I think the difficulty here is that the optimizer is oriented toward the
"low level" and mainly concerned with choosing indexes, processing order
etc (see https://www.sqlite.org/optoverview.html ) and has sort of a
narrow view of the task.
Approaching it from the other end and breaking down the S
I have the following (simplified and generalized) schema and query:
CREATE TABLE nodes (
id INTEGER NOT NULL PRIMARY KEY
);
CREATE TABLE edges (
a INTEGER NOT NULL REFERENCES nodes (id),
b INTEGER NOT NULL REFERENCES nodes (id),
PRIMARY KEY (a, b)
);
WITH reachable (a, b) AS (
SELECT a
On 5/27/15, Matthias-Christian Ott wrote:
> It looks like the optimization can be generalized because you can pull
> the selection into the initial select query of the CTE. Is there a
> chance that SQLite could perform this optimization in one of the next
> releases?
>
Are you offering to fund t
3 matches
Mail list logo
