> -----Original Message----- > From: D. Richard Hipp [mailto:[EMAIL PROTECTED] > Sent: Sunday, July 10, 2005 6:12 AM > To: sqlite-users@sqlite.org > Subject: RE: [sqlite] Binding a column name? > > On Sun, 2005-07-10 at 00:23 -0700, Brown, Dave wrote: > > That is what I do. But that also means I have to call > sqlite_prepare() > > each time, instead of just once. I was originally hoping I could > > prepare() once and just bind. > > > > If another thread or process VACUUMs the database or creates > a new table or makes any other structure changes to the > database file, all of your prepared statements will be > invalided and you will have to rerun sqlite3_prepare(). > Since you generally have no control over when another process > might VACUUM the database, you should always be prepared to > rerun sqlite3_prepare() if necessary. This is true even if > you are only running your SQL statement once and then > finalizing it because another process might VACUUM and > invalidate your statement in the very brief window of time > between your calls to sqlite3_prepare() and sqlite3_step(). > > Your best bet it to use a wrapper class of some sort that > automates the task of rerunning sqlite3_prepare() when necessary. >
Does sqlite store the SQL text passed into sqlite3_prepare? If not, then I assume this means that any time we use sqlite3_prepare, we should cache the SQL text "in the wrapper" in case we need to re-prepare it. Along the same line, I suppose we have to cache all the bound parameters, since they will have to re-bound as well. Is it possible to get the SQLITE_SCHEMA error after the first sqlite3_step call, while iterating throw the rows? Tim
