Per-Arne Hellarvik wrote:
> I need to convert a database from NE8ISO8859P10 to UTF8.
>
> I tried running csscan with terrible results: ~270.000 rows with lossy conversion.
> I can write some procs which does the conversion from the lossy conversion characters to legal chars.
> But I need to get a mapping between NE8ISO8859P10<->UTF8.
>
> Maybe I'm missing out on something here, but I haven't found any such mappings on the net.
> And csscan just returns the first 30 chars of the lossy-row.
>
> Anyone done something like this before?
>
It wasn't NE8ISO8859P10 fault.
It was that M$oft 1252 codepage darn thing....
CHR(128) to CHR(159), which broke the whole thing.
Received on Mon Dec 29 2003 - 01:54:18 CST
