Adding option to overwrite "single null" values

[ecqiu]

RcppSimdJson has the option to overwrite single null values with the 'single_null' option in RcppSimdJson::fparse, this can be useful e.g. in overwriting NULLs to NAs when reading the JSON rather than later for speed purposes.

Should yyjsonr also have this option?

Here is an example of how the option works:

> RcppSimdJson::fparse('{"x":null,"y":null,"z":3}')
$x
NULL

$y
NULL

$z
[1] 3

> RcppSimdJson::fparse('{"x":null,"y":null,"z":3}',single_null = NA)
$x
[1] NA

$y
[1] NA

$z
[1] 3

Related PR:
#29

[m-muecke]

I'm not sure if this issue covers the same issue, but it would be lovely if a list with NULLs has the option of getting converted to NAs like {jsonlite} does it with the simplify = TRUE flag, for reference the implementation: https://github.com/jeroen/jsonlite/blob/86f481d34eb4e7893e32a88608ce220421f59c41/R/list_to_vec.R#L1

x <- "[{\"transaction\":[{\"price\":[0.2715,0.1030,null,null],\"missing\":[null,null,null,null]}]}]"
jsonlite::fromJSON(x, simplifyDataFrame = FALSE)
#> [[1]]
#> [[1]]$transaction
#> [[1]]$transaction[[1]]
#> [[1]]$transaction[[1]]$price
#> [1] 0.2715 0.1030     NA     NA
#> 
#> [[1]]$transaction[[1]]$missing
#> [1] NA NA NA NA
yyjsonr::read_json_str(x, opts = yyjsonr::opts_read_json(obj_of_arrs_to_df = FALSE, arr_of_objs_to_df = FALSE))
#> [[1]]
#> [[1]]$transaction
#> [[1]]$transaction[[1]]
#> [[1]]$transaction[[1]]$price
#> [1] 0.2715 0.1030     NA     NA
#> 
#> [[1]]$transaction[[1]]$missing
#> [[1]]$transaction[[1]]$missing[[1]]
#> NULL
#> 
#> [[1]]$transaction[[1]]$missing[[2]]
#> NULL
#> 
#> [[1]]$transaction[[1]]$missing[[3]]
#> NULL
#> 
#> [[1]]$transaction[[1]]$missing[[4]]
#> NULL

^{Created on 2024-12-06 with reprex v2.1.1}