Hadley Wickham
Hadley Wickham
We already have to shim `library()` so I'm not sure what this would get us.
Ooooh it's because it's not `library()` generating the errors directly, it's because duckdplyr requires dplyr. I retract my objection 😄
Hard to see how to fix without breaking this test: ```R test_that("primitive functions are wrapped", { expect_identical(as_mapper(`-`)(.y = 10, .x = 5), -5) expect_identical(as_mapper(`c`)(1, 3, 5), c(1, 3, 5)) })...
I think this should be a simple fix given that this behaviour is unlikely to affect any code in practice — just delete the `typeof(.f) %in% c("special", "builtin")` branch and...
With zero recollection of the code my first guess this is because we use `ls()` somewhere.
Duplicate of #1572. Thanks for working so hard to create a minimal reprex!
This seems like a reasonable approach to me. Are you still interested in finishing it off?
Hmmm, that's because `R CMD check` runs checks in a special library that only includes the packages explicitly listed in your `DESCRIPTION`. So from testthat's perspective, curl isn't installed. But...
FYI purrr is now scheduled for release to CRAN on Dec 19