Make `epi_slide_opt` work with `!!varname`, maybe `!!varnames`, maybe `!!!varnames`

[brookslogan]

library(epiprocess)
#> Registered S3 method overwritten by 'tsibble':
#>   method               from 
#>   as_tibble.grouped_df dplyr
#> 
#> Attaching package: 'epiprocess'
#> The following object is masked from 'package:stats':
#> 
#>     filter
varname <- "case_rate"
varnames <- c("case_rate", "death_rate")
covid_case_death_rates_extended %>% epi_slide_sum(!!varname, .window_size=7)
#> Error in !varname: invalid argument type
covid_case_death_rates_extended %>% epi_slide_sum(!!varnames, .window_size=7)
#> Error in !varnames: invalid argument type
covid_case_death_rates_extended %>% epi_slide_sum(!!varnames, .window_size=7)
#> Error in !varnames: invalid argument type

^{Created on 2024-11-03 with reprex v2.1.1}

We probably don't need the if branch here and can probably just use the else branch code in any situation.

  if (inherits(quo_get_expr(enquo(.col_names)), "character")) {
    pos <- eval_select(dplyr::all_of(.col_names), data = .x, allow_rename = FALSE)
  } else {
    pos <- eval_select(enquo(.col_names), data = .x, allow_rename = FALSE)
  }

• !!varnames probably will work, but I'm not sure if this is intended.
• !!! may or may not work; it may be more "proper" for it not to work, as it looks like splatting multiple arguments into the function, while our .col_names is just a single arg slot. (If we look at the similar across function, it will balk with a confusing message if we attempt to use !!!, because it is actually splatting multiple fn args. But I don't think we have fn arg splatting here so it might actually work in our case.)

Workaround is to use manually use all_of(varname) or all_of(varnames) instead of something with !! or !!!.