GH-39987: [R] Make it possible to use a rtools libarrow on windows

[assignUser]

This enables the use of libarrow from rtools. This is currently only possible by cross compiling manually but will be part of a future rtools version.

These changes can't be tested, there are no user facing changes for now.

• Closes: [R] Make it possible to use an existing (rtools) install of libarrow on Windows #39987

[github-actions]

Thanks for opening a pull request!

If this is not a minor PR. Could you open an issue for this pull request on GitHub? https://github.com/apache/arrow/issues/new/choose

Opening GitHub issues ahead of time contributes to the Openness of the Apache Arrow project.

Then could you also rename the pull request title in the following format?

GH-${GITHUB_ISSUE_ID}: [${COMPONENT}] ${SUMMARY}

or

MINOR: [${COMPONENT}] ${SUMMARY}

In the case of PARQUET issues on JIRA the title also supports:

PARQUET-${JIRA_ISSUE_ID}: [${COMPONENT}] ${SUMMARY}

See also:

• Other pull requests
• Contribution Guidelines - Contributing Overview

[assignUser]

@github-actions crossbow submit r-binary-packages

[github-actions]

Revision: f405c66

Submitted crossbow builds: ursacomputing/crossbow @ actions-e8f1c270de

Task	Status
r-binary-packages

[nealrichardson]

Am I missing it, but where do we download from rwlinlib (specifically, https://github.com/rwinlib/arrow/archive/refs/tags/v13.0.0.zip right?)

[nealrichardson]

Does Rtools always have pkg-config? If so, we can probably simplify things here since there's only one way to build on Windows.

[assignUser]

rtools43 does, not sure about older versions

[github-actions]

⚠️ GitHub issue #39987 has been automatically assigned in GitHub to PR creator.

[assignUser]

but where do we download from rwlinlib

We don't, since we added checksum verification in #37941 we use the artifactory for windows too.

[nealrichardson]

but where do we download from rwlinlib

We don't, since we added checksum verification in #37941 we use the artifactory for windows too (as we generate checksums during the release).

I think for CRAN submission we should use rwinlib.

[conbench-apache-arrow]

After merging your PR, Conbench analyzed the 5 benchmarking runs that have been run so far on merge-commit e1241e7.

There was 1 benchmark result with an error:

• Commit Run on ursa-i9-9960x at 2024-02-08 02:58:21Z
  • tpch (R) with engine=arrow, format=parquet, language=R, memory_map=False, query_id=TPCH-21, scale_factor=10

There was 1 benchmark result indicating a performance regression:

• Commit Run on ursa-i9-9960x at 2024-02-08 02:58:21Z
  • dataset-selectivity (Python) with dataset=nyctaxi_multi_ipc_s3, selectivity=1%

The full Conbench report has more details.