Hin-Tak Leung
2011-Aug-04 00:38 UTC
[Rd] R check mis-identifying Illumina data file format as executable and other thoughts.
This is somewhat a summary/continuation of an R bug report: (https://bugs.r-project.org/bugzilla3/show_bug.cgi?id=14645) Illumina's cluster definition files (*.egt) are one of the proprietary and undocumented file formats used by their GenomeStudio line of products for genomic studies. snpMatrix 1.17.0.7 onwards (http://sourceforge.net/projects/outmodedbonsai/files/snpMatrix%20next/) contains codes for reading that file format, as well as two example files of such type generated from public data, and also a vignette demonstrating their usefulness and relevance to genomic studies. R svn check (the upcoming 2.14) mis-identifies those bundled files as undeclared binary executable files and aborts with that as error. (1) the files are not executables, they just happened to be mis-identified as such (2) even if they are genuine binary executables, there might be legitimate reasons to bundle them with a package? (e.g. I have R code to look at their content, and just treat them as arbitrary proprietary undocumented formats). I can think of some rather interesting possible enhancement to R core's bytecode-compiler for parsing and morphing genuine binary executables - that might be useful for just-in-time compilation for R on the android platform. So surely there are genuine/legitimate needs for shipping binary executables in an R package?
Apparently Analagous Threads
- Sweave, cairo_pdf, CJK, ghostscript
- Thai vignette, cross-compile for Mac OS X, universal/multiarch (Fwd: Mac OS X builds of CelQuantileNorm, vcftools/samtools/tabix, and snpStats)
- updated R-cairo bridge, official R-3.1.*-mavericks.pkg crippled, snpMatrix 1.19.0.20
- patch to add cairo support to Sweave (Re: Sweave, cairo_pdf, CJK, ghostscript)
- updated R-cairo bridge, official R-3.1.*-mavericks.pkg crippled, snpMatrix 1.19.0.20