On May 22, 2009, at 6:19 PM, Marek Szuba wrote:

> By the way, do you think we will still need
> classutil.default_tmp_path() after this change?

No, better to use Python's standard way of doing things.

> in the "filepopen_cwd" branch attempts to make it possible for called
> processes to use a different working directory than the current one
> even if the subprocess module (whose Popen() method has got a  
> dedicated
> keyword for this) is not present. This will be needed if, as Titus
> suggested, we are to make formatdb run in the same directory as its
> data files.

I'm not convinced that we want to go this route.  It seems klugey --  
the user who tells Pygr to run formatdb or blastall is definitely not  
expecting Pygr to go changing the working directory as a side  
effect...  Correct me if I'm wrong, but this temporary directory  
change would affect all threads running within our Python process,  
right?

Istvan suggested we only release a Windows package for Python 2.5.  In  
that case the standard library subprocess module takes care of the cwd  
keyword argument for us.  Therefore the only possible concern is  
Python 2.3 on unix.  If we really want to handle that case, we can do  
it trivially within the shell command we pass to system() -- this  
works on both unix and Windows/cygwin.  Then the directory change  
happens in the child process, without affect the Python process at all.


> 3. The "blast_test" branch has been extended with more parser-only
> tests. Now every existing BLAST test which parses blastall output has
> its canned-output counterpart.

Great!  We talked about adding a multiseq blast test to check that the  
multiseq results match the combined results of running all the  
sequences individually.  Did you add that?

-- Chris

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"pygr-dev" group.
To post to this group, send email to pygr-dev@googlegroups.com
To unsubscribe from this group, send email to 
pygr-dev+unsubscr...@googlegroups.com
For more options, visit this group at 
http://groups.google.com/group/pygr-dev?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to