Re: SQLite weirdness
| From: | Wez Furlong | Date: | Fri, 09 Jul 2004 18:04:09 +0000 |
| Subject: | Re: SQLite weirdness | ||
| References: | 1 2 | Groups: | php.internals |
| Request: | Send a blank email to internals+get-11034@lists.php.net to get a copy of this message | ||
Stas already came up with a solution for the non-oo api; it probably
just needs porting to the oo interface.
--Wez.
On Fri, 9 Jul 2004 09:06:30 +0200 (CEST), Derick Rethans <derick@php.net> wrote:
> On Thu, 8 Jul 2004, Derick Rethans wrote:
>
> > After some investigation I found that if I removed the "$ret = " before
> > the query() call, then all queries are run just fine, and throw also a
> > warning when I put back in the typoes.
>
> I did some more investigations and it seems that as soon you try to
> assign the result of sqlite_query() to a variable it always only execute
> the first query in a multi-query query-string. Not only for CREATE as I
> wrote before but it's also a problem for multiple INSERTs, or a mixed
> INSERT/SELECT. In the code it's the difference between sqlite_compile
> (sqlite.c:1472) (used when there is no assignment) and sqlite_exec
> (sqlite.c:1559) when there is an assignment.
>
> A slightly modified "expected" list:
> >
> > I would have expected something else though:
> > 1. Without the "$ret = " it should throw a PHP warning for each broken
> > query. (That's current what it's doing).
> > 2. With the "$ret = " it should NOT throw any warnings, and set the
> > $db->lastError() value accordingly whether there were errors
> > (with the last query) or not.
>
> IMO this should be fixed ASAP, I can perhaps try to come up with a patch
> myself too.
>
>
>
> regards,
> Derick
>
> --
> PHP Internals - PHP Runtime Development Mailing List
> To unsubscribe, visit: http://www.php.net/unsub.php
>
>