Allow non-R solution code #290
Conversation
* Don't try to parse non-R solution code * .solution isn't available in these cases * Update mock_this_exercise() to follow same flow
gadenbuie
force-pushed
the
parse-check-r-only
branch
from
e1be1ca
to
c622ec4
Compare
instead we can get it from `.result` if that's an error condition
| @@ -169,7 +169,7 @@ check_exercise <- function( | |||
| ) | |||
|
|
|||
| # Skip parse checking if it was already done in {learnr} | |||
| if (!learnr_includes_parse_check(stage)) { | |||
| if (!learnr_includes_parse_check(stage) && identical(learnr_args$engine %||% "r", "r")) { | |||
There was a problem hiding this comment.
In the future, I'll think we'll want to standardize how we discover what language a piece of code is, but this is fine for now
| engine <- knitr_engine_caption(learnr_args[["engine"]]) | ||
| if (!identical(engine, "R")) { |
There was a problem hiding this comment.
Same here, I'd rather have a separate function or pattern to check for the exercise/code language, but I think that's a larger amount of work we'll want to tackle later on
There was a problem hiding this comment.
Quick question: why does engine == "R", but learnr_args$engine == "r"?
There was a problem hiding this comment.
yeah that's totally one of the reasons to standardize. knitr_engine_caption() turns engine strings like r, sql, javascript into text suitable for the exercise caption, e.g. R, SQL or JavaScript. So "r" or NULL comes out as "R".
There was a problem hiding this comment.
(In this case we're prepping the string for use in the message, so it makes sense to compare with "R".)
| @@ -81,11 +85,11 @@ mock_this_exercise <- function( | |||
| .label = "mock", | |||
| .engine = "r", | |||
| .stage = "check", | |||
| .result = rlang::missing_arg(), | |||
There was a problem hiding this comment.
What's the difference between .result = rlang::missing_arg() and just .result with no default value?
There was a problem hiding this comment.
If .result isn't provided, we want to create .result from the .user_code, provided that it's R code that we can evaluate.
If we don't have R code though, we need the user to tell us what the result will be. That result might be NULL, so I'm using rlang::missing_arg() to tell the difference.
We could leave .result without a default value and use missing(.result) to catch that, but I prefer being explicit. I also prefer using "no default value" as a strong "required argument" signal; in this case .result isn't required.
.solutionisn't available in these casesmock_this_exercise()to follow same flow