[FR] Allow showing expression instead of values in arg_match() error message

[mikmart]

Sometimes I need to match an argument against a long (100+) list of valid values. Currently the arg_match() error message will be overwhelmed by listing out all possible valid values. In such cases, it would be useful to have a way to include the expression from the formal argument in the message instead.

For example with:

foo <- function(dataset = ls("package:datasets")) {
  rlang::arg_match(dataset)
}

foo("Titanci")
#> Error in `foo()`:
#> ! `dataset` must be one of "ability.cov", "airmiles", "AirPassengers",
#>   "airquality", "anscombe", "attenu", "attitude", "austres", "beaver1",
#>   "beaver2", "BJsales", "BJsales.lead", "BOD", "cars", "ChickWeight",
#>   "chickwts", "co2", "CO2", "crimtab", "discoveries", "DNase", "esoph", "euro",
#>   "euro.cross", "eurodist", "EuStockMarkets", "faithful", "fdeaths",
#>   "Formaldehyde", "freeny", "freeny.x", "freeny.y", "HairEyeColor",
#>   "Harman23.cor", "Harman74.cor", "Indometh", "infert", "InsectSprays", "iris",
#>   "iris3", "islands", "JohnsonJohnson", "LakeHuron", "ldeaths", "lh",
#>   "LifeCycleSavings", "Loblolly", "longley", "lynx", "mdeaths", "morley",
#>   "mtcars", "nhtemp", "Nile", "nottem", "npk", "occupationalStatus", "Orange",
#>   "OrchardSprays", "PlantGrowth", "precip", "presidents", "pressure",
#>   "Puromycin", "quakes", "randu", "rivers", "rock", "Seatbelts", "sleep",
#>   "stack.loss", "stack.x", "stackloss", "state.abb", "state.area",
#>   "state.center", "state.division", "state.name", "state.region", "state.x77",
#>   "sunspot.month", "sunspot.year", "sunspots", "swiss", "Theoph", "Titanic",
#>   "ToothGrowth", "treering", "trees", "UCBAdmissions", "UKDriverDeaths",
#>   "UKgas", "USAccDeaths", "USArrests", "UScitiesD", "USJudgeRatings",
#>   "USPersonalExpenditure", "uspop", "VADeaths", "volcano", "warpbreaks",
#>   "women", "WorldPhones", or "WWWusage", not "Titanci".
#> ℹ Did you mean "Titanic"?

It would be nice to be able to instead get:

foo("Titanci")
#> Error in `foo()`:
#> ! `dataset` must be one of `ls("package:datasets")`, not "Titanci".
#> ℹ Did you mean "Titanic"?

[olivroy]

Alternatively,

Could arg_match() also trim down the number of choices if there are many, like cli does with its {.or }

arg_match0("28", as.character(1:27))
#> Error:
#> ! `"28"` must be one of "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "17", "18", "19", "20",
  "21", "22", "23", "24", "25", "26", or "27", not "28".
#> ℹ Did you mean "18"?
cli::cli_text("{.or {.str {as.character(1:27)}}}")
#> "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "17", "18", …, "26", or "27"