| Commit message (Collapse) | Author | Age | Files | Lines |
| ... | |
| |
|
|
|
|
|
| |
Commit f0152dc3cc9a42aa4dc3c0014b4c29381e9b39d6 removed situation
where `get()` in a `GetWithoutIdMixin` based class could return `None`
Update the type-hints to no longer return `Optional` AKA `None`
|
| |
|
|
|
|
|
|
|
|
|
| |
Previously in commit 233b79ed442aac66faf9eb4b0087ea126d6dffc5 I had
used the `name` argument for `ModuleNotFoundError()`. This basically
is the equivalent of not passing any message to
`ModuleNotFoundError()`. So when the exception was raised it wasn't
very helpful.
Correct that and add a unit-test that shows we get the message we
expect.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
| |
`list()`
`as_list=False` is confusing as it doesn't explain what is being
returned. Replace it with `iterator=True` which more clearly explains
to the user that an iterator/generator will be returned.
This maintains backward compatibility with `as_list` but does issue a
DeprecationWarning if `as_list` is set.
|
| | |
|
| | |
|
| |\
| |
| | |
test: increase projects coverage
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Python 3.11 added an additional check in the argparse libary which
detected duplicate subparsers being added. We had duplicate subparsers
being added.
Make sure we don't add duplicate subparsers.
Closes: #2015
|
| |/ |
|
| |\
| |
| |
| |
| | |
Sineaggi/add-chunked-to-list-of-retryable-exceptions
Add ChunkedEncodingError to list of retryable exceptions
|
| | | |
|
| |\ \
| | |
| | | |
feat(objects): support getting project/group deploy tokens by id
|
| | | | |
|
| |\ \ \
| | | |
| | | | |
feat(user): support getting user SSH key by id
|
| | |/ / |
|
| | |/
|/|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
A common cause of issues filed and questions raised is that a user
will call a `list()` method and only get 20 items. As this is the
default maximum of items that will be returned from a `list()` method.
To help with this we now emit a warning when the result from a
`list()` method is greater-than or equal to 20 (or the specified
`per_page` value) and the user is not using either `all=True`,
`all=False`, `as_list=False`, or `page=X`.
|
| |/ |
|
| | |
|
| |
|
|
| |
This reverts commit e3035a799a484f8d6c460f57e57d4b59217cd6de.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
Create a custom `warnings.warn` wrapper that will walk the stack trace
to find the first frame outside of the `gitlab/` path to print the
warning against. This will make it easier for users to find where in
their code the error is generated from
|
| |
|
|
|
|
|
|
|
|
|
|
| |
When pagination occurs python-gitlab uses the URL provided by the
GitLab server to use for the next request.
We had previously set the GitLab server configuraiton to say its URL
was `http://gitlab.test` which is not in DNS. Set the hostname
in the URL to `http://127.0.0.1:8080` which is the correct URL for the
GitLab server to be accessed while doing functional tests.
Closes: #1877
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
The usage of deleting was incorrect according to the current API.
Remove custom `delete()` method as not needed.
Add tests to show it works with labels needing to be encoded.
Also enable the test_group_labels() test function. Previously it was
disabled.
Add ability to do a `get()` for group labels.
Closes: #1867
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Create a new ArrayAttribute class. This is to indicate types which are
sent to the GitLab server as arrays
https://docs.gitlab.com/ee/api/#array
At this stage it is identical to the CommaSeparatedListAttribute class
but will be used later to support the array types sent to GitLab.
This is the second step in a series of steps of our goal to add full
support for the GitLab API data types[1]:
* array
* hash
* array of hashes
Step one was: commit 5127b1594c00c7364e9af15e42d2e2f2d909449b
[1] https://docs.gitlab.com/ee/api/#encoding-api-parameters-of-array-and-hash-types
Related: #1698
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
| |
This name more accurately describes what the type is. Also this is the
first step in a series of steps of our goal to add full support for
the GitLab API data types[1]:
* array
* hash
* array of hashes
[1] https://docs.gitlab.com/ee/api/#encoding-api-parameters-of-array-and-hash-types
|
| |
|
|
|
| |
Add a test to make sure that all of the `gitlab/v4/objects/` files are
imported in `gitlab/v4/objects/__init__.py`
|
| | |
|
| | |
|
| |
|
|
|
| |
In the `responses` library the usage of `match_querystring` is
deprecated. Convert to using `match`
|
| |
|
|
|
| |
Convert all usage of the `httpmock` library to using the `responses`
library.
|
| |\
| |
| | |
chore: rename `gitlab/__version__.py` to `gitlab/_version.py`
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
It is confusing to have a `gitlab/__version__.py` because we also
create a variable `gitlab.__version__` which can conflict with
`gitlab/__version__.py`.
For example in `gitlab/const.py` we have to know that
`gitlab.__version__` is a module and not the variable due to the
ordering of imports. But in most other usage `gitlab.__version__` is a
version string.
To reduce confusion make the name of the version file
`gitlab/_version.py`.
|
| | |
| |
| |
| |
| |
| | |
Convert from httmock to responses in test_mixins_methods.py
This leaves only one file left to convert
|
| |/
|
|
|
|
| |
When doing the functional tests use the new function
`projects.transfer` instead of the deprecated function
`projects.transfer_project()`
|
| | |
|
| | |
|
| |\
| |
| | |
fix: use url-encoded ID in all paths
|