[mythtv] SQL prepare / bindValue patch for mythtv
Isaac Richards
ijr at case.edu
Tue Dec 7 00:02:21 UTC 2004
On Monday 06 December 2004 06:27 pm, David Härdeman wrote:
> A sage old MythTV wizard once said...
>
> On Sat, Nov 20, 2004 at 12:46:10AM -0500, Isaac Richards wrote:
> >Looks like a very good start.
>
> [...]
>
> >Instead of changing all those "s to 's in the queries, I'd also really
> >prefer if things were switched to use the prepare/bindValue methods.
> >I think it's cleaner, and it takes care of proper quoting/etc..
>
> I've now completed part one of this *phew*
Excellent.
> The result is a 200k patch which changes almost all code under mythtv/libs
> to prepare/bindValue functions, available here:
> http://www.hardeman.nu/~david/mythpatch/mythtv-libs-preparebindvalue.patch
>
> The patch obviously touches alot of files, and has been lightly tested,
> so thread carefully...on the other hand, the number of files also means
> that the patch might bitrot quite quickly (so I'd be happy if people
> tested it and gave feedback/commited it as quickly as possible).
Does this do the right thing wrt testing for NULL? I noticed in
mythcontext.cpp, you've got a
+ query.bindValue(":HOSTNAME", "NULL");
IIRC, Qt is just going to substitute NULL for the placeholder on exec, but we
need to do 'IS NULL' and not '= NULL', right?
Isaac
More information about the mythtv-dev
mailing list