fix apc processing for W2101805856

[trangdata]

Closes #305.

library(openalexR)
paper <- oa_fetch(identifier = "W2101805856", output = "list")
paper$apc_list
#> $value
#> [1] 3500000
#> 
#> $currency
#> [1] "IRR"
#> 
#> $value_usd
#> NULL
#> 
#> $provenance
#> [1] "doaj"
if (any(lengths(paper[c("apc_list", "apc_paid")]) > 0)) {
  apc_fields <- list(value = NA, currency = NA, value_usd = NA, provenance = NA)
  apc <- list(rbind.data.frame(list(
    c(type = "list", utils::modifyList(apc_fields, as.list(paper$apc_list))),
    c(type = "paid", utils::modifyList(apc_fields, as.list(paper$apc_paid)))
  )))
}
#> Error in rbind.data.frame(list(c(type = "list", utils::modifyList(apc_fields, : invalid list argument: all variables should have the same length

# my current fix is to use `replace_w_na` but open to more elegant solution
if (any(lengths(paper[c("apc_list", "apc_paid")]) > 0)) {
  apc_fields <- list(value = NA, currency = NA, value_usd = NA, provenance = NA)
  apc <- list(rbind.data.frame(
    c(type = "list", utils::modifyList(apc_fields, openalexR:::replace_w_na(paper$apc_list))),
    c(type = "paid", utils::modifyList(apc_fields, openalexR:::replace_w_na(paper$apc_paid)))
  ))
}
apc
#> [[1]]
#>   type   value currency value_usd provenance
#> 1 list 3500000      IRR        NA       doaj
#> 2 paid      NA     <NA>        NA       <NA>

^{Created on 2024-12-20 with reprex v2.1.1}

[yjunechoe]

Thanks! Agreed that NULL needs special handling here.

I pushed a minor edit: instead of resolving them to the logical NA via replace_w_na(), I'm attempting a (more verbose) option of removing NULL entries and coalescing them with appropriately-typed NAs (i.e., making apc_fields more robust).

LMK how this looks!

tibble::as_tibble(oa_fetch(identifier = "W2101805856")$apc[[1]])
#> # A tibble: 2 × 5
#>   type    value currency value_usd provenance
#>   <chr>   <dbl> <chr>        <dbl> <chr>     
#> 1 list  3500000 IRR             NA doaj      
#> 2 paid       NA <NA>            NA <NA>

[trangdata]

Looks great @yjunechoe. I considered this as well (and also something like apc_list[sapply(apc_list, is.null)] <- NULL) and ended up just picking one but I'm glad you have a preference. I can merge when you approve!

[yjunechoe]

Let's merge!