False positive in extraction_operator_linter() with plotmath expressions

[IndrajeetPatil]

One can use plotmath to display mathematical expressions in a graphic:

library(ggplot2)
expr <- list(quote(italic("x")["i"]))
ggplot() + labs(title = parse(text = expr))

^{Created on 2022-07-29 by the reprex package (v2.0.1)}

But the code for subscript produces an incorrect lint:

library(lintr)

lint(
  text = "list(quote(italic('x')['i']))",
  linters = extraction_operator_linter()
)
#> <text>:1:23: warning: [extraction_operator_linter] Use `[[` instead of `[` to extract an element.
#> list(quote(italic('x')['i']))
#>                       ^

^{Created on 2022-07-29 by the reprex package (v2.0.1)}

The suggestion is misleading as it won't produce the same output:

library(ggplot2)
expr <- list(quote(italic("x")[["i"]]))
ggplot() + labs(title = parse(text = expr))

^{Created on 2022-07-29 by the reprex package (v2.0.1)}

[AshesITR]

Maybe we could optionally skip quote() and friends.

[IndrajeetPatil]

Yeah, I agree. We can also skip its close cousins from {rlang} package.

[MichaelChirico]

I think there are many such cases where extraction_operator_linter() is giving false positives (e.g., [[ method is not defined for the object, or it really is intentional to maintain the class but subset to one element, e.g. iris["Species"] is perfectly valid).

Not sure how worth effort it is to maintain exceptions here, authors dinged by this linter often should simply not use it IMO :)

[MichaelChirico]

Perhaps a better choice is to add parameters: allow_single_extraction and allow_dollar_extraction that toggle which of the usages are linted.