thisisnic commented on issue #35844: URL: https://github.com/apache/arrow/issues/35844#issuecomment-1570532107
Thanks for making this suggestion to simplify things in our codebase! Agreed that the package does have a huge number of tests; I guess it's a consequence of trying to replicate behaviour from so many other packages + base R, and the fact that Arrow has a lot of different functionality. My concern here would be that using a tool which isn't widely known or adopted in the R community would reduce the number of lines of code in our tests, but ultimately further raise the barrier to entry for new contributors. I've also personally found that some previous attempts to reduce the amount of code (e.g. the use of `for` loops in some tests) has actually made things harder in the longer term when debugging failing tests, as it takes more effort to find the specific source of the test failure. That said though, we do have wrapper functions like `compare_dplyr_binding()`, which previously tested functions on both Tables and RecordBatches (before those interfaces had more in common), so something like that could be an alternative option. Would you mind giving some examples of the places where the duplication seems most problematic? -- 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: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
