Add c, seq, mean methods so ITime retains class

[MichaelChirico]

Closes #3628

Also deleted split.IDate and cut.IDate methods. They are untested with no examples, and I really don't know what the expected behavior is. split returns a list and really the important part of the method is how to handle the splitting factor f, not the first argument. split.default already retains the IDate class:

dput(split.default(as.IDate(0L), 'foo'))
# list(foo = structure(0L, class = c("IDate", "Date")))

And cut returns a factor so I don't know how as.IDate is supposed to operate on that.

[codecov]

Codecov Report

Merging #3630 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #3630      +/-   ##
==========================================
+ Coverage   98.19%   98.19%   +<.01%     
==========================================
  Files          66       66              
  Lines       12923    12924       +1     
==========================================
+ Hits        12690    12691       +1     
  Misses        233      233

Impacted Files	Coverage Δ
R/IDateTime.R	98.75% <100%> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 421f672...e1767f6. Read the comment docs.

[codecov]

Codecov Report

Merging #3630 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff            @@
##           master   #3630      +/-   ##
=========================================
+ Coverage   98.19%   98.2%   +<.01%     
=========================================
  Files          66      66              
  Lines       12923   12980      +57     
=========================================
+ Hits        12690   12747      +57     
  Misses        233     233

Impacted Files	Coverage Δ
R/IDateTime.R	98.75% <100%> (ø)	⬆️
src/nafill.c	100% <0%> (ø)	⬆️
src/init.c	100% <0%> (ø)	⬆️
R/shift.R	96.42% <0%> (+0.13%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 421f672...e46f7c6. Read the comment docs.

[jangorecki]

Ideally to raise warning about functions to be deprecated, then in next version remove. Also mention in NEWS.

[MichaelChirico]

In this case, I don't think those functions work at all, so I'm not sure how anybody could have used them... a Note in NEWS maybe

[jangorecki]

I understand but it is unwise just to drop exported functions without any warning.