Github user culler commented on the pull request:
https://github.com/apache/spark/pull/3158#issuecomment-62480405
Hi @marmbrus and @liancheng, I have now tried what seems to be a very
simple and direct solution to the `===` problem. My patch no longer uses
`===`. I added a new equality operator to the DSL called `-=-`. It is
identical to the `===` operator (which is still in place) with the exception
that it also works for DSL expressions which begin with a literal. And it does
not cause any conflicts with scalatest.
I reverted all of the changes to the various test files that were intended
to work around the `===` issue. That brings this PR down to 5 files. All of
the changes related to the DSL are in one file:
sql/hive/src/test/scala/org/apache/spark/sql/hive/StatisticsSuite.scala. All
of the changes related to RichDate and RichTimestamp are in the other 4 files.
This means that the two different changes have effectively been isolated
and the structure of the PR should be more transparent now. It would not be
hard to remove the one file and start another PR for it, except for the
technical detail that github only allows one fork at a time. I suppose I could
fork myself, so to speak, and open another github account for my alter ego.
PS The `===` issue is not just about having too many implicit conversions
in scope -- it is also about combining those conversions with macros.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
