`na` argument ignored in read_sheet

[antoine-sachet]

Specifying custom NA values in a non-character column does not work: the NA values are treated as character cells instead of being set to NA.

This is related to other opened issues about a better column type interface (#51 and in particular #64), but it is a sufficiently distinct use case to warrant a separate issue IMO.

I have created a public spreadsheet (1D98U_4wx2pm2-0hUuOOJNM0NKDv-WouP2IRuBFnWrRg) containing:

A	B	C	D
1	1	1	1
Missing	NA		2
3	3	3	3

Reprex:

googlesheets4::read_sheet("1D98U_4wx2pm2-0hUuOOJNM0NKDv-WouP2IRuBFnWrRg", 
    na = c("", "NA", "Missing"))
#> Reading from 'Reprex for guessed ctype in presence of NAs'
#> 
#> # A tibble: 3 x 4
#>   A         B             C     D
#>   <list>    <list>    <dbl> <dbl>
#> 1 <dbl [1]> <dbl [1]>     1     1
#> 2 <chr [1]> <chr [1]>    NA     2
#> 3 <dbl [1]> <dbl [1]>     3     3

^{Created on 2019-12-09 by the reprex package (v0.3.0)}

Of course, I was expecting the behaviour from readr.

  
csv_txt <- "A   B   C   D
1   1   1   1
Missing NA      2
3   3   3   3"

readr::read_delim(csv_txt, na = c("", "Missing", "NA"), delim = "\t")
#> # A tibble: 3 x 4
#>       A     B     C     D
#>   <dbl> <dbl> <dbl> <dbl>
#> 1     1     1     1     1
#> 2    NA    NA    NA     2
#> 3     3     3     3     3

^{Created on 2019-12-09 by the reprex package (v0.3.0)}

[jennybc]

I now see this

devtools::load_all()
#> Loading googlesheets4

ssid <- "1D98U_4wx2pm2-0hUuOOJNM0NKDv-WouP2IRuBFnWrRg"
sheets_browse(ssid)
read_sheet(ssid, na = c("", "NA", "Missing"))
#> Reading from 'Reprex for guessed ctype in presence of NAs'
#> 
#> # A tibble: 3 x 4
#>       A     B     C     D
#>   <dbl> <dbl> <dbl> <dbl>
#> 1     1     1     1     1
#> 2    NA    NA    NA     2
#> 3     3     3     3     3

^{Created on 2019-12-10 by the reprex package (v0.3.0.9001)}