Dear R-devel,

Question:

How can I get Sweave to recognize that the \thebibliography section is already created and generate the vignette pdf with a single pass? Or is there a way to let R CMD build know that Sweave needs to be run twice?

If (1) is not possible, any suggestions how to auto-generate a crude generic bibtex .bib file from my existing document?

Background:

I maintain serveral R packages that include package vignettes. The vignettes include the reference citations inline as\bibitem lines inside a thebiblilography section. For some reason, this requires running Sweave twice on the document (the first pass is presumably preformatting the citation information?) to produce a pdf. This seems to be the source of problems when included in R's preferred vignette location (/vignettes/myPackage.Rnw) because the R package builder just runs Sweave once during the check process, so no pdf is produced and warnings are generated like

checking package vignettes in ‘inst/doc’ ... WARNING
 Package vignette without corresponding PDF/HTML

In the past, it has been possible for me to pre-build the pdf, and include it in the /inst/doc/ directory. But it seems the upcoming versions of R may prohibit this.

It appears that it is possible to avoid all this by including a bibtex .bib file along with the vignette instead of doing the citations inline. If the .bib files exists, it seems to generate with a single pass. But reformatting all of the citations for each document into bibtex will be a huge pain. Including a blank \bibliography{} command above my \thebibliography environment also allows the document to be build in a single pass, but generates a warning about duplicate biliography and writes the section header twice (not surprisingly).


Note: I've cross-posted this question at:
http://stackoverflow.com/questions/22237675/how-to-get-sweave-to-recognize-thebibliography-environment-and-build-r-package

thanks for your help,
 best,
 -skye

______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

Reply via email to