Yes, it's inside one of my entries in ChangeLog as TOFIX. I didn't want to touch that, but already searched for toupper/tolower/islower/etc references, and there isn't a great number of them fortunately.
I've also added Harbour API replacement for these. Some of them currently using above functions. First we should fix these by replacing it with simple ASCII char comparisons/operations, and as a next phase change above APIs to the Harbour ones. Does this sound right? Brgds, Viktor On Thu, Feb 5, 2009 at 5:24 PM, Przemyslaw Czerpak <[email protected]>wrote: > On Thu, 05 Feb 2009, Przemyslaw Czerpak wrote: > [...] > Hi, > > And one thing else much more important for me because it's > a real problem in current code which may cause unpredictable > results, f.e. some index corruption. > We should eliminate functions which uses locale settings because > they can badly change the behavior of existing code. > If you link your code with some external library which enables > locale then it effects all isalpha()/isupper()/tolower() and similar > functions. If someone has index on UPPER(<Sth>) then you have problem. > Our UPPER() function may give different results when locale are enabled > (it depends on used language and encoding). If you have two applications > (f.e. one written as daemon without user interface or with Harbour CUI > and second using GTK (GTK enables locale during initialization) and both > operates on the same index with UPPER then this index will be corrupted. > This have to be fixed. > > best regards, > Przemek > _______________________________________________ > Harbour mailing list > [email protected] > http://lists.harbour-project.org/mailman/listinfo/harbour >
_______________________________________________ Harbour mailing list [email protected] http://lists.harbour-project.org/mailman/listinfo/harbour
