avocado-framework/avocado

View on GitHub
docs/source/releases/81_0.rst

Summary

Maintainability
Test Coverage
===========================
81.0 Avengers: Infinity War
===========================

The Avocado team is proud to present another release: Avocado 81.0,
AKA "Avengers: Infinity War", is now available!

This release introduces many exciting new features.  We can't even
wait to get to the more specific sections below to talk about some
of the highlights:

* A new test runner architecture, previously known as the "N(ext)
  Runner", now available as the "nrunner" plugin.  It currently allows
  tests to be run in parallel in either processes or into Podman based
  containers.  In the near future, it should include LXC, Kata
  Containers, QEMU/KVM based virtual machines, etc.  It also includes
  the foundation of a requirement resolution mechanism, in which tests
  can declare what they need to run (specific Operating System
  versions, architectures, packages, etc).  Expect the Avocado feature
  set to evolve around this new architecture.

* A fully usable Job API, making most of Avocado's functionalities
  programmable and highly customizable.  Expect the Job API to be
  declared public soon, that is, to be available as ``avocado.Job``
  (instead of the current :class:`avocado.core.job.Job`) just like
  the Avocado Test API is available at :class:`avocado.Test`.

* A new settings API that is tightly linked to the Job API.  You can
  see all the existing configurations at runtime by running ``avocado
  config reference``.  To integrate Avocado to an existing project or
  a CI environment, a custom job with a few configurations will give
  you a lot of flexibility with very little need to write Python code.
  Some examples are available at ``examples/jobs``.

* Support for multiple test suites in a Job, so that each test suite
  can be configured differently and independently of each other.
  Fulfill your use case easily (or let your imagination go wild) and
  define different runners for different test suites, different
  parameters to different test suites, or run some test suites
  locally, while others isolated on containers.  Anything that is
  configurable with the new settings API should be transparently
  configurable in the context of a test suite (provided the test suite
  deals with that feature).

This release is also a "pre-LTS release".  Development sprint #82 will
focus on stabilization, culminating in the 82.0 LTS release.

Release documentation: `Avocado 81.0
<http://avocado-framework.readthedocs.io/en/81.0/>`_

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

* The ``remote``, ``vm`` and ``docker`` runners (which would run jobs
  on remote, vm and docker containers) were removed, after having being
  deprecated on version 78.0.

* The "standalone job" feature, in which a test could be run as a
  standalone job was removed after having being deprecated on version
  80.0.  The alternative is to use an Avocado Job (using the Job API),
  with a test defined on the same file, as can be seen on the example
  file ``examples/jobs/passjob_with_test.py`` in the source tree.

* The ``yaml_to_mux`` varianter plugin now attempts to respect the
  type of the value given to ``--mux-inject``.  For example, ``1`` is
  treated as integer, a value of ``1,2`` is treated as list a value of
  ``abc`` is treated as string, and a value of ``1,2,5-10`` is treated
  as list of integers as ``1,2,-5`` (as it is evaluated by
  :func:`ast.literal_eval`).

* Python unittests names are now similar to Avocado's own instrumented
  tests names, that is, they list the file name as a path, followed by
  the class and method name.  The positive aspect of this change is
  that that they can be reused again as a test reference (which means
  you can copy and paste the name, and re-run it).

* The ``avocado-runner-*`` standalone runners can now look for a
  suitable Python class to handle a given test kind by using
  ``setuptools`` entrypoints.

* For users of the Job API, a "dictionary based" varianter was
  introduced, that allows you to describe the variations of tests in a
  test suite directly via a Python dictionary.

* The output produced on the human UI for failed ``SIMPLE`` tests is
  now much more straightforward and contains more relevant data.

* Users attempting to use both the ``--loader`` and the
  ``--external-runner`` features will be warned against it, because of
  its inherent incompatibility with each other.

* A new ``avocado replay`` command supersedes the ``avocado run
  --replay`` command/option.

* The previous experimental command ``nlist`` has been removed, and
  its functionality can now be activated by using ``avocado list
  --resolver``.  This is part of promotion of the N(ext) Runner
  architecture from experimental to being integrated into Avocado.

Bug Fixes
=========

* Filtering using tags while listing the tests (but not while running
  them) was broken on the previous release, and has now been fixed.

* Result event plugins were misbehaving because they were instantiated
  too early.  Now they're loaded later and lazily.

* Failure to load and run the Python unittest with the nrunner's
  :class:`avocado.core.nrunner.PythonUnittestRunner` depending on
  the directory it was called from is now fixed.

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

* The :mod:`avocado.utils.vmimage` now contains an auxiliary
  documentation (:ref:`avocado.utils.vmimage.supported_images`) that
  lists the exact Operating System names, versions and architectures
  that have been tested with an Avocado release.

* The :mod:`avocado.utils.pmem` library can now check if a given
  command is supported by the underlying ``ndctl`` binary.

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

* Improvements to the selftests, including a collection of jobs that
  are run as tests, and a job that tests a good number of Job API
  features using variants.

* The :mod:`avocado.core.settings` is a completely redesigned module,
  and central to Avocado's future set and Job API.  It was present as
  ``avocado.core.future.settings`` on previous versions. All module
  and plugins have been migrated to the new API.

* The :mod:`avocado.utils.software_manager` module has been split into
  a finer grained directory and module structure.

* Various documentation content improvements, and various build
  warnings were addressed.

* The ``avocado_variants`` attribute is no longer kept in the job
  configuration as an instance of a
  :class:`avocado.core.varianter.Varianter`, instead, the
  configuration for the various variants are kept in the configuration
  and it's instantiated when needed.

* :mod:`avocado.utils.wait` now uses :func:`time.monotonic`, which
  makes it more reliable and less susceptible to errors when the
  system clock changes while this utility function is running.

* Refactors resulting in more code being shared among Avocado
  Instrumented and Python unittest handling on the
  :mod:`avocado.core.safeloader` module.

* The :mod:`avocado.core.safeloader` module now supports relative
  imports when attempting to follow imports to find valid classes with
  tests.

* A new :mod:`avocado.core.suite` was introduced, which is the basis
  of the multiple test suite support in a Job.

* Codeclimate.com is now being used for code coverage services.

* Codeclimate.com now has the bandit plugin enabled, which means that
  security related alerts are also caught and shown on the analysis.

For more information, please check out the complete
`Avocado changelog
<https://github.com/avocado-framework/avocado/compare/80.0...81.0>`_.