If package installation has an error after the package is install (e.g.
when running tags), then there will be output on stderr and the fallback
regex will match; however, since pkg_add exited non-zero, changed is
never added as a key to the dictionary. As a result the code at the end
of main that checks if anything has changed raises a KeyError.
* pythonific!! no camel cases, bitte
* simplified iface attributes parsing
* some improvements, passing tests
* simplified set_interface_option()
* further simplifications
* remove unreachable stmt
* pythonified a file open
* added changelog fragment
* adjustment per PR
* PR: fixed the auto- case
* PR: added testcase and chglog frag for the misleading change report
* extra line removed
* integration is not destructive
* Adding "Id" to the add user function
Some implementations of Redfish (e.g. the one in Cisco's CIMC) seem to require the id of the new user for account creation.
I'm not that firm with Python but lines 982 and 983 should fix it.
* changed indention
* created changelog fragment
* Update changelogs/fragments/3343-redfish_utils-addUser-userId.yml
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update change type
* supplemented the description of the ID parameter
* Update plugins/modules/remote_management/redfish/redfish_command.py
Co-authored-by: Felix Fontein <felix@fontein.de>
Co-authored-by: Felix Fontein <felix@fontein.de>
* Initial commit
* Adding changelog fragment
* Further refactoring
* Fixing bad copy/paste and adding task for psutil >= 5.7.0 install
* Inverting psutil installation order to reduce duplication
* Optimizing regex compilation
* Adds few more gitlab group options
* Update plugins/modules/source_control/gitlab/gitlab_group.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Removes default for new string options
* Removes default from argument_spec
* Adds changelog fragment
* Update plugins/modules/source_control/gitlab/gitlab_group.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group.py
Co-authored-by: Felix Fontein <felix@fontein.de>
Co-authored-by: Maik Schueller <maik.schueller@sap.com>
Co-authored-by: Felix Fontein <felix@fontein.de>
* Added redis_data_info module
Added:
- redis_data_info module and suggested 'exists' return flag.
- module_utils for redis with a base class that handles database connections.
- inhereited unit tests and added some new ones for the exit flag
* Docfix and sanity
* typo
* Suggested doc changes and ssl option
* TLS and validate_certs fix
* Set support_check_mode for info plugin
* Docfix and import errors
* Redis versioning Fix
* version bump and append fixes
* django_manage - fix fixures
* docs formatting adjustments
* param apps also in need of splitting
* oops, the splitted version was not being properly added to the command args
* added changelog fragment
* check for None
* moving to shlex.split()
* Update changelogs/fragments/3334-django_manage-split-params.yaml
Co-authored-by: Felix Fontein <felix@fontein.de>
Co-authored-by: Felix Fontein <felix@fontein.de>
* Some vendors surround header etag with quotes, which need to be cleaned before sending a patch
* Minor change fragment
* Add etag strip quote option
* Rebase
* Cleanup fragment
* Apply suggestions from code review
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/remote_management/redfish/redfish_command.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Description update
* Update plugins/modules/remote_management/redfish/redfish_config.py
Co-authored-by: Felix Fontein <felix@fontein.de>
Co-authored-by: Kyle Williams <kyle.williams@thetradedesk.com>
Co-authored-by: Felix Fontein <felix@fontein.de>
* - fix to issue 3041
- add func to work with user lists
- add func to set members to the ones give
* Added version_added to new parameter
* fixed elements in definition of gitlab_users nad wrong import in gitlab_users
* linter issues fixed
* added list elelements to argument_spec
* More whitspeaces for the linter
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* adapted changelog
* removed bugfix (other PR), changes due to review
* changed input handling according to review
* Fixed test findings
* Added list of dict to allow for specifying user/access_level tuples
* corrected doc section
* fixed parameter definitions
* removed strange additional import
* Update changelogs/fragments/3041-gitlab_x_members_fix_and_enhancement.yml
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_project_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* fixed required if
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Added suggestions from PR
* fixed identation problem
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Zainab Alsaffar <za5775@rit.edu>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Zainab Alsaffar <za5775@rit.edu>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Recommended changes from discussionst
* Fixed issues from automatic tests
* added missing metaclass due to test finding
* added integration tests
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* fixed optimization for only one user
* Reverted gitlab_project_members to original version - changes will be done in a separate branch
* added examples for new functionality
* - fixed changelog after reverting gitlab_project_memebers
- fully reverted gitlab_project_members
* Fixed error handling: when single users are not updateable in bulk mode the exception should not stop the code flow but document the problem in the result.
* Better error handling
* on error give username, not gitlab numeric userid
* Fixed broken check_mode
* Update plugins/modules/source_control/gitlab/gitlab_group_members.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Change from review
Co-authored-by: Max Bidlingmaier <Max-Florian.Bidlingmaier@sap.com>
Co-authored-by: Felix Fontein <felix@fontein.de>
Co-authored-by: Zainab Alsaffar <za5775@rit.edu>
* zfs.py: treated received properties as local and added diff mode support
If you use "zfs set" to explicitly set ZFS properties, they are marked as from source "local". If ZFS properties are implicitly set by using "zfs send" and "zfs receive", for example as part of a template based installation, they are marked as from source "received". But as there is no technical difference between both types of them, the “received” ZFS properties should also be considered “local”. Otherwise Ansible would detect changes, which aren’t actual changes. Therefore I changed line 202/207 to reflect this.
For us it’s quite important, that Ansible modules support the diff mode in order to qualify changes. Therefore I added some code lines to address this.
* added changelog fragment for PR #502
* fixed typos in changelog fragment for PR #502
* minor changes in changelog fragment for PR #502
* added link to pull request in changelog fragment for PR #502
* extended the diff data structure to always include the name of the zfs filesystem
* added code to also maintain the diff data structure after a change
* reverted back some code lines for better code readability
* added an extra dict in the diff data structure to hold the zfs properties
* udm_dns_record: Fix handling of PTR records (#3244)
Before, it was not possible to manage PTR records in Univention DNS,
due to broken zone lookups and improper used parameters of the object.
This patch fixes the PTR handling, allowing both v4 and v6 entries.
* udm_dns_record: [doc] add changelog fragment
* udm_dns_record: [fix] validation errors
* udm_dns_record: import ipaddress module conditionally (#3244)
* udm_dns_record: fix sanity check error, improve doc (#3244)
* udm_dns_record: Improve changes to meet community standards (#3244)
* snap - improved error handling
* added changelog fragment
* removed experiments left overs
* rolled back the smaller list of params for commands other than install
* Fixed incorrect VMID: cloning to an existing VM
During a cloning operation, if the destination VM already exists the VMID returned is not correct.
The VMID returned should be that of the destination VM and not that of the source VM (consistent with line 1230).
A playbook that relies on the returned VMID, for example, to perform other operations on the destination VM, will not work properly if it is unexpectedly interrupted.
* Add files via upload
* moved 3266-vmid-existing-target-clone.yml to changelogs/fragments/
replaced line separator CRLF -> LF
* storing vmid list in variable to avoid multiple API calls
* nmcli: Disallow Wi-Fi options not supported by nmcli
By querying nmcli directly
* Added changelog fragment
* Added tests
* Simplify `get_available_options()`
* Update changelogs/fragments/3141-disallow-options-unsupported-by-nmcli.yml
Co-authored-by: Felix Fontein <felix@fontein.de>
* Remove redundant `802-11-wireless` settings from test show outputs
* Update `mocked_wireless_create(mocker)`
* Update plugins/modules/net_tools/nmcli.py
Co-authored-by: Ajpantuso <ajpantuso@gmail.com>
* Address comment re. creating function & use nmcli naming conventions
I.E. `setting`.`property` = `value`
```
nmcli> help set
set [<setting>.<prop> <value>] :: set property value
This command sets property value.
Example: nmcli> set con.id My connection
```
* Added `ignore_unsupported_suboptions` option & improved `wifi(_sec)` doc
* Corrected pep8 issues
```
ERROR: Found 2 pep8 issue(s) which need to be resolved:
ERROR: plugins/modules/net_tools/nmcli.py:342:161: E501: line too long
(236 > 160 characters)
ERROR: plugins/modules/net_tools/nmcli.py:359:161: E501: line too long
(237 > 160 characters)
```
* Fixed remaining sanity check issues and added even more docs
* No need to split Note
* Update plugins/modules/net_tools/nmcli.py
3.5.0 has already been released.
Co-authored-by: Felix Fontein <felix@fontein.de>
* Followed uniformity guideline for format macros from Ansible's dev guide
* Addressed comment
https://github.com/ansible-collections/community.general/pull/3141#discussion_r689098383
* Documentation cleanup continuation
* Replace `NM_SETTING_*`s having a description with their numeric value
* Splitting up long paragraphs.
Also removed `wifi`.`seen-bssids` as it "`is only meant for reading`"
* Addressed remaining comments and clarified `wake-on-lan` note
* Update plugins/modules/net_tools/nmcli.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/net_tools/nmcli.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/net_tools/nmcli.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Update plugins/modules/net_tools/nmcli.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* Finishing addressing documentation comments.
* Update plugins/modules/net_tools/nmcli.py
Co-authored-by: Ajpantuso <ajpantuso@gmail.com>
* Update plugins/modules/net_tools/nmcli.py
Co-authored-by: Ajpantuso <ajpantuso@gmail.com>
* Update nmcli.py
* Added wifi-related `list` type options to `settings_type` method
* Moved `edit_commands` `execution` logic into its own method
* Move `unsupported_property` deletion into `main` function
* Missing `.items()`
* Resolved missing proper `nmcli conn edit` arguments
* Resolve pylint issue `dangerous-default-value`
Co-authored-by: Felix Fontein <felix@fontein.de>
Co-authored-by: Ajpantuso <ajpantuso@gmail.com>
Co-authored-by: David Hummel <dhummel@Fingerling>
* module_helper - implemented classmethod to start the module plus minor change
* rolled back the __changed__() method
* added changelog fragment
* Update plugins/module_utils/mh/base.py
Co-authored-by: Sviatoslav Sydorenko <wk.cvs.github@sydorenko.org.ua>
* no capt Piccards allowed in the base class
* removed extra piccards
Co-authored-by: Sviatoslav Sydorenko <wk.cvs.github@sydorenko.org.ua>
* slack - minor refactoring and pythonifying
* added changelog fragment
* Update changelogs/fragments/3205-slack-minor-refactor.yaml
Co-authored-by: Felix Fontein <felix@fontein.de>
Co-authored-by: Felix Fontein <felix@fontein.de>
* refactor to vdo
* adjusted if condition
* added changelog fragment
* Update plugins/modules/system/vdo.py
Co-authored-by: Felix Fontein <felix@fontein.de>
* adjustements per the PR
* more occurrences of bool compared with yes or no
* Update changelogs/fragments/3191-vdo-refactor.yml
Co-authored-by: Felix Fontein <felix@fontein.de>
Co-authored-by: Felix Fontein <felix@fontein.de>
* ini_file - prepare for fixing #273
- restructure tests
- fix error message call: fail_json() takes 1 positional argument but 2 were given
* ini_file - multiple values for one option (#273)
- add module option 'exclusive' (boolean) for the abbility to add
single option=value entries without overwriting existing options
with the same name but different values
- add abbility to define multiple options with the same name but
different values
* ini_file - add more tests for ini_file
* ini_file - fix sanity tests
* apply suggested changes:
- rename 03-regressions.yml to 03-encoding.yml
- fix typos
- fix documentation
* apply suggested changes:
- test errors also for result is failed
* apply suggested changes:
- make state=absent also work with module option exclusive
- add more tests for state=absent and module option exclusive
* fix sanity test:
- 02-values.yml:251:9: hyphens: too many spaces after hyphen
* apply proposed changes
* apply proposed changes from review
- adjust version_added to 3.6.0
- small syntax change in changelog fragment
* Initial Commit
* Comparing with tar file checksums rather than tar header checksums
* Added changelog fragment
* Revert "Comparing with tar file checksums rather than tar header checksums"
This reverts commit bed4b17107.
* Restricting idempotency tests by format
* Applying review suggestions
* zypper: support transactional-updates
- Check if transactional updates are in use by checking for the
existence of /var/lib/misc/transactional-update.state
- Prefix zypper-commands with /sbin/transactional-update --continue --drop-if-no-change --quiet run
if this is the case
fixesansible-collections/community.general#3159
* re-add get_bin_path for executables
* fix typo
* Support older version of psutil (RHEL7 and RHEL6)
The psutil python module is a true mess, they changed the API twice. The function arguments, as well as the objects that are returned.
The documentation does not make it clear which version supports what so the safest implementation is this waterfall approach.
A better approach would be to inspect the returned information, rather than trust a version, but that would not be any more efficient.
In the end it is better to have something that at least works out-of-the-box on all platforms than something that requires custom updates to system packages before it works as expected. Especially for something as basic as `pids`.
* A little bit more concise
* Apply suggestions from code review
* Add changelog fragment.
Co-authored-by: Felix Fontein <felix@fontein.de>
When a package name contains special characters (e.g. "g++"), they are
interpreted as part of the regexp.
This can lead to a crash with an error in the python re module, for
instance with "g++":
sre_constants.error: multiple repeat
Fix this by escaping the package name.
Co-authored-by: Baptiste Jonglez <git@bitsofnetworks.org>
* nmcli: use `stdin` for setting private `wifi_sec` options
I.E.:
* `802-11-wireless-security.leap-password`
* `802-11-wireless-security.psk`
* `802-11-wireless-security.wep-key0`
* `802-11-wireless-security.wep-key1`
* `802-11-wireless-security.wep-key2`
* `802-11-wireless-security.wep-key3`
* Changelog fragement formatting.
* Update changelogs/fragments/3160-pass-wifi-secrets-via-stdin-to-nmcli-module.yml
Co-authored-by: Felix Fontein <felix@fontein.de>
* Make `wifi_sec_secret_options()` into a constant
* Minor cleanup
`'set ' + key + ' ' + value`
=>
`'set %s %s' % (key, value)`
* Change `casing`
* Change `WIFI_SEC_SECRET_OPTIONS` from `list` to `tuple`
* Update `edit_connection()` to not reset `edit_commands`
It will just re`set` them if `edit_connection()` is called more than
once.
* Do not call `edit_connection()` if `connection_update(*)` fails
* Fixed `pep8` issue `E713` in tests
`test for membership should be 'not in'`
* Simplify `create_connection()`/`modify_connection()` logic
* `WIFI_SEC_SECRET_OPTIONS`=>`SECRET_OPTIONS`, options are prefixed
* Moved `if key in self.SECRET_OPTIONS` into `if value is not None` check
We don't need to do anything is the value is None
Co-authored-by: Felix Fontein <felix@fontein.de>
* Fixed `wifi_sec` option changes are not detected
Also updated `docs` URL and formatting to match that of the `wifi`
option
* Removed extraneous `appends` to `cmd` in `connection_update`
These really should have only been added to `connection_options` whose
return values get `extended` onto `cmd`
