I have a number of files which can be read and converted into a series of
lines of lext - however the means of reading the
file is not known to the standard Hadoop splitters. I understand that I can
Override FileInputFormat to set isSplitable to false -
I am a little unclear on how to get the Job to Use my version of
that FileInputFormat  and nowhere do I see a place to
override the code for reading the file and converting it to lines of text.
Anyone know how to do this??

--
Steven M. Lewis PhD
Institute for Systems Biology
Seattle WA

Steven,

> I have a number of files which can be read and converted into a series of
> lines of lext - however the means of reading the
> file is not known to the standard Hadoop splitters. I understand that I can
> Override FileInputFormat to set isSplitable to false -
> I am a little unclear on how to get the Job to Use my version of
> that FileInputFormat  and nowhere do I see a place to
> override the code for reading the file and converting it to lines of text.
> Anyone know how to do this??

Could you look at JobConf.setInputFormat() API to set your input format ?

Thanks
Hemanth