commit b93b128cb8d4ba6ee57f946123cbba1a9de4b432 · treybastian.com/1brc

+19

calculate_average_ianopolous.sh

···

       1
       +
       #!/bin/sh

     

       2
       +
       #

     

       3
       +
       #  Copyright 2023 The original authors

     

       4
       +
       #

     

       5
       +
       #  Licensed under the Apache License, Version 2.0 (the "License");

     

       6
       +
       #  you may not use this file except in compliance with the License.

     

       7
       +
       #  You may obtain a copy of the License at

     

       8
       +
       #

     

       9
       +
       #      http://www.apache.org/licenses/LICENSE-2.0

     

       10
       +
       #

     

       11
       +
       #  Unless required by applicable law or agreed to in writing, software

     

       12
       +
       #  distributed under the License is distributed on an "AS IS" BASIS,

     

       13
       +
       #  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.

     

       14
       +
       #  See the License for the specific language governing permissions and

     

       15
       +
       #  limitations under the License.

     

       16
       +
       #

     

       17
       +
       

     

       18
       +
       JAVA_OPTS="--enable-preview"

     

       19
       +
       java $JAVA_OPTS --class-path target/average-1.0.0-SNAPSHOT.jar dev.morling.onebrc.CalculateAverage_ianopolous

+201

src/main/java/dev/morling/onebrc/CalculateAverage_ianopolous.java

···

       1
       +
       /*

     

       2
       +
        *  Copyright 2023 The original authors

     

       3
       +
        *

     

       4
       +
        *  Licensed under the Apache License, Version 2.0 (the "License");

     

       5
       +
        *  you may not use this file except in compliance with the License.

     

       6
       +
        *  You may obtain a copy of the License at

     

       7
       +
        *

     

       8
       +
        *      http://www.apache.org/licenses/LICENSE-2.0

     

       9
       +
        *

     

       10
       +
        *  Unless required by applicable law or agreed to in writing, software

     

       11
       +
        *  distributed under the License is distributed on an "AS IS" BASIS,

     

       12
       +
        *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.

     

       13
       +
        *  See the License for the specific language governing permissions and

     

       14
       +
        *  limitations under the License.

     

       15
       +
        */

     

       16
       +
       package dev.morling.onebrc;

     

       17
       +
       

     

       18
       +
       import java.io.*;

     

       19
       +
       import java.nio.*;

     

       20
       +
       import java.nio.channels.*;

     

       21
       +
       import java.util.stream.*;

     

       22
       +
       import java.util.*;

     

       23
       +
       

     

       24
       +
       /* A simple implementation that memory maps the file, reads chunks in parallel and minimises allocation without any unsafe.

     

       25
       +
        *

     

       26
       +
        * Timings on 4 core i7-7500U CPU @ 2.70GHz:

     

       27
       +
        * average_baseline: 4m48s

     

       28
       +
        * ianopolous:         48s

     

       29
       +
       */

     

       30
       +
       public class CalculateAverage_ianopolous {

     

       31
       +
       

     

       32
       +
           public static final int MAX_LINE_LENGTH = 107;

     

       33
       +
           public static final int MAX_STATIONS = 10000;

     

       34
       +
       

     

       35
       +
           public static void main(String[] args) {

     

       36
       +
               File input = new File("./measurements.txt");

     

       37
       +
               long filesize = input.length();

     

       38
       +
               long chunkSize = 256 * 1024 * 1024;

     

       39
       +
               int nChunks = (int) ((filesize + chunkSize - 1) / chunkSize);

     

       40
       +
               List<HashMap<String, Stat>> allResults = IntStream.range(0, nChunks).mapToObj(i -> {

     

       41
       +
                   HashMap<String, Stat> results = new HashMap(512);

     

       42
       +
                   parseStats(i * chunkSize, Math.min((i + 1) * chunkSize, filesize), results);

     

       43
       +
                   return results;

     

       44
       +
               }).parallel().toList();

     

       45
       +
               HashMap<String, Stat> result = allResults.getFirst();

     

       46
       +
               for (int i = 1; i < allResults.size(); ++i) {

     

       47
       +
                   for (Map.Entry<String, Stat> entry : allResults.get(i).entrySet()) {

     

       48
       +
                       Stat current = result.putIfAbsent(entry.getKey(), entry.getValue());

     

       49
       +
                       if (current != null) {

     

       50
       +
                           current.merge(entry.getValue());

     

       51
       +
                       }

     

       52
       +
                   }

     

       53
       +
               }

     

       54
       +
       

     

       55
       +
               System.out.println(new TreeMap<>(result));

     

       56
       +
           }

     

       57
       +
       

     

       58
       +
           public record Station(String name, ByteBuffer buf) {

     

       59
       +
           }

     

       60
       +
       

     

       61
       +
           public static boolean matchingStationBytes(int start, int end, MappedByteBuffer buffer, Station existing) {

     

       62
       +
               buffer.position(start);

     

       63
       +
               for (int i = start; i < end; i++) {

     

       64
       +
                   if (existing.buf.get(i - start) != buffer.get(i))

     

       65
       +
                       return false;

     

       66
       +
               }

     

       67
       +
               return true;

     

       68
       +
           }

     

       69
       +
       

     

       70
       +
           public static Station parseStation(int start, int end, int hash, MappedByteBuffer buffer, List<List<Station>> stations) {

     

       71
       +
               int index = Math.floorMod(hash, MAX_STATIONS);

     

       72
       +
               List<Station> matches = stations.get(index);

     

       73
       +
               if (matches == null) {

     

       74
       +
                   List<Station> value = new ArrayList<>();

     

       75
       +
                   byte[] stationBuffer = new byte[end - start];

     

       76
       +
                   buffer.position(start);

     

       77
       +
                   buffer.get(stationBuffer);

     

       78
       +
                   String name = new String(stationBuffer);

     

       79
       +
                   Station res = new Station(name, ByteBuffer.wrap(stationBuffer));

     

       80
       +
                   value.add(res);

     

       81
       +
                   stations.set(index, value);

     

       82
       +
                   return res;

     

       83
       +
               }

     

       84
       +
               else {

     

       85
       +
                   for (int i = 0; i < matches.size(); i++) {

     

       86
       +
                       Station s = matches.get(i);

     

       87
       +
                       if (matchingStationBytes(start, end, buffer, s))

     

       88
       +
                           return s;

     

       89
       +
                   }

     

       90
       +
                   byte[] stationBuffer = new byte[end - start];

     

       91
       +
                   buffer.position(start);

     

       92
       +
                   buffer.get(stationBuffer);

     

       93
       +
                   Station res = new Station(new String(stationBuffer), ByteBuffer.wrap(stationBuffer));

     

       94
       +
                   matches.add(res);

     

       95
       +
                   return res;

     

       96
       +
               }

     

       97
       +
           }

     

       98
       +
       

     

       99
       +
           public static void parseStats(long startByte, long endByte, Map<String, Stat> results) {

     

       100
       +
               try {

     

       101
       +
                   RandomAccessFile file = new RandomAccessFile("./measurements.txt", "r");

     

       102
       +
                   long maxEnd = Math.min(file.length(), endByte + MAX_LINE_LENGTH);

     

       103
       +
                   long len = maxEnd - startByte;

     

       104
       +
                   if (len > Integer.MAX_VALUE)

     

       105
       +
                       throw new RuntimeException("Segment size must fit into an int");

     

       106
       +
                   int maxDone = (int) (endByte - startByte);

     

       107
       +
                   MappedByteBuffer buffer = file.getChannel().map(FileChannel.MapMode.READ_ONLY, startByte, len);

     

       108
       +
                   int done = 0;

     

       109
       +
                   // read first partial line

     

       110
       +
                   if (startByte > 0) {

     

       111
       +
                       for (int i = 0; i < MAX_LINE_LENGTH; i++) {

     

       112
       +
                           byte b = buffer.get(i);

     

       113
       +
                           if (b == '\n') {

     

       114
       +
                               done = i + 1;

     

       115
       +
                               break;

     

       116
       +
                           }

     

       117
       +
                       }

     

       118
       +
                   }

     

       119
       +
       

     

       120
       +
                   List<List<Station>> stations = new ArrayList<>(MAX_STATIONS);

     

       121
       +
                   for (int i = 0; i < MAX_STATIONS; i++)

     

       122
       +
                       stations.add(null);

     

       123
       +
                   int lineStart = done;

     

       124
       +
                   int lineSplit = 0;

     

       125
       +
                   long temperature = 0;

     

       126
       +
                   int hash = 1;

     

       127
       +
                   boolean negative = false;

     

       128
       +
                   while (done < maxDone) {

     

       129
       +
                       Station station = null;

     

       130
       +
                       for (int i = done; i < done + MAX_LINE_LENGTH && i < maxEnd; i++) {

     

       131
       +
                           byte b = buffer.get(i);

     

       132
       +
                           if (b == '\n') {

     

       133
       +
                               done = i + 1;

     

       134
       +
                               Stat res = results.get(station.name);

     

       135
       +
                               temperature = negative ? -temperature : temperature;

     

       136
       +
                               if (res != null) {

     

       137
       +
                                   res.add(temperature);

     

       138
       +
                               }

     

       139
       +
                               else {

     

       140
       +
                                   res = new Stat();

     

       141
       +
                                   res.add(temperature);

     

       142
       +
                                   results.put(station.name, res);

     

       143
       +
                               }

     

       144
       +
                               lineStart = done;

     

       145
       +
                               station = null;

     

       146
       +
                               hash = 1;

     

       147
       +
                               break;

     

       148
       +
                           }

     

       149
       +
                           else if (b == ';') {

     

       150
       +
                               lineSplit = i;

     

       151
       +
                               station = parseStation(lineStart, lineSplit, hash, buffer, stations);

     

       152
       +
                               temperature = 0;

     

       153
       +
                               negative = false;

     

       154
       +
                           }

     

       155
       +
                           else if (b == '-' && station != null) {

     

       156
       +
                               negative = true;

     

       157
       +
                           }

     

       158
       +
                           else if (b != '.' && station != null) {

     

       159
       +
                               temperature = temperature * 10 + (b - 0x30);

     

       160
       +
                           }

     

       161
       +
                           else {

     

       162
       +
                               hash = 31 * hash + b;

     

       163
       +
                           }

     

       164
       +
                       }

     

       165
       +
                   }

     

       166
       +
               }

     

       167
       +
               catch (IOException e) {

     

       168
       +
                   throw new RuntimeException(e);

     

       169
       +
               }

     

       170
       +
           }

     

       171
       +
       

     

       172
       +
           public static class Stat {

     

       173
       +
               long min = Long.MAX_VALUE, max = Long.MIN_VALUE, total = 0, count = 0;

     

       174
       +
       

     

       175
       +
               public void add(long value) {

     

       176
       +
                   if (value < min)

     

       177
       +
                       min = value;

     

       178
       +
                   if (value > max)

     

       179
       +
                       max = value;

     

       180
       +
                   total += value;

     

       181
       +
                   count++;

     

       182
       +
               }

     

       183
       +
       

     

       184
       +
               public void merge(Stat value) {

     

       185
       +
                   if (value.min < min)

     

       186
       +
                       min = value.min;

     

       187
       +
                   if (value.max > max)

     

       188
       +
                       max = value.max;

     

       189
       +
                   total += value.total;

     

       190
       +
                   count += value.count;

     

       191
       +
               }

     

       192
       +
       

     

       193
       +
               private static double round(double value) {

     

       194
       +
                   return Math.round(value) / 10.0;

     

       195
       +
               }

     

       196
       +
       

     

       197
       +
               public String toString() {

     

       198
       +
                   return round((double) min) + "/" + round(((double) total) / count) + "/" + round((double) max);

     

       199
       +
               }

     

       200
       +
           }

     

       201
       +
       }