Skip to content

Improve Avro implementation #332

@wgtmac

Description

@wgtmac

The current Avro writer and reader implementations are not that efficient because we need to use avro::GenericDatum as the intermediate layer:

  • Avro writer: ArrowArray -> avro::GenericDatum -> avro::Encoder
  • Avro reader: avro::Decoder -> avro::GenericDatum -> ArrowArray

It would be more efficient if we can get rid of conversion from/to avro::GenericDatum by directly using the encoder/decoder in the Avro writer/reader. It can greatly improve the performance of manifest file I/O.

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or requesthelp wantedExtra attention is needed

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions