[spark] Support custom Paimon lake table paths

### Search before asking

- [x] I searched in the [issues](https://github.com/apache/fluss/issues) and found nothing similar.

### Description

### Motivation

Fluss tables can use `table.datalake.database-name` and
`table.datalake.table-name` to map a Fluss table to a Paimon table with a
different database or table name. https://github.com/apache/fluss/pull/3476

Spark lake reads currently reject tables whose resolved Paimon path differs
from the Fluss table path. This issue tracks adding full Spark support for
custom Paimon lake table paths.

### Scope

- Resolve and use the mapped Paimon database and table name when creating the
Spark lake source.
- Support custom Paimon paths in Fluss/Paimon union reads.
- Ensure predicate pushdown continues to work when the lake source uses a
mapped database or table name.
- Add coverage for custom database mapping, custom table mapping, and combined
database/table mapping.
- Cover both direct lake reads and union reads.

### Current behavior

Spark throws the following exception when a lake read resolves to a different
Paimon table path:

```text
Custom lake table path is not supported for Spark lake reads yet.

### Willingness to contribute

- [ ] I'm willing to submit a PR!

Start by tracing Spark lake source creation and the existing rejection for resolved Paimon paths that differ from the Fluss table path. Check direct lake reads and Fluss/Paimon union reads, then add coverage for custom database mapping, custom table mapping, combined mappings, and predicate pushdown; done means all listed reads work with mapped paths.

Written by the indexing model from the issue text.