Re: WITHOUT ROWID tables

Ok, not ideal. Still confusing, but I see the difference.
For my code, I know the schemas. I guess a SQL builder could offer up query
options to the user browsing new databases.

On Sat, Feb 15, 2020 at 11:26 AM Simon Slavin <[email protected]> wrote:

> On 15 Feb 2020, at 3:14pm, [email protected] wrote:
>
> >> To determine if table XYZ is a WITHOUT ROWID table, run "PRAGMA
> >> index_info('XYZ');".  If you get back one or more rows, then XYZ is a
> >> WITHOUT ROWID table.  If you get back no rows, then XYZ is a rowid
> >> table.
> >>
> > Confused...What if I made an index on a ROWID table?
> > CREATE INDEX "Z" ON "DOC" ("n_id");
>
> The parameter in index_info() is normally the name of an index.  So if you
> create an index "Z" and ask for index_info("Z") you will get information on
> that index.
>
> If you create a WITHOUT ROWID table with name 'Y", and ask for
> index_info("Y") you will get information on the primary key of that table.
>
> If both exist, you get information about the index.
>
> Simon
> _______________________________________________
> sqlite-users mailing list
> [email protected]
> http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
>
_______________________________________________
sqlite-users mailing list
[email protected]
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users