Uploaded image for project: 'Spark'
  1. Spark
  2. SPARK-26151

Return partial results for bad CSV records

    XMLWordPrintableJSON

    Details

    • Type: Improvement
    • Status: Resolved
    • Priority: Minor
    • Resolution: Fixed
    • Affects Version/s: 2.4.0
    • Fix Version/s: 3.0.0
    • Component/s: SQL
    • Labels:
      None

      Description

      Currently, CSV datasource and from_csv returns a rows with all nulls for bad CSV records in the PERMISSIVE mode even some of fields were parsed and converted successfully. For example, the CSV input:

      0,2013-111-11 12:13:14
      1,1983-08-04
      

      for the first line returned row is Row(null, null) but value 0 can be parsed and converted successfully. And result can be Row(0, null). This ticket aims to change implementation of UnivocityParser and return the partial result.

        Attachments

          Activity

            People

            • Assignee:
              maxgekk Max Gekk
              Reporter:
              maxgekk Max Gekk
            • Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved: