Increase test coverage of ArrowWriter

[tustvold]

Which issue does this PR close?

Part of #1764

Rationale for this change

Whilst working on #1764 I realised we didn't have very good test coverage of writing arrow arrays with different encodings to parquet. This is because the arrow_reader fuzz tests write data using the ColumnWriter API and not the arrow writer API.

In the past these have boiled down to much the same thing, and so this hasn't been an issue. However, as we move to writing directly from arrow arrays, we introduce new codepaths that need testing

What changes are included in this PR?

Ensures good coverage of various encodings, writer versions, etc...

Are there any user-facing changes?

No, this just changes some tests

[alamb]

💯 for adding tests first before changing code 👍

For any other reviewer, I found https://github.com/apache/arrow-rs/pull/2220/files?w=1 without whitespace easier to review

[ursabot]

Benchmark runs are scheduled for baseline = 6ce4c4e and contender = 3505afa. 3505afa is a master commit associated with this PR. Results will be available as each benchmark for each run completes.
Conbench compare runs links:
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ec2-t3-xlarge-us-east-2] ec2-t3-xlarge-us-east-2
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on test-mac-arm] test-mac-arm
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ursa-i9-9960x] ursa-i9-9960x
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ursa-thinkcentre-m75q] ursa-thinkcentre-m75q
Buildkite builds:
Supported benchmarks:
ec2-t3-xlarge-us-east-2: Supported benchmark langs: Python, R. Runs only benchmarks with cloud = True
test-mac-arm: Supported benchmark langs: C++, Python, R
ursa-i9-9960x: Supported benchmark langs: Python, R, JavaScript
ursa-thinkcentre-m75q: Supported benchmark langs: C++, Java