[GitHub] drill issue #936: DRILL-5772: Add unit tests to indicate how utf-8 support c...

[arina-ielchiieva <gi...@git.apache.org>]

Github user arina-ielchiieva commented on the issue:

    https://github.com/apache/drill/pull/936
  
    @paul-rogers 
    agree with you that charsets used in saffron properties should be defaulted in Drill to `UTF-8` since Drill can read UTF-8 data and it's strange that it would fail by default when Calcite will attempt to parse string into literal used in query.
    
    I have looked into Calcite code and there is no option to hard-code charset values for Calcite but charset can be changed using properties.
    There are two options of setting saffron properties:
    1. as system property;
    2. using `saffron.properties` file.
    
    I don't really like passing them as `-D` when starting the drillbit 9since there are at least two), so I am more inclined to use `saffron.properties` file. Unfortunately, in Calcite code `saffron.properties` location is expected to be working folder [1], i.e. the place where java process was started. I have created Jira and pull request in Calcite to allow `saffron.properties` to be present in classpath since it's more convenient [2]. I'll keep you updated on Calcite community feedback.
    
    [1] https://github.com/apache/calcite/blob/master/core/src/main/java/org/apache/calcite/util/SaffronProperties.java#L113
    
    [2] https://issues.apache.org/jira/browse/CALCITE-2014


---