-
Notifications
You must be signed in to change notification settings - Fork 452
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
This package is preventing adding support for new SQL dialects #716
Comments
Please push a new release to CRAN ASAP. I cannot create a new SqlRender release until this is resolved. @clairblacketer @fdefalco |
I filed PR #717 adding the pregenerated DDL files for the IRIS dialect, hoping that helps. |
thanks both. To start I will remove the offending unit tests and resubmit to CRAN. I'll let you know when it is complete |
I resubmitted to CRAN v1.0.1 of the package |
@schuemie v1.0.1 of the CommonDataModel package was accepted and is on its way to CRAN with the offending test removed. |
Thanks! |
One of the unit tests is to see if the generated SQL files match those that were pre-generated. This is done for all dialects supported by SqlRender as defined here. We just added the 'iris' dialect to SqlRender, which of course isn't one of the pre-rendered translations.
When I push SqlRender to CRAN, because CommonDataModel depends on SqlRender its unit tests are run with the new version. Because the pre-rendered SQL files do not contain the IRIS translations the unit tests fail and CRAN rejects the new SqlRender version.
In general I think this unit test is far too restrictive. It does not allow any changes to SqlRender. Even if the list of dialects stays the same, if SqlRender comes with a different but semantically equivalent translation it is blocked by CRAN. Please remove the unit test, or make it a lot more lenient.
Adding @bdeboe and @alex-odysseus for awareness of the first of probably many consequences of adding a new dialect.
The text was updated successfully, but these errors were encountered: