Treat .Rmarkdown files as .Rmd files

[JohannesNE]

.Rmarkdown files are like .Rmd files, except they render to .md instead of .html (https://bookdown.org/yihui/blogdown/output-format.html)

Unfortunately they are rejected by styler (Error: Can only style .R, .Rmd and .Rnw files.)

[JohannesNE]

If you would rather implement this differently, of if I've overlooked something, feel free to just reject at implement it properly.

[codecov-commenter]

Codecov Report

Merging #824 (2e64b07) into master (d31ee93) will increase coverage by 0.03%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #824      +/-   ##
==========================================
+ Coverage   90.46%   90.49%   +0.03%     
==========================================
  Files          47       47              
  Lines        2432     2441       +9     
==========================================
+ Hits         2200     2209       +9     
  Misses        232      232              

Impacted Files	Coverage Δ
R/set-assert-args.R	92.68% <100.00%> (ø)
R/ui-styling.R	100.00% <100.00%> (ø)
R/utils-files.R	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d31ee93...2e64b07. Read the comment docs.

[github-actions]

This is how benchmark results would change (along with a 95% confience interval in relative change) if eede456 is merged into master:

• cache_applying: 0.09s -> 0.09s [-3.12%, +3.97%]
• cache_recording: 0.96s -> 0.96s [-0.83%, +2.66%]
• without_cache: 2.4s -> 2.38s [-1.09%, +0.05%]

Further explanation regarding interpretation and methodology can be found in the documentation.

[lorenzwalthert]

Thanks @JohannesNE. Closes: #480. Before a merge, we should:

• add tests: tests/testthat/test-public-api.R. Look for Rmd and add one test for every API function (style_file(), style_dir(), style_pkg()). Just look for 'Rmd' in that file (starting around line 134).
• Add a news bullet: Files with .Rmarkdown extensions are now recognised as an R markdown files in style_file() and friends (Treat .Rmarkdown files as .Rmd files #824).
• In the documentation, state explicitly that the new extension is treated as R markdown, i.e. adapt the documentation of the parameter filetype in prettify_pkg().

Do you want to give this a try?

[JohannesNE]

My experience is limited, but I have given it a try.

[lorenzwalthert]

@JohannesNE looks great overall, thanks a lot. Let's keep our fingers crossed for the tests to pass 🙏.

[JohannesNE]

Sorry. I actually meant to make a comment that i was unsure about committing the updated reference files, but forgot about it.

[lorenzwalthert]

There is something strange going on with some unrelated reference files I think. When I locally run the tests, I get modified files (so tests fail), but on CI/CD, the same reference files pass (I know from previous experience). I reverted to the old ref files in 2e64b07, let's hope we can make tests pass. Maybe I'll open a separate issue. The problem persists for some time already.

[lorenzwalthert]

Are you working on macOS as well?

[JohannesNE]

I'm on Ubuntu 21.04. I am afraid won't be of much help with that problem. I usually "fix" encoding problems by trial-and-error 🤷‍♂️

[github-actions]

This is how benchmark results would change (along with a 95% confience interval in relative change) if 49c52ed is merged into master:

• cache_applying: 0.1s -> 0.1s [-3.61%, +3.75%]
• cache_recording: 1.06s -> 1.07s [-0.07%, +0.44%]
• without_cache: 2.69s -> 2.69s [-0.13%, +0.4%]

Further explanation regarding interpretation and methodology can be found in the documentation.

[lorenzwalthert]

I'm on Ubuntu 21.04. I am afraid won't be of much help with that problem. I usually "fix" encoding problems by trial-and-error 🤷‍♂️

Ok. Seems not a problem in CI/CD, neither on CRAN, so I guess we can leave it for now.

[lorenzwalthert]

Thanks @JohannesNE 🥳