[BioC] Affymetrix file parsers released under LGPL open source license

[Richard Finney]

How I got the Affy file parser stuff to compile 
on Linux ...

First, I downloaded and unziped the zip files.
I created a file called istream 
in the current working directory that just included
the line "#include <stdio.h>".  I tried including
the place where the "istream" file  lives with a -I
but kept
having to put in more and more -I's option 
to g++ but at the  end I had several screenfuls of
"already defineds".
Yuk.  So I punted.  The "#include <stdio.h>" in the
fake "istream" file is
needed to provide a prototype for the sscanf()
function.

Then I put a "int main(){return 0;}" in
ReadAffxFiles.cpp
so that it had a "main" to get happy about.
The code appears to be written with Microsoft "C"
which as we all know is ... well ... different.

Then I created and ran the following batch file.

___ batch file for unix to compile, cut here ____
g++ -w -I./ -I/usr/include/ \
-I/home/gcc/gcc-3.0.2/libstdc++-v3/libsupc++/ \
BARFileData.cpp \
BPMAPFileData.cpp \
CDFFileData.cpp \
CELFileData.cpp \
CHPFileData.cpp \
FileIO.cpp \
ReadAffxFiles.cpp

## Note, make a file called istream in the current
working directory, it should have
## this next one line (including the # which doubles a
comment line in this script)....  
#include <stdio.h>
____ end cut _____

Great.  It creates an a.out file.

Note the -I./ directive so that the compiler finds
the "istream" file in the current working directory
and notice the -w
parameter.  The -w parameter apparently shuts up
all the warnings.  You'll get several "miscasting"
warnings if you don't use it.  Somebody should
do the proper castings if they get real motivated.
Note also the "-I/...libsupc++/" include directive.
That's where libsupc++  is on my Clinton era home
redhat box and it will probably live in a different
place on your more correctly
configured deluxe 4GHz gentoo machine.
Use "find / -type d -name 'libsupc++' -print" to 
find it.

The actual invocation of useful functions ?
I'll leave that to the eager and desperate readers
of this post.  The writer/poster of the original
code at Affy provides a documentation file on
how to talk to his code among the 4 downloadable
zip files.

Somebody (Ben Bolstad, I think) wrote some binary
CEL file parsing routines in Bioconductor.
This provides an alternate starting point
for anybody interested in hacking around with
CEL files.

-rich



--- Harry Mangalam <hjm at tacgi.com> wrote:

> Thanks very much Mike - this is a very welcome
> addition.
> 
> I've read the docs and scanned the code for
> comments, but I could not 
> find which version of gcc/g++ this can be built
> with.  
> Running Debian Sid, both g++ 2.95 and g++ 3.4 give
> errors right off 
> the bat, mostly in FileIO.cpp  Before I plunge into
> a debug session, 
> is there more guidance you can give as to preferred
> compilers and 
> flags?
> 
> 
> If anyone else has been successful with this, I'd be
> interested in the 
> details. I'm adding a skeleton autoconf setup, so
> any info would be 
> welcome.
> 
> hjm
> 
> On Monday 13 December 2004 2:44 pm, Lelivelt, Mike
> wrote:
> > Dear Bioconductor community:
> >
> >
> >
> > Affymetrix is pleased to release another component
> of its software
> > development kit to the open source community.  We
> have released C++
> > source code to parse many of our files, including
> our CEL v4 binary
> > format, under an LGPL open source license.  Please
> visit the URL
> > below to download the code and support materials
> directly.
> >
> >
> >
> > Thanks, mike
> >
> >
> >
> >
>
http://www.affymetrix.com/support/developer/filesdk/index.affx
> >
> >
> >
> > Mike Lelivelt, Ph.D. | Sr. Manager, Informatics
> Applications |
> > Office 608-985-6440  | Mobile 608-239-6937
> >
> > AFFYMETRIX, INC. | 3380 Central Expressway | Santa
> Clara, CA 95051
> > | eFax 253-461-6969
> >
> >
> >
> >
> >  [[alternative HTML version deleted]]
> >
> > _______________________________________________
> > Bioconductor mailing list
> > Bioconductor at stat.math.ethz.ch
> > https://stat.ethz.ch/mailman/listinfo/bioconductor
> 
> -- 
> Cheers, Harry
> Harry J Mangalam - 949 856 2847 (vox; email for fax)
> - hjm at tacgi.com 
>             <<plain text preferred>>
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
>