Antoine Fabri

The last case however is a bug, I have code that attempts to map `?`to the standard help when it's relevant, and this code is clunky. Side note, please don't...

Before that we also have a bug due to `modify_qm_calls()` not recursing (calling `?` when finding typed nested functions.

To add to the list: * `for(p in periods) {}` loops on the seconds * `ifelse(TRUE, period1, period2)` returns the seconds (`dplyr::if_else()` works fine). * `period_1 | period_2` and `period_1...

Relevant: https://github.com/tidyverse/dplyr/issues/1045 A few notes: * `bind_rows()` is not a generic so to support lazy tables work needs to be done in {dplyr} * It would be `union_all()` for sure,...

FWIW `rlang::expr_deparse(quote({}))` is `"{ }"` with a space, `base::deparse()` inserts a line break.

the keyword is source markers: see https://github.com/ThinkR-open/golem/pull/595/commits/398b5bb2a3a28cd81f6c60df3397ce3571685637

Use case: https://stackoverflow.com/questions/73833284/r-merging-data-frame-and-concatenating-values/73833671#73833671