[spark] Support custom Paimon lake table paths

[BackendArchitectX]

Purpose

Linked issue: close #3832

Support Spark reads from Paimon lake tables whose database and/or table name is configured independently from the Fluss table path.

Previously, Spark rejected lake reads when the resolved Paimon table path differed from the Fluss table path.

Brief change log

• Resolve the configured lake table path before creating the Spark lake source.
• Use the resolved Paimon path for lake reads while preserving the original Fluss table path.
• Enable lake predicate pushdown for tables using custom lake paths.
• Add coverage for:
  • custom database mapping
  • custom table mapping
  • combined database and table mapping
  • lake-only reads
  • lake + log union reads
  • predicate pushdown with a custom lake path
  • no-lake-snapshot fallback

Tests

Passed:

• SparkLakePaimonLogTableReadTest
  • 18 tests passed
• SparkLakePaimonPrimaryKeyTableReadTest
  • 10 tests passed
• spotless:check for fluss-spark-common and fluss-spark-ut
• checkstyle:check for fluss-spark-common and fluss-spark-ut
• git diff --check

API and Format

No public API or storage format changes.

Documentation

No documentation changes are required. This change enables Spark support for the existing custom Paimon lake table path configuration.