add assignment_as_infix = TRUE to indentation_linter()

[AshesITR]

fixes #1800

Currently, this only suppresses indentation for descendants of <--triggered indents and only if these descendants are

• infix operators
• not within a function call

I'm not 100% sure this fixes all cases, but I added a more complex test case to ensure it works as intended.
Let me know if you can think of other test cases that should be included.

Also, let me know if the exception should apply to other assignment-like operators, such as EQ_ASSIGN, EQ_SUB, EQ_FORMALS, ...

One case where we probably still have a false-positive is within lapply() and friends. That seems hard if not impossible to fix:

lapply(x,
  function(e) {
    temp_var <-
      e +
      42
  }
)

[codecov-commenter]

Codecov Report

Merging #1812 (e36addf) into main (e0ebf6b) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #1812   +/-   ##
=======================================
  Coverage   98.86%   98.86%           
=======================================
  Files         112      112           
  Lines        4841     4852   +11     
=======================================
+ Hits         4786     4797   +11     
  Misses         55       55           

Impacted Files	Coverage Δ
R/indentation_linter.R	100.00% <100.00%> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[AshesITR]

Link rot check is getting HTTP 500s at the moment. All other checks pass ✔️

[MichaelChirico]

thanks!

[AshesITR]

Thanks for your feedback.
I added OP-LEFT-BRACE to the resetting tokens and EQ_SUB, EQ_ASSIGN and EQ_FORMALS to the suppressing tokens.
The new tests cover all the new exceptions.

I hope the comment is more clear on how the XPath works now.

[AshesITR]

@MichaelChirico LMK when you have the time to review. I want to avoid spamming merges.

[MichaelChirico]

nit: i usually prefer each argument with defaults on its own line once >1 line is used, WDYT? not sure it's a stated rule, but it's similar to the long lines rule for calls.

[MichaelChirico]

nit: maybe a switch() is more appropriate? change_type feels pretty enum-like

[MichaelChirico]

that's some horrible code 😄

[MichaelChirico]

Only nits left, great work!