* Properly load module_prefix for collections
Now that we are using collections for tasks, we need to properly split
the name and load the prefix properly.
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
* Add unit tests for task_executor
This commit adds missing unit tests for action handler in test_task_executor.
Signed-off-by: Daniel Mellado <dmellado@redhat.com>
* Support using importlib on py>=3 to avoid imp deprecation
* Add changelog fragment
* importlib coverage for py3
* Ansiballz execute should use importlib too
* recursive module_utils finder should utilize importlib too
* don't be dumb
* Fix up units
* Clean up tests
* Prefer importlib.util in plugin loader when available
* insert the module into sys.modules
* 3 before 2 for consistency
* ci_complete
* Address importlib.util.find_spec returning None
* Move check_type_str() out of basic.py
* Move check_type_list() out of basic.py
* Move safe_eval() out of basic.py
* Move check_type_dict() out of basic.py
* Move json importing code to common location
* Move check_type_bool() out of basic.py
* Move _check_type_int() out of basic.py
* Move _check_type_float() out of basic.py
* Move _check_type_path() out of basic.py
* Move _check_type_raw() out of basic.py
* Move _check_type_bytes() out of basic.py
* Move _check_type_bits() out of basic.py
* Create text.formatters.py
Move human_to_bytes, bytes_to_human, and _lenient_lowercase out of basic.py into text.formatters.py
Change references in modules to point to function at new location
* Move _check_type_jsonarg() out of basic.py
* Rename json related functions and put them in common.text.converters
Move formatters.py to common.text.formatters.py and update references in modules.
* Rework check_type_str()
Add allow_conversion option to make the function more self-contained.
Move the messaging back to basic.py since those error messages are more relevant to using this function in the context of AnsibleModule and not when using the function in isolation.
* Add unit tests for type checking functions
* Change _lenient_lowercase to lenient_lowercase per feedback
* Rename method and make private
* Use is_iterable, combine transformations
* Remove unused return_values from network modules
* Improve docstrings in new functions
* Add new PASS_VAR
* Add unit tests for list_no_log_values
* Fix unit tests for Python 2.6
Refinements:
- return legal_inputs and update class properties
- remove redundant arguments from method and handle in caller
- add better exception types to method
* Add unit tests for handle_aliases
* Python interpreter discovery
* No longer blindly default to only `/usr/bin/python`
* `ansible_python_interpreter` defaults to `auto_legacy`, which will discover the platform Python interpreter on some platforms (but still favor `/usr/bin/python` if present for backward compatibility). Use `auto` to always use the discovered interpreter, append `_silent` to either value to suppress warnings.
* includes new doc utility method `get_versioned_doclink` to generate a major.minor versioned doclink against docs.ansible.com (or some other config-overridden URL)
* docs revisions for python interpreter discovery
(cherry picked from commit 5b53c0012ab7212304c28fdd24cb33fd8ff755c2)
* verify output on some distros, cleanup
User module can contain Indentation errors or syntax errors.
Handle AST exceptions rather than showing traceback while importing such module.
Fixes: #21707
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* Note: Python2 is not as intelligent at detecting false import loops as
Python3. context_objects.py cannot be added to cli/arguments because it
would set up an import loop between cli/__init__.py,
cli/arguments/context_objects.py, and context.py on Python2.
ci_complete
The goal of breaking apart the base_parser() function is to get rid of
a bunch of conditionals and parameters in the code and, instead, make
code look like simple composition.
When splitting, a choice had to be made as to whether this would operate
by side effect (modifying a passed in parser) or side effect-free
(returning a new parser everytime).
Making a version that's side-effect-free appears to be fighting with the
optparse API (it wants to work by creating a parser object, configuring
the object, and then parsing the arguments with it) so instead, make it
clear that our helper functions are modifying the passed in parser by
(1) not returning the parser and (2) changing the function names to be
more clear that it is operating by side-effect.
Also move all of the generic optparse code, along with the argument
context classes, into a new subdirectory.
* Once cli args are parsed, they're constant. So, save the parsed args
into the global context for everyone else to use them from now on.
* Port cli scripts to use the CLIARGS in the context
* Refactor call to parse cli args into the run() method
* Fix unittests for changes to the internals of CLI arg parsing
* Port callback plugins to use context.CLIARGS
* Got rid of the private self._options attribute
* Use context.CLIARGS in the individual callback plugins instead.
* Also output positional arguments in default and unixy plugins
* Code has been simplified since we're now dealing with a dict rather
than Optparse.Value
* Move get_all_subclasses out of sys_info as it is unrelated to system
information.
* get_all_subclasses now returns a set() instead of a list.
* Don't port get_platform to sys_info as it is deprecated. Code using
the common API should just use platform.system() directly.
* Rename load_platform_subclass() to get_platform_subclass and do not
instantiate the rturned class.
* Test the compat shims in module_utils/basic.py separately from the new
API in module_utils/common/sys_info.py and module_utils/common/_utils.py
* Move ansible.compat.tests to test/units/compat/.
* Fix unit test references to ansible.compat.tests.
* Move builtins compat to separate file.
* Fix classification of test/units/compat/ dir.
* module_common: set required parameter templar
Fix the following error (related to b455901):
$ ./hacking/test-module -m ./lib/ansible/modules/system/ping.py -I ansible_python_interpreter=/usr/bin/python
Traceback (most recent call last):
File "./hacking/test-module", line 268, in <module>
main()
File "./hacking/test-module", line 249, in main
(modfile, modname, module_style) = boilerplate_module(options.module_path, options.module_args, interpreters, options.check, options.filename)
File "./hacking/test-module", line 152, in boilerplate_module
task_vars=task_vars
File "ansible/lib/ansible/executor/module_common.py", line 910, in modify_module
environment=environment)
File "ansible/lib/ansible/executor/module_common.py", line 736, in _find_module_utils
shebang, interpreter = _get_shebang(u'/usr/bin/python', task_vars, templar)
File "ansible/lib/ansible/executor/module_common.py", line 452, in _get_shebang
interpreter = templar.template(task_vars[interpreter_config].strip())
AttributeError: 'NoneType' object has no attribute 'template'
* module_common.modify_module: templar is required
* Cache tasks as they are queued instead of en masse
This also moves the task caching from the PlayIterator to the
StrategyBase class, where it makes more sense (and makes it easier
to not have to change the strategy class methods leading to an API
change).
Fixes#31673
* Cleaning up unit tests due to 502ca780
* Adding aliases for field attributes and renaming async attribute
As of Python 3.7, the use of async raises an error, whereas before the use
of the reserved word was ignored. This adds an alias field for field attrs
so that both async and async_val (interally) work. This allows us to be
backwards-compatible with 3rd party plugins that may still reference Task.async,
but for the core engine to work on Py3.7+.
* Remove files fixed for 'async' usage from the python 3.7 skip list
* better cleanup on task results display
callbacks get 'clean' copy of result objects
moved cleanup into result object itself
removed now redundant callback cleanup
moved no_log tests
* moved import as per feedback
- old functionality is still available direct lookup use, the following are equivalent
with_nested: [[1,2,3], ['a','b','c']]
loop: "{{lookup('nested', [1,2,3], ['a','b','c'])}}"
- avoid squashing with 'loop:'
- fixed test to use new intenal attributes
- removed most of 'lookup docs' as these now reside in the plugins
* Ansible Config part2
- made dump_me nicer, added note this is not prod
- moved internal key removal function to vars
- carry tracebacks in errors we can now show tracebacks for plugins on vvv
- show inventory plugin tracebacks on vvv
- minor fixes to cg groups plugin
- draft config from plugin docs
- made search path warning 'saner' (top level dirs only)
- correctly display config entries and others
- removed unneeded code
- commented out some conn plugin specific from base.yml
- also deprecated sudo/su
- updated ssh conn docs
- shared get option method for connection plugins
- note about needing eval for defaults
- tailored yaml ext
- updated strategy entry
- for connection pliugins, options load on plugin load
- allow for long types in definitions
- better display in ansible-doc
- cleaned up/updated source docs and base.yml
- added many descriptions
- deprecated include toggles as include is
- draft backwards compat get_config
- fixes to ansible-config, added --only-changed
- some code reoorg
- small license headers
- show default in doc type
- pushed module utils details to 5vs
- work w/o config file
- PEPE ATE!
- moved loader to it's own file
- fixed rhn_register test
- fixed boto requirement in make tests
- I ate Pepe
- fixed dynamic eval of defaults
- better doc code
skip ipaddr filter tests when missing netaddr
removed devnull string from config
better becoem resolution
* killed extra space with extreeme prejudice
cause its an affront against all that is holy that 2 spaces touch each other!
shippable timing out on some images, but merging as it passes most
* Enable the pylint no-name-in-module check. Checks that identifiers in
imports actually exist. When we do this, we also have to ignore
_MovedItems used in our bundled six. This means pylint won't check
for bad imports below ansible.module_utils.six.moves but that's
something that pylint punts on with a system copy of six so this is
still an improvement.
* Remove automatic use of system six. The exec in the six code which
tried to use a system library if available destroyed pylint's ability
to check for imports of identifiers which did not exist (the
no-name-in-module check). That test is important enough that we
should sacrifice the bundling detection in favour of the test.
Distributions that want to unbundle six can replace the bundled six in
ansible/module_utils/six/__init__.py to unbundle. however, be aware
that six is tricky to unbundle. They may want to base their efforts
off the code we were using:
2fff690caa/lib/ansible/module_utils/six/__init__.py
* Update tests for new location of bundled six Several code-smell tests
whitelist the bundled six library. Update the path to the library so
that they work.
* Also check for basestring in modules as the enabled pylint tests will
also point out basestring usage for us.
So I thought I fixed it before, but there's still one location where
the `rc` value is influential to decide whether a task failed or not.
We already established in #24867 that it is up to the module to decide
what the return code actually means, not the task executor. We modified
the existing modules to move that logic into the module (eg. for
command, shell, etc.)
This relates to the integration tests of win_robocopy, where different
return codes have different meanings:
- 0 -- No files copied.
- 1 -- Files copied successfully! (changed)
- 2 -- Some Extra files or directories were detected. No files were copied. (warning)
- 3 -- (2+1) Some files were copied. Additional files were present. (changed)
- 4 -- Some mismatched files or directories were detected. Housekeeping might be required! (changed + warning)
- 5 -- (4+1) Some files were copied. Some files were mismatched. (changed + warning)
- 6 -- (4+2) Additional files and mismatched files exist. No files were copied. (warning)
- 7 -- (4+1+2) Files were copied, a file mismatch was present, and additional files were present. (changed + warning)
- 8 -- Some files or directories could not be copied! (changed + failed)
- 9 - 15 -- Fatal error. Check log message! (failed)
- 16 -- Serious Error! No files were copied! Do you have permissions to access $src and $dest? (failed)
This also fixes#24652
* test/: PEP8 compliancy
- Make PEP8 compliant
* Python3 chokes on casting int to bytes (#24952)
But if we tell the formatter that the var is a number, it works
Just after release of 2.0.0 (in 2.0.0.1) we had a change to the API of
callbacks without bumping the API version. We added the playbook to the
arguments passed to the callbacks.
This wasn't in the Tower callback at the time. In order to prevent
breaking that callback we added a temporary hack to inspect the
callback's API to decide if we needed to call it with arguments or not.
We scheduled the hack for removal in January 2017. Since that's now
past, removing the hack.
Change signed off by matburt on the Tower side.
* Update module_utils.six to latest
We've been held back on the version of six we could use on the module
side to 1.4.x because of python-2.4 compatibility. Now that our minimum
is Python-2.6, we can update to the latest version of six in
module_utils and get rid of the second copy in lib/ansible/compat.
This version just gets the relevant paths from PluginLoader and then
uses the existing imp.find_plugin() calls in the AnsiballZ code to load
the proper module_utils.
Modify PluginLoader to optionally omit subdirectories (module_utils
needs to operate on top level dirs, not on subdirs because it has
a hierarchical namespace whereas all other plugins use a flat
namespace).
Rename snippet* variables to module_utils*
Add a small number of unittests for recursive_finder
Add a larger number of integration tests to demonstrate that
module_utils is working.
Whitelist module-style shebang in test target library dirs
Prefix module_data variable with b_ to be clear that it holds bytes data
* Unittests for some of module_common.py
* Port test_run_command to use pytest-mock
The use of addCleanup(patch.stopall) from the unittest idiom was
conflicting with the pytest-mock idiom of closing all patches
automatically. Switching to pytest-mock ensures that the patches are
closed and removing the stopall stops the conflict.
Rather than trying to enumerate tasks or track an ever changing cur_role
flag in PlayIterator, this change simply sets a flag on the last block in
the list of blocks returned by Role.compile(). The PlayIterator then checks
for that flag when the cur_block number is incremented, and marks the role
as complete if the given host had any tasks run in that role.
Fixes#20224
In order to support legacy plugins, the following two method signatures
are allowed for `CallbackBase.v2_playbook_on_start`:
def v2_playbook_on_start(self):
def v2_playbook_on_start(self, playbook):
Previously, the logic to handle this divergence checked to see if the
callback plugin being called supported an argument named `playbook`
in its `v2_playbook_on_start` method. This was fragile in a few ways:
- if a plugin author did not use the literal `playbook` to name their
method argument, their plugin would not be called correctly
- if a plugin author wrapped their `v2_playbook_on_start` method and
by doing so changed the argspec to no longer expose an argument
with that literal name, their plugin would not be called correctly
In order to continue to support both types of callback for backwards
compatibility while making the call more robust for plugin authors,
the logic can be reversed in order to have a positive check for the old
method signature instead of a positive check for the new one.
Signed-off-by: Steve Kuznetsov <skuznets@redhat.com>
Rather than repeatedly searching for tasks by uuid via iterating over
all known blocks, cache the tasks when they are added to the PlayIterator
so the lookup becomes a simple key check in a dict.
This feature changes the scalar value of `serial:` to a list, which
allows users to specify a list of values, so batches can be ramped
up (commonly called "canary" setups):
- hosts: all
serial: [1, 5, 10, "100%"]
tasks:
...
Ansible when there was a percentage that was calculated to be less than
1.0 would run all hosts as the value for a rolling update.
The error is due to the fact that Python will round a
float that is under 1.0 to 0, which will trigger the case of
0 hosts. The 0 host case tells ansible to run all hosts.
The fix will see if the percentage calculation after int
conversion is 0 and will else to 1 host.
* Unit tests exposed a problem where nested blocks did not correctly
hit rescue/always portions of parent blocks
* Cleaned up logic in PlayIterator
* Unfortunately fixing the above exposed a potential problem in the
block integration tests, where a failure in an "always" section may
always lead to a failed state and the termination of execution
beyond that point, so certain parts of the block integration test
were disabled.
- now workers passes queue to task_executor so it can send back events per item and on retry attempt
- updated result class to pass along events to strategy
- base strategy updated to forward new events to callback
- callbacks now remove 'items' on final result but process them directly when invoked per item
- new callback method to deal with retry attempt messages (also now obeys nolog)
- updated tests to match new signature of task_executor
fixes#14558fixes#14072
The full error was
======================================================================
ERROR: test_task_executor_execute (units.executor.test_task_executor.TestTaskExecutor)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/mg/src/ansible/test/units/executor/test_task_executor.py", line 252, in test_task_executor_execute
mock_action.run.return_value = dict(ansible_facts=dict())
File "/home/mg/src/ansible/lib/ansible/executor/task_executor.py", line 317, in _execute
if self._task.async > 0:
TypeError: unorderable types: MagicMock() > int()
----------------------------------------------------------------------
Experiments show that Python 2 MagicMock() > 0 is true, so I'm setting
the async property on mock_task to 1. (If I set it to 0, the test fails
anyway.)
Also making PlayContext a child class of the Playbook Base class,
which gives it access to all of the FieldAttribute code to ensure
field values are correctly typed after post_validation
Fixes#11381
Paul Belanger