If you’ve ever needed to just dump a GAE model back to the browser (through JSON serialisation for example) then you know it’s a bit of a pain in the arse. Have no fear though, I’ve written a little class for you:
#!/usr/bin/env python
def serialise_model(m, d):
# let's serialise this bad boy
ps = m.properties()
# check if we're dealing with a list or not
if isinstance( d, list ):
# prepare list for return
r = [ ]
# cycle through the list
for i in d:
r.append( _serialise_model_cycle( ps, i ) )
# return list
return r
else:
# just cycle through the data
return _serialise_model_cycle( ps, d )
def _serialise_model_cycle(ps, d):
# prepare the return dict
r = { }
# we cycle through the properties of the model
for k, p in ps.iteritems():
v = getattr( d, k )
# if the value isn't empty, we add it to the list
if v is not None and v != '':
# set the value
r[ k ] = v
# don't forget the key
r[ 'key' ] = str( d.key() )
# return the data
return r
def serialise_model(m, d):
# let's serialise this bad boy
ps = m.properties()
# check if we're dealing with a list or not
if isinstance( d, list ):
# prepare list for return
r = [ ]
# cycle through the list
for i in d:
r.append( _serialise_model_cycle( ps, i ) )
# return list
return r
else:
# just cycle through the data
return _serialise_model_cycle( ps, d )
def _serialise_model_cycle(ps, d):
# prepare the return dict
r = { }
# we cycle through the properties of the model
for k, p in ps.iteritems():
v = getattr( d, k )
# if the value isn't empty, we add it to the list
if v is not None and v != '':
# set the value
r[ k ] = v
# don't forget the key
r[ 'key' ] = str( d.key() )
# return the data
return r
(It’s also up on Github as a gist: https://gist.github.com/946538. This is how you use it:
Step 1: import it
from serialise import serialise_model
Step 2: get your model (it can be a list too)
model = Model().get( ... )
Step 3: serialise it’s ass
serial = serialise_model( model )
And then you can take ‘serial’ and serialise it ready to send to the browser 
