The reporting of formats failing to parse is a bit inconsistent. When all values passed to parse_date_time are NA, the warning of "all formats failed to parse" is given. When some (but not all) are NA, no warning is given for the NA values even though technically it's the same type of issue.
Feature request: If all values are NA, do not warn that all formats failed to parse. (I know I could use the quiet argument to silence the warning, but I want the warning in the case of actual parsing errors.)
> parse_date_time(NA,
+ orders=c("%Y-%m-%d %H:%M", "%d%b%Y:%H:%M:%S", "%d%b%y:%H:%M"))
Warning: All formats failed to parse. No formats found.
[1] NA
> parse_date_time(c(NA, "2017-01-01 01:01"),
+ orders=c("%Y-%m-%d %H:%M", "%d%b%Y:%H:%M:%S", "%d%b%y:%H:%M"))
[1] NA "2017-01-01 01:01:00 UTC"
> parse_date_time(c("foo", NA, "2017-01-01 01:01"),
+ orders=c("%Y-%m-%d %H:%M", "%d%b%Y:%H:%M:%S", "%d%b%y:%H:%M"))
Warning: 1 failed to parse.
[1] NA NA "2017-01-01 01:01:00 UTC"
The reporting of formats failing to parse is a bit inconsistent. When all values passed to parse_date_time are NA, the warning of "all formats failed to parse" is given. When some (but not all) are NA, no warning is given for the NA values even though technically it's the same type of issue.
Feature request: If all values are NA, do not warn that all formats failed to parse. (I know I could use the
quietargument to silence the warning, but I want the warning in the case of actual parsing errors.)