do not export as.Date.IDate anymore #4782
Conversation
|
FWIW i've got a mirror of CRAN locally, statically checked if any revdeps (recursive depends/imports) were using the previous export, didn't find any. |
|
FWIW comes up too when letting GitHub search in the CRAN mirror: https://github.com/search?q=org%3Acran+as.Date.IDate |
NAMESPACE
Outdated
| S3method(as.Date, IDate) # note that zoo::as.Date masks base::as.Date. Both generic. | ||
| export(as.Date.IDate) # workaround for zoo bug, see #1500. Removing this export causes CI pipeline to fail on others.Rraw test 6, but I can't reproduce locally. | ||
| if (getRversion() >= "3.6.0") { | ||
| S3method(as.Date, IDate) |
There was a problem hiding this comment.
@jangorecki Why >= "3.6.0"? Won't as.Date(x) break in R < 3.6.0 if the method is not exported? At least a comment on why >= 3.6.0 should be added.
Also could the comments be put back so we know there's some history on these lines pointing to #1500 and reminding us that zoo::as.Date masks base::as.Date.
There was a problem hiding this comment.
Higher up in the file the switch on 3.6.0 is about delayed registration (i.e. the pkg:: prefix on the first argument to S3method) which was new in R 3.6.0. But that's not the case here with as.Date.
There was a problem hiding this comment.
I figured the >=3.6 was copied from higher up. So I just went ahead and removed it in 6a20166, to merge and let's see if it passes GLCI. If this was wrong then we can always do a follow up PR.
There was a problem hiding this comment.
When I added the suggest version for zoo, I did bit64 and bit at the same time (although nothing to do with this PR).
Codecov Report
@@ Coverage Diff @@
## master #4782 +/- ##
=======================================
Coverage 99.47% 99.47%
=======================================
Files 73 73
Lines 14557 14557
=======================================
Hits 14481 14481
Misses 76 76 Continue to review full report at Codecov.
|
closes #4777