The 4235b329 changed the
stdout/stderr/output logging handlers to not to add tailing '\n's, but
these are processed by LoggingFile, which is designed to remove the
'\n's and to logged them per line. As a result the unittest was
incorrectly changed as the first outputs actually add the '\n' which was
removed in the expected output.

This patch keeps the raw approach as it is necessary for the last
buffered line and instead of buffering puts all lines including the
tailing '\n' and not '\n' for the remaining last line (if it's
available). Consequently this should improve the situation where output
is being produced by stderr and stdout concurrently as no buffering of
the non-tailing-newline output is done on LoggingFile level.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

* https://github.com/avocado-framework/avocado:
  Removed port related information while returing location

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

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

This changes uses the six compatibility equivalents for iterkeys,
iteritems and itervalues.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

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

The main goal here is to make string_to_safe_path behave properly
under Python 3.  The big change here is the added code that selects
the module where the maketrans() function is located, which in Python
3 is a static method of bytes.

Additionally, it removes the module level definitions
(FS_UNSAFE_CHARS, FS_TRANSLATE) which don't have any other use so far.

FS_USNAFE_CHARS has become a simple variable, now expressed as a
sequence of bytes.  If other characters are added to the translation
table, maketrans() will make sure the they are the same size, so
I believe the code to generate the second argument of the same size
can be dropped.

Now, a side effect is that, on Python 2, the result of the json
parsing (json.loads()) is a dict with unicode keys and values.  When
that is passed to the updated string_to_safe_path() implementation, a
causes TypeError is raised (from Objects/unicodeobj.c,
charmaptranslate_lookup()).
Signed-off-by: NCleber Rosa <crosa@redhat.com>

And its related functions (callers and callees).

In Python 3, strings are unicode, so an encode operation will
return a byte object, while a decode (on a byte or bytearray)
will return a string (unicode).

To start with, string_safe_encode() wants to solve the problem
of users mising unicode and other encoded strings, which doesn't
make sense for Python 3.

Additionally, it uses encode and decode in a way that is
very specific to Python 2, such as assuming the result of a
decode() can then be "encode()d" and still be a string.

For Python 3, the idea here is to *not* deal with any
encode/decode operations attempting to get a "safe" string.
The first part of the function, which converts non-string
objects to string, is still being relied upon by some code,
so it's kept (at least for now).
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Earlier function used to return the entire physical slot location
including port information if any. Now it returns slot information
only till adapter level, not the port level.
Signed-off-by: NVenkat R B <vrbagal1@linux.vnet.ibm.com>

This is a mass change for something that has proved right (it was
applied on previous commits on a as-needed basis).  To avoid having to
hit the same issues over and over again, let's change all occurences.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

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

Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

* https://github.com/avocado-framework/avocado:
  Project infrastructure removal of flexmock
  selftests/unit/test_vm.py: port from flexmock to mock
  selftests/unit/test_utils_partition.py: port TestMtabLock.test_lock to mock
  selftests/unit/test_test.py: port test_all_dirs_exists_no_hang to mock
  selftests/unit/test_test.py: remote flexmock requirement of test_long_name
  selftests/unit/test_datadir.py: port test_settings_dir_alternate_dynamic to mock
  selftests/unit/test_datadir.py: port test_unique_log_dir to mock
  selftests/unit/test_datadir.py: port test_datadir_from_config to mock
  selftests/unit/test_distro.py: port from flexmock to mock (unittest.mock)
  selftests/unit/test_remote.py: port from flexmock to mock (unittest.mock)
  RemoteTestRunner: fix description of setup() method

This commit revomes flexmock from the various requirements file, from
the RPM SPEC file.  Starting here, it's officially no longer an
Avocado dependency.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The approach used on this port is to avoid excessive mocking, like the
test_remote.py port, and use the real code, unless when insteracting
with dynamic and non-existing components (such as a the hypervisor or
running domain).

A partial list of the changes when compared to the previous version:

* Args was made into a real argparse.Namespace instance because
  there's no need for it to be a mocked object.  Actually, it is
  better to have an exact object instance, rather than a mock when
  it's possible to.

* The "references" attribute being set on the, then mocked, job instance
  has been dropped, because the natural way of passing that to the job
  is using the "reference" key in args.  Also, while purely cosmetic,
  the references that look like the example instrumented tests have
  been renamed (passtest -> passtest.py).

* Removal of JSON_RESULTS, which was not being used in the test.

* Use dry_run and clean up job results.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

And remove module flexmock requirements.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

And remove the requirement on flexmock on this module.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Basically, this is not about switching to mock instead, because
mocking properties is more challeging with mock, and mocking
a property of an existing instance is, to the best of my knowledge,
not possible.

So, it seemed like a good idea to use a different test class with an
overloaded property that returns the fake filename, and while at it,
separate split that into two other tests.

It should be easier to understand what each block (now a test) is
about.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Pretty straightforward port, not really any remarks besides another
test file becoming flexmock free.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Besides the obvious port, I reckon that using a more obvious fake
return value for the return of `time.strftime` improves the legibility
of the test.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The context manager feature of mock.patch replaces a lot of the code
that restores the original state.  Besides that, one explicit assert
was added to make sure that the non-patched state is what we expect.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The first line says it all, this is a pretty straightforward port.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

This is a port to the mock library that is in the standard in Python 3, and
it's available in Python 2, and already a requirement for Avocado.

The approach used on this port is to avoid excessive mocking, and use
the real code, unless when insteracting with dynamic and non-existing
components (such as a remote system).

A partial list of the changes when compared to the previous version:

* Args was made into a real argparse.Namespace instance because
  there's no need for it to be a mocked object.  Actually, it is
  better to have an exact object instance, rather than a mock when
  it's possible to.

* The "references" attribute being set on the, then mocked, job instance
  has been dropped, because the natural way of passing that to the job
  is using the "reference" key in args.  Also, while purely cosmetic,
  the references that look like the example instrumented tests have
  been renamed (passtest -> passtest.py).

* The "unique_id" attribute being set on the, then mocked, job instance
  had a value (1-sleeptest;0) completely incompatible with job IDs, and
  was removed from being set manually.

* This version does *not* mock "low level" remote execution of commands,
  such as running "env" or "avocado -v" on the remote host.  Instead, it
  mocks the end result of check_remote_avocado completely.

* The extra test on RemoteTestRunner.setup() seems unnecessary, because
  that is part of the run_suite() workflow.

* Checking if `__init__` was called is not be possible with mock. See
  https://docs.python.org/3/library/unittest.mock.html#mocking-magic-methodsSigned-off-by: NCleber Rosa <crosa@redhat.com>

The goal of this change is to make the behavior of the Avocado command
line application as similar as possible to the behavior when running
under Python 2.  This bit is about what happens when the simplest
possible Avocado command line is run:

 $ avocado

On Python 2, it prints the complete usage, along with the error
message "too few arguments".  While we could go employ extreme changes
to make the very same error message be given on Python 3, this seems
wrong.  The reason is that the message is generated by the Python 2
standard library.  The Python 3 standard library relies on "required"
arguments instead.  When settings the subcommand sub parser as a
required option, the behavior is exactly the same, minus the error
message which becomes "the following arguments are required:
subcommand".
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Let's make this test more in line with the unittest.TestCase style,
making use of the assertions, whose failures can be point us to the
exact location of the failure.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The current test loops over the result directory versions, and while
it aggregates and prints the individual failures, the unittest
framework has the same purpose.

To me, as a user, I'd rather know that test "xyz" failed right away,
than having to parse the output of a longer text message.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

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

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

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

Mock is slightly lighter than MagicMock, which is only necessary
when mocking "magic" methods.  So, let's revert to the lighter
Mock class.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

This module already uses mock, but doesn't import it in a way
that is usable on Python 3.  Also, let's make the style the
same as in the other modules that import conditionally for
Python 2 and 3.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

Mock is slightly lighter than MagicMock, which is only necessary
when mocking "magic" methods.  So, let's revert to the lighter
Mock class.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

This module already uses mock, but doesn't import it in a way
that is usable on Python 3.  Also, let's make the style the
same as in the other modules that import conditionally for
Python 2 and 3.
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>

six is already a stevedore requirement, but since we'll now be
using it directly, it's a good idea to list it explicitly as
a dependency.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

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

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

Avocado process module has a routine to drain the std* buffer with a
per-line basis loop. This loop makes the new-line character to be
dropped.

On the other hand, the Python logging system is hard-coded to add a new-line
character at the end of the msg string.

The problem takes place when `process` writes a line which originally
had new-line character at the end but the logging system still adds the
new-line character to that line.

In this patch:
- avocado.utils.process to include the new-line character in the lines
  that originally had one.
- avocado.core.test to use for raw outputs a custom logging FileHandler
  that does not add a new-line character at the end of each line.

Verified with (checking `.../test-results/1-.../[stdout|stderr|output]`):

    $ avocado run --external-runner '/bin/echo' 'foo'

    $ avocado run --external-runner '/bin/echo -n' 'foo'

    $ avocado run --external-runner '/bin/ls' '.'

    $ avocado run --external-runner '/bin/ls' 'nonexistingfile

Reference: https://trello.com/c/dH7FLT75Signed-off-by: NAmador Pahim <apahim@redhat.com>