In that case, I would put onto hackage a library "BiobaseBLAST" which has the Biobase.BLAST.Tabulated.* modules (.hs /.Import.hs / .Export.hs) with an Iteratee interface.
You factor out the XML part and then we check if we should put everything into one library? Or should we try out having one library? ... It would be easier if hackage had a good way to mark librarier as deprecated so they don't show up anymore. Or do they have and I just don't know it? Gruss, Christian >Christian Höner zu Siederdissen <choe...@tbi.univie.ac.at> writes: > >> yes I have a definite need to parse /huge/ amounts of tabulated BLAST >> data (order of gigabytes, don't ask ;-) So, BLAST parsing would >> definitely be useful. I think, we should discuss this a bit, though >> -- I have become a fan of the Iteratee library and parsing such >> amount of data would make Iteratee a good choice. > >Yes - I've only done this by streaming through, and only handled the >XML format, so it's not likely any of my stuff is very useful to you, >anyway. > >> But if your current approach handles this as well, I'd be happy too. >> And if, on the other hand you would want to concentrate on the xml >> part, I could write the tabulated part. > >It's possible that this is better handled by separate libraries - BLAST >makes different information available in the different formats, so it's >likely that the data structures will need to reflect these differences >- although common stuff could be factored out, of course. > >-k >-- >If I haven't seen further, it is by standing in the footprints of >giants _______________________________________________ Biohaskell mailing list Biohaskell@biohaskell.org http://malde.org/cgi-bin/mailman/listinfo/biohaskell
