You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@arrow.apache.org by GitBox <gi...@apache.org> on 2020/09/10 13:03:06 UTC

[GitHub] [arrow] pitrou commented on a change in pull request #8156: ARROW-9810: [C++] Generalized nested reconstruction helpers

pitrou commented on a change in pull request #8156:
URL: https://github.com/apache/arrow/pull/8156#discussion_r486317787



##########
File path: cpp/src/parquet/level_comparison.h
##########
@@ -0,0 +1,91 @@
+// 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.
+#pragma once
+
+#include <algorithm>
+#include <cstdint>
+
+#include "arrow/util/bit_util.h"
+
+namespace parquet {
+namespace internal {
+
+// These APIs are likely to be revised as part of ARROW-8494 to reduce duplicate code.
+// They currently represent minimal functionality for vectorized computation of definition
+// levels.
+
+/// Builds a bitmap by applying predicate to the level vector provided.
+///
+/// \param[in] levels Rep or def level array.
+/// \param[in] num_levels The number of levels to process (must be [0, 64])
+/// \param[in] predicate The predicate to apply (must have the signature `bool
+/// predicate(int16_t)`.
+/// \returns The bitmap using least significant "bit" ordering.
+///
+/// N.B. Correct byte ordering is dependent on little-endian architectures.
+///
+template <typename Predicate>
+inline uint64_t LevelsToBitmap(const int16_t* levels, int64_t num_levels,
+                               Predicate predicate) {
+  // Both clang and GCC can vectorize this automatically with SSE4/AVX2.
+  uint64_t mask = 0;
+  for (int x = 0; x < num_levels; x++) {
+    mask |= static_cast<uint64_t>(predicate(levels[x]) ? 1 : 0) << x;
+  }
+  return ::arrow::BitUtil::ToLittleEndian(mask);
+}
+
+/// Builds a  bitmap where each set bit indicates the corresponding level is greater

Review comment:
       I don't understand why you're doing this. Integer comparisons are cheap, while writing and reading bitmaps is expensive. While go through an intermediate bitmap (and then spend some time trying to optimize it?).




----------------------------------------------------------------
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