[Bioperl-l] Re: problem with alignment test data & parsers

Peter Schattner schattner@alum.mit.edu
Thu, 24 May 2001 08:48:29 -0700


Heikki Lehvaslaiho wrote:

> Peter,
>
> I innocently added code into Bio::LocatableSeq::end to test if the
> value given is supported by start() and seq(). Almost all of the test
> data sets in t/data turned out to be wrong.
>
> It seems to be that the problem got started from the Stockholm format
> file.
> The parser seems to read in only the first line from each sequence and
> ignores the rest. Apparently these parse results have then been
> written into other test files in various formats. Being not familiar
> with the Stockholm format I can see two options:
>
> 1. The alignment file is not correct and the extra new line
>    characters should be removed.
> 2. The regexp doing the parsing needs to do quite clever look ahead
>    tricks to see when the next sequence starts.
>
> I've committed the changed LocatableSeq::end into bioperl-live to help
> in sorting out this problem. The SimpleAlign-related tests will fail
> until the problem is fixed.
>

Heikki

I can't tell whether you are saying that you intend to investigate this
problem yourself (which would make me happy since my plate is quite full
at the moment. ) or whether you would like me to look into it - in which
case  do e-mail the script that's demonstrating the problem and I'll add
this to my "to do" list.

Cheers

Peter