Those two are interconnected, let's keep them in one file instead of
local imports.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Avoid unnecessary complexity of quering for enabled logs and only support
os.environ to show the full traceback.

Additionally avoid problems with redirected sys.stderr by writing
directly to fd 2, which is STDERR.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

This patch changes the way logging is initialized in avocado app:

1. stdout/stderr/logging is replaced with in-memory handlers
2. when arg parsing is done, logging is reconfigured accordingly to
   setting and re-logs stored messages (when streams are enabled)
3. when job starts and "test" stream is enabled, "" and "avocado.test"
   streams are enabled
4. when job finishes and "test" stream enabled the additional outputs,
   the additional outputs are disabled again.

The 3 and 4 is necessarily as the "" logger contains all logs including
avocado initialization (stevedore, PIL, ...). This way the "test" stream
still outputs everything, but only during the job-execution and not
during avocado initialization and cleanup. (one can use "early" log to
see those).

Last but not least it keeps "avocado.app" output in stdout (and stderr),
but it moves all other outputs to stderr instead. This is IMO better
handling as one can distinguish between debug and "additional" output.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Refactor logging initialization [subset/first round]

No actual changes, only reorder imports alphabetically (system, utils,
relative).
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The format_help was used to show "avocado multiplex -h" when
incompatible options were used, but it requires the parser to be stored
and as this is the only plugin doing this, let's just get rid of it
instead fixing the bug.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The function "add_console_handler" is quite useful. This patch extends
it's capabilities and renames it to fit the new purpose.

Because the new default logging level for "add_log_handler" is INFO,
we add the "avocado.test" and root logger to level DEBUG.

Then finally a few adjustments to the tests that rely on specific
log format and level:

1) selftests/functional/test_multiplex.py: adapt to changes in
test logger That means that extra prefixes ('avocado.test') are
printed out on each line when avocado is run with '--show-job-log'.
While at it, be more precise and check for the specific params at the
specific run with the given variant, instead of looking for a given
param in the output of the three tests variants.

2) selftests/functional/test_standalone.py: adapt to the change of
log prefix ('avocado.test') and extra line because of DEBUG level.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Lets share this log with all debug info in avocado not enabled by default.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The "start_file_logging" is used when job results are created to add
job file-handlers to logging.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

avocado.core.html: Fix the unicode handling for older pystache [v2]

Python 2.6 works fine with both, but let's use the specific version
of pystache we support (in html-plugin) but don't test in CI.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The older versions of pystache encodes the result into default encoding
(ascii), which causes exceptions when writing the results into
utf8 file using codecs.

The newer pystache doesn't require this as it already returns unicode.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The sysinfo files can contain non-ascii characters.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

This hook is used by older pystache (pystache==0.4.1). We don't
use it so let's simply ignore it.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The --upgrade forces all mentioned packages into the highest possible
version, even though the existing ones match the prescription. Without
it the packages that match prescription are kept and only non-matching
ones are upgraded.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

We use features of pip-6.0.1 in our requirement files, therefor we
depend on pip>=6.0.1. This commit adds this dependency before
requirements are processed without "--upgrade". When the user has
pip>=6.0.1, nothing happens. When he has older pip it's upgraded.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

avocado.plugins.replay fix crashes when using --replay-test-status [v2]

Options --multiplex-files and/or urls are incompatible with
--replay-test-status, since the resulting variants will differ from the
original job.

This patch adds a test to fail avocado when those incompatible options are
used togheter.

Also, some minor fixes in replay messages; and in return code when --remote
is used.

Reference: https://trello.com/c/OJpRv36iSigned-off-by: NAmador Pahim <apahim@redhat.com>

Managing different requirements files for different python version
requires much more efforts when the requirements became too complex.

This issue emerged when a required package `stevedore` decide not
supporting Python 2.6 since version `1.11.0`.

```
>>> import stevedore
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib/python2.6/site-packages/stevedore/__init__.py", line 23, in <module>
    LOG.addHandler(logging.NullHandler())
AttributeError: 'module' object has no attribute 'NullHandler
```

Pip provides a feature named [Environment
Markers](https://www.python.org/dev/peps/pep-0496/) for requirements
file to include specific packages conditionally which could properly
solve this issue.

This fix:
1. Use environment markers and merge python version specific
requirements files;
2. Remove python version specific requirements files references in
`Makefile` and Travis configuration;
3. Use `pip install -r requirements.txt` to replace per-line package
installation;
4. Restrain package `stevedore` on Python 2.6 to `1.10.0`.
Signed-off-by: NHao Liu <hliu@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>

Move remote/vm result functionality to runners [v1]

avocado.core.html: Add extensive debug on UnicodeDecodeError

Two bugfixes regarding the testplan

Currently we store job urls in job.urls, while letting people override
them in "job.run()". This is unnecessary and IMO wrong. Let's simply
set the "job.urls" in "job.__init__()" and use them throughout the
whole workflow.

This requires very minor change in standalone tests, everything else
works as before. Additionally we don't need the "preprocess_urls"
which was not actually used.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The functionality to run tests on remote machines and on VMs are
implemented using a mixture of TestRunner and TestResult classes.

In all fairness, it looks like the TestResult classes were abused
because non-standard attributes were used as a foundation for that
functionality. Now that `setup` and `tear_down` have been added
to the TestRunner interface, let's move those custom features from
the TestResult classes to the TestRunner ones.
Signed-off-by: NCleber Rosa <crosa@redhat.com>
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Some users are getting UnicodeDecodeError, but we're unable to reproduce
it. This commit lists the full template and repr() of the failure so we
can at least see what was rendered and identify the source.

No one should actually see this message and we should remove it once we
identify and fix this problem, but without such extensive debug we can't
do anything about it.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

The parsed (and possibly processed) list of urls given on the command
line is quite useful to other implementation of test runners (see
the upcoming changes on the remote/vm runners). So, let's add a proper
interface that can be used to customize/sanitize the urls given on
the command line, and keep that as a reference in the `urls` attribute.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Instead of getting the application arguments from the test result,
let's get them from from job object itself. The idea is that, once
fnctionality is moved from the test result class to the test runner
class, the same source of arguments can be used for the incoming
code.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

We're adding extra functionality to test runners, were they seem to
be more appropriate than in the test result classes. Let's add empty
implementations on the base class so that their re-implementation
is optional on inherited classes.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

This implementation of __getattr__ has allowed users of the test
result proxy, most importantly the result test runner, to ask for
attributes, including important methods, that would only be available
on the RemoteTestResult implementation.

Things have been working more or less by pure chance, based on the
lucky fact that the RemoteTestResult would be at the right place on
the proxy `output_plugins`.

Let's forbid users of the test result proxy to look for attributes
that are not part of the test result "contract" by removing this
magical (and rather evil) method.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The "ps" binary is used in avocado.utils.process library.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

This unittest patches find_command to return /usr/bin/sudo and then runs
it and expects some results. This works only with correctly set sudo,
but it does not work without sudo (as the binary is not found) or with
incorrectly set sudo. Let's use "/bin/true" instead.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

avocado.core.runner: Support sigtstp to interrupt tests [v3]

Makefile: cleanup Python compiled files properly

When we abort the test and the test ignores SIGTERM, avocado hangs for
infinity. Let's wait a while and send SIGKILL.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>