Use different log levels depending on result code

Different result codes mean different things, and thus logging all of
them at INFO level can be confusing when one is scrolling through many
lines of logs looking for errors.

This commit changes the level at which command results are logged
depending on their result codes.

This work was performed as part of YAN-688.

Signed-off-by: Rodrigo Tobar <rtobar@icrar.org>

src/ska_tango_base/commands.py

@@ -174,6 +174,14 @@ class ResponseCommand(BaseCommand):
     tuple.
     """
 
+    RESULT_LOG_LEVEL = {
+        ResultCode.OK: logging.INFO,
+        ResultCode.STARTED: logging.INFO,
+        ResultCode.QUEUED: logging.INFO,
+        ResultCode.FAILED: logging.ERROR,
+        ResultCode.UNKNOWN: logging.WARNING
+    }
+
     def __call__(self, argin=None):
         """
         What to do when the command is called. This base class simply
@@ -208,7 +216,8 @@ class ResponseCommand(BaseCommand):
         else:
             (return_code, message) = self.do(argin=argin)
 
-        self.logger.info(
+        self.logger.log(
+            self.RESULT_LOG_LEVEL.get(return_code, logging.ERROR),
             f"Exiting command {self.name} with return_code {return_code!s}, "
             f"message: '{message}'"
         )