Hi List,
I was looking at the query plan of a rather simple query, but I don't
understand why sqlite would choose this query plan.
...I was surprised that sqlite came up with the inferior query plan...
Note: After an analyze aaa (on a decently populated table) sqlite chooses
the full
an automatic index (but our application
never uses 'analyze' to avoid that other (bad performing) query plans are used
during operation than during testing).
Note 2: Adding NOT INDEXED to aa1 gives the desired query plan, but of course
I prefer that sqlite choses the right query plan.
Regards,
Rob
Regards
Rob Golsteijn
Hi all,
I have some rows in a table (not very many, typically less than 20) and I
want to generate a unique, sequential number for each row. In another dbms
I've used a row_number function (amongst others) to achieve this but I can't
see anything with equivalent
when typing an incomplete sql statement in the shell and
then press CTRL-D.
Sqlite version is 3.8.4.3.
Regards,
Rob Golsteijn
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
Dan Kennedy Wed, 04 Jun 2014 03:56:33 -0700
On 06/04/2014 05:06 PM, Rob Golsteijn wrote:
Hi List,
I noticed that the sqlite shell core dumps when it is started with an init
file that ends with an incomplete statement. Example:
Init file called my_init.sql
JOIN A ON A.a*A.a + B.b*B.b = C.c*c.c JOIN B ON A.a*A.a +
B.b*B.b = C.c*c.c;
c a b
-- -- --
5 3 4
5 4 3
-Urspr?ngliche Nachricht-
Von: Rob Golsteijn [mailto:rob.golste...@mapscape.eu]
Gesendet
cannot
handle or is this just invalid SQL?
In the latter case it would be nice if SqLite complained about it).
Rob Golsteijn
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
);
For Examples 1-3 I would expect an error for duplicate table aliases.
Regards,
Rob Golsteijn
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
.
My question is whether the change is a bug or an intended change?
From the release history's text it looks like it could have been introduced by
the changes in Ticket 2500cdb9be05
Regards,
Rob Golsteijn
___
sqlite-users mailing list
sqlite
On Thu, Jun 20, 2013 at 10:26 AM, Rob Golsteijn
rob.golste...@mapscape.euwrote:
SELECT a.col2,
b.col2,
a.b.col2,-- invalid column name
b.a.col2,-- invalid column name
a.a.col2,-- invalid column name
b.b.col2
to reject such invalid names and report an error,
instead of silently ignore the first part of the column name? Or did I miss
something?
Regards,
Rob Golsteijn
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin
Output:
PRAGMA foreign_keys=OFF;
BEGIN TRANSACTION;
CREATE TABLE test (r REAL);
INSERT INTO test VALUES(Inf);
INSERT INTO test VALUES(-Inf);
COMMIT;
Error: near line 4: no such column: Inf
Error: near line 5: no such column: Inf
---
Rob Golsteijn
The following code is intended to change text values 'UND' into NULL values.
However, I see that string values different from 'UND' get corrupted. In the
example below the value 'ENG' is corrupted and the new value is a string of
three non-printable characters (ascii values 0x03 0x17 0x13) in
13 matches
Mail list logo