This may affect other concurrent apps which use this adapter. Not sure.
On Monday, 14 January 2013 19:31:52 UTC-6, DenesL wrote:
>
>
> This also seems to work:
>
> db._adapter.parse_blob=lambda value,field_type: value
>
> Any cons?.
>
--
This also seems to work:
db._adapter.parse_blob=lambda value,field_type: value
Any cons?.
--
Thank you.
On Sunday, January 13, 2013 11:52:19 PM UTC-5, Massimo Di Pierro wrote:
>
> web2py uses b64encode/decode to store/retrieve data in blobs. Legacy
> tables probably do not do it.
>
> You need you make your own custom adapter to override a method. For
> example for PostgreSQL:
>
> from
web2py uses b64encode/decode to store/retrieve data in blobs. Legacy tables
probably do not do it.
You need you make your own custom adapter to override a method. For example
for PostgreSQL:
from gluon.dal import ADAPTERS, UseDatabaseStoredFile,PostgreSQLAdapter
class MyPostgresAdapter(PostgreS
Suggestions to override parse_blob are welcome.
--
It seems to be from the data, it is a legacy table.
Using 4607 trunk a manual select gives:
Traceback (most recent call last):
File "", line 1, in
File "C:\w2p\trunk\gluon\dal.py", line 9181, in select
return adapter.select(self.query,fields,attributes)
File "C:\w2p\trunk\gluon\dal.py"
I have fixed the first error too but not sure where the second comes from.
Can you help?
On Sunday, 13 January 2013 09:25:13 UTC-6, DenesL wrote:
>
> Thank you, it still needs more work:
>
> Traceback (most recent call last):
> File "C:\w2p\trunk\gluon\restricted.py", line 212, in restricted
>
Thank you, it still needs more work:
Traceback (most recent call last):
File "C:\w2p\trunk\gluon\restricted.py", line 212, in restricted
exec ccode in environment
File "C:\w2p\trunk\applications\test\controllers/rest.py", line 28, in
File "C:\w2p\trunk\gluon\globals.py", line 193, in
I have implemented what you requested:
patterns = [ ('/dogs[pet]', db.pet.kind=='dog') ]
Can you please check it?
On Monday, 31 December 2012 11:30:34 UTC-6, DenesL wrote:
>
> +1 for an extended syntax instead of routes.
>
> Blog post about http://blog.2partsmagic.com/restful-uri-design/ which
I think this can be implemented. Will try later today.
On Monday, 31 December 2012 11:30:34 UTC-6, DenesL wrote:
>
> +1 for an extended syntax instead of routes.
>
> Blog post about http://blog.2partsmagic.com/restful-uri-design/ which
> discusses some schemes (about halfway down, with the title
^bump^
--
+1 for an extended syntax instead of routes.
Blog post about http://blog.2partsmagic.com/restful-uri-design/ which
discusses some schemes (about halfway down, with the title "Choosing a URI
schemes for resource hierarchies" (sic)).
One thing to think about is how to define the creation of new r
It could be possible via routes but it would be nice it the syntax were to
allow it. I stopped developing it because I though instead of reinventing
the wheel with this syntax we may want to support some standard. Is there
any?
On Sunday, 30 December 2012 00:01:03 UTC-6, DenesL wrote:
>
> Yes,
Yes, from the docs it is clear that one could use such syntax but the idea
is to hide it somehow.
So instead of /pets/dog I want just /dogs
Note that I am providing all parameters for the query in the pseudo-code
{pet.kind.eq.dog} to avoid having another arg.
On Saturday, December 29, 2012 12:4
you can say something like:
"/pets[pet]/{pet.kind.eq}"
and call /pets/dog or /pets/cat
It is not documented because I am convinced this syntax is definitive. It
needs more etsting and as you suggest it may be richer.
massimo
On Saturday, 29 December 2012 11:32:25 UTC-6, DenesL wrote:
>
> Hi
15 matches
Mail list logo
