%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %% Name: tunicode.tex %% Purpose: Overview of the Unicode support in wxWindows %% Author: Vadim Zeitlin %% Modified by: %% Created: 22.09.99 %% RCS-ID: $Id$ %% Copyright: (c) 1999 Vadim Zeitlin %% Licence: wxWindows license %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \section{Unicode support in wxWindows}\label{unicode} This section briefly describes the state of the Unicode support in wxWindows. Read it if you want to know more about how to write programs able to work with characters from languages other than English. \subsection{What is Unicode?} Starting with release 2.1 wxWindows has support for compiling in Unicode mode on the platforms which support it. Unicode is a standard for character encoding which addresses the shortcomings of the previous, 8 bit standards, by using at least 16 (and possibly 32) bits for encoding each character. This allows to have at least 65536 characters (what is called the BMP, or basic multilingual plane) and possible $2^{32}$ of them instead of the usual 256 and is sufficient to encode all of the world languages at once. More details about Unicode may be found at {\tt www.unicode.org}. % TODO expand on it, say that Unicode extends ASCII, mention ISO8859, ... As this solution is obviously preferable to the previous ones (think of incompatible encodings for the same language, locale chaos and so on), many modern operating systems support it. The probably first example is Windows NT which uses only Unicode internally since its very first version. Writing internationalized programs is much easier with Unicode and, as the support for it improves, it should become more and more so. Moreover, in the Windows NT/2000 case, even the program which uses only standard ASCII can profit from using Unicode because they will work more efficiently - there will be no need for the system to convert all strings the program uses to/from Unicode each time a system call is made. \subsection{Unicode and ANSI modes} As not all platforms supported by wxWindows support Unicode (fully) yet, in many cases it is unwise to write a program which can only work in Unicode environment. A better solution is to write programs in such way that they may be compiled either in ANSI (traditional) mode or in the Unicode one. This can be achieved quite simply by using the means provided by wxWindows. Basically, there are only a few things to watch out for: \begin{itemize} \item Character type ({\tt char} or {\tt wchar\_t}) \item Literal strings (i.e. {\tt "Hello, world!"} or {\tt '*'}) \item String functions ({\tt strlen()}, {\tt strcpy()}, ...) \item Special preprocessor tokens ({\tt \_\_FILE\_\_}, {\tt \_\_DATE\_\_} and {\tt \_\_TIME\_\_}) \end{itemize} Let's look at them in order. First of all, each character in an Unicode program takes 2 bytes instead of usual one, so another type should be used to store the characters ({\tt char} only holds 1 byte usually). This type is called {\tt wchar\_t} which stands for {\it wide-character type}. Also, the string and character constants should be encoded using wide characters ({\tt wchar\_t} type) which typically take $2$ or $4$ bytes instead of {\tt char} which only takes one. This is achieved by using the standard C (and C++) way: just put the letter {\tt 'L'} after any string constant and it becomes a {\it long} constant, i.e. a wide character one. To make things a bit more readable, you are also allowed to prefix the constant with {\tt 'L'} instead of putting it after it. Of course, the usual standard C functions don't work with {\tt wchar\_t} strings, so another set of functions exists which do the same thing but accept {\tt wchar\_t *} instead of {\tt char *}. For example, a function to get the length of a wide-character string is called {\tt wcslen()} (compare with {\tt strlen()} - you see that the only difference is that the "str" prefix standing for "string" has been replaced with "wcs" standing for "wide-character string"). And finally, the standard preprocessor tokens enumerated above expand to ANSI strings but it is more likely that Unicode strings are wanted in the Unicode build. wxWindows provides the macros {\tt \_\_TFILE\_\_}, {\tt \_\_TDATE\_\_} and {\tt \_\_TTIME\_\_} which behave exactly as the standard ones except that they produce ANSI strings in ANSI build and Unicode ones in the Unicode build. To summarize, here is a brief example of how a program which can be compiled in both ANSI and Unicode modes could look like: \begin{verbatim} #ifdef __UNICODE__ wchar_t wch = L'*'; const wchar_t *ws = L"Hello, world!"; int len = wcslen(ws); wprintf(L"Compiled at %s\n", __TDATE__); #else // ANSI char ch = '*'; const char *s = "Hello, world!"; int len = strlen(s); printf("Compiled at %s\n", __DATE__); #endif // Unicode/ANSI \end{verbatim} Of course, it would be nearly impossibly to write such programs if it had to be done this way (try to imagine the number of {\tt \#ifdef UNICODE} an average program would have had!). Luckily, there is another way - see the next section. \subsection{Unicode support in wxWindows} In wxWindows, the code fragment from above should be written instead: \begin{verbatim} wxChar ch = wxT('*'); wxString s = wxT("Hello, world!"); int len = s.Len(); \end{verbatim} What happens here? First of all, you see that there are no more {\tt \#ifdef}s at all. Instead, we define some types and macros which behave differently in the Unicode and ANSI builds and allows us to avoid using conditional compilation in the program itself. We have a {\tt wxChar} type which maps either on {\tt char} or {\tt wchar\_t} depending on the mode in which program is being compiled. There is no need for a separate type for strings though, because the standard \helpref{wxString}{wxstring} supports Unicode, i.e. it stores either ANSI or Unicode strings depending on the compile mode. Finally, there is a special {\tt wxT()} macro which should enclose all literal strings in the program. As it is easy to see comparing the last fragment with the one above, this macro expands to nothing in the (usual) ANSI mode and prefixes {\tt 'L'} to its argument in the Unicode mode. The important conclusion is that if you use {\tt wxChar} instead of {\tt char}, avoid using C style strings and use {\tt wxString} instead and don't forget to enclose all string literals inside {\tt wxT()} macro, your program automatically becomes (almost) Unicode compliant! Just let us state once again the rules: \begin{itemize} \item Always use {\tt wxChar} instead of {\tt char} \item Always enclose literal string constants in {\tt wxT()} macro unless they're already converted to the right representation (another standard wxWindows macro {\tt \_()} does it, so there is no need for {\tt wxT()} in this case) or you intend to pass the constant directly to an external function which doesn't accept wide-character strings. \item Use {\tt wxString} instead of C style strings. \end{itemize} \subsection{Unicode and the outside world} We have seen that it was easy to write Unicode programs using wxWindows types and macros, but it has been also mentioned that it isn't quite enough. Although everything works fine inside the program, things can get nasty when it tries to communicate with the outside world which, sadly, often expects ANSI strings (a notable exception is the entire Win32 API which accepts either Unicode or ANSI strings and which thus makes it unnecessary to ever perform any conversions in the program). To get a ANSI string from a wxString, you may use the mb\_str() function which always returns an ANSI string (independently of the mode - while the usual \helpref{c\_str()}{wxstringcstr} returns a pointer to the internal representation which is either ASCII or Unicode). More rarely used, but still useful, is wc\_str() function which always returns the Unicode string. % TODO describe fn_str(), wx_str(), wxCharBuf classes, ... \subsection{Unicode-related compilation settings} You should define {\tt wxUSE\_UNICODE} to $1$ to compile your program in Unicode mode. Note that it currently only works in Win32 and that some parts of wxWindows are not Unicode-compliant yet (ODBC classes, for example). If you compile your program in ANSI mode you can still define {\tt wxUSE\_WCHAR\_T} to get some limited support for {\tt wchar\_t} type. This will allow your program to perform conversions between Unicode strings and ANSI ones (\helpref{wxEncodingConverter}{wxencodingconverter} depends on this partially) and construct wxString objects from Unicode strings (presumably read from some external file or elsewhere).