All the Avocado deps are now either on Fedora or EPEL, so there's no
need for any extra setup, and consequently, for that documentation
section.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

mock has switched to using systemd-nspawn for its chroots.  It looks
like it's breaking the Avocado workflow for building packages.

Until we can better evaluate what's missing there, let's revert back
to the old implementation.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Similar to make link, the requirements-plugins target walks through
the plugins and install dependencies.

It checks if the plugin has either:

 * A Makefile
 * A requirements.txt file

https://trello.com/c/R9rxte6o/608-add-make-requirements-pluginsSigned-off-by: NGreeshma Gopinath <ggopinat@redhat.com>
Signed-off-by: NCleber Rosa <crosa@redhat.com>

When switching between 36lts and master branch, content from 36lts
(the .egg-info dir) may be kept.  The reason for that is that 36lts
used "avocado" as the module name, and thus produces
"avocado.egg-info" dir.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Make clean is missing the egg directories. This patch includes in make
clean a command to remove them.

Reference: https://trello.com/c/VgF0fkEaSigned-off-by: NAmador Pahim <apahim@redhat.com>

A finer level of granularity may help to extend our testing coverage
while reducing the amount of false positives.  This introduces the
three different levels, mapped like this:

   * level 0, AKA "make check"
   * level 2, AKA "make check-full"

There are no changes of test assignment, that is, tests previously
being run under "check-full" will continue to be run only at that
target.  The same is true for tests that would run under "make check"
before this.

The big change is that there's now a middle ground, that can be
activated by manually setting the AVOCADO_CHECK_LEVEL variable.  Level
1 is intended to be used be used on environments that are halfway
between a dedicated machine and a really low powered environment.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

This commit modifies the optional_plugins to depend on VERSION files
located in their main directories and adds "propagate-version" target to
Makefile to update the VERSION files for optional as well as "linkable"
plugins, if VERSION file was present.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Downstream (Fedora and EPEL) RPM packages have had a different way
of building release (by default) and snapshot packages by editing
the SPEC file.

Let's sync with them, while also allowing snapshot packages to be
built from our usual `make rpm` command.  Release builds are still
supported with `make rpm-release`.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Now that the optional plugins have been excluded from the main
git archive, let's create them (currently with setuptools sdist
command).

Also, let's adjust the upload instruction a bit to signal that
other packages may be in PYPI_UPLOAD too.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

PyPI understands (rightfully so) one Python module per archive.  So,
it doesn't make sense to bundle all the plugins there for users of
`pip install avocado-framework`.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

This is a step towards reducing the delta between Avocado "upstream"
packages, that is, the packages offered in our own custom repo, and
packages provided by "dowsntream" distros such as Fedora and (EP)EL.

The package rename is probably the change that could affect users
the most, so let's try to do this early (sprint-wise).  The proper
"Provides" and "Obsoletes" tags should, though, make the update
seamless.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Our users have pointed out that the basic Avocado test runner, and its
related packaged form, has too many dependencies.  They're absolutely
right, and the whole goal of having plugins in Avocado is that
dependencies can be made optional according to the features required
by a given user.

This specific change creates three new optional plugins, containing
the extended test runners: remote, vm and docker.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

We already tweaked the "make link" to not fail when "make link" fails
but I forgot about unlink, which is now failing. Let's allow the "make
link" failure as there might be non-related modules not implementing the
"make unlink" functionality.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The "make link" feature goes through subdirs trying to either use
`setup.py` or `Makefile` to install all plugins. The problem is that
even when `Makefile` is present it might not support `make link` target
which results in failure. This patch instead of failure in this case
just lists "FAIL" followed by the module name.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Since we favor the creation of a tarball with "git archive", PyPI gets
sensitive every now and then with our uploads.  Since the 42.0
release, pushing the new packages to PyPI would result in Server
Errors.

This adds the PKG-INFO to the tarball that will be uploaded, which
seems to fix the current errors.  Also, since it's now a single file,
it can be pushed with the "twine" command line application.

Alternatively, we could double check our file manifest, and rely
entirely on setuptools to generate the tarball.  But, I'm not a big
fan of adding files to version control, and then to a manifest of any
sort.

Note: in the past, this was also necessary because of the tweaks to
the information inside the PKG-INFO file, namely the package name
(avocado -> avocado-framework).
Signed-off-by: NCleber Rosa <crosa@redhat.com>

We renamed make check-long to make check-full. Let's reflect that change
in the help message.
Signed-off-by: NAmador Pahim <apahim@redhat.com>

We aim to mark more tests that are not suitable to run everywhere, so
it looks like a good idea to rename this to check-full.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

And not based on Python version.  It was observed that platforms based
on Python 2.6, such as EL6, do ship with setuptools that has support
for `--user`.  This eases considerably the development environment
on these platforms.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

This will change the behavior one gets with `make develop` so that
optional (but in-tree) plugins will be enabled by default.  For now,
this means enabling the HTML result plugin.

This also means that `make check` will also test those by default,
as it's already been done on Travis CI.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The `checkall` script should not try to develop/install/clean anything.
It's suppose to run the tests and it's upto the user to prepare the
environment they want to check.

This won't affect `make check` as it already cleans/develops before
executing checkall.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The avocado package was renamed in 760d82c9 to `avocado-framework`,
but the cleanup was left unchanged. This commit fixes it.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

We have been using "avocado" as the name of our package, as seen by
core Python tools such as pkg_resources and repositories uch as
pypi.python.org.  The fact is that, even though our libraries modules
live under the "avocado" namespace, there's no need to forge the
package name.

With this change, external packages can declare "avocado-framework"
as a requirement.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Signed-off-by: NCleber Rosa <crosa@redhat.com>

Signed-off-by: NCleber Rosa <crosa@redhat.com>

The `link` and `clean` targets, very useful under development
environments, have supported directories containing `Makefiles`.

It turns out that for some plugins it may be enough (and more
Pythonic) to just have `setup.py` files, so let's support that.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The goal of the `optional_plugins` directory is to hold all plugins
that ship with Avocado, but are not considered core functionality.
Plugins that are considered to be core still live under
`avocado/plugins`.

This also adds all directories to be found under `optional_plugins` to
the list of candidate plugins when running `make clean` or `make
link`.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The current wildcard expansion will look list every file (no matter
what type) as a plugin candidate.  It may be just my setup, but I have
a few files on the same common directory that shouldn't be treated as
plugin dirs.

While at it, also limit to directories that actually have a Makefile.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

When generating RPM packages for different distros, it makes sense
to override the chosen mock config (the target distro) in the command
line, such as:

 $ make MOCK_CONFIG=epel-6-x86_64 rpm

So let's add the default value ("default" itself) to every mock command
and let the user running make to override it.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Signed-off-by: NCleber Rosa <crosa@redhat.com>

The support for Debian packages have already been moved to a contrib
dir, their Makefile rules were kept but not even updated.  Let's
move those to the contrib dir.

Notice: the status of these rules are equal to the other Debian package
files, that is, they're unsupported and we do not guarantee they're
actually working.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

As the Makefile gets more complex and has more variables, I found it
useful to display how it's being driven.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Some warnings may be generated by setuptools code when running
`setup.py --version`.  Let's silence those.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Some useful (usually functional) tests can take a long time to run.
Let's add a new target that developers can use to run long tests every
once in a while.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Under some circumstances, getting the version indirectly through
`avocado/core/version.py` can return `unknown.unknown`.  Let's use the
authoritative source for version numbers: `setup.py` itself.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

This patch adds the `make spell` command so developers can easily
check the spelling on docstrings and comments. False positives
words should be added to spell.ignore file.
Signed-off-by: NAmador Pahim <apahim@redhat.com>

Avocado needs a bit of custom massaging to be able to properly live
under "avocado-framework" on PyPI.  This adds a target that will
make the upload of new releases easier.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

While testing `make requirements*` on a fresh machine, I found that
we rely on pip itself, which may not be available.  Its bootstrap
process is pretty simple, atomic and well documented, so it seems
like a good idea to add it as a convenience.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Support `make requirements-plugins` to walk all sub-directories
like the same way `make link` works.
Signed-off-by: NWei Jiangang <weijg.fnst@cn.fujitsu.com>
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

When running the per-commit checks on CI, the previous environment is
not as clean as it should be.

One example is when a given work introduces a new plugin. First the
HEAD of the branch is tested, which includes the installation of the
new plugin.  Then, the per-commit checks would return to the first
commit in the branch, which doesn't contain the plugin, but the
environment still knows about it.  This is a matter of forcing the
clean up of setuptools entrypoints (or any setuptools generated info
for that matter).

While adding the deeper cleaning, it seemed useful to split the
Makefile "link" target into "develop", which runs "python setup.py
develop" on the Avocado source tree alone.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Python 2.6 does not support --user in `python setup.py develop`. Let's
use system-wide develop when link/unlink is executed with python 2.6.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>