Known Restrictions of the Current VAC++

┌──────────────────────────────────────────────────────────────────────┐
│Some of IBM Visual Age C++ for OS/2 functions applies MSE             │
│specifications rather than XPG4 traditional specifications.           │
└──────────────────────────────────────────────────────────────────────┘

ISO/IEC 9899:1990/Amendment 1:1993(E), so-called MSE (Multibyte Support Extensions) is basically a super set of XPG4 C functions. However, several functions are slightly different having extra parameters or different parameters so that they will work more flexibly.

MSE interfaces are applied for the following functions:

XPG4:
   size_t wcsftime( wchar_t *wcs, size_t maxsize, const char*format,
                    const struct tm *timptr );

MSE/VAC++:
   size_t wcsftime( wchar_t *wcs, size_t maxsize, const wchar_t*format,
                    const struct tm *timptr );

XPG4:
   wchar_t *wcstok( wchar_t *ws1, const wchar_t *ws2 );

MSE/VAC++:
   wchar_t *wcstok( wchar_t *ws1, const wchar_t ws2, wchar_t **ptr);

XPG4:
   wchar_t *wcswcs( const wchar_t *ws1, const wchar_t *ws2 );

MSE/VAC++:
   wchar_t *wcsstr( const wchar_t *ws1, const wchar_t *ws2 );

%lc and %ls specifiers are used for wchar_t strings and characters instead of %C and %S specifiers in formatted I/O (printf and scanf family).

┌──────────────────────────────────────────────────────────────────────┐
│XPG4 extensions of formatted I/O (printf and scanf family) specifiers │
│and message catalog facilities are not currently supported, but are in│
│plan.                                                                 │
└──────────────────────────────────────────────────────────────────────┘

• n$ (numbered arguments).

• Apostrophe flag character (printf family).

• The cat* functions and message catalog facilities

┌──────────────────────────────────────────────────────────────────────┐│Era name is not supported.                                            │
└──────────────────────────────────────────────────────────────────────┘

The following constants are not defined in langinfo.h

• ERA

• ERA_D_FMT

• ERA_D_T_FMT

• ERA_T_FMT

• ALT_DIGITS

┌──────────────────────────────────────────────────────────────────────┐│iconv does not support code sets for DBCS languages other than        │
│Japanese.                                                             │
└──────────────────────────────────────────────────────────────────────┘

At present, the supported code sets for DBCS languages by the iconv() are as follows:

IBM-932 <-> IBM-930

PC-EBCDIC Katakana, Mixed

PC-EBCDIC Extended Latin, Mixed

PC-EBCDIC Katakana, SBCS only

PC-EBCDIC Extended Latin, SBCS only

EUC-EBCDIC Katakana, Mixed

EUC-EBCDIC Extended Latin, Mixed

These are the combinations of Japanese code conversion. Character code conversion between code page 942 (Japanese PC SAA code page) and EBCDIC can be substituted by those of IBM-932. It is still too early to use iconv() of VAC++ because of the following restrictions:

• Some DBCS conversions are invalid.

• Though conversion for IBM-932 can be used for code page 942, it requires special logic and may be considered as a deviation from Internationalization concept.

• Other DBCS languages are not supported.