On Mon, Feb 2, 2015 at 9:00 AM, <sqlite-users-requ...@sqlite.org> wrote:
> From: RSmith <rsm...@rsweb.co.za> > To: sqlite-users@sqlite.org > Subject: Re: [sqlite] Encoding question > Message-ID: <54cebb71.8060...@rsweb.co.za> > Content-Type: text/plain; charset=windows-1252; format=flowed > > In short, the UTF-8 Pragma settings /allows/ your data to be interpreted > as such. It doesn't /force/ it, nor magically /converts/ > the data into UTF-8, and it most certainly does not under any > circumstances *guarantee* the UTF-8-ness of data. (Though it does > guarantee that /IF/ you put valid UTF-8 data in there, it will be handled > and returned correctly). > Thanks for this and the other responses, makes sense. I suppose it's similar to putting non-integer data into an INTEGER column. This is in the context of an SQLite utility I sell which I'm trying to make unicode compatible so I have no control over the data in the database, just have to interpret it the best I can. I've seen that there are algorithms out there that will detect different encodings but it seems that the algorithms are not 100% reliable. I should also have mentioned that the question also included table names, column names, constraint names, etc, but I'll assume the same applies to them as for the data. Pete lcSQL Software <http://www.lcsql.com> Home of lcStackBrowser <http://www.lcsql.com/lcstackbrowser.html> and SQLiteAdmin <http://www.lcsql.com/sqliteadmin.html> _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
