It would be nice to show result url in output.
ResultEvent class is not good for that because it is called too early,
and makes output looks ugly. Let's use Result.render hook which is
specially designed for this stage
Signed-off-by: NDmitry Monakhov <dmonakhov@openvz.org>

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

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

On `make rpm`, docs build fails with:

    docs/source/optional_plugins/varianter_pict.rst:33: WARNING: Could
    not lex literal_block as "ini".

This patch changes the way to define a block of code for a
proven-to-be-good one.
Signed-off-by: NAmador Pahim <apahim@redhat.com>

Signed-off-by: NAmador Pahim <apahim@redhat.com>

It looks like Travis does some different handling of process, and
we need to manually force the writes to the stdout.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

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

- Revert "travis: Use parallel check in travis"

    This reverts commit 4b296996.

    Parallel check was enabled in Travis, but it's overloading the system
    and making the time sensitive tests to FAIL.

- avocado.utils.process: fix default prefix

    Log prefix needs at least the %-style string formatter. Let's include
    that as the default so we don't break under 'no prefix'.
Signed-off-by: NAmador Pahim <apahim@redhat.com>

Signed-off-by: NAmador Pahim <apahim@redhat.com>

Signed-off-by: NAmador Pahim <apahim@redhat.com>

Signed-off-by: NAmador Pahim <apahim@redhat.com>

Signed-off-by: NAmador Pahim <apahim@redhat.com>

PICT is a free software (MIT licensed) tool that implements
combinatorial testing.  Information about it can be found here:

   https://github.com/Microsoft/pict/

This is a very simple, but functional varianter implementation.  It
performs the basic tasks of creating variants and making parameters
accessible to tests.  To test this version, it's recommended to
download PICT, build it (it only takes a C compiler such as GCC or
clang and make) and put it in a location included in your PATH (or use
`--pict-binary`).

To list the variants generated with the default order (2, that is,
do a pairwise idenpendent combinatorial testing):

 $ avocado variants --pict-parameter-file=examples/params.pict
 Pict Variants (11):
 Variant amd-scsi-rtl8139-windows-rhel6-acff:    /run
 ...
 Variant amd-ide-e1000-linux-rhel6-eb43:    /run

To list the variants generated with a 3-way combination:

 $ avocado variants --pict-parameter-file=examples/params.pict \
   --pict-order-of-combinations=3

 Pict Variants (28):
 Variant intel-ide-virtio-windows-rhel7-aea5:    /run
 ...
 Variant intel-scsi-e1000-linux-rhel7-9f61:    /run

It's also possible to run tests with variants generated by the same
command line switches.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The official fabric project doesn't support Python 3 yet.  Let's use
the unofficial Python 3 port for now.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

We have never used the subprocess32 backport module, and probably
wouldn't even support if users were doing that.  Let's remove the
the possibility of confusion and byte rot.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The snakefood package is not available on Python 3.  Let's skip
this check then.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

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

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

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

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

And the support for combined mode.  So far, there was no need to
determine how processes would be run, and how their output would be
collected: the stdout and stderr would always be collcted separately.

Now, with the need to support the "combined" mode, the
`--output-check-record` command line switch not only has to control
which files are populated (stdout.expected, stderr.expected) but also
it needs to control the execution of process, because it's what will
generate the "raw material" for a possible "output.expected" file.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

This is refactor of the current locks/file/threads created in the
SubProcess class to a more contained an isolated (and tested)
block of functionality.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The stream_prefix does nothing, let's remove it.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Which simplifies the code and makes it more Pythonic.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

There are a few active players in the handling of the threads that
handle the reading from PIPEs and writing into files (and logging
streams), including files, locks and thread objects.

Let's define them in advance at instance initialization time, make
them protected instead of public.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Besides a small optmization, it seems more correct a single I/O
readiness decision in each loop, instead of possible different
results in the two calls.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The ignore_bg_processes parameter is unnecessary, because it's also
available as an attribute.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

And raise an exception early if an invalid value is given.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Let's document the best practices on subclassing Avocado.
Signed-off-by: NAmador Pahim <apahim@redhat.com>

The parallel check should be faster and now it should not produce
false-negatives, let's use it in travis.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Currently we check all commits between the first commit and
origin/master, which frequently fails on PRs based on older commits.
Let's check only the PR commits in PR checker while still checking all
the commits when checking pushes as failures there are emailed to
maintainers who should resolve the issue manually.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Let's allow aborting the execution with automatic cleanup.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

This script can be used after one merges a PR to verify __all__ the
commits pass `make check` and publishes the results to github.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The test_basic.RunnerOperationTest.test_early_latest_result waits for
link creation for 0.5s and then it simply fails, without waiting for the
process. The problem is that in such case the tmpdir can be cleaned
before it's used by the created process and therefor left behind.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Occasionally the parallel check reports "pid XXX is not a child of this
shell", let's try to avoid this issue.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The parallel checker is a great tool, but frequently fails with
false-negatives. Let's re-run the failed tests in series in case less
than 10 failures were present and consider check successful in case they
pass.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Signed-off-by: NAmador Pahim <apahim@redhat.com>

Recent changes to remote runner use only stderr to get Avocado version,
but the docker plugin does not return separate stream for stdout and
stderr, therefor it always fails to get the version string. This patch
uses combination of stdeerr + stdout, which should be safe as we only
care about the first occurrence (so in case both stderr and stdout
produce version string, the stderr one will be used, anyway such
question is only hypothetical)
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Signed-off-by: NAmador Pahim <apahim@redhat.com>

The IMAGE_PROVIDERS stores possible classes to provide images, let's
avoid unnecessary multiple occurrences.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>