"David Moore" davidm@sjsoft.com wrote:
The patch below first checks to see if the codepage is one of the exceptions, and if it is, passes in NULLs for the last two arguments.
I wonder how did you build the list of exceptions: why CP_UTF7 is there but CP_UTF8 is not? Why 50xxx,52xxx,57xxx code pages are there?
I'd suggest as a start to have only CP_UTF7, CP_UTF8 and CP_SYMBOL in the exception list.