9 Replies Latest reply: May 13, 2013 6:06 PM by 151291 RSS

    DG4ODBC to MSSQL and nvarchar2 problem

    ha*735103*ns
      Hello all,

      We have been using HS for several years on a 10gR2 on a Windows 2003 32bit platform for connections to mssql.

      Now we are testing with DG4ODBC on a Windows2003 64bits platform with 11gR2 and are experiencing the following issue:
      When using the HS_PASSTHROUGH facility to select data from the mssql database all varchar field are now multibyte.
      Same receiving database and same mssql database!

      From the DG4ODBC trace:
      hoadaMOD bit-values found (0x40:TREAT_AS_NCHAR,0x400:UNICODE_COLUMN)
      DTY NULL-OK LEN MAXBUFLEN PR/SC CST IND MOD NAME
      12 VARCHAR N 256 256 128/128 1000 0 440 name

      When using hs_passthrough@remote.get_value(c,1,x) and x is a varchar2 there a spaces between the received characters
      When x is a nvarchar2 the result is as expected and usable.
      But we are unwilling to change the application to use nvarchars instead on varchars.

      I have been fiddling with several settings as: HS_LANGUAGE, HS_KEEP_REMOTE_COLUMN_SIZE, FS_FDS_MAP_NCHAR, HS_NLS_LENGTH_SEMANTICS and HS_FDS_REMOTE_DB_CHARSET but still unable to get DG4ODBC to behave exactly as old old-style HS.

      Has anybody a idea how to fix this.

      Regards,
      Hans