Support FixedSizeBinary in Row format

[tustvold]

Which issue does this PR close?

Closes #.

Rationale for this change

We claim support for this data type, with the implementation actually panicking on unreachable if given such an array. It is fairly trivial to add support for this, so lets do so

What changes are included in this PR?

Are there any user-facing changes?

[alamb]

Looks good to me

Shall we add a mention of fixed length byte encoding in the docs:

arrow-rs/arrow/src/row/mod.rs

Lines 215 to 261 in 6f41b95

	/// ## Variable Length Bytes (including Strings) Encoding
	///
	/// A null is encoded as a `0_u8`.
	///
	/// An empty byte array is encoded as `1_u8`.
	///
	/// A non-null, non-empty byte array is encoded as `2_u8` followed by the byte array
	/// encoded using a block based scheme described below.
	///
	/// The byte array is broken up into 32-byte blocks, each block is written in turn
	/// to the output, followed by `0xFF_u8`. The final block is padded to 32-bytes
	/// with `0_u8` and written to the output, followed by the un-padded length in bytes
	/// of this final block as a `u8`.
	///
	/// Note the following example encodings use a block size of 4 bytes,
	/// as opposed to 32 bytes for brevity:
	///
	/// ```text
	/// ┌───┬───┬───┬───┬───┬───┐
	/// "MEEP" │02 │'M'│'E'│'E'│'P'│04 │
	/// └───┴───┴───┴───┴───┴───┘
	///
	/// ┌───┐
	/// "" │01 |
	/// └───┘
	///
	/// NULL ┌───┐
	/// │00 │
	/// └───┘
	///
	/// "Defenestration" ┌───┬───┬───┬───┬───┬───┐
	/// │02 │'D'│'e'│'f'│'e'│FF │
	/// └───┼───┼───┼───┼───┼───┤
	/// │'n'│'e'│'s'│'t'│FF │
	/// ├───┼───┼───┼───┼───┤
	/// │'r'│'a'│'t'│'r'│FF │
	/// ├───┼───┼───┼───┼───┤
	/// │'a'│'t'│'i'│'o'│FF │
	/// ├───┼───┼───┼───┼───┤
	/// │'n'│00 │00 │00 │01 │
	/// └───┴───┴───┴───┴───┘
	/// ```
	///
	/// This approach is loosely inspired by [COBS] encoding, and chosen over more traditional
	/// [byte stuffing] as it is more amenable to vectorisation, in particular AVX-256.
	///
	/// ## Dictionary Encoding

I realize the actual encoding is straightforward

[alamb]

Maybe it is worth noting that this value is chosen to be neither 0 or 0xff (the null sentinels)?

[alamb]

I double checked and there is no such thing as FixedSizeLargeBinary

arrow-rs/arrow-schema/src/datatype.rs

Lines 155 to 157 in 6f41b95

	FixedSizeBinary(i32),
	/// Opaque binary data of variable length and 64-bit offsets.
	LargeBinary,

👍

[ursabot]

Benchmark runs are scheduled for baseline = 733d32e and contender = ab3f384. ab3f384 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