[SPARK-57211][SQL] Cast strings to TIMESTAMP_NTZ(p)/TIMESTAMP_LTZ(p)

[MaxGekk]

What changes were proposed in this pull request?

This PR wires Cast to support casting StringType to the nanosecond-capable timestamp types TimestampNTZNanosType(p) and TimestampLTZNanosType(p) with fractional-seconds precision p in [7, 9], on both the interpreted and codegen paths and across all eval modes (LEGACY, ANSI, TRY):

• CAST(<string> AS TIMESTAMP_NTZ(p))
• CAST(<string> AS TIMESTAMP_LTZ(p))

Concretely, in Cast.scala:

• Add StringType -> TimestampNTZNanosType(p) / TimestampLTZNanosType(p) arms to canCast and canAnsiCast. Try-cast is covered automatically (canTryCast delegates to canAnsiCast, and canUseLegacyCastForTryCast already matches (StringType, DatetimeType), which the nanos types extend).
• Add (StringType, TimestampLTZNanosType) to Cast.needsTimeZone. The NTZ string is zone-independent, mirroring the micro TIMESTAMP_NTZ cast.
• Add interpreted castToTimestampLTZNanos / castToTimestampNTZNanos and matching codegen, dispatched from castInternal / nullSafeCastFunction with the precision taken from the target type. The result is a TimestampNanosVal (or null in legacy/try mode on malformed input).
• The NTZ cast adopts allowTimeZone = true to match the existing micro TIMESTAMP_NTZ string cast, and resolves the TODO(SPARK-57032) left on stringToTimestampNTZNanosAnsi.

This reuses the parse entry points added in SPARK-57032 on SparkDateTimeUtils (inherited by DateTimeUtils), which already return a normalized TimestampNanosVal and apply per-precision truncation, so no separate normalization module is required for the string path.

Existing preview gating is unchanged: Cast.checkInputDataTypes calls TypeUtils.failUnsupportedDataType, which throws FEATURE_NOT_ENABLED when spark.sql.timestampNanosTypes.enabled is off.

Why are the changes needed?

This is a sub-task of SPARK-56822 (SPIP: Timestamps with nanosecond precision).

The logical types, the TIMESTAMP_NTZ(p) / TIMESTAMP_LTZ(p) SQL syntax, the physical row value TimestampNanosVal, and the string-to-nanos parse helpers all exist, but Cast had zero arms for the nanos types. As a result CAST(s AS TIMESTAMP_NTZ(9)) failed type-check with CAST_WITHOUT_SUGGESTION even when the preview flag spark.sql.timestampNanosTypes.enabled was on. String ingestion is the most common entry point for these types and unblocks typed literals, filters, and CTAS once coercion lands.

Does this PR introduce any user-facing change?

Yes, but only when the preview flag spark.sql.timestampNanosTypes.enabled is enabled (it defaults to off in production). With the flag on, CAST(<string> AS TIMESTAMP_NTZ(p)) and CAST(<string> AS TIMESTAMP_LTZ(p)) for p in [7, 9] now produce correct nanosecond values in LEGACY, ANSI, and TRY modes; previously they failed type-checking. With the flag off, the behavior is unchanged (FEATURE_NOT_ENABLED). Existing microsecond timestamp string casts are unchanged.

How was this patch tested?

• CastSuiteBase: success cases for both types over p in [7, 9] and a 7-9 digit fractional corpus; LTZ parameterized over time zones, NTZ zone-independent (including a discarded zone suffix). Plus a flag-off guard asserting FEATURE_NOT_ENABLED.
• CastWithAnsiOnSuite: malformed-input parse errors (DateTimeException / CAST_INVALID_INPUT).
• CastWithAnsiOffSuite / TryCastSuite: malformed input returns NULL.
• Golden-file checks added to cast.sql (regenerated with SPARK_GENERATE_GOLDEN_FILES=1): positive cases assert the result type via typeof (the reverse direction, nanos -> string rendering, is not wired yet and is tracked under SPARK-57162); negative cases exercise the ANSI parse-error path (and NULL in non-ANSI mode).

Verified locally:

$ build/sbt 'catalyst/testOnly *CastSuite *CastWithAnsiOnSuite *CastWithAnsiOffSuite *TryCastSuite'
$ build/sbt 'sql/testOnly org.apache.spark.sql.SQLQueryTestSuite -- -z cast.sql'
$ ./dev/scalastyle

Was this patch authored or co-authored using generative AI tooling?

Generated-by: Cursor (Claude Opus 4.8)