Sprint 11

Since we are not running short anytime of error code definitions, I
believe we can give a more descriptive return code when jobs are
actually interrupted.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

That is, within a second, if the test process is running, the user
will see a "full cycle" of the throbber.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The command now is 'datadir'.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Inside the job result instance dir, introduce a 'test-results'
directory that holds the results of all the other test instances.
This organizes things better and is a standard dir name,
since now we mimic a the origin test directory structure.

Update the docs accordingly.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

synctest fails on the CI VM, let's use the simpler,
bogus datadir test to do that functional verification.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Create a test that uses files from the data dir, as
a functional test, so we notice when that functionality
is broken.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Verify if plugins are correctly listed and none are
disabled.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Order the available plugins and use colors to better
organize the information displayed:

$ scripts/avocado plugins
Plugins loaded:
    datadir - Implements the avocado 'datadir' subcommand (Enabled)
    json - JSON output (Enabled)
    journal - Test journal (Enabled)
    multiplexer - Implements the avocado 'multiplex' subcommand (Enabled)
    plugins_list - Implements the avocado 'plugins' subcommand (Enabled)
    run_vm - Run tests on a Virtual Machine (Enabled)
    sysinfo - Collect system information (Enabled)
    test_lister - Implements the avocado 'list' subcommand (Enabled)
    test_runner - Implements the avocado 'run' subcommand (Enabled)
    xunit - xUnit output (Enabled)
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Add a test that checks the behavior of the 'list' subcommand.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Show test aliases and full paths on the output of
the 'list' command:

$ scripts/avocado list
Tests dir: /home/lmr/Code/avocado.lmr/tests
    Alias         Path
    gendata       /home/lmr/Code/avocado.lmr/tests/gendata.py
    synctest      /home/lmr/Code/avocado.lmr/tests/synctest.py
    linuxbuild    /home/lmr/Code/avocado.lmr/tests/linuxbuild.py
    fiotest       /home/lmr/Code/avocado.lmr/tests/fiotest.py
    skiptest      /home/lmr/Code/avocado.lmr/tests/skiptest.py
    warntest      /home/lmr/Code/avocado.lmr/tests/warntest.py
    doublefail    /home/lmr/Code/avocado.lmr/tests/doublefail.py
    timeouttest   /home/lmr/Code/avocado.lmr/tests/timeouttest.py
    doublefree    /home/lmr/Code/avocado.lmr/tests/doublefree.py
    trinity       /home/lmr/Code/avocado.lmr/tests/trinity.py
    multiplextest /home/lmr/Code/avocado.lmr/tests/multiplextest.py
    sleeptenmin   /home/lmr/Code/avocado.lmr/tests/sleeptenmin.py
    sleeptest     /home/lmr/Code/avocado.lmr/tests/sleeptest.py
    errortest     /home/lmr/Code/avocado.lmr/tests/errortest.py
    failtest      /home/lmr/Code/avocado.lmr/tests/failtest.py
    abort         /home/lmr/Code/avocado.lmr/tests/abort.py
    whiteboard    /home/lmr/Code/avocado.lmr/tests/whiteboard.py
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Execute a test that uses the data dir (synctest), both
with alias and non alias. It will help us to capture
when there is a regression using their data files.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

So avocado can correctly find data files for a test that
is referenced by its alias.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Instead of starting tags at 1 and compulsorily putting
them on tests (even if it is the 1st execution of the test),
start with 0 and omit it if it is only the first test.

Also, reproduce the directory structure of tests in the
logs directory. This means that a test execution of

mydir/firstlevel.py
mydir/category/secondlevel.py

Would translate in the final directory results as:

mydir/firstlevel.py/
mydir/category/secondlevel.py/

Preserving the directory structure makes things more
consistent and intuitive, while preserving a unique
test ID.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Use a job ID that is a SHA1, similar to git, instead
of the UUID. Also, use the first 7 characters of the
job ID in the directory structure, to mirror the git
short SHA1 identifier.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Separate the code that generates a job SHA1 on its own
library. It'll make things easier for unittesting and other
needs.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

This is a workaround for the functional suite running
slow, until there's a more proper solution.

Reduce the polling time from 1 second to 0.01 second,
so short lived tests end quicker.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Try as much as possible to turn timestamps into the
ISO format. In fact, iso requires hours and minutes
to be separated by a ':', but that makes URLs far
less convenient, requiring escape characters. So
let's replace it with a dot.

Also, put the short ID after the date, so date sort
can be properly done.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Make some changes to bring the concept of job ID as a
user visible information:

* Make the human output to display the job ID
* Make the json output to have the job ID
* Record the job id in a file inside the results dir

Adapt functional tests to that new schema.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Another change to clarify the purpose of each directory.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Instead of using 'debug log' we can use 'job log', as it
is a human readable log of the job execution.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

We used 'run' as a prefix for avocado job output directories
after virt-test (basically, inertia). Let's use something
better:

job-[first 8 characters of job id]-timestamp

This way the job id starts to be more prominent in the
human interface as well.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

We can aim for a more terse output than what we have
today. This is my proposal to simplify things a bit,
while still being clear.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Run avocado in a long test, interrupt it and verify if
the runner displays the appropriate behavior, in terms
of exit code and output.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

If we press Ctrl+C, the test that was being executed
still does not have a status, and that reflects in the
job log with a None printed in the final status.

So, if there's no status yet, we should safely (?)
assume that the test was interrupted.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Handle KeyboardInterrupt inside avocado, making
Ctrl+C to terminate the program, instead of generating
a python backtrace.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Add a special exception for when tests are interrupted
by Ctrl + C, and the required mapping in status.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Add send_signal(), that lets us to send arbitrary signals
to subprocesses, and rewrite kill() and terminate() in
terms of it.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

avocado.plugins: Simplify commands --xunit/--json.

avocado: Bump to version 0.10.1

Bugfix release.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

avocado: Fix renamed attribute from avocado.core.output

The reference was incorrect, causing some of the plugins
to crash.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Simplify commands for xUnit and JSON output plugins:

* Remove options --xunit-output and --json-output.
* Use '-' to stdout or filename to options --xunit and --json.

Usage examples:

$ avocado --xunit - run sleeptest  # goes to stdout
$ avocado --xunit /tmp/foo.xml run sleeptest
$ avocado --xunit - --json /tmp/foo.json run sleeptest.
Signed-off-by: NRuda Moura <rmoura@redhat.com>

Version 0.10.0

Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>