Data.table: [bug] possible regression in as.IDate.

[arunsrinivasan]

require(data.table) # built 2017-10-31
as.IDate("20170929", "%Y%m%d")
# Error in charToDate(x) : 
#   character string is not in a standard unambiguous format
as.Date("20170929", "%Y%m%d")
# [1] "2017-09-29"

This worked fine in 1.10.4 and also on current CRAN version, but fails on devel.

Output of sessionInfo()

# R version 3.3.2 (2016-10-31)
# Platform: x86_64-w64-mingw32/x64 (64-bit)
# Running under: Windows Server >= 2012 x64 (build 9200)
# 
# locale:
# [1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252   
# [3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C                          
# [5] LC_TIME=English_United States.1252    

# attached base packages:
# [1] stats     graphics  grDevices utils     datasets  methods   base     
# 
# other attached packages:
# [1] data.table_1.10.5
# 
# loaded via a namespace (and not attached):
# [1] tools_3.3.2 yaml_2.1.14

[MichaelChirico]

%Y%m%d is being passed as the tz argument:

as.IDate("20170929", format = "%Y%m%d")
# [1] "2017-09-29"

Here's the commit that changed, I think that was from @jangorecki's PR #1393

Personally I'm in the habit of always writing out format =, what do you think?

[arunsrinivasan]

I don't think there's a need for tz argument for .default method. That's what breaks things in this scenario. There's a tz argument added for .POSIXct method in the same PR, which is also consistent with as.Date and should be sufficient IMHO.

[MichaelChirico]

Agreed. Perhaps Jan had something in mind...

[mattdowle]

Might be my fault as I had a hand in that PR. I'll take a look ...

[arunsrinivasan]

Awesome! Thanks for the super-quick fix!

[aadler]

@MichaelChirico I also write out format; in some testing I've found it to be slightly faster, which adds up in data.table-sized data structures.