[Variant]: Implement `DataType::Time32/Time64` support for `cast_to_variant` kernel

[alamb]

• Part of [EPIC] [Parquet] Implement Variant type support in Parquet #6736

Is your feature request related to a problem or challenge? Please describe what you are trying to do.
We implemented cast_to_variant for some primitive types in

• implement cast_to_variant kernel to cast native types to VariantArray #8044

Now we need to complete out type support for other Arrow types.

Describe the solution you'd like
Add support for the types named in the title as well.

Describe alternatives you've considered

1. Add the corresponding match arms in the cast_to_variant kernel:

   arrow-rs/parquet-variant-compute/src/cast_to_variant.rs

   Line 70 in c25c5a7

   match input_type {

2. add tests (e.g.

   arrow-rs/parquet-variant-compute/src/cast_to_variant.rs

   Lines 125 to 142 in c25c5a7

   	fn test_cast_to_variant_int8() {
   	run_test(
   	Arc::new(Int8Array::from(vec![
   	Some(i8::MIN),
   	None,
   	Some(-1),
   	Some(1),
   	Some(i8::MAX),
   	])),
   	vec![
   	Some(Variant::Int8(i8::MIN)),
   	None,
   	Some(Variant::Int8(-1)),
   	Some(Variant::Int8(1)),
   	Some(Variant::Int8(i8::MAX)),
   	],
   	)
   	}

   )

[klion26]

take

[klion26]

@alamb As Variant::TimeNTZ has not been implemented, do we need to wait for TimeNTZ to be implemented first or convert Time32 to other types (such as Int32)?

If we need to wait for TimeNTZ to be implemented, maybe I can help implement TimeNTZ after reaching consensus on how to implement it.

[alamb]

It would be great to figure out how to implement TimeNTZ 🙏

[klion26]

@alamb From the iceberg implementation and the parquet testing pr, the TimeNTZ will store a long in the file.

iceberg time read logic sample code

....

case TIME:    
case TIMESTAMPTZ_NANOS:    
case TIMESTAMPNTZ_NANOS:    
  return VariantUtil.readLittleEndianInt64(value, PRIMITIVE_OFFSET);
....
static long readLittleEndianInt64(ByteBuffer buffer, int offset) {  
  return buffer.getLong(buffer.position() + offset);  
}

When decoding, we can read from the file with u64 and parse it to NaiveTime, and write the u64 back to file when serializing.

But, currently, there is no test file in parquet-testing(primitive_time.metadata and primitive_time.value). I'm playing with the iceberg code to generate the test file, and will file a pr to support TimeNTZ when the testing files are ok.

[alamb]

label_issue.py automatically added labels {'parquet'} from #8044

[alamb]

label_issue.py automatically added labels {'parquet-variant'} from #8114