Joe Wilson <[EMAIL PROTECTED]> wrote: > > The following question is purely for curiosity's sake - I don't > advocate doing this... I just want to get a better understanding > of the sqlite3 file format. > > Let's assume the most trivial case - a table without indexes of > any kind (implicit or explicit). Just arguing from a point of view > of the sqlite3 file format, if the row for the table/btree is deleted > in sqlite_master and the freelist is not updated, I can't see the > database being corrupted. The orphaned pages will not be reused until > VACUUM, and will needlessly occupy space on the disk, but they don't > seem to affect the other tables' and indexes' btree pages. In this > scenario, won't the new pages of other btrees avoid these orphaned > pages altogether? > > I appreciate that should you run pragma integrity_check you'll see > something like this: > > *** in database main *** > Page 2 is never used > > but that's not corruption, per se, it's just space inefficiency, is > it not? >
Correct -- D. Richard Hipp <[EMAIL PROTECTED]> ----------------------------------------------------------------------------- To unsubscribe, send email to [EMAIL PROTECTED] -----------------------------------------------------------------------------
