Fixed dlmsw.cpp compilation with older SDKs and at run-time for systems prior to...

[wxWidgets.git] / interface / wx / chartype.h

diff --git a/interface/wx/chartype.h b/interface/wx/chartype.h
index 474cfd044dbe2692227ac67fe2c043c87673ae64..f67471002091e35dfb54afd660908dfbf01396b2 100644 (file)
--- a/interface/wx/chartype.h
+++ b/interface/wx/chartype.h
@@ -3,10 +3,10 @@
 // Purpose:     interface of global functions
 // Author:      wxWidgets team
 // RCS-ID:      $Id$
-// Licence:     wxWindows license
+// Licence:     wxWindows licence
 /////////////////////////////////////////////////////////////////////////////
 
-/** @ingroup group_funcmacro_string */
+/** @addtogroup group_funcmacro_string */
 //@{
 
 /**
@@ -15,13 +15,16 @@
     builds of wxWidgets. This macro simply returns the value passed to it
     without changes in ASCII build. In fact, its definition is:
 
-@code
-#ifdef UNICODE
-#   define wxT(x)  L##x
-#else // !Unicode
-#   define wxT(x)  x
-#endif
-@endcode
+    @code
+    #ifdef UNICODE
+    #   define wxT(x)  L##x
+    #else // !Unicode
+    #   define wxT(x)  x
+    #endif
+    @endcode
+
+    Note that since wxWidgets 2.9.0 you shouldn't use wxT() anymore in your
+    program sources (it was previously required if you wanted to support Unicode).
 
     @see @ref overview_unicode, wxS()
 
@@ -30,11 +33,42 @@
 #define wxT(string)
 
 /**
-    wxS is macro which can be used with character and string literals (in other words,
-    @c 'x' or @c "foo") to either convert them to wide characters or wide strings
-    in @c wchar_t-based (UTF-16) builds or keep them unchanged in @c char-based
+    Compatibility macro which expands to wxT() in wxWidgets 2 only.
+
+    This macro can be used in code which needs to compile with both
+    wxWidgets 2 and 3 versions, in places where the wx2 API requires a Unicode string
+    (in Unicode build) but the wx3 API only accepts a standard narrow
+    string, as in e.g. wxCmdLineEntryDesc structure objects initializers.
+
+    Example of use:
+    @code
+    const wxCmdLineEntryDesc cmdLineDesc[] =
+    {
+        { wxCMD_LINE_SWITCH, wxT_2("q"), wxT_2("quiet"),
+          wxT_2("Don't output verbose messages") },
+        wxCMD_LINE_DESC_END
+    };
+    @endcode
+
+    Without @c wxT_2 the code above wouldn't compile with wxWidgets 2, but using @c
+    wxT instead, it wouldn't compile with wxWidgets 3.
+
+    @see wxT()
+
+    @since 2.8.12, 2.9.2
+
+    @header{wx/chartype.h}
+ */
+#define wxT_2(string)
+
+/**
+    wxS is a macro which can be used with character and string literals (in other words,
+    @c 'x' or @c "foo") to convert them either to wide characters or wide strings
+    in @c wchar_t-based (UTF-16) builds, or to keep them unchanged in @c char-based
     (UTF-8) builds.
 
+    Basically this macro produces characters or strings of type wxStringCharType.
+
     The use of this macro is optional as the translation will always be done at
     run-time even if there is a mismatch between the kind of the literal used
     and the string or character type used in the current build.
@@ -55,6 +89,11 @@
 
     Don't confuse this macro with _()!
 
+    Note that since wxWidgets 2.9.0 the use of _T() is discouraged just like
+    for wxT() and also that this macro may conflict with identifiers defined in
+    standard headers of some compilers (such as Sun CC) so its use should
+    really be avoided.
+
     @header{wx/chartype.h}
 */
 #define _T(string)
@@ -88,9 +127,11 @@ typedef wxUSE_UNICODE_dependent wxUChar;
 
     The @c wxUSE_UNICODE_WCHAR symbol is defined to @c 1 when building on
     Windows while it's defined to @c 0 when building on Unix, Linux or OS X.
+    (Note that @c wxUSE_UNICODE_UTF8 symbol is defined as the opposite of
+    @c wxUSE_UNICODE_WCHAR.)
 
-    Note that wxStringCharType is the type used by wxString for internal storage
-    of the characters.
+    Note that wxStringCharType (as the name says) is the type used by wxString
+    for internal storage of the characters.
 */
 typedef wxUSE_UNICODE_WCHAR_dependent wxStringCharType;