>  <name of author>

    This library is free software; you can redistribute it and/or
    modify it under the terms of the GNU Lesser General Public
    License as published by the Free Software Foundation; either
    version 2.1 of the License, or (at your option) any later version.

    This library is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
    Lesser General Public License for more details.

    You should have received a copy of the GNU Lesser General Public
    License along with this library; if not, write to the Free Software
    Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA

Also add information on how to contact you by electronic and paper mail.

You should also get your employer (if you work as a programmer) or your
school, if any, to sign a "copyright disclaimer" for the library, if
necessary.  Here is a sample; alter the names:

  Yoyodyne, Inc., hereby disclaims all copyright interest in the
  library `Frob' (a library for tweaking knobs) written by James Random Hacker.

  <signature of Ty Coon>, 1 April 1990
  Ty Coon, President of Vice

That's all there is to it!