[wb_to_df] fix date1904 conversion

[JanMarvin]

⚠️ Previously wb_to_df() was not using date1904 detection, hence conversions via wb_to_df() were always assuming an origin of 1900-01-01 contrary to 1904-01-01. This file setting is somewhat exotic and we did provide tools to let the user detect the error, it is nevertheless a shameful mistake. Sorry.

DATE  <- as.Date("2015-02-07") + -10:10
POSIX <- as.POSIXct("2022-03-02 19:27:35") + -10:10
time <- data.frame(DATE, POSIX)

library(openxlsx2)
tmp <- temp_xlsx()
wb <- wb_workbook()
wb$workbook$workbookPr <- '<workbookPr date1904="true"/>'
wb$add_worksheet()$add_data(x = time)
wb$save(tmp)

wb_to_df(wb) %>% head()
#>         DATE               POSIX
#> 2 2015-01-28 2022-03-02 19:27:25
#> 3 2015-01-29 2022-03-02 19:27:26
#> 4 2015-01-30 2022-03-02 19:27:27
#> 5 2015-01-31 2022-03-02 19:27:28
#> 6 2015-02-01 2022-03-02 19:27:29
#> 7 2015-02-02 2022-03-02 19:27:30
wb_load(tmp)$to_df() %>% head()
#>         DATE               POSIX
#> 2 2015-01-28 2022-03-02 19:27:25
#> 3 2015-01-29 2022-03-02 19:27:26
#> 4 2015-01-30 2022-03-02 19:27:27
#> 5 2015-01-31 2022-03-02 19:27:28
#> 6 2015-02-01 2022-03-02 19:27:29
#> 7 2015-02-02 2022-03-02 19:27:30
wb_to_df(tmp) %>% head()
#>         DATE               POSIX
#> 2 2015-01-28 2022-03-02 19:27:25
#> 3 2015-01-29 2022-03-02 19:27:26
#> 4 2015-01-30 2022-03-02 19:27:27
#> 5 2015-01-31 2022-03-02 19:27:28
#> 6 2015-02-01 2022-03-02 19:27:29
#> 7 2015-02-02 2022-03-02 19:27:30