chore: upgrade code style for python3.10

[jonburdo]

What this PR does / why we need it:

Primarily updates code style after upgrading the min Python version to 3.10 to change;

• Union[T, U] → T | U
• Optional[T] → T | None

as suggested here: #282 (comment)

Also includes a second commit to handle such union types in convert_value utility with a test.

Checklist:

• Docs included if any changes are user facing

[google-oss-prow]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign kramaranya for approval. For more information see the Kubernetes Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[Copilot]

Pull request overview

Updates the codebase to use Python 3.10+ type-hint syntax (PEP 604 unions) and enhances the Katib/Kubernetes optimizer utility to better handle Optional[T] vs T | None type metadata when converting string values.

Changes:

• Enable Ruff/pyupgrade enforcement of T | U and T | None by removing the ignores for UP007/UP045.
• Convert many public and internal type hints from Optional[...] / Union[...] to PEP 604 | syntax across trainer/optimizer/common modules.
• Update convert_value to unwrap union-with-None types and add a targeted unit test.

Reviewed changes

Copilot reviewed 31 out of 31 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
pyproject.toml	Removes UP007/UP045 ignores so Ruff enforces modern union/optional syntax.
kubeflow/trainer/types/types.py	Rewrites many dataclass field annotations to `T
kubeflow/trainer/test/common.py	Updates shared test utility dataclass annotations to PEP 604 style.
kubeflow/trainer/options/kubernetes.py	Updates option call signatures/types to PEP 604 unions.
kubeflow/trainer/options/common.py	Updates common option call signatures/types to PEP 604 unions.
kubeflow/trainer/backends/localprocess/utils.py	Updates helper function annotations to PEP 604 optionals.
kubeflow/trainer/backends/localprocess/types.py	Updates localprocess model annotations to PEP 604 optionals.
kubeflow/trainer/backends/localprocess/job.py	Updates command parameter union annotation to PEP 604 style.
kubeflow/trainer/backends/localprocess/backend.py	Updates backend API method annotations to PEP 604 unions/optionals.
kubeflow/trainer/backends/kubernetes/utils.py	Updates Kubernetes backend utility function annotations to PEP 604 style.
kubeflow/trainer/backends/kubernetes/backend_test.py	Updates test helpers’ annotations to PEP 604 style.
kubeflow/trainer/backends/kubernetes/backend.py	Updates Kubernetes backend API method annotations to PEP 604 unions/optionals.
kubeflow/trainer/backends/container/types.py	Updates container backend config model annotations to PEP 604 optionals.
kubeflow/trainer/backends/container/runtime_loader.py	Updates runtime loader helper return types to PEP 604 optionals.
kubeflow/trainer/backends/container/backend_test.py	Updates test adapter/type annotations to PEP 604 optionals/unions.
kubeflow/trainer/backends/container/backend.py	Updates container backend API method annotations to PEP 604 unions/optionals.
kubeflow/trainer/backends/container/adapters/podman.py	Updates adapter method signatures to PEP 604 optionals.
kubeflow/trainer/backends/container/adapters/docker.py	Updates adapter method signatures to PEP 604 optionals.
kubeflow/trainer/backends/container/adapters/base.py	Updates abstract adapter signatures to PEP 604 optionals.
kubeflow/trainer/backends/base.py	Updates base backend interface annotations to PEP 604 unions/optionals.
kubeflow/trainer/api/trainer_client.py	Updates client API annotations to PEP 604 unions/optionals.
kubeflow/optimizer/types/search_types.py	Updates search space type hints to PEP 604 unions.
kubeflow/optimizer/types/optimization_types.py	Updates optimization types to PEP 604 unions/optionals.
kubeflow/optimizer/types/algorithm_types.py	Updates algorithm types to PEP 604 optionals.
kubeflow/optimizer/backends/kubernetes/utils_test.py	Adds unit tests covering union/optional handling in convert_value.
kubeflow/optimizer/backends/kubernetes/utils.py	Improves convert_value union-with-None handling; updates return type hints elsewhere.
kubeflow/optimizer/backends/kubernetes/backend.py	Updates optimizer backend method annotations to PEP 604 optionals.
kubeflow/optimizer/backends/base.py	Updates optimizer base interface annotations to PEP 604 optionals.
kubeflow/optimizer/api/optimizer_client.py	Updates optimizer client API annotations to PEP 604 optionals.
kubeflow/common/utils.py	Updates common helper signature annotations to PEP 604 optionals.
kubeflow/common/types.py	Updates shared config model annotations to PEP 604 optionals.

Comments suppressed due to low confidence (2)

kubeflow/trainer/backends/container/adapters/podman.py:214

• list_containers accepts filters: ... | None = None but immediately calls filters.items(), which will raise when filters is omitted/None; either make filters required (no default) or guard against None (e.g., treat None as an empty dict).

    def list_containers(self, filters: dict[str, list[str]] | None = None) -> list[dict]:
        """List Podman containers with optional filters."""
        # Work-around for https://github.com/containers/podman-py/issues/542
        for k, v in filters.items():
            if len(v) == 1:

kubeflow/trainer/backends/kubernetes/backend_test.py:226

• pip_index_urls is annotated as list[str] | None but the function unconditionally indexes/slices it, so passing None will raise; either drop | None from the type or handle None by falling back to constants.DEFAULT_PIP_INDEX_URLS inside the function.

def get_custom_trainer(
    env: list[models.IoK8sApiCoreV1EnvVar] | None = None,
    pip_index_urls: list[str] | None = constants.DEFAULT_PIP_INDEX_URLS,
    packages_to_install: list[str] = ["torch", "numpy"],
    image: str | None = None,
) -> models.TrainerV1alpha1Trainer:
    """
    Get the custom trainer for the TrainJob.
    """
    pip_command = [f"--index-url {pip_index_urls[0]}"]
    pip_command.extend([f"--extra-index-url {repo}" for repo in pip_index_urls[1:]])

[Copilot]

In convert_value, the boolean parsing lowercases the input but compares against a capitalized token ("True"), so values like "True"/"true" will parse as False; compare against lowercase tokens (e.g., "true") and consider adding a test case for "true"/"false" in addition to "1"/"0".

[coveralls]

Pull Request Test Coverage Report for Build 21925778716

Details

• 151 of 154 (98.05%) changed or added relevant lines in 28 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+5.2%) to 73.144%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
kubeflow/optimizer/backends/kubernetes/utils_test.py	12	15	80.0%

Totals
Change from base Build 21919105330:	5.2%
Covered Lines:	3045
Relevant Lines:	4163

---

💛 - Coveralls

[jonburdo]

For the first commit I just removed these and ran:

uv run ruff check --fix
uv run ruff format

[jonburdo]

get_origin isn't perfectly consistent across python versions

$ python3.13 - <<'EOF'
from typing import get_origin, Optional
print(get_origin(int | None))
print(get_origin(Optional[int]))
EOF

<class 'types.UnionType'>
typing.Union


$ python3.14 - <<'EOF'
from typing import get_origin, Optional
print(get_origin(int | None))
print(get_origin(Optional[int]))
EOF

<class 'typing.Union'>
<class 'typing.Union'>

[astefanutti]

That's great, thanks @jonburdo!

/lgtm