Inconsistency exists between transcoding in SAS^® 9.1.3 and SAS 9.2 for some of SJIS punctuation characters.

The example below uses the KCVT function and illustrates the problem. In SAS 9.2, transcoding does not handle the Unicode x301c character properly:

 
  data;
   x='8160'x ; * '～';
   un1=kcvt(x,'sjis','ucs2b');
   sj1=kcvt(un1,'ucs2b','sjis');
   put un1= $hex4. sj1=;
   un2=unicodec(x);
   sj2=unicode(un2);
   put un2= sj2=;
   y='301c'x ; * unicode tilde ;
   sj3=kcvt(y,'ucs2b','');
   put sj3 = sj3 ;
   sj4=unicode(y,'utf16b');
   put sj4= sj4;
  run;

  9.1.3 Result:

  un1=301C sj1=～
  un2=\uFF5E sj2=～
  sj3=～ ～
  sj4=～ ～

  9.2 Result:

  un1=FF5E sj1=～
  un2=\uFF5E sj2=～
  sj3=　 　
  sj4=～ ～

The results in SAS 9.1.3 are correct. The results in SAS 9.2 are incorrect.

Operating System and Release Information

A fix for SAS 9.1.3 (9.1 TS1M3) for this issue is available at:

A fix for SAS 9.1.3 (9.1 TS1M3) with Asian Language Support (DBCS) for this issue is available at: