Add Index support in the read path

The scope of this Jira is to add support for reading indexes from a Parquet file.

The changes will involve de-serializing indexes and adding API to return them.

To test the implementation, we can get Parquet files with indexes generated via Impala or parquet-mr and see that the parquet-cpp tools can print them.

**Reporter**: [Deepak Majeti](https://issues.apache.org/jira/secure/ViewProfile.jspa?name=mdeepak) / @majetideepak
**Assignee**: [Deepak Majeti](https://issues.apache.org/jira/secure/ViewProfile.jspa?name=mdeepak) / @majetideepak

_{**Note**: *This issue was originally created as [PARQUET-1848](https://issues.apache.org/jira/browse/PARQUET-1848). Please see the [migration documentation](https://issues.apache.org/jira/browse/PARQUET-2502) for further details.*}

Inizia individuando il percorso di lettura dei metadati Parquet e la superficie dell'API in cui verrebbero restituiti gli indici deserializzati. Usa file Parquet con indici generati tramite Impala o parquet-mr, quindi verifica che gli strumenti parquet-cpp possano stamparli.

Scritto dal modello di indicizzazione a partire dal testo della issue.