[FLINK-17224][table] Support precision of TIME type

[raminqaf]

What is the purpose of the change

Add precision supports for the TIME function. Currently, Flink only takes precision 0 and ignores any other type of precision.

Brief change log

• Pass precision to TIME funciton.
• Fixed the TIME conversion formula in JsonToRowDataConverters.convertToTime() from localTime.toNanoOfDay() * 1000_000 to localTime.toNanoOfDay() / 1000_000
• Enhanced test coverage in JsonRowDataSerDeSchemaTest.testSerDe() to include comprehensive testing of all supported data types
• Updated test assertions to use correct field indexing and proper row size

Verifying this change

This change is already covered by existing tests, such as:

• JsonRowDataSerDeSchemaTest.testSerDe() - Enhanced to comprehensively test TIME type conversion for both JsonParser and non-JsonParser deserializers
• The test now verifies correct handling of TIME(0) and TIME(3) precision levels
• Added test coverage for the previously failing scenario that caused the DateTimeException

This change added tests and can be verified as follows:

• The enhanced testSerDe() method now tests the complete round-trip serialization/deserialization for TIME types
• Verified that the fix resolves the DateTimeException: Invalid value for HourOfDay error
• Confirmed that both isJsonParser=true and isJsonParser=false code paths work correctly
• Manually verified that TIME values like "12:12:43" and "12:12:43.123" are properly converted without precision loss

Does this pull request potentially affect one of the following parts:

• Dependencies (does it add or upgrade a dependency): no
• The public API, i.e., is any changed class annotated with @Public(Evolving): no
• The serializers: yes - affects JSON format serialization/deserialization of TIME types
• The runtime per-record code paths (performance sensitive): no - this is format-specific conversion logic
• Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn, ZooKeeper: no
• The S3 file system connector: no

Documentation

• Does this pull request introduce a new feature? no
• If yes, how is the feature documented? not applicable - this is a bug fix for existing functionality

[raminqaf]

The TimeLongConvert converts the time to millisecond precision. I was wondering if we need to change this to TimeLongConverter implements DataStructureConverter<Long, Long> so that we internally support nanosecond precision.

[flinkbot]

CI report:

• d90a4e2 Azure: FAILURE

Bot commands

The @flinkbot bot supports the following commands:

• @flinkbot run azure re-run the last Azure build

[davidradl]

If we change the way that the json format parses times, then the output will change for existing users, which has implications to the logic of their processing. I suggest that this way of parsing be enabled by a new json format configuration option - so it can be knowingly adopted.

[raminqaf]

A similar PR exists:
#22775

[raminqaf]

We are not really supporting TIME(5). As far as I understood, the time converters are storing the time as integer and not as long. We lose the precession (>3) there.