> collation-sequence is default NO-CASE Default collation is BINARY. So either yours is default or NOCASE - not both.
> A question, however, on the Latin-1, ASCII range requirement: this is a > column requirement and not a database requirement, correct? It's not a requirement at all. It's just the fact that LIKE will compare ASCII characters case-insensitive and all other characters case-sensitive when case-sensitive comparison is off. Pavel On Mon, Nov 16, 2009 at 7:47 AM, Tim Romano <tim.rom...@yahoo.com> wrote: > After reading http://www.sqlite.org/optoverview.html, I think my query > meets the requirements for index use with the LIKE operator: > > The column is varchar(75) and so TEXT affinity. > The column uses Latin-1 characters exclusively. > The wildcard appears at the far right end of the string literal, e.g. > myColumn LIKE 'foo%' > The escape clause does not appear. > Case-sensitivity=false; > collation-sequence is default NO-CASE > > QUESTION: > A question, however, on the Latin-1, ASCII range requirement: this is a > column requirement and not a database requirement, correct? I have > several columns with text affinity; one is strict ASCII and represents > characters outside the ASCII range as html-entities (e.g. "ü") and > the others store the unicode characters. The database encoding is UTF-8. > > My query with the LIKE operator worked instantaneously in MS-Access, > BTW, where I originally had the database. After exporting to delimited > text and reimporting into SQLite, most queries in SQLite are just as > fast, executing in under a second. But this query with the LIKE operator > takes 40 seconds because of the full-table scan. > > Thanks > > > _______________________________________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
