On 11/16/2013 10:46 AM, RSmith wrote:
Perhaps we should make the allowed DDL subset a part of the spec.
That way we make explicit what is allowed and anything outside of
that is forbidden. Pepijn
Perhaps.
It would involve a rather large document though, one which an average
user is sure to skip over but at least it provides indemnity, plus I
don't see an enormous or sudden uptake of WITHOUT ROWID and/or partial
Index tables in the general populous. By the time this is a
generality, you should be a couple of versions of your system further
down the product line.
Ryan
It might be simpler to simply specify a minimum release of SQLite that
must be supported, and forbid any DDL statements that could not be
parsed at that release, giving partial indexes (and maybe without rowid
tables) as an example.
(Pepijn, when a you first access any object in database containing an
object created using an unsupported DDL command, the prepare will fail
with:
SQLITE_CORRUPT 11 /* The database disk image is malformed */
As that will trigger a parse of the contents of the sqlite_master
table. Note that this does not actually happen when you first open the
database, unless you have a statically loaded extension that accesses a
database object.)
Peter
_______________________________________________
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
