Github user paul-rogers commented on a diff in the pull request:
    --- Diff: 
    @@ -0,0 +1,183 @@
    + * Licensed to the Apache Software Foundation (ASF) under one
    + * or more contributor license agreements.  See the NOTICE file
    + * distributed with this work for additional information
    + * regarding copyright ownership.  The ASF licenses this file
    + * to you under the Apache License, Version 2.0 (the
    + * "License"); you may not use this file except in compliance
    + * with the License.  You may obtain a copy of the License at
    + *
    + *
    + *
    + * Unless required by applicable law or agreed to in writing, software
    + * distributed under the License is distributed on an "AS IS" BASIS,
    + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    + * See the License for the specific language governing permissions and
    + * limitations under the License.
    + */
    +package org.apache.drill.exec.physical.impl.protocol;
    +import org.apache.drill.common.exceptions.UserException;
    +import org.apache.drill.exec.ops.OperatorContext;
    +import org.apache.drill.exec.record.RecordBatch.IterOutcome;
    + * State machine that drives the operator executable. Converts
    + * between the iterator protocol and the operator executable protocol.
    + * Implemented as a separate class in anticipation of eventually
    + * changing the record batch (iterator) protocol.
    + */
    +public class OperatorDriver {
    +  public enum State { START, SCHEMA, RUN, END, FAILED, CLOSED }
    --- End diff --
    Next, the question about the `END` and `FAILED` states. `END` means we 
reached EOF and have no data left to deliver. Calling `next()` in the `END` 
state simply returns `DONE` (there is *still* no more data.)
    On the other hand, `FAILED` indicates that an error occurred. This seemed 
like a useful thing to know. But, as I review this version of the code, I see 
we are not actually using this state. So, rather than add a `CANCELLED` state, 
maybe we can collapse `FAILED` into `END`: which is just a signal to `next()` 
to return `DONE`.
    Cleanup should be no different in the three cases: we must release all 
resources regardless of the reason that `close()` is called.


Reply via email to