[ESS] noweb file and assignment key
maechler at stat.math.ethz.ch
Wed Oct 31 10:48:16 CET 2012
>>>>> "RM" == Richard Morrisroe <richie.morrisroe at gmail.com>
>>>>> on Tue, 30 Oct 2012 20:57:32 +0000 writes:
RM> Hi All, While attempting to avoid real work, I appear to
RM> have found a possible cause for this kind of error.
RM> See the .Rnw file below.
now committed to the ESS development sources as
RM> Essentially, it appears that an apostrophe ' will
RM> confuse ess-smart-underscore and it then believes that
RM> it is in a string,
RM> where the normal behaviour does not occur.
exactly: You diagnosed perfectly.
RM> Another apostrophe fixes the problem. I hope this proves
it does, thank you.
The bad news is that this bug is still in the ESS development
sources ... which I had wanted to release as ESS 12.09-1
Now this small nastyness is really something we (ESS core
developers) should address before release.
Martin Maechler, ETH Zurich
RM> <<packages, echo=FALSE, results=hide>>=
RM> i.can.haz.assginment <- TRUE @
RM> Now, we will introduce an extraneous apostrophe, for no
RM> reason '
RM> <<test2>>= #ess-smart-underscore-no-longer-works @
RM> ' <<test3, echo=TRUE, results=verbatim >>= nowitdoes <-
RM> cat("Note the apostrophe above which closes the earlier
RM> one\n") @
RM> I have attempted to localise the problem more precisely,
RM> but have not had much success.
>>>>> Denis Haine <denis.haine at gmail.com>
>>>>> on Mon, 29 Oct 2012 11:26:36 -0400 writes:
>> Hello, After upgrading to emacs 24 (emacs 24.2.1, ess
>> 12.09) I noticed the underscore key is not mapped anymore
>> to the R/S assignment sign when into a Rnw buffer. The
>> underscore key is still correctly giving the assignment
>> sign when into a R buffer. How can I restore the usual
>> underscore behavior in a noweb buffer? Thanks for your
> I'm pretty sure this was not a universal problem. Our
> group here uses ESS 12.09 (I'm using the development
> version), and we use Sweave ( *.Rnw ) very very often.
> Nobody has told me about such a problem.
> Couldn't it be that it happens exactly in the situation
> where The switch from R mode ("code chunk") to latex
> mode ("text chunk") does not quite work... and we know
> that happens not so infrequently. But you can always
> work around that by "re syncing" or sometimes easiest by
> re-opening the *.Rnw file.
More information about the ESS-help