Conference azur::mcc

    I'm not saying we shouldn't do this but the AES structure is
    supposed to be opaque with regard to the caller of the AES routines.
    If we released a tool to dump the AES *fields* we would break
    the opaqueness.  A tool which logically dumped the AES in an
    implementation-independent format would be a better idea.
    

    What version of MCC are you running?  If you are running the latest
    version (UT0.1.1 ?) try :
    
    DBG> SET IMAGE MCC$KERNEL_SHR
    DBG> SET MODULE MCC$AES
    DBG> CALL MCC$AES_DUMP( address-of-AES-goes-here )
    
    As Jim noted in the previous note, this is a generic non-implementation
    specific display of the AES.  So it may not be *exactly* what you want.
    
    -Matt.
    A

	Thanks, that's exactly what I wanted.

	I'm unfortunatly running T1.0.0 version and did not found any
	MCC$AES_DUMP routine in the MCC$KERNEL_SHR image.

	So the best is wait for the next update...


						Pierre.

    Mike,
    
    TBD can do some of it but the problem is if you are running from
    another PM (like the TRM) it is useful to be able to turn on 
    a logical and have your code dump the information; or from the
    debugger be able to do a set image and call the routine directly
    to validate a parameter. I know this later is used heavily in our
    debugging.
    
    Also note that the TRM dump (under the control of the MCC$TRM_PM
    logical in UT1) will be udated in the next release to handle the
    dumping of constructed data types as instances.
    
    jill
    

/*
 *                   D U M P _ A E S
 *
 * Call this with the address of the handle. That is, if you have
 *
 *   MCC_A_AES   entity;
 *
 *   DBG>eval/addr entity
 *   DBG>call dump_aes (xxxxxxxxxx)
 */
static dump_aes (handle)
   MCC_A_AES *handle;
{
   unsigned long depth, status, i, class, wild_type, wild_depth;
   MCC_T_FullName val_buf;
   MCC_T_Descriptor value = 
   {
      sizeof (MCC_T_FullName), DSC_K_DTYPE_T, DSC_K_CLASS_S, val_buf,
      sizeof (MCC_T_FullName), 0, 1, 0, 0, 0
      }
   ;
   
   /* Display the depth of the entity spec and any wildcard */
   status = mcc_aes_depth (handle, &depth);
   if (BAD(status)) return (status);
   printf ("%d CI Pairs in Entity Spec, ", depth);

   wild_depth = 0;
   status = mcc_aes_wild (handle, &wild_depth, &wild_type);
   if (BAD(status)) return (status);
   switch (wild_type)
   {
      case MCC_K_AES_NOT_WILD:
         printf ("No Wildcard\n");
         break;
      case MCC_K_AES_CLASS_WILD:
	    printf ("Wildcarded Class at Depth %d\n", wild_depth);
	    break;
      case MCC_K_AES_INSTANCE_FULL:
	    printf ("Full Instance Wildcard at Depth %d\n", wild_depth);
	    break;
      case MCC_K_AES_INSTANCE_PARTIAL:
	    printf ("Partial Instance Wildcard at Depth %d\n", wild_depth);
	    break;
      default:
	    printf ("Unknown Wildcard Type\n");
	    break;
   }

   /* For each CI pair, display everything we know about it */
   for (i=0;i<depth;i++)
   {
      value.mcc_w_curlen = value.mcc_w_maxstrlen;
      status = mcc_aes_get (handle, &i, &class, &value);
      if (BAD(status)) return (status);

      printf ("  %d: Class: %d DataType: %d Length: %d Id: %d",
	      i, class, value.mcc_l_dt, value.mcc_w_curlen,
	      value.mcc_l_id);
      printf (" Flags: %d\n", value.mcc_b_flags);

   }

   return (status);
}

	Very good example.
	It's a very good help...
	Regards, Philippe.