Jefffrey commented on issue #13408:
URL: https://github.com/apache/datafusion/issues/13408#issuecomment-3370224483

   > at dbt, we also had a problem of hard-to-provide test coverage for Utf8, 
Utf8View. A `varchar` can be one of them, but the code (eg UDFs) ultimately 
needs to accept both what we ended up doing is fuzzing in SLT tests. At the 
test start we generate a random number. Then every string literal is either 
Utf8, Utf8View or LargeUtf8 -- deterministically based on its value and this 
random number. This helped identify many places too rigid or with missing 
coercion logic (often not only string related), while still allowing us to 
reproduce any test failures observed on CI.
   
   Should we raise a new issue to track this idea? To fuzz not only strings but 
also lists (largelist, list, fixedsize) and binary/fixedbinary?


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to