[GH-3257] GeoPandas: Implement distributed hilbert_distance

[jiayuasu]

Did you read the Contributor Guide?

• Yes, I have read the Contributor Rules and Contributor Development Guide.

Is this PR related to a ticket?

• Yes, and the PR name follows the format [GH-XXX] my subject.

Closes #3257.
Part of #2230.

What changes were proposed in this PR?

• Add a native ST_HilbertDistance(geometry, xmin, ymin, xmax, ymax, level) implementation that maps a geometry-envelope midpoint onto a Hilbert curve over a supplied extent.
• Represent the unsigned 32-bit Hilbert address as a non-negative Long, including the level-16 maximum value 4294967295.
• Clip midpoint coordinates outside the supplied extent, map zero-width axes to grid coordinate zero, propagate null SQL inputs, reject empty geometries, and preserve GeoPandas-compatible level behavior.
• Register and expose ST_HilbertDistance through Spark SQL and DataFrame APIs, Flink SQL, Snowflake WKB and GeoJSON UDF paths, and the Python Spark SQL API.
• Implement distributed GeoSeries.hilbert_distance and GeoDataFrame.hilbert_distance through the active geometry column.
• Use one eager distributed aggregate to validate geometries and, when total_bounds is omitted, derive the extent of all envelope midpoints. Only the single aggregate row reaches the driver; geometry rows and returned keys remain distributed.
• Preserve duplicate and MultiIndex indexes, the hilbert_distance result name, explicit-bound behavior, empty-series behavior, and GeoPandas parity without Python UDF execution.
• Document the SQL function for Spark, Flink, and Snowflake and add the distributed method to the English and Chinese GeoPandas tutorials.

How was this patch tested?

• Full common test suite: 1,292 tests passed.
• Focused Spark ST_HilbertDistance suite: 4 tests passed.
• Focused Flink function test: passed.
• Snowflake native-function and DDL suites: 4 tests passed.
• Focused Python GeoPandas, parity, and DataFrame API coverage: 15 tests passed.
• Regression coverage includes Hilbert-address goldens at levels 2, 3, and 16; the unsigned level-16 maximum; envelope-midpoint bounds; zero-width axes; clipping; null and typed-empty geometries; empty sources; explicit and inferred bounds; level validation; named duplicate and MultiIndex indexes; active-geometry delegation; exactly one metadata action; and the absence of Python execution nodes.
• The full documentation build, repository formatting checks, and git diff --check passed.
• The commit passed the repository pre-commit suite.

Did this PR include necessary documentation updates?

• Yes, I am adding a new API. I used the current SNAPSHOT version, v2.0.0.
• Yes, I added Spark, Flink, and Snowflake SQL function documentation, release notes, and English and Chinese GeoPandas tutorial updates.