[python] Add opt-in Parquet OffsetIndex reads for row windows

[XiaoHongbo-Hope]

What

Add opt-in Parquet OffsetIndex reads for one contiguous row-ID window per row group. PyPaimon fetches the selected data pages and required dictionary pages, then decodes them with PyArrow. This supports scalar and nested STRUCT/ARRAY/MAP fields, including nested projections.

Enable it per read:

table = table.copy({"parquet.filter.columnindex.enabled": "true"})

The option is off by default in PyPaimon. Disjoint windows, VARIANT, missing indexes, decoded row-group cache reads, and unsupported or unprofitable selections use the existing reader. This does not change the table format or add ColumnIndex predicate filtering.

Trade-off

Reading fewer bytes can require more range requests, so lower latency is not guaranteed. A read-only OSS A/B on an isolated table confirmed identical results and this byte/request trade-off; it is not an end-to-end throughput claim.

Validation

Tests passed on PyArrow 19 and 24, including nested fields, projections, fallback and error paths. Java parquet-mr nested fixtures matched full reads; an older PyArrow fixture exercised the fallback. PyArrow 16 was not tested locally.