Duration columns

[tvatter]

Hi,

I am aware of #549 and #1037, but it does not really address my problem. Additionally, I already filed a similar bug report on the lubridate repo as I am unsure whether the responsibility to handle the following issue falls on readr or lubridate...

Here's a regex of my issue:

df <- tibble(x = duration(1, "years"))     # x is a duration
tmp <- tempfile(); write_csv(df, tmp)    # write the df to file
df2 <- read_csv(tmp)                           # x is now a character which can't be easily coerced

The main issue is that as.duration(as.character(duration(1, "years")))) returns NA because the parser from the duration class is designed with a different purpose in mind.

Now, I get that one could just convert a duration as a numeric before writing, via mutate(across(where(is.duration), as.numeric)).
But while I understand that readr can't handle every class out there, I was expecting the duration class, being provided by a tidyverse package, to be handled "properly" by write_csv.

Thoughts?

[vspinu]

Fixed in tidyverse/lubridate#1005