[Parquet] Split `ParquetMetadataReader` into IO/decoder state machine and thrift parsing

[alamb]

Is your feature request related to a problem or challenge? Please describe what you are trying to do.

• Related to [Epic] Parquet Reader Improvement Plan / Proposal - July 2025 #8000
• Similar to Decouple IO and CPU operations in the Parquet Reader (push decoder) #7983, but for the metadata reader

The current ParquetMetadataReader intermixes three things:

1. The state machine for decoding parquet metadata (footer, then metadata, then (optional) indexes)
2. orchestrating IO (aka calling read, etc)
3. Decoding thrift encoded byte into objects

This makes it almost impossible to add features like "only decode a subset of the columns in the ColumnIndex" and other potentially advanced usecases

Describe the solution you'd like

Now that we have a "push" style API for metadata decoding that avoids IO, I would like to separate out these three parts so that we can add better features

Describe alternatives you've considered

Additional context

[alamb]

This will also assist @etseidl while he works on

• Use custom thrift decoder to improve speed of parsing parquet metadata #5854