avocado-framework/avocado

=============
80.0 Parasite
=============

The Avocado team is proud to present another release: Avocado 80.0,
AKA "Parasite", is now available!

This release (and the previous one) contains mainly internal changes
in preparation for the N(ext) Runner architecture to replace the
current one, and for the Job API to become a fully supported feature.

It's expected that release 81.0 will be the last release containing
major changes before a "pre-LTS release".  This way, development
sprint #82 will focus on stabilization, culminating in an 82.0 LTS
release.

Release documentation: `Avocado 80.0
<http://avocado-framework.readthedocs.io/en/80.0/>`_

Users/Test Writers
==================

* The Avocado configuration that is logged during a job execution is
  now the dictionary that is produced by the
  :mod:`avocado.core.future.settings` module, instead of the
  configuration file(s) content.  This is relevant because this
  configuration contains the result of everything that affects a job,
  such as defaults registered by plugins, command line options, all
  in addition to the configuration file.  The goal is to have more
  consistent behavior and increased job "replayability".

* As explained in the previous point, an Avocado Job is now configured
  by the configuration set by the :mod:`avocado.core.future.settings`
  code.  Because of the way this module works, options need to be
  registered, before the content on the config files can be considered
  valid values for a given option.  This has been done for a large
  number of Avocado features, but be advised that some configuration
  may not yet be seen by the job, because of the lack of option
  registration. We're working to identify and enable complete feature
  configuration on the next release.

* The "log level" of an Avocado is now defined using the standard
  Python level names.  If you have a custom configuration for this
  setting, you may need to adjust it (usually only a matter of
  lowercase to uppercase).

* The runner that will be used in a job can now be defined in the
  command line (in addition to being previously supported by a
  configuration entry).  If you want to try out the experimental
  N(ext) Runner, for instance, you should be able to use a command
  such as ``avocado run --test-runner=nrunner /path/to/my/tests``.

* The N(ext) Runner received support for job timeouts, and won't
  run further tests if the timeout expires.

* The N(ext) Runner now users the same Test ID that the current test
  runner uses, both in the to-be-removed ``avocado nrun`` and in the
  ``avocado run --test-runner=nrunner`` scenario.

* A brand new command, ``jobs`` enables users to, among other things
  to list information about previously executed jobs.  A command such
  as ``avocado jobs show`` will show the latest job information.

* The "standalone jobs" feature has been **deprecated**.  This feature
  allows users to write a test, that contains a builtin job executor
  for such a test that allows the test file to be executable.  This
  will be replaced by the Job API, which transparently supports the
  specification of the **same** file as a source of tests.

* The ``avocado run --loaders ?`` command to list available loaders
  has been removed.  This command line usage pattern is not consistent
  across Avocado (or follows the POSIX guidelines), and with the
  N(ext) Runner architecture depending on the
  :mod:`avocado.core.resolver` feature set, one will be able to see
  the resolvers with the ``avocado plugins`` command.

* The lower level :class:`avocado.core.job.Job`, instead of the
  ``avocado run`` command, is now responsible for generating result
  files, such as the JSON (``results.json``), xUnit (``results.xml``),
  etc.  This allows users of the Job API, as well as users of the
  ``avocado run`` command to have results generated as intended.

* The lower level :class:`avocado.core.job.Job`, instead of the
  ``avocado run`` command, is now also responsible for collecting the
  job-level system information (AKA ``sysinfo``).  This allows users
  of the Job API, as well as users of the ``avocado run`` command to
  have this feature available.

Bug Fixes
=========

* The ``avocado sysinfo`` command reverts to the pre-regression
  behavior, and now creates a directory following the
  ``sysinfo-$TIMESTAMP`` pattern and uses that for the content of the
  sysinfo output, instead of using the current directory by default.

* An incorrect configuration key name of the ``result_upload`` command,
  as part of the "results_upload" plugin, was fixed.

* :func:`avocado.utils.disk.get_disks` now supports all block devices,
  like multipaths, LVs, etc. Previously it used to return only
  ``/dev/sdX`` devices.

Utility APIs
============

* All of the :class:`avocado.utils.gdb` APIs are now back to a working
  state, with many fixes related to bytes and strings, as well as
  buffered I/O caching fixes.

* :mod:`avocado.utils.pmem` now supports the ``all`` namespace behavior
  for newer versions of the ``ndctl`` utility.

* :mod:`avocado.utils.software_manager` support for the Zypper package
  manager was improved to support the installation of package build
  dependencies.

Internal Changes
================

* Refactors for the :class:`avocado.core.nrunner.BaseRunnerApp` that
  made the list of commands available as a class attribute avoiding
  multiple resolutions and string manipulation when a command needs to
  be resolved.

* The N(ext) Runner received some foundation work for the persistence
  and retrieval of test generated artifacts.  The work takes into
  consideration that tests may be run disconnected of the the overall
  test job, and the job can retrieve those at a later time.

* The N(ext) Runner spawner selection is on the ``avocado nrun``
  command is now done by means of the ``--spawner=`` option that takes
  a spawner name, instead of the previous ``--podman-spawner`` option.
  This logic should be kept on the ``avocado run`` implementation and
  allow for new spawners to be used transparently.

* Internal reliability improvements to the N(ext) Runner status server
  implementation.

* The ``avocado nrun`` command now respects the ``--verbose`` command
  line option, producing less output if it's not given.

* The core sysinfo implementation received cleanups and now makes now
  distinction between collection at job or test time, and works on
  both or at any other moment.

* The :mod:`avocado.core.future.settings` now allows command line
  parsers to be added to previously registered options.  This allows
  features that don't require a command line to register options, and
  plugins that want to control such options with a command line to do
  so in a decoupled and extensive way.

* A new plugin interface, :class:`avocado.core.plugin_interfaces.Init`,
  was introduced to allow plugins that need to initialize themselves
  very early (and automatically) on Avocado.  Such plugins have no
  knowledge of the current configuration, but may use that interface
  to register new options (among other things).

* An Avocado Job is now run as part of the selftests suite, and more
  can be added.  This is intended to avoid breakage of the Job API as
  it gets closer to become a supported feature.

For more information, please check out the complete
`Avocado changelog
<https://github.com/avocado-framework/avocado/compare/79.0...80.0>`_.