Re: [Gnumed-devel] ? top down faster for emr browser

[Syan Tan]

I think it's spent doing the processing, as I'm using python os.system ('openssl enc -aes128 -pass file:pass.txt')

etc .. and starting a new process and new input and output streams for each word marked as

requiring encryption after parsing the narrative. The narrative parsing isn't too fast either,

as I remember when just replacing marked words with 'xxxx' .

I think I'm also doing a sql lookup on the word / marked  table for every word parsed in the narrative to see if it is marked.

The updates are written into a file
as a batch update of 101 rows , and then another system process of "psql -f batchxxx "is run at each iteration

of 101 rows. I am only using one connection.

On Mon May 1 0:07 , Karsten Hilbert sent:

> > narratives in a duplicate database. ( this
> > takes a long time, about 12 hours or more for  360,000 rows). 
> Thanks for hammering the schema. Is most of this time spent
> reading/updating the database ? Does it use a lot of new
> connections.