Yaesu newcat style command response reading improvements

After testing on an FT-450D which tends to send busy responses "?;" to
subsequent commands after a band  changing command it appears that the
undoumented "?;"  rig busy response  should be  treated as a  wait and
subsequent reads may return the  expected response i.e. the command is
still in progress and should not be resent.

As this is  all based on observations of failure  modes rater than any
doumentation from Yaesu, there may be further enhancements necessary.

yaesu/newcat.c

@@ -3941,9 +3941,15 @@ int newcat_vfomem_toggle(RIG * rig)
 }
 
 /*
- * Writes a null terminated command string in priv->cmd_str to the CAT
- * port and returns a response from the rig in priv->ret_data which is
- * also null terminated.
+ * Writes a null  terminated command string from  priv->cmd_str to the
+ * CAT  port and  returns a  response from  the rig  in priv->ret_data
+ * which is also null terminated.
+ *
+ * Honors the 'retry'  capabilities field by resending  the command up
+ * to 'retry' times until a valid response is received. In the special
+ * cases of receiving  a valid response to a different  command or the
+ * "?;" busy please wait response; the command is not resent but up to
+ * 'retry' retries to receive a valid response are made.
  */
 int newcat_get_cmd (RIG *rig)
 {
@@ -3951,46 +3957,55 @@ int newcat_get_cmd (RIG *rig)
   struct newcat_priv_data *priv = (struct newcat_priv_data *)rig->state.priv;
   int retry_count = 0;
   int rc = -RIG_EPROTO;
-  int bytes_read;
 
   while (rc != RIG_OK && retry_count++ <= state->rigport.retry)
     {
-      /* send the command */
-      rig_debug(RIG_DEBUG_TRACE, "cmd_str = %s\n", priv->cmd_str);
-      if (RIG_OK != (rc = write_block(&state->rigport, priv->cmd_str, strlen(priv->cmd_str))))
+      if (rc != -RIG_BUSBUSY)
         {
-          return rc;
+          /* send the command */
+          rig_debug(RIG_DEBUG_TRACE, "cmd_str = %s\n", priv->cmd_str);
+          if (RIG_OK != (rc = write_block(&state->rigport, priv->cmd_str, strlen(priv->cmd_str))))
+            {
+              return rc;
+            }
         }
 
       /* read the reply */
-      if ((bytes_read = read_string(&state->rigport, priv->ret_data, sizeof(priv->ret_data),
-                                    &cat_term, sizeof(cat_term))) <= 0)
+      if ((rc = read_string(&state->rigport, priv->ret_data, sizeof(priv->ret_data),
+                            &cat_term, sizeof(cat_term))) <= 0)
         {
-          continue;             /* retry */
+          continue;             /* usually a timeout - retry */
         }
       rig_debug(RIG_DEBUG_TRACE, "%s: read count = %d, ret_data = %s\n",
-                __func__, bytes_read, priv->ret_data);
+                __func__, rc, priv->ret_data);
+      rc = RIG_OK;              /* received something */
 
-      /* Check that command termination is correct */
+      /* Check that command termination is correct - alternative is
+         response is longer that the buffer */
       if (!strchr(&cat_term, priv->ret_data[strlen(priv->ret_data) - 1]))
         {
           rig_debug(RIG_DEBUG_ERR, "%s: Command is not correctly terminated '%s'\n",
                     __func__, priv->ret_data);
+          rc = -RIG_BUSBUSY;    /* don't write command again */
+                                /* we could decrement retry_count
+                                   here but there is a danger of
+                                   infinite looping so we just use up
+                                   a retry for safety's sake */
           continue;             /* retry */
         }
 
       /* check for error codes */
       if (2 == strlen(priv->ret_data))
         {
-          /* The following error responses are documented for Kenwood
-             but not for Yaesu, but at least one of them is known to
-             occur in that the FT-450 certainly responds to "IF;"
-             occasionally with "?;". The others are harmless even of
+          /* The following error responses  are documented for Kenwood
+             but not for  Yaesu, but at least one of  them is known to
+             occur  in that  the  FT-450 certainly  responds to  "IF;"
+             occasionally with  "?;". The others are  harmless even of
              they do not occur as they are unambiguous. */
           switch (priv->ret_data[0])
             {
             case 'N':
-              /* Command recognised by rig but invalid data entered. */
+              /* Command recognized by rig but invalid data entered. */
               rig_debug(RIG_DEBUG_VERBOSE, "%s: NegAck for '%s'\n", __func__, priv->cmd_str);
               return -RIG_ENAVAIL;
 
@@ -4007,11 +4022,12 @@ int newcat_get_cmd (RIG *rig)
               break;            /* retry */
 
             case '?':
-              /* Command not understood by rig */
-              rig_debug(RIG_DEBUG_ERR, "%s: Unknown command '%s' or rig busy\n", __func__, priv->cmd_str);
-              rc = -RIG_ERJCTED;
-              break;            /* retry */
+              /* Rig busy wait please */
+              rig_debug(RIG_DEBUG_ERR, "%s: Rig busy\n", __func__, priv->cmd_str);
+              rc = -RIG_BUSBUSY;
+              break;            /* retry read only */
             }
+          continue;
         }
 
       /* verify that reply was to the command we sent */
@@ -4024,7 +4040,7 @@ int newcat_get_cmd (RIG *rig)
            */
           rig_debug(RIG_DEBUG_ERR, "%s: wrong reply %.2s for command %.2s\n",
                     __func__, priv->ret_data, priv->cmd_str);
-          rc = -RIG_BUSBUSY;    /* retry */
+          rc = -RIG_BUSBUSY;    /* retry read only */
         }
     }