Upgrade to microgrid client v0.7

[llucax]

Use the new microgrid client ComponentId and MicrogridId.

[llucax]

Draft to avoid merging accidentally because it depends on frequenz-floss/frequenz-client-microgrid-python#122 and a release, but it is really ready for a review.

[Copilot]

Pull Request Overview

This PR upgrades the project to use microgrid client v0.7 by replacing many usages of raw integer component IDs with the new ComponentId type and updating dependency/version references. Key changes include:

• Updating dependency constraints in pyproject.toml and documentation in RELEASE_NOTES.md and mkdocs.yml.
• Replacing int‑based component IDs with ComponentId in multiple modules throughout the SDK.
• Adjusting type annotations, function signatures, and caches accordingly to support the new microgrid client types.

Reviewed Changes

Copilot reviewed 67 out of 67 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
pyproject.toml	Updated dependency for frequenz-client-microgrid to use a git reference for v0.7.
RELEASE_NOTES.md	Documented upgrade to v0.7 and changes in how IDs are handled.
mkdocs.yml	Commented out the previous version’s documentation objects inventory for microgrid client.
Multiple SDK modules	Converted int types to ComponentId in function signatures, caches, documentation, and type hints.
benchmarks/power_distribution/power_distributor.py	Updated benchmark function signatures to use ComponentId instead of int.

Comments suppressed due to low confidence (3)

pyproject.toml:32

• Consider replacing the temporary git dependency with the official v0.7 release reference once it is published to ensure reproducible builds.

#  "frequenz-client-microgrid >= 0.7.0, < 0.8.0",

src/frequenz/sdk/microgrid/_power_distributing/_component_managers/_battery_manager.py:188

• [nitpick] Since the method now returns ComponentId instead of int, please verify that all dependent modules and downstream usages are updated accordingly.

def component_ids(self) -> collections.abc.Set[ComponentId]:

benchmarks/power_distribution/power_distributor.py:40

• [nitpick] Updating the batteries type from int to ComponentId may affect benchmark logic; please ensure these changes are validated across all benchmark tests.

async def send_requests(batteries: set[ComponentId], request_num: int) -> list[Result]:

[llucax]

Ready for review now.

[shsms]

You could make a few more commits; it's fine if they don't pass the tests individually. Or at least make fixup commits for review and squash before merging. Otherwise, it is too hard to keep track of where you are if you have to take several breaks while reviewing.

[llucax]

I can only see breaking the commit on a per-file basis, but that doesn't really bring much, if you want to resume the review after a while and splitting on a file-by-file basis is enough, then you can just use GitHub feature of marking files as read in the review.

If you think there was another way to split the commit, I'm interested!

[shsms]

I can only see breaking the commit on a per-file basis, but that doesn't really bring much, if you want to resume the review after a while and splitting on a file-by-file basis is enough, then you can just use GitHub feature of marking files as read in the review.

I see, didn't know that. I'll try.

[shsms]

Now an amd test is also failing :-|

[llucax]

Oh dear, failing in Python 3.12 which I'm not using to test locally. Is the component graph tests, so I wonder if it is the set[int] ordering coming to bite us in a different way in 3.12 (maybe I missed some fix that worked in 3.11). I will need to have a look.