Hi, With the usage of python-sql, I measure an increase for the execution of the unittest by about 37%. It was expected to have such increase because python-sql adds a new layer which of course have a cost for the benefit it provides such:
- more optimized queries
- more database independency
- better modularity
I made some measurement and it seems that the extra time come mainly
from the convertion to string and parameters. So the creation of the
objects that constitute the query seems to have the same cost as when it
was done by string concatination.
I already done some improvement since the release 0.2 [1]:
- add array support as arrays are like list but enforce the type,
the string/parameter can be optimized. Patch for trytond is
comming for reduce_ids
- using list internally for params
- reduce call to Flavor.get
I made also some optimization test which did not succeed:
- using generator to compute paramters
- using list to generate string before concate it with: ''.join()
but the code became too much difficult to read.
So I'm calling for help to optimize further python-sql.
I'm even not against adding some C optimization if required.
[1] https://code.google.com/p/python-sql/source/list
Thanks,
--
Cédric Krier
B2CK SPRL
Rue de Rotterdam, 4
4000 Liège
Belgium
Tel: +32 472 54 46 59
Email/Jabber: cedric.kr...@b2ck.com
Website: http://www.b2ck.com/
pgpBeRFL3Lxip.pgp
Description: PGP signature
