Deprecate extraction_operator_linter()

[IndrajeetPatil]

Closes #1485
Closes #1484

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (3fd90ba) 98.64% compared to head (a96cb77) 98.64%.

❗ Current head a96cb77 differs from pull request most recent head fdbaa1b. Consider uploading reports for the commit fdbaa1b to get more accurate results

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2409   +/-   ##
=======================================
  Coverage   98.64%   98.64%           
=======================================
  Files         126      125    -1     
  Lines        5623     5628    +5     
=======================================
+ Hits         5547     5552    +5     
  Misses         76       76           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[MichaelChirico]

thanks!

[IndrajeetPatil]

Ah, there are new warnings detected by the vigilant workflow that need to be dealt with before we can merge this.

[IndrajeetPatil]

I don't understand what's going on here.

In the dummy_packages/package we are using in test-cache.R (e.g.), we don't specify extraction_operator_linter in the config and just set linters: linters_with_defaults(), so why do we see the following warning?

Trying to remove 'extraction_operator_linter', which is not in defaults.

Full traceback:

── Warning (test-cache.R:446:3): cache = TRUE workflow works ───────────────────
Warning encountered while loading config:
  Warning from config setting 'linters' in 'NULL':
    Trying to remove 'extraction_operator_linter', which is not in `defaults`.
Backtrace:
     ▆
  1. ├─lintr::lint_package(pkg, cache = TRUE) at test-cache.R:446:3
  2. │ └─lintr:::read_settings(pkg_path) at lintr/R/lint.R:249:5
  3. │   └─lintr:::read_config_file(config_file) at lintr/R/settings.R:74:3
  4. │     ├─base::withCallingHandlers(...) at lintr/R/settings.R:140:3
  5. │     ├─base::tryCatch(load_config(config_file), error = malformed)
  6. │     │ └─base (local) tryCatchList(expr, classes, parentenv, handlers)
  7. │     │   └─base (local) tryCatchOne(expr, names, parentenv, handlers[[1L]])
  8. │     │     └─base (local) doTryCatch(return(expr), name, parentenv, handler)
  9. │     └─lintr (local) load_config(config_file)
 10. │       ├─base::withCallingHandlers(...) at lintr/R/settings.R:113:9
 11. │       ├─base::tryCatch(...)
 12. │       │ └─base (local) tryCatchList(expr, classes, parentenv, handlers)
 13. │       │   └─base (local) tryCatchOne(expr, names, parentenv, handlers[[1L]])
 14. │       │     └─base (local) doTryCatch(return(expr), name, parentenv, handler)
 15. │       ├─base::eval(parsed_setting)
 16. │       │ └─base::eval(parsed_setting)
 17. │       └─lintr::all_linters(...)
 18. │         └─lintr::linters_with_tags(tags = NULL, packages = packages, ...) at lintr/R/with.R:141:3
 19. │           └─lintr::modify_defaults(..., defaults = tagged_linters) at lintr/R/with.R:124:3
 20. │             └─base::warning(...) at lintr/R/with.R:48:5
 21. ├─base::.signalSimpleWarning(...)
 22. │ └─base::withRestarts(...)
 23. │   └─base (local) withOneRestart(expr, restarts[[1L]])
 24. │     └─base (local) doWithOneRestart(return(expr), restart)
 25. ├─lintr (local) `<fn>`(`<smplWrnn>`)
 26. │ └─base::warning(...) at lintr/R/settings.R:125:13
 27. ├─base::.signalSimpleWarning(...)
 28. │ └─base::withRestarts(...)
 29. │   └─base (local) withOneRestart(expr, restarts[[1L]])
 30. │     └─base (local) doWithOneRestart(return(expr), restart)
 31. └─lintr (local) `<fn>`(`<smplWrnn>`)

[AshesITR]

Traceback no 17. shows that it runs all_linters(). Seems to pick up our config, where the warning is correct because all_linters() excludes deprecated and defunct linters by default.

TL;DR: Remove extraction_operator_linter = NULL from our .lintr.

[IndrajeetPatil]

TL;DR: Remove extraction_operator_linter = NULL from our .lintr.

This is what I had thought, but then I realized that this is going to cause the lint workflow to fail IINM.
I hope we don't have to add # nolint directives to everywhere $ is appearing.