GH-54489: [CI][Integration] Install jpype1 from conda-forge in conda-integration image

[CurtHagenlocher]

Rationale for this change

The conda-integration image doesn't pin Python, so it now gets Python 3.15. ci/scripts/integration_arrow.sh installs jpype1 with pip, and PyPI has no cp315 wheel for jpype1 yet. pip falls back to building from source, which fails because the image doesn't include Apache Ant. See #54489 for details.

What changes are included in this PR?

Install jpype1 from conda-forge in conda-integration.dockerfile, as conda-python-jpype.dockerfile already does. conda-forge has builds for Python 3.15 on both linux-64 and linux-aarch64.

integration_arrow.sh is unchanged: its pip install jpype1 finds the conda package already installed and does nothing. It still installs jpype1 when the script runs outside this image.

Are these changes tested?

Yes. The Integration job in apache/arrow-dotnet builds this image from scratch and fails with the error above. I ran that job on a fork with the Arrow checkout pointed at this branch, and it got past the jpype1 installation that used to fail. When I opened this PR the run was still in progress: https://github.com/CurtHagenlocher/arrow-dotnet/actions/runs/39150917107

Are there any user-facing changes?

No.

Was AI used for this PR?

In accordance to the AI generation guidelines, please disclose below whether and how AI was used in this PR.

PR code and description written by:

• Human
• AI

Reviewed before submission by:

• Human
• AI
• Not reviewed

🤖 Generated with Claude Code

• GitHub Issue: [CI][Integration] conda-integration fails to install jpype1 with Python 3.15 #54489

[github-actions]

⚠️ GitHub issue #54489 has been automatically assigned in GitHub to PR creator.

[CurtHagenlocher]

The verification run linked in the description has finished, and the Integration job passed: https://github.com/CurtHagenlocher/arrow-dotnet/actions/runs/39150917107/job/117744207024

• The image was built with Python 3.15.0, and conda installed jpype1-1.7.1-py315h62b6238_1. pip install jpype1 then reported Requirement already satisfied.
• The Java C Data Interface tests ran in both directions (.NET → Java and Java → .NET).
• Result: 0 failures, 337 skips. That skip count matches the last successful run before the Python 3.15 change (Oct 12).

[Copilot]

🟢 Approval recommended

The narrow change follows the existing JPype image pattern, with successful integration validation and no unresolved findings.

No open issues 🎉

What changed in this PR

Preinstalls jpype1 from conda-forge in the integration image, avoiding source-build failures when PyPI lacks wheels for the selected Python version.

Changes:

• Adds jpype1 to the existing mamba installation.
• Documents the reason; leaves the integration script unchanged.

File	Description
ci/​docker/​conda-integration.dockerfile	Installs and documents the conda-provided JPype dependency.

_{Review effort: Balanced}

[kou]

+1