8d93ba9120
* Tag return value docs if they are a dict (and not str/None). * Try to parse return docs as YAML. * Properly dump return values in ansible-doc. * Adjust plugin formatter. * Add changelog fragment. * Don't add 'default' for return values. * Fix plugin_formatter. * Only try to parse return docs if they are still a string. * Add tests. * Warn if RETURN cannot be parsed. * Adjust tests. Also test for warning. * if -> elif (otherwise EXAMPLE will be parsed too). * Always parse return documentation, and fail if it is invalid YAML. * Polishing. * Mostly re-enable ansible-doc tests. Listing from the local collection seems to be somewhat broken. I assume this is why the test was disabled. * Lint and make tests work with Python 2. * Keep FQCNs in plugins (not modules), i.e. restore previous state. |
5 years ago | |
---|---|---|
.. | ||
backport | 5 years ago | |
build_library | 5 years ago | |
shippable | 5 years ago | |
tests | 5 years ago | |
ticket_stubs | 5 years ago | |
README.md | 5 years ago | |
ansible-profile | 5 years ago | |
build-ansible.py | 5 years ago | |
deprecated_issue_template.md | 6 years ago | |
env-setup | 5 years ago | |
env-setup.fish | 5 years ago | |
fix_test_syntax.py | 7 years ago | |
get_library.py | 6 years ago | |
report.py | 5 years ago | |
return_skeleton_generator.py | 5 years ago | |
test-module | 5 years ago | |
test-module.py | 5 years ago |
README.md
'Hacking' directory tools
env-setup
The 'env-setup' script modifies your environment to allow you to run ansible from a git checkout using python 2.6+. (You may not use python 3 at this time).
First, set up your environment to run from the checkout:
$ source ./hacking/env-setup
You will need some basic prerequisites installed. If you do not already have them and do not wish to install them from your operating system package manager, you can install them from pip
$ easy_install pip # if pip is not already available
$ pip install -r requirements.txt
From there, follow ansible instructions on docs.ansible.com as normal.
test-module.py
'test-module.py' is a simple program that allows module developers (or testers) to run a module outside of the ansible program, locally, on the current machine.
Example:
$ ./hacking/test-module.py -m lib/ansible/modules/commands/command.py -a "echo hi"
This is a good way to insert a breakpoint into a module, for instance.
For more complex arguments such as the following yaml:
parent:
child:
- item: first
val: foo
- item: second
val: boo
Use:
$ ./hacking/test-module.py -m module \
-a '{"parent": {"child": [{"item": "first", "val": "foo"}, {"item": "second", "val": "bar"}]}}'
return_skeleton_generator.py
return_skeleton_generator.py helps in generating the RETURNS section of a module. It takes JSON output of a module provided either as a file argument or via stdin.
fix_test_syntax.py
A script to assist in the conversion for tests using filter syntax to proper jinja test syntax. This script has been used to convert all of the Ansible integration tests to the correct format for the 2.5 release. There are a few limitations documented, and all changes made by this script should be evaluated for correctness before executing the modified playbooks.