On Thu, 20 Nov 2008 15:25:35 +0100, Christophe Leske <[EMAIL PROTECTED]> wrote in General Discussion of SQLite Database <sqlite-users@sqlite.org>:
> >>> Any chance to spare the ID field and get an index on the rowid for a >>> given table? >> I do not understand the question. Please rephrase. Use more words. >> >Most of the ID fields are primary integer keys which also coincident with the >value of the rowid for a given table. Yes, you cannot count on it, but as long >as you do not change the content of a given table, the value of an ID field >declared as a integer primary key is usually the same as the rowid of the >record: > >rowid=id > >I was wondering if there is a cheap way (memorywise) to index the rowid and >thus omit the ID field. http://www.sqlite.org/lang_createtable.html says: "Specifying a PRIMARY KEY normally just creates a UNIQUE index on the corresponding columns. However, if primary key is on a single column that has datatype INTEGER, then that column is used internally as the actual key of the B-Tree for the table. " So, in CREATE TABLE ( ID INTEGER PRIMARY KEY, ... ); , ID is used as / identical with ROWID. Note: Use the exact words: INTEGER PRIMARY KEY >Christophe Leske HTH -- ( Kees Nuyt ) c[_] _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
