You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 4 Next »

Proposer

Approver

Status

Current state: Under Discussion

Discussion thread: here

JIRA: here

Released: N/A

Prior doc link : https://docs.google.com/document/d/1bj-xpkRomVtbzvLb_4BRngDIGkkMR5yzxXRRzkA7QVo/edit#heading=h.di66rda5xhp2

Abstract

Hudi delta Streamer does not have direct support for pulling data in CSV format from kafka/HDFS-logs. The only possible alternative to ingesting CSV data to Hudi dataset is to first convert them into json/avro before pulling in through delta-streamer. This RFC proposes a mechanism to directly support sources in CSV format.

Background

Introduce any much background context which is relevant or necessary to understand the feature and design choices.

Implementation

Extend the DeltaStreamer by implementing a CSV Source(kafka/hdfs)

  • We can use existing  FilebasedSchemaProvider class for decoding csv data.
  • If the Csv data does not  contains header then, for field names in source  schema we can use _c0,_c1,_c2..etc as the field names according to the record position in the csv data. (For kakfa data & non header csv files we can use this method).
  • If header is present in the csv data then  need to use header information for the field names.
  • Need to introduce a configuration property to inform whether header is present  or not.
  • Need to introduce a new configuration property to support any configurable delimiters.

Rollout/Adoption Plan

  • There won’t be any impact for existing users. This is just a new feature

Test Plan

            Unit and Manual Integration test





  • No labels