Module parquet 

Functions

combine_parquet_files

Combines multiple Parquet files using object store with storage options

combine_parquet_files_from_object_store

Combines multiple Parquet files from object store

create_object_store_from_path

Creates an object store from a URI string with optional storage options.

min_max_from_parquet_metadata

Extracts the minimum and maximum i64 values for the specified column_name from a Parquet file’s metadata using object store with storage options.

min_max_from_parquet_metadata_object_store

Extracts the minimum and maximum i64 values for the specified column_name from a Parquet file’s metadata in object store.

normalize_path_to_uri

Normalizes a path to URI format for consistent object store usage.

read_parquet_from_object_store

Reads a Parquet file from an object store and returns all record batches plus the Arrow schema from the builder. The builder’s schema includes metadata restored from the file’s ARROW:schema key_value_metadata; use it for decoding instead of each batch’s schema (which has metadata stripped).

write_batch_to_parquet

Writes a RecordBatch to a Parquet file using object store, with optional compression.

write_batches_to_object_store

Writes multiple RecordBatch items to an object store URI, with optional compression, row group sizing, and key_value_metadata (e.g. for instrument “class” so it survives roundtrip).

write_batches_to_parquet

Writes multiple RecordBatch items to a Parquet file using object store, with optional compression, row group sizing, and storage options.