🤖 I have created a release *beep* *boop*
---


<details><summary>2.54.1-SNAPSHOT</summary>

### Updating meta-information for bleeding-edge SNAPSHOT release.
</details>

---
This PR was generated with [Release
Please](https://github.com/googleapis/release-please). See
[documentation](https://github.com/googleapis/release-please#release-please).

Co-authored-by: release-please[bot] <55107282+release-please[bot]@users.noreply.github.com>

LTS 8 release branch

This PR manually update arrow.version to v18.2.0

Fix a few links that does not work propely.

Fixes #2470

### Approach
#### In `showcase/`
* Revamp the generation and verification scripts used by
`-Penable-golden-tests` and `-Pupdate` to build and use the Hermetic
Build Docker image

### Effects
 * New `samples/` folder added in `showcase`
 * Reformatted all of grpc, proto and gapic libraries in `showcase/`

---------

Co-authored-by: cloud-java-bot <[email protected]>

This PR eliminates the issue where call credentials get attached twice
to a RPC in DirectPath cases. Particularly, when user credentials get
used, the problem causes the RPC to fail due to the duplication of the
quota project ID (internal-only context:
[b/364288002](http://b/364288002)).

The approach is to strip the credentials in the callOptions returned by
the `GrpcCallContext` if the `TransportChannel` is DirectPath. The side
effect is that users won't be able to configure call credentials via the
`ApiCallContext` if DirectPath is used.

We think this is acceptable because:

1. Users can still configure the credentials via
`TransportChannelProvider.withCredentials()`. At a higher level, this is
done by configuring the `CredentialsProvider` in the `StubSettings` or
the `ServiceOptions`.
2. As of now, DirectPath has its own special authentication flow, in
which the service account attached to the GCE VM or GKE Pod will be
used. Although in some special cases, the call credentials will be used
to authenticate the client's identity, the peculiar nature of DirectPath
should justify us limiting the flexibility of how call credentials can
be configured in this case.

## Tested DirectPath using Spanner 
Headers sent
```
[:authority: spanner.googleapis.com, :path: /google.spanner.v1.Spanner/BatchCreateSessions, :method: POST, :scheme: https, content-type: application/grpc, te: trailers, user-agent: spanner-java/6.86.0 grpc-java-netty/1.69.0, ..., grpc-accept-encoding: gzip, authorization: Bearer ya29.****, ..., authorization: Bearer 1234, grpc-timeout: 56962080u]
```
Bearer token is sent twice (first `ya29.***` value is valid and second
`1234` is invalid). The second one was attached by customizing the
ApiCallContext to send an invalid CallCredentials as part of the
CallOptions. The call still succeeded as the first Bearer token in the
Metadata is used.

observed in
https://github.com/googleapis/sdk-platform-java/actions/runs/13730578942/job/38406676526?pr=3684

#3568 is merged.

)

#3568 updated the
name `showcase` module to `java-showcase`. This PR updates a few places
that still reference the old name to the new name.

fixes: #3277

I run the flaky test locally 1000 times but couldn't reproduce the
issue. Despite that, reduce the LRO timeout values in the Showcase test
`ITLongRunningOperation.testHttpJson_LROUnsuccessfulResponse_exceedsTotalTimeout_throwsDeadlineExceededException`
could make the test scenario fail fast and less flaky.

This should be consistent with logback-core version. 1.3.x used because
we test java 11+ in showcase

(typo in commit message, logback-classic is the one fixed)

b/368579018

Adding alias to repository-path and api-definitions-path to be
compatible with Librarian
[options](https://github.com/googleapis/librarian/blob/d3f78d51103c523da60251e790c05a681331192a/internal/command/flags.go#L40)

[click options](https://click.palletsprojects.com/en/stable/options/)
allows alias directly, so this change is trivial and does not affect any
existing usage cases.

This pull request is generated with proto changes between
[googleapis/googleapis@fbbbf50](googleapis/googleapis@fbbbf50)
(exclusive) and
[googleapis/googleapis@66ab6ce](googleapis/googleapis@66ab6ce)
(inclusive).

BEGIN_COMMIT_OVERRIDE
BEGIN_NESTED_COMMIT
chore: [common-protos] update copyright year for auto-generated protos

PiperOrigin-RevId: 732130682

Source Link:
[googleapis/googleapis@9415ba0](googleapis/googleapis@9415ba0)
END_NESTED_COMMIT
END_COMMIT_OVERRIDE

…onfig to v1.15.0 (#3690)

RenovateBot is not creating a pull request for the version. I created
this manually based on the previous pull request
https://github.com/googleapis/sdk-platform-java/pull/3655/files.

🤖 I have created a release *beep* *boop*
---


<details><summary>2.55.0</summary>

##
[2.55.0](v2.54.0...v2.55.0)
(2025-03-12)


### Features

* generate showcase using docker image
([#3568](#3568))
([3857e3f](3857e3f))
* next release from main branch is 2.55.0
([#3668](#3668))
([1eda55f](1eda55f))


### Bug Fixes

* remove call credentials from call options if DirectPath
([#3670](#3670))
([5ede29c](5ede29c))


### Dependencies

* update arrow.version to v18.2.0
([#3675](#3675))
([5a555e5](5a555e5))
</details>

---
This PR was generated with [Release
Please](https://github.com/googleapis/release-please). See
[documentation](https://github.com/googleapis/release-please#release-please).

Co-authored-by: release-please[bot] <55107282+release-please[bot]@users.noreply.github.com>