Remove pipe on end of commented code prior to checking parsability

[jcken95]

Closes #2671

The PR removes pipes from the end of extracted_code when checking parsability of code. A test to check that a lint is detected on commented lines ending with a pipe (which are otherwise parsable) are indeed detected.

This solves a common false negative occurrence in, for example, data wrangling

mtcars |>
# dplyr::filter(cyl > 4) |>
   dplyr::summarise(m = mean(wt))

The above snippet detect no commented code lints, but changes in this PR would lead to a commented code linter being detected

[MichaelChirico]

@AshesITR / @IndrajeetPatil WDYT?

[AshesITR]

I have often seen commented pipe steps in code I reviewed. I think it's a useful addition unlikely to produce many false positives.

Maybe we could also change the linters behavior to check if the code parses when removing the comment character from a consecutive block of comments as well, i.e. make it "context-aware". That would necessitate making it file-level, though and might break some stuff, so that's definitively out-pf-scope here.

[MichaelChirico]

Thanks!