It's called the Yihuiverse style guide. It's official now. haha. No, but it was already called kind of Yihui's style guide in #340 (comment) 😄. Of course, we can also go for a different name if @yihui does not agree and I think it could be more informative too.
Thanks for the PR @Robinlovelace. I am not sure what the current status of the issue is. We had
#340 (comment) where we agreed to add these two style guides to styler but without exporting them. Alternatively, we could export and add a life-cycle badge "experimental" to this style guide. Another question is if it should inherit all other rules from the tidyverse style guide or if the <- -> = rule is the only rule in the style guide. As you can note from the referenced issues, in particular mlr-org/mlr#2278 (comment), I am rather conservative about the topic of adding other style guides to styler. I think I need to discuss again with @krlmlr sometime soon. Also, we'd need to add unit tests to this PR in case we'd plan to merge this in styler.

Hehe I like the yihuiverse_style() function name! Seriously though, I agree it's worth thinking about an appropriate name. Originally I called the equals_style() but changed it to a more fun name for this PR. I think @yihui should definitely get a say over this. So far I can see/think of 5 options for the style name:

• equals_style()
• eq_assign_style()
• equals_assign_style()
• yihui_style()
• yihuiverse_style()

I can think of various +s and -s of each name. I'd be happy with any of them and there may be more viable options I've not seen/thought of. I think you guys (primarily @lorenzwalthert and @yihui ) should have a greater say than me over the name so up to you: I'll happily change the name in the PR. Also happy to add tests and make any further changes / put the code elsewhere pending feedback.

Regarding whether

it should inherit all other rules from the tidyverse style guide

I'm not sure. Are there any other changes I should make for this to be 'yihuiverse-compliant' ✔️ @yihui hehe?

In summary: thanks for the feedback; I'll make changes pending more feedback, particular relating to the style's name.

I'm fine with kidding, but please don't make "Yihui style" or "Yihuiverse" official.

Sure. I think we better find another name for the style guide that does not involve your name.

Codecov Report

Merging #449 into master will decrease coverage by 0.46%.
The diff coverage is 37.93%.

@@            Coverage Diff             @@
##           master     #449      +/-   ##
==========================================
- Coverage   91.02%   90.55%   -0.47%     
==========================================
  Files          36       36              
  Lines        1626     1631       +5     
==========================================
- Hits         1480     1477       -3     
- Misses        146      154       +8

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 c7e9e91...8256cf5. Read the comment docs.

Thanks for the feedback @yihui I've updated the style name in-line with my original thinking - used equals_style() as it's more concise and I think equally understandable as equals_assign_style() and more so than eq_assign_style().

Another issue with this PR I've discovered is that the style does not work as expected:

txt = "x = c(
  1
)"
styler::style_text(txt)
#> x <- c(
#>   1
#> )
styler::style_text(txt, style = styler::equals_style)
#> x = c(
#> 1
#> )

^{Created on 2018-12-05 by the reprex package (v0.2.1)}

@lorenzwalthert any suggestions on how to fix this? At present I'm setting the tokens as follows - was thinking of naming the list but not sure if that will help. Here's the function at present in any case:

equals_style <- function() {
 create_style_guide(token = list(force_assignment_op_equals))
}

Regarding tests, I've had a first bash but not sure if this fits with your previous tests. Advice on this also appreciated @lorenzwalthert:

library(styler)
library(testthat)
txt_left_assign = "x <- 1"
txt_equals_assign = "x = 1"
res_equals_style = style_text(txt_left_assign, style = equals_style)
txt_equals_style = as.character(res_equals_style)
test_that("equals_style replaces <- with = in one-liner",{
  expect_identical(
    txt_equals_style,
    txt_equals_assign)
})

^{Created on 2018-12-05 by the reprex package (v0.2.1)}

As far as tests go, We have a testing framework powered by test_collection(). Essentially, there is an *-in.R file and a *-out.R file. The *-in.R file is the input that is transformed and - if it matches the *-out.R file, the test has passed. You can create an *-in.R file, run devtools::test(f = "[your file]")
and an *-out.R file is generated. If the file matches your expectation, you can commit it. Note that files are overwritten and version control should be used to track failed tests. The files are placed in tests/testthat under the category they fit. Please have a look at the documentation for test_collection() and see other unit tests.
Also, I hope that you can understand that one or two tests are not sufficient to test the new style guide. There are always unexpected cases such as the one you are presenting above. styler currently has almost 150 tests for the tidyverse style guide and it fooled my over and over again. We need more comprehensive testing. Because you can't think of all cases possible, I think it is helpful to use the source code of another package, style it with the style guide you are suggesting, look at the git diff. Then, pick a few other cases and add them as to the *-in.R test suite. In your case it is pretty easy because everything should be left unchanged for a package that does not contain arrow assignment. We can't add styling of whole packages as unit tests, but we should at least do it once and cover a few cases. That's at least how I did it for oneliner, maybe you have better suggestions. In that case, let me know. At some point, we may want to build infrastructure to do that automatically with travis, but I guess its not a priority now.

As far as the indention problem goes we need use_raw_indention = TRUE:

library(styler)
force_assignment_op_equals <- function(pd) {
  to_replace <- pd$token == "LEFT_ASSIGN"
  pd$token[to_replace] <- "EQ_ASSIGN"
  pd$text[to_replace] <- "="
  pd
}

equals_style <- function() {
  create_style_guide(
    token = list(force_assignment_op_equals),
    use_raw_indention = TRUE
  )
}

txt <- "x = c(
  1
)"
style_text(txt, style = equals_style)
#> x = c(
#>   1
#> )

^{Created on 2018-12-05 by the reprex package (v0.2.1)}

I also updated CONTRIBUTING.md to cover testing. It may contain other interesting information for you about contributing to styler.

Good news on the testing front: equals_style() has already proved useful in tidying-up style in a package with equals assignment after commits adding <- assignment: itsleeds/stats19@327e436

Some of those examples could be good in the tests I imagine.

See my comments in the repo there. Maybe there is an unexpected interaction with #381. Also, you could rebase on master to make the (unrelated) CI failure on R 3.2 go away now.

Think I've borked the rebase. Can create a new clean PR if that would make sense - suggested way forward unless there are better suggestions.