[GitHub] [lucene-solr] uschindler commented on a change in pull request #2176: Initial rewrite of MMapDirectory for JDK-16 preview (incubating) PANAMA APIs

[GitBox <gi...@apache.org>]

uschindler commented on a change in pull request #2176:
URL: https://github.com/apache/lucene-solr/pull/2176#discussion_r551061081



##########
File path: lucene/core/src/java/org/apache/lucene/store/MemorySegmentIndexInput.java
##########
@@ -0,0 +1,525 @@
+/*
+ * 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
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * 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.lucene.store;
+
+import java.io.EOFException;
+import java.io.IOException;
+import java.lang.invoke.VarHandle;
+import java.nio.ByteOrder;
+import java.util.Arrays;
+import java.util.Objects;
+
+import org.apache.lucene.util.IOUtils;
+
+import jdk.incubator.foreign.MemoryHandles;
+import jdk.incubator.foreign.MemorySegment;
+
+/**
+ * Base IndexInput implementation that uses an array of MemorySegments to represent a file.
+ *
+ * <p>For efficiency, this class requires that the segment size are a power-of-two
+ * (<code>chunkSizePower</code>).
+ */
+public abstract class MemorySegmentIndexInput extends IndexInput implements RandomAccessInput {
+  // We pass 1L as alignment, because currently Lucene file formats are heavy unaligned: :(
+  static final VarHandle VH_getByte = MemoryHandles.varHandle(byte.class, 1L, ByteOrder.BIG_ENDIAN).withInvokeExactBehavior();
+  static final VarHandle VH_getShort = MemoryHandles.varHandle(short.class, 1L, ByteOrder.BIG_ENDIAN).withInvokeExactBehavior();
+  static final VarHandle VH_getInt = MemoryHandles.varHandle(int.class, 1L, ByteOrder.BIG_ENDIAN).withInvokeExactBehavior();
+  static final VarHandle VH_getLong = MemoryHandles.varHandle(long.class, 1L, ByteOrder.BIG_ENDIAN).withInvokeExactBehavior();
+  
+  final boolean isClone;
+  final long length;
+  final long chunkSizeMask;
+  final int chunkSizePower;
+  final MemorySegment[] segments;
+  
+  int curSegmentIndex = -1;
+  MemorySegment curSegment; // redundant for speed: segments[curSegmentIndex], also marker if closed!
+  long curPosition; // relative to curSegment, not globally
+
+  public static MemorySegmentIndexInput newInstance(
+      String resourceDescription,
+      MemorySegment[] segments,
+      long length,
+      int chunkSizePower) {
+    if (segments.length == 1) {
+      return new SingleSegmentImpl(resourceDescription, segments[0], length, chunkSizePower, false);
+    } else {
+      return new MultiSegmentImpl(resourceDescription, segments, 0, length, chunkSizePower, false);
+    }
+  }
+
+  private MemorySegmentIndexInput(
+      String resourceDescription,
+      MemorySegment[] segments,
+      long length,
+      int chunkSizePower,
+      boolean isClone) {
+    super(resourceDescription);
+    this.segments = segments;
+    this.length = length;
+    this.chunkSizePower = chunkSizePower;
+    this.chunkSizeMask = (1L << chunkSizePower) - 1L;
+    this.isClone = isClone;
+    this.curSegment = segments[0];
+  }
+  
+  void ensureOpen() {
+    if (curSegment == null) {
+      throw alreadyClosed();
+    }
+  }
+
+  RuntimeException wrapAlreadyClosedException(RuntimeException e) {
+    if (e instanceof NullPointerException) {
+      return alreadyClosed();
+    }
+    // TODO: maybe open a JDK issue to have a separate, more
+    // meaningful exception for unmapped segments:
+    if (e.getMessage() != null && e.getMessage().contains("closed")) {
+      return alreadyClosed();
+    }
+    return e;
+  }
+  
+  // return value only for compiler!

Review comment:
       no, we need some return value so we can call it and return a value. But as it is throwing an exception anyways, the return value does not matter.
   
   These tricks are often used for exception handlers to delegate "exception handling" to a method where you know it will never return and just bubble up the exception:
   - if the exception is known before (by type), you use `throw callSomeMethod(param)`. The pro thing: it's clean as you delegate the instantiation of the exception to the factory, but throw it on your own. The problem here is: only works with a known exception type - and a single one, because the method needs it as return type. A generic `Exception` is a desaster, as it reuires to declare that on throws! 
   - an alternative is (and that's used here): put the throw clause also into the method, so the only sense of our method is throwing some exception, it never returns cleanly. Backside: the method needs a return type, so you can do: `return callSomeMethod()` from the caller. Of course you can add a separate return statement after the method call, but that looks more bullshitty.
   
   Here we use the second approch. The most common return type to all callers of this method is `byte`. Compiler is happy, voila!




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org