Re: [XMLPipeDB-developer] UniProtDatabaseProfile submits

[John D. N. D. <do...@lm...>]

OK, thanks.  I will look into issue #1 concurrently and will report back.  Regarding issue #2, I confirmed with Kam that we do need the per-species value for each ID.  As we are prioritizing database approaches first, here you have a distinct exercise in SQL formation that is relevant to the tutorial course:

- Modify the formed query so that it returns, for each ID, the corresponding species name.  If successful, you will then have both ID and species name to provide to tableManager.submit, and you will not need an outer loop for every species profile.

The way I suggest you do this is to form a valid query directly in PostgreSQL (i.e., via pgAdmin or psql).  That should be pretty easy to get since you are logging the final query string --- just copy and paste that.  Experiment with the query directly within PostgreSQL until you have one that works.  Then transplant that into the code.

John David N. Dionisio, PhD
Associate Professor, Computer Science
Associate Director, University Honors Program
Loyola Marymount University



On Sep 2, 2011, at 11:04 AM, Richard Brous wrote:

> OK, so still two outstanding issues regarding the last 2 modified methods:
>  
> 1. getInfoTableManager()
>  
> Still returning 3 species names when only two are selected. Returns one species name when only one is selected.
> - I will review the loop again and see if I can determine what is going on here.
>  
> 2. getSystemTableManager()
>  
> I added a for loop within the while condition which after discussions we know is incorrect. Will pull that out but we need to determine how to move forward with the issue of determining which species each id belongs to. removed loop and submitted to SourceForge but now build is essentially broken till we figure out the wrinkle. Adding a for speciesprofile loop at the top was discussed as a solution of last resort.
>  
> UniProtDatabaseProfile - while loop begins at line # 585
>  
> ---------------------------------------------------------------------------------------------------------------
>  
> Also made modifications to getSystemsTableManager() which applies the species customizations for each species profile selected. Seems correct to me but please confirm.
>  
> --------------------------------------------------------------------------------------------------------------
>  
> Discussions surrounding getPrimarySystemTableManager() assumed likely bad coding since it seems to call for a customized schema although it is identical to the default UniProt schema.
>  
> It calls getPrimarySystemTableManagerCustomizations() which is specific to Arabidopsis thaliana and Ecoli only.
>  
> Didn't want to dig into this yet until other higher priority modifications were done first plus I believe a discussion with Dr. D was going to occur as well. 
> 
> ---------------------------------------------------------------------------------------------------------------
>  
> getRelationsTableManager() is the last method which remains untouched but I will start reviewing and come up with a plan of attack.
>  
> --------------------------------------------------------------------------------------------------------------
>  
> These are all the loose ends I can think of to document.
>  
> Thanks!
>  
> Richard
>  
>  
>  
> On Wed, Aug 31, 2011 at 12:52 PM, Richard Brous <rbr...@gm...> wrote:
> Submitted changes to UniProtDatabaseProfile based on the logical corrections from yesterdays convo with Dondi.
>  
> These do not yet include the discussed optimizations, just wanted to get it working and do some test exports... which completed successfully.
>  
> I also added a for loop under the while loop in getSystemTableManager() to ensure each query result is processed for each selected species profile.
> The loop is on line 587 and I added it because it seemed a solution to have each species profile included... is this correct or should I be submitting all at once versus each species?
>  
> Richard
> 
> <ATT00001..txt><ATT00002..txt>