Review: Needs Fixing
Hello Yogesh,
I don't think this fix is right or sufficient. We should not fix the low-level
functions like get_nice_size(), otherwise we will replicate this dict-aware
code in many places.
The functions we use to post-process the field values should only be concerned
about the value for one field, and should be very simple.
The get() method should be the one doing this work: instead of directly
processing the results depending on the type, it should first check if this is
a multi result, and if it is, then handle each result in the appropriate manner
according to its type.
Remember that a multi function can compute fields of different types.
The code for handling a multi-result and taking individual results in it should
only be in one place, and then code for handling one result of a certain type
should also only be in one place.
In pseudo-code:
def get(...):
....
if is_multi:
for field,value in result.iteritems():
# post-process according to type of field
result[id][field] = postprocess(obj.columns[field]._type,value)
else:
# post-process according to type of field
result[id] = postprocess(self._type, result[id])
The pseudo-code is dumb but I hope this illustrates that the code for handling
a multi-result dict is only in one place (get), and the code for
post-processing the values is only in one place (here in a "postprocess" method
that can do the right thing according to the type)
I hope is clear, otherwise don't hesitate to ask ;-)
--
https://code.launchpad.net/~openerp-dev/openobject-server/trunk-bug-781247-ysa/+merge/61933
Your team OpenERP R&D Team is subscribed to branch
lp:~openerp-dev/openobject-server/trunk-bug-781247-ysa.
_______________________________________________
Mailing list: https://launchpad.net/~openerp-dev-gtk
Post to : [email protected]
Unsubscribe : https://launchpad.net/~openerp-dev-gtk
More help : https://help.launchpad.net/ListHelp
