Uploaded image for project: 'Apache Drill'
  1. Apache Drill
  2. DRILL-6835 Schema Provision using File / Table Function
  3. DRILL-6965

Adjust table function usage for all storage plugins and implement schema parameter

    XMLWordPrintableJSON

Details

    Description

      Schema can be used while reading the table into two ways:
      a. schema is created in the table root folder using CREATE SCHEMA command and schema usage command is enabled;
      b. schema indicated in table function.
      This Jira implements point b.

      Schema indication using table function is useful when user does not want to persist schema in table root location or when reading from file, not folder.

      Schema parameter can be used as individual unit or in together with for format plugin table properties.

      Usage examples:

      Pre-requisites:
      V3 reader must be enabled: set `exec.storage.enable_v3_text_reader` = true;

      Query examples:

      1. There is folder with files or just one file (ex: dfs.tmp.text_table) and user wants to apply schema to them:
      a. indicate schema inline:

      select * from table(dfs.tmp.`text_table`(
      schema => 'inline=(col1 date properties {`drill.format` = `yyyy-MM-dd`}) 
      properties {`drill.strict` = `false`}'))
      

      To indicate only table properties use the following syntax:

      select * from table(dfs.tmp.`text_table`(
      schema => 'inline=() 
      properties {`drill.strict` = `false`}'))
      

      b. indicate schema using path:
      First schema was created in some location using CREATE SCHEMA command. For example:

      create schema 
      (col int)
      path '/tmp/my_schema'
      

      Now user wants to apply this schema in table function:

      select * from table(dfs.tmp.`text_table`(schema => 'path=`/tmp/my_schema`'))
      

      2. User wants to apply schema along side with format plugin table function parameters.
      Assuming that user has CSV file with headers with extension that does not comply to default text file with headers extension (ex: cars.csvh-test):

      select * from table(dfs.tmp.`cars.csvh-test`(type => 'text', 
      fieldDelimiter => ',', extractHeader => true,
      schema => 'inline=(col1 date)'))
      

      More details about syntax can be found in design document:
      https://docs.google.com/document/d/1mp4egSbNs8jFYRbPVbm_l0Y5GjH3HnoqCmOpMTR_g4w/edit

      Attachments

        Issue Links

          Activity

            People

              arina Arina Ielchiieva
              arina Arina Ielchiieva
              Vova Vysotskyi Vova Vysotskyi
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: