aboutsummaryrefslogtreecommitdiff
path: root/src/main
diff options
context:
space:
mode:
authorKidlike <steve.tcmg@gmail.com>2024-01-13 22:22:36 +0100
committerGitHub <noreply@github.com>2024-01-13 22:22:36 +0100
commitc01668403a13d86dcd2305884198387ee1f073f0 (patch)
tree55c0dab97b81b57d11e1a1ad25bf843af5f1182f /src/main
parent32143b2a4c0f929c362c8e80527ef47fb2daa63b (diff)
submission for kidlike (#294)
* first version * second version (0m59s) * third version (0m46s) * fourth version (0m39s) * fifth version (0m18s) * follow naming conventions from project structure * fix rounding (see /issues/49) * formatting changes from build * name should case-match github username * sixth version (14s) * seventh version (11s) * potential fix for other systems? * no need for sdk install * binary should go to ./target * building native-image only when not existing yet
Diffstat (limited to 'src/main')
-rw-r--r--src/main/java/dev/morling/onebrc/CalculateAverage_Kidlike.java230
1 files changed, 230 insertions, 0 deletions
diff --git a/src/main/java/dev/morling/onebrc/CalculateAverage_Kidlike.java b/src/main/java/dev/morling/onebrc/CalculateAverage_Kidlike.java
new file mode 100644
index 0000000..903da31
--- /dev/null
+++ b/src/main/java/dev/morling/onebrc/CalculateAverage_Kidlike.java
@@ -0,0 +1,230 @@
+/*
+ * Copyright 2023 The original authors
+ *
+ * Licensed 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 dev.morling.onebrc;
+
+import static java.nio.charset.StandardCharsets.UTF_8;
+
+import java.io.File;
+import java.io.IOException;
+import java.io.RandomAccessFile;
+import java.nio.MappedByteBuffer;
+import java.nio.channels.FileChannel.MapMode;
+import java.text.DecimalFormat;
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.Map;
+import java.util.TreeMap;
+import java.util.concurrent.Executors;
+import java.util.stream.Collectors;
+
+/**
+ * Versions (correlate to git commits):
+ * <ol>
+ * <li>2m34s: parallel file read -> load byte chunks in memory -> sequentially process bytes for result</li>
+ * <li>0m59s: process byte chunks in parallel (had to introduce smarter byte chunking so it splits only on newlines)</li>
+ * <li>0m46s: smaller numeric types for MeasurementAggregator</li>
+ * <li>0m39s: implement custom byte[] to int parsing, instead of Double.parseDouble(new String(bytes))</li>
+ * <li>0m18s: run with GraalVM native-image</li>
+ * <li>0m14s: remove ConcurrentHashMap</li>
+ * <li>0m11s: remove Map#compute</li>
+ * </ol>
+ *
+ * <p>
+ * Hardware:
+ * <ul>
+ * <li>CPU: https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+7+5800H&id=3907</li>
+ * <li>NVMe: https://www.harddrivebenchmark.net/hdd.php?hdd=SKHynix%20HFS001TDE9X084N&id=28560</li>
+ * </ul>
+ * </p>
+ */
+public class CalculateAverage_Kidlike {
+
+ public static void main(String[] args) {
+ File file = new File("./measurements.txt");
+ long fileSize = file.length();
+ int processors = (fileSize < 1_000_000) ? 1 : Runtime.getRuntime().availableProcessors();
+ long chunkSize = fileSize / processors;
+
+ MappedByteBuffer[] byteBuffers = new MappedByteBuffer[processors];
+
+ try (var executor = Executors.newVirtualThreadPerTaskExecutor()) {
+ for (int i = 0; i < processors; i++) {
+ long start = i * chunkSize;
+ long length = (i == processors - 1) ? fileSize - chunkSize * (processors - 1) : chunkSize;
+ int processor = i;
+ executor.execute(() -> {
+ long realStart = start;
+ try (RandomAccessFile raf = new RandomAccessFile(file, "r")) {
+ if (start != 0) {
+ raf.seek(start);
+ while (raf.readByte() != '\n') {
+ // move pointer to newline, so that we can process the byte chunks in parallel later.
+ }
+ realStart = raf.getFilePointer();
+ }
+ long realLength = fileSize - realStart;
+ if (realStart + length < fileSize) {
+ raf.seek(realStart + length);
+ while (raf.readByte() != '\n') {
+ // move pointer to newline, so that we can process the byte chunks in parallel later.
+ }
+ realLength = raf.getFilePointer() - realStart;
+ }
+
+ MappedByteBuffer byteBuffer = raf.getChannel().map(MapMode.READ_ONLY, realStart, realLength);
+ byteBuffer.load();
+ byteBuffers[processor] = byteBuffer;
+ }
+ catch (IOException e) {
+ throw new RuntimeException(e);
+ }
+ });
+ }
+ }
+
+ System.out.println(new TreeMap(calculateMeasurements(byteBuffers)));
+ }
+
+ private static Map<String, MeasurementAggregator> calculateMeasurements(MappedByteBuffer[] buffers) {
+ return Arrays.stream(buffers).parallel()
+ .map(buffer -> {
+ var results = new HashMap<String, MeasurementAggregator>();
+
+ var state = State.NEXT_READ_CITY;
+ var citySink = new CheapByteBuffer(100);
+ var measurementSink = new CheapByteBuffer(10);
+
+ while (buffer.hasRemaining()) {
+ byte b = buffer.get();
+ char c = (char) b;
+
+ if (c == ';') {
+ state = State.NEXT_READ_MEASUREMENT;
+ continue;
+ }
+
+ if (c == '\n') {
+ String city = new String(citySink.getBytes(), UTF_8);
+ int measurement = bytesToInt(measurementSink.getBytes());
+ var entry = results.get(city);
+ if (entry == null) {
+ entry = new MeasurementAggregator();
+ results.put(city, entry);
+ }
+
+ entry.count++;
+ entry.sum += measurement;
+ entry.min = (short) Math.min(entry.min, measurement);
+ entry.max = (short) Math.max(entry.max, measurement);
+
+ citySink.clear();
+ measurementSink.clear();
+ state = State.NEXT_READ_CITY;
+ continue;
+ }
+
+ switch (state) {
+ case NEXT_READ_CITY -> citySink.append(b);
+ case NEXT_READ_MEASUREMENT -> measurementSink.append(b);
+ }
+ }
+
+ return results;
+ })
+ .flatMap(m -> m.entrySet().stream())
+ .collect(Collectors.toMap(Map.Entry::getKey, Map.Entry::getValue, MeasurementAggregator::merge));
+ }
+
+ /**
+ * Removes decimal points and returns an integer. For example -12.3 would return -123
+ */
+ static int bytesToInt(byte[] bytes) {
+ short index = (short) (bytes.length - 1);
+
+ boolean isNegative = (bytes[0] == '-');
+ int number = (bytes[index] - '0');
+ index -= 2;
+
+ number += (10 * (bytes[index--] - '0'));
+
+ if (index == 1 || (!isNegative && index == 0)) {
+ number += (100 * (bytes[index] - '0'));
+ }
+
+ if (isNegative) {
+ return -number;
+ }
+ else {
+ return number;
+ }
+ }
+
+ private enum State {
+ NEXT_READ_CITY,
+ NEXT_READ_MEASUREMENT
+ }
+
+ /**
+ * Numbers are stored as integers, because of {@link #bytesToInt}, and then divided by 10.0 to restore their decimal point.
+ */
+ private static class MeasurementAggregator {
+
+ private static final DecimalFormat rounder = new DecimalFormat("0.0");
+ short min = Short.MAX_VALUE;
+ short max = Short.MIN_VALUE;
+ long sum;
+ int count;
+
+ private MeasurementAggregator merge(MeasurementAggregator other) {
+ min = (short) Math.min(min, other.min);
+ max = (short) Math.max(max, other.max);
+ sum += other.sum;
+ count += other.count;
+
+ return this;
+ }
+
+ @Override
+ public String toString() {
+ return rounder.format(min / 10.0)
+ + "/" + rounder.format(Math.round((double) sum / count) / 10.0)
+ + "/" + rounder.format(max / 10.0);
+ }
+ }
+
+ private static class CheapByteBuffer {
+
+ private final byte[] data;
+ private int length;
+
+ public CheapByteBuffer(final int startSize) {
+ this.data = new byte[startSize];
+ this.length = 0;
+ }
+
+ public void append(final byte b) {
+ data[length++] = b;
+ }
+
+ public void clear() {
+ this.length = 0;
+ }
+
+ public byte[] getBytes() {
+ return Arrays.copyOf(this.data, this.length);
+ }
+ }
+}