[Bioc-devel] affy's cleancdfname

[Seth Falcon]

On 26 Sep 2005, ririzarr at jhsph.edu wrote:

> i believe that is what was intended. if the cdf name comes from an
> enviromnent or an object in the workspace then it is assumed that
> the correct name is given and it does not have to be inferred from
> the information in the celfile.
>
> is someone having problems because of this?

No user that I know of.  But I am having problems cleaning this bit of
code up.  The main thing I'm trying to accomplish is to remove the
dependency on reposTools.  

reposTools will be going away and so the more packages that don't
depend on it, the better.

It seems to me that cdfFromBioC() should only have the code to
download the package and then call cdfFromLibPath() --- otherwise the
code is duplicated.  But this means a double call of cleancdfname.

How about option #2 then?

+ seth

>
> On Mon, 26 Sep 2005, Seth Falcon wrote:
>
>> There seems to be some inconsistencies in when cleancdfname is
>> called in getCDFenv.R.
>>
>> getCdfInfo() calls the following helpers:
>> cdfFromEnvironment() ## no call to cleancdfname
>> cdfFromData()        ## no call to cleancdfname
>> cdfFromLibPath()     ## calls cleancdfname
>> cdfFromBioC()        ## calls cleancdfname
>>
>> I have two alternatives to propose and would like comments:
>>
>> 1. remove cleancdfname calls from cdfFrom* funcs and put the call
>>    in
>> getCdfInfo().  This feels right in that we call the func once, but
>> it changes the behavior of two of the cdfFrom* funcs where to use
>> them a user would now have to call cleancdfname themselves.
>>
>> 2. Change behavior of cleancdfname so that repeated calls don't add
>> additional "cdf" to the end.  This would make the assumption that
>> no valid cdf package name ends "cdfcdf".
>>
>> + seth
>>
>> _______________________________________________
>> Bioc-devel at stat.math.ethz.ch mailing list
>> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>>