From: Joseph Wang <joe@gn...> - 2006-06-28 04:48:29
Unless someone tells me otherwise, I'm planning on committing the R interface
to SWIG tomorrow evening. I've put the suggested changes that have been
posted to swig-devel, and it looks good to go at this end.
From: William S Fulton <wsf@fu...> - 2006-06-29 21:45:07
Joseph Wang wrote:
> Unless someone tells me otherwise, I'm planning on committing the R interface
> to SWIG tomorrow evening. I've put the suggested changes that have been
> posted to swig-devel, and it looks good to go at this end.
Your checkin looks good and doesn't break anything, which is unusual
because new modules usually break something:) However, saying that,
configure.in is missing parts. You need to add something in to detect R
in order to run the test-suite and examples. The @SKIP_R@ in Makefile.in
will not expand until you do this.
A couple of other minor points. Can you keep the consistency in naming
conventions, eg we don't use lower camel case naming conventions for
testcases, so copyStruct_runme.R / copyStruct.i and simpleArray.i needs
renaming. Actually, although these appear in your
test-suite/r/Makefile.in file, these .i files do not exist. Right now it
looks like the test-suite won't run, but I don't have R installed to
check but it looks to me as though the test-suite prints out a pass or
fail message (unittest.R) If so, please make this work like the other
languages, where silence is golden and only problems make themselves known.
Get latest updates about Open Source Projects, Conferences and News.