Change:
- This was removed in 2014 in 122a7021bc.
- The option still exists and is enabled by default and can lead to user
confusion when people aren't expecting packages (or updated
dependencies for it) to get installed and they do.
- Add the option documentation back with a few notes to make it clear what
is happening.
Test Plan:
N/A, no code change, just documentation
Tickets:
- Refs #69497
Signed-off-by: Rick Elrod <rick@elrod.me>
* Fix sanity errors
Signed-off-by: Rick Elrod <rick@elrod.me>
* Enable installing collections from git repositories
* Add tests for installing individual and multiple collections from git repositories
* Test to make sure recursive dependencies with different syntax are deduplicated
* Add documentation
* add a changelog
* Skip Python 2.6
* Only fail if no collections are located in a git repository
Add support for a 'type' key for collections in requirement.yml files.
Update the changelog and document the supported keys and allowed values for the type.
Add a note that the collection(s) in the repo must contain a galaxy.yml
* Add a warning about embedding credentials in SCM URLs
* Update with review suggestions
* suppress sanity compile failure for Python 2.6
Gets rid of the unknown field names in the return data.
Allows the plugin return docs to format under the new docs pipeline.
* Expect that package_facts will pass return-syntax-error now.
* `meta/` directory in collections
* runtime metadata for redirection/deprecation/removal of plugin loads
* a compatibility layer to keep existing content working on ansible-base + collections
* a Python import redirection layer to keep collections-hosted (and otherwise moved) content importable by things that don't know better
* supported Ansible version validation on collection loads
* Add multipart/form-data functionality
* Fix some linting issues
* Fix error message
* Allow filename to be provided with content
* Add integration test
* Update examples
* General improvements to multipart handling
* Use prepare_multipart for galaxy collection publish
* Properly account for py2 vs py3, ensuring no max header length
* Address test assumptions
* Add unit tests
* Add changelog
* Ensure to use CRLF instead of NL
* Ignore line-endings in fixture
* Consolidate code, add comment
* Bump fallaxy container version
* ci_complete
* Fix filedescriptor out of range in select() when running commands
* Simplify the run_command() code
Now that we're using selectors in run_command(), we can simplify some of
the code.
* Use fileobj.read() instead of os.read()
* No longer use get_buffer_size() as we can just slurp all of the data
instead.
Also use a simpler conditional check of whether the selector map is
empty
Co-authored-by: Toshio Kuratomi <a.badger@gmail.com>
Left hand side slicing is confusing and slower but maybe more memory
efficient in some circumstances. There is one case where it adds to
code safety: when it's used to substitute a different list in place of a
slice of the original list and the original list could have been bound
to a different variable in some other code. (The most likely case of
this is when it's a global variable and some other code might import
that variable name).
Because of the confusion factor we think it should only be used for the
safety case or where it's been benchmarked and shown to have some sort
of documentatble improvement. At the moment, only one piece of code
falls into those categories so this PR removes all the other instances
of left hand side slicing.
Change:
Rather than hardcoding .pyo and .pyc, filter on all BLACKLIST_EXTS in
the non-legacy logic of PluginLoader (_find_fq_plugin). The two harcoded
extensions are part of BLACKLIST_EXTS already and this simply adds the
rest of the blacklisted extensions to the check.
In addition, check .endswith() instead of an exact match of the suffix,
like everywhere else that uses BLACKLIST_EXTS. This allows for
blacklisting, for example, emacs's backup files which can appear after
any extension, leading to things like `foo.py~`.
Test Plan:
Ran `ansible-playbook` against a collection where a `foo.py~` module was
getting executed instead of `foo.py` which also appeared in the same
directory. `foo.py~` is no longer executed.
Tickets:
Fixes#22268
Refs #27235
Signed-off-by: Rick Elrod <rick@elrod.me>
With https://github.com/pallets/jinja/pull/1190 merged our short-circuit
is no longer valid (has it ever been?) as now data like ' True ' may go
through our ansible_native_concat function as opposed to going through
intermediate call to Jinja2's native_concat before. Now we need to always
send data through literal_eval to ensure native types are returned.
* remove azure extras and extras_require support
* Since Azure will be collectionized, the requirements will float more frequently than Ansible releases; the Azure collection needs to host the requirements now.
* Removed the dynamic extras support as well, since Azure was the only thing using it. If we need it again, it's easy to pull back from history.
* Mark azure-requirements as orhpaned.
This keeps the docs around so that existing links from old test runs remain valid.
Co-authored-by: Matt Clay <matt@mystile.com>
* Rename `tests` test to match plugin type.
* Rename `test_infra` test to avoid confusion.
This test target is not a test for test plugins.
* Rename `vars_prompt` test to avoid confusion.
* Update sanity ignores.
* Create a dedicated windows-minimal test target.
The windows-minimal target is a copy of the win_ping test, taking the place of that test as the windows minimal test run on multiple python versions.
It includes a private copy of the win_ping module so it will work after migration.
This will keep tests passing during the migration prep process.
* Update sanity ignores.
* Split out cache plugin unit tests.
* Rename unit tests to match code under test.
* Relocate unit test code to match code under test.
* Another rename.
* Update sanity ignores.
