Conversation
Updated remote repository reference for formatters. Signed-off-by: Joe Zhu <sha.joe.zhu@gmail.com>
Signed-off-by: Joe Zhu <sha.joe.zhu@gmail.com>
…j_templates_scripts
331456d to
8133017
Compare
Signed-off-by: David Muñoz Tord <david.munoztord@mailbox.org>
Unit Tests Summary 1 files 274 suites 24m 2s ⏱️ For more details on these errors, see this check. Results for commit 6b7ec8a. ♻️ This comment has been updated with latest results. |
Unit Test Performance DifferenceTest suite performance difference
Additional test case details
Results for commit 5edc4a6 ♻️ This comment has been updated with latest results. |
|
hi @munoztd0 , thanks for submitting the PR, i would suggest to snapshot the ascii tables instead of rtf, the table is much more readable for comparing on github, rtf is less ideal in this case. i am guessing you would like to use rtf due to output formatting? |
@shajoezhu I 100% agree but you're also right on your guess. If that is a blocker I could put it up the food chain but I have no hope for that. |
|
New error on tsfae13 appeared but only on the CRAN pass, still need to investigate |
|
@munoztd0 https://github.com/insightsengineering/scda.test/blob/main/tests/testthat/test-table_jnj_tbl1.R |
|
i suggest we not include rtf files here. otherwise it is going to fail always if minor changes which is not efficent, and we wont be able to test for anything. i think just create simple static snapshot is good. thanks guys! |
I could do that but that would be quite a mess and also not have the same names that in internal, but lets discuss it. |
There was a problem hiding this comment.
Thanks for the PR! I would like to echo Joe about rtf files. It seems a bit of a fragile test considering that not everyone has the same rtf software on their machine. I strongly suggest to snapshot-test the final table and not the rtf file. Decoration details can be tested as strings if not included directly in the table.
It is important to have readable snapshots as possible changes are easier to understand and review that way. Furthermore, simple changes may have a large line changes count that would add to the .git file size. Reducing that size down after some time would be a non-trivial task.
I 100% agree, I will try to discuss that internally and see if I can make a case for it.. Changing the PR to draft for now, |
As discussed we want to integrate JnJ innovative medecine newly release open-source TLG catalog scripts.
Any feedback would be very welcomed.