Cart
[Home] [By Thread] [By Date] [Recent Entries]
Joe English scripsit: > The 'codePoint' typedef may be problematic: > > // Unicode code points (4-byte int on most systems) > typedef wchar_t codePoint; > > The C standard makes no useful guarantees about > the size or interpretation of 'wchar_t'. On some > systems it's identical to plain 'char', and even > on systems where it's big enough to hold all of > Unicode, there's no guarantee about what encoding > the wcs* and *wcs functions use. wchar_t should > not be used in programs that are meant to generate > portable data and be portable themselves; you just > don't know what you're going to get. I have argued privately that wchar_t is in fact the Right Thing here despite its variability in size (UTF-32 on Unix platforms, UTF-16 on Windows), because it makes genx compatible with both standardized and non-standardized facilities, most especially "..."L strings. Some conditional logic will be needed to interpret the input as UTF-16 or UTF-32, which can be based on sizeof(wchar_t). Hypothetical platforms where sizeof(wchar_t) == 1 can be neglected. -- He made the Legislature meet at one-horse John Cowan tank-towns out in the alfalfa belt, so that jcowan@r... hardly nobody could get there and most of http://www.reutershealth.com the leaders would stay home and let him go http://www.ccil.org/~cowan to work and do things as he pleased. --Mencken, _Declaration of Independence_
|
