Stephan Witt wrote: > Am 03.12.2010 um 04:36 schrieb Enrico Forestieri: > >> It's becoming a mess. We have isdigit(), isDigit(), isDigitASCII(), >> and now iswdigit() enters the scene. > > iswdigit() is POSIX standard for wchar_t arguments - our char_type > is a typedef of wchar_t. But I don't insist here.
Not on windows. I agree with Enrico: iswdigit is wrong (and it is not used in my patch in trac). This is even more general: None of the wide string functions should be used, because they would only work on OSes that use UCS4 encoding for wstring and a 32bit wchar_t >> Too many of those functions, too much confusion. >> We should start clearing up things. > ... >> So, IMHO, we should: >> 1) ban isdigit() [use isDigitASCII() instead] That was the first part of my patch. >> 2) replace all occurrences of isDigit() with isDigitASCII() I don't know. Unless somebody goes through all usages and confirms that indeed an ASCII digit is wanted I would rather keep it as it is. >> 3) use QChar::isNumber() instead of iswdigit(). > > I'd have no problem with this. My goal was to get the thing fixed. > IMHO, a wrapper around QChar::isNumber() in textutils.h and lstrings.cpp > should be added then. If you want to use QChar::isNumber() for the spell checking test: Why not. Unless somebody comes up with an example I believe that it basically does not matter whether isDigit(), isDigitASCII() or QChar::isNumber() is used there. Whatever you decide, please ensure that hasDigit() gets renamed properly. Georg
