Staffan Tylen wrote: > I must admit that I'm a bit confused here. If I'm not wrong UTF-8 differs > from ascii when the value is higher than '7f'x, but storing data in sqlite > as text with character values beteen 'x80'x and 'ff'x seems to be no > problem. I previously thought that this could only be done in blob format. > > create table t (a); > insert into t values(cast(x'ff' as text)); > select a,length(a),hex(a) from t; > |1|FF > > My conclusion is that storing single-byte characters of any value is > allowed, is this true?
SQLite assumes that all strings you give it are encoded in UTF-8, and does not actually check the encoding. It gives you the same string back, so you could, in theory, use a different encoding, as long as you do not use any database string processing functions. In practice, I would not recommend this. > At the bottom it says: *Note to Windows users:* The encoding used for the > filename argument of sqlite3_open() and sqlite3_open_v2() must be UTF-8, > not whatever codepage is currently defined. Filenames containing > international characters must be converted to UTF-8 prior to passing them > into sqlite3_open() or sqlite3_open_v2(). > > So what does "must be converted" mean? I don't know how sqlite3.exe works > here but if I do > > sqlite3 ?.db > > where '?' as we've seen is '82'x it happily creates a file with a > non-displayable character in the first position that seems to be 1 byte > long. The shell assumes that its arguments are UTF-8, and gives the filename unchanged to sqlite3_open*(). When you've entered the filename in the Windows console with the default settings, it is not encoded in UTF-8. Regards, Clemens
