#53: Consider avoiding encoding/decoding bytea in queries
--------------------------+-------------------
Reporter: cito | Owner:
Type: enhancement | Status: new
Priority: major | Milestone: 6.0
Component: C Module | Version: 4.1
Resolution: | Keywords: bytea
--------------------------+-------------------
Changes (by cito):
* milestone: 5.1 => 6.0
Comment:
I see two problems if we want to use binary ''only for bytea values'':
1. For input, it is possible to send only some of the parameters as
binary. In Python 3, maybe we could do this for parameters sent in as
bytes. But in Python 2, we can't discern strings from bytes in the input
values. So I suggested postponing this to a future Python-3-only-version
of !PyGres.
2. For output, it is currently not even possible to send only some of the
result columns as parameters. Also, we would need to analyze the query
first to find which columns are bytea value.
If we decide to use binary ''for all values'', this is difficult because
the binary representations must match and according to the PostgreSQL docs
it also requires knowledge of the internal representation expected by the
backend.
Postponing this to 6.0 for now.
--
Ticket URL: <http://trac.pygresql.org:8000/pgtracker/ticket/53#comment:1>
PyGreSQL <http://www.pygresql.org/>
PyGreSQL Tracker
_______________________________________________
PyGreSQL mailing list
[email protected]
https://mail.vex.net/mailman/listinfo/pygresql
