Added na.rm for Summary.numeric_version

[marberts]

https://bugs.r-project.org/show_bug.cgi?id=19000

[llrs]

I'm sure this will help with the summary() but how will it help with max(), min(), range()? They will be called independently of the summary.numeric_version() method.

[marberts]

Thanks for taking a look @llrs. This is for the Summary.numeric_version() method for the group generic (capital S), not summary.numeric_version(). (There actually isn't a method for summary() for numeric_version objects.) Each of min(), max(), and range() dispatch to Summary.numeric_version() when passed a numeric_version object, and are the only such functions that don't immediately throw an error for not being applicable.

Hopefully this clarifies the patch, but please let me know if the patch/bugzilla report aren't clear so that I can fix it.

[llrs]

Hi, I am interested on improving support for package_version and saw your enhancement (my PR is on #228 but somehow it breaks the CI).

Now that I think of it it would be nice if a test is added on tests/reg-test-1e.R to make sure this works as intended.
I didn't follow up how max(), min(), range() are dispatched internally to Summary But if the tests pass then it will be easier to merge (and I'll be reassured that it happens without checking too deep the code :).

[llrs]

Great! One last things is to directly attach the diff to bugzilla as it might make easier for R core members to pick that than check the link and copy the patch from the web.

[marberts]

Good idea. I've updated the bugzilla.