1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
|
/*
* 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 java.io.*;
import java.nio.file.Files;
import java.nio.file.Paths;
import java.util.*;
import java.util.concurrent.Callable;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.Executors;
import java.util.concurrent.Future;
public class CalculateAverage_maeda6uiui {
record RecordCollectorResult(
Map<String, Double> mins,
Map<String, Double> maxes,
Map<String, Double> sums,
Map<String, Integer> counts) {
}
static class RecordCollector implements Callable<RecordCollectorResult> {
private String inputFilepath;
private long startByteIndex;
private long numBytesToRead;
private char delimiter;
private int byteBufferSize;
private int bisBufferSize;
private Map<String, Double> mins;
private Map<String, Double> maxes;
private Map<String, Double> sums;
private Map<String, Integer> counts;
public RecordCollector(
String inputFilepath,
long startByteIndex,
long numBytesToRead,
char delimiter,
int byteBufferSize,
int bisBufferSize) {
this.inputFilepath = inputFilepath;
this.startByteIndex = startByteIndex;
this.numBytesToRead = numBytesToRead;
this.delimiter = delimiter;
this.byteBufferSize = byteBufferSize;
this.bisBufferSize = bisBufferSize;
mins = new HashMap<>();
maxes = new HashMap<>();
sums = new HashMap<>();
counts = new HashMap<>();
}
private int byteToInt(byte b) {
return switch (b) {
case '0' -> 0;
case '1' -> 1;
case '2' -> 2;
case '3' -> 3;
case '4' -> 4;
case '5' -> 5;
case '6' -> 6;
case '7' -> 7;
case '8' -> 8;
case '9' -> 9;
default -> -1;
};
}
private double parseDouble(byte[] bs) {
// Get the sign
int valSign;
if (bs[0] == '-') {
valSign = -1;
}
else {
valSign = 1;
}
// Get the dot position
int dotPos = -1;
for (int i = 0; i < bs.length; i++) {
if (bs[i] == '.') {
dotPos = i;
break;
}
}
if (dotPos == -1) {
return Double.NaN;
}
// Get the integer part
int valIntPart;
int intPartStartIndex = (valSign == -1) ? 1 : 0;
int intPartLength = dotPos - intPartStartIndex;
// One-digit value
if (intPartLength == 1) {
valIntPart = this.byteToInt(bs[dotPos - 1]);
}
// Two-digit value
else if (intPartLength == 2) {
int valTens = this.byteToInt(bs[dotPos - 2]);
int valOnes = this.byteToInt(bs[dotPos - 1]);
valIntPart = valTens * 10 + valOnes;
}
else {
return Double.NaN;
}
// Get the decimal part
double valDecPart = this.byteToInt(bs[dotPos + 1]) * 0.1;
return valSign * (valIntPart + valDecPart);
}
@Override
public RecordCollectorResult call() {
// Start and end indices are most likely pointing to the middle of a line
// Therefore, actual start and end indices should be determined
// before proceeding to actual reading of the file
long actualStartByteIndex = -1;
long actualEndByteIndex = -1;
try (var bis = new BufferedInputStream(new FileInputStream(inputFilepath))) {
int b;
int readCount = 0;
long firstLFPos;
// If start index specified is 0, actual start index is also 0
if (startByteIndex == 0) {
actualStartByteIndex = 0;
}
else {
// Skip until the preceding byte of the start index specified
bis.skipNBytes(startByteIndex - 1);
// Get the preceding byte
b = bis.read();
// If the preceding byte is LF,
// actual start index is the start index specified
// because it is the start of a new line
if (b == '\n') {
actualStartByteIndex = startByteIndex;
}
}
if (actualStartByteIndex != -1) {
// Skip until the end byte specified
bis.skipNBytes(numBytesToRead);
}
// Start index specified is pointing to the middle of a line
// In that case, actual start index is the one following the LF of that line
// (Start index of the next line)
else {
firstLFPos = startByteIndex;
while ((b = bis.read()) != -1) {
readCount++;
if (b == '\n') {
break;
}
firstLFPos++;
}
actualStartByteIndex = firstLFPos + 1;
// Skip until the end byte specified
bis.skipNBytes(numBytesToRead - readCount);
}
// Actual end index is the first LF encountered
readCount = 0;
firstLFPos = startByteIndex + numBytesToRead;
while ((b = bis.read()) != -1) {
readCount++;
if (b == '\n') {
break;
}
firstLFPos++;
}
actualEndByteIndex = firstLFPos;
}
catch (IOException e) {
System.err.println(e);
return null;
}
// Get actual number of bytes to read
long actualNumBytesToRead = actualEndByteIndex - actualStartByteIndex + 1;
// Read bytes from the range obtained above
try (var bis = new BufferedInputStream(new FileInputStream(inputFilepath), bisBufferSize)) {
// Skip until the start byte
bis.skipNBytes(actualStartByteIndex);
final int EXTENSION_SIZE = 64;
var buffer = new byte[byteBufferSize];
var extendedBuffer = new byte[byteBufferSize + EXTENSION_SIZE];
// Read bytes in chunk
long numTotalBytesRead = 0;
while (true) {
int chunkSize;
if (actualNumBytesToRead - numTotalBytesRead < byteBufferSize) {
chunkSize = (int) (actualNumBytesToRead - numTotalBytesRead);
}
else {
chunkSize = byteBufferSize;
}
if (chunkSize <= 0) {
break;
}
Arrays.fill(buffer, (byte) 0);
bis.read(buffer, 0, chunkSize);
numTotalBytesRead += chunkSize;
// Copy read content to another buffer
Arrays.fill(extendedBuffer, (byte) 0);
System.arraycopy(buffer, 0, extendedBuffer, 0, chunkSize);
// Read until next LF is found
// if end of buffer read above does not correspond to end of line
for (int i = 0; i < EXTENSION_SIZE; i++) {
int b = bis.read();
if (b == -1) {
break;
}
else if (b == '\n') {
extendedBuffer[chunkSize + i] = '\n';
numTotalBytesRead++;
break;
}
extendedBuffer[chunkSize + i] = (byte) b;
numTotalBytesRead++;
}
int currentDelimPos = -1;
int currentLFPos = -1;
int nextLineStartPos = 0;
for (int i = 0; i < extendedBuffer.length; i++) {
if (extendedBuffer[i] == 0) {
break;
}
if (extendedBuffer[i] == delimiter) {
currentDelimPos = i;
}
else if (extendedBuffer[i] == '\n') {
currentLFPos = i;
}
if (currentLFPos != -1) {
// Error
if (currentDelimPos == -1) {
System.err.printf(
"Error near byte index %d\n",
actualStartByteIndex + numTotalBytesRead);
}
else {
String stationName = new String(
Arrays.copyOfRange(extendedBuffer, nextLineStartPos, currentDelimPos));
// Parse string to double by myself
// because Double.parseDouble() is slow...
double temperature = this.parseDouble(
Arrays.copyOfRange(extendedBuffer, currentDelimPos + 1, currentLFPos));
// Populate the maps
if (!mins.containsKey(stationName)) {
mins.put(stationName, temperature);
maxes.put(stationName, temperature);
sums.put(stationName, temperature);
counts.put(stationName, 1);
}
else {
double currentMin = mins.get(stationName);
double currentMax = maxes.get(stationName);
double currentSum = sums.get(stationName);
int currentCount = counts.get(stationName);
if (temperature < currentMin) {
mins.put(stationName, temperature);
}
else if (temperature > currentMax) {
maxes.put(stationName, temperature);
}
sums.put(stationName, currentSum + temperature);
counts.put(stationName, currentCount + 1);
}
}
nextLineStartPos = currentLFPos + 1;
currentDelimPos = -1;
currentLFPos = -1;
}
}
}
}
catch (IOException e) {
System.err.println(e);
return null;
}
return new RecordCollectorResult(mins, maxes, sums, counts);
}
}
private static double round(double d) {
return Math.round(d * 10.0) / 10.0;
}
public static void main(String[] args) {
final String INPUT_FILEPATH = "./measurements.txt";
final int DESIRED_NUM_THREADS = 20;
final char DELIMITER = ';';
final int BIS_BUFFER_SIZE = 1024 * 1024;
final int BYTE_BUFFER_SIZE = 1024;
final int MULTI_THREAD_NUM_LINES_THRESHOLD = DESIRED_NUM_THREADS * 10;
// First get the number of total bytes in the input file
long numTotalBytes;
try {
numTotalBytes = Files.size(Paths.get(INPUT_FILEPATH));
}
catch (IOException e) {
e.printStackTrace();
return;
}
// Make sure the input file has enough lines
// for this multithreading approach to work efficiently
int actualNumThreads = 1;
try (var br = new BufferedReader(new FileReader(INPUT_FILEPATH))) {
int lineCount = 0;
while (br.readLine() != null) {
lineCount++;
if (lineCount >= MULTI_THREAD_NUM_LINES_THRESHOLD) {
actualNumThreads = DESIRED_NUM_THREADS;
break;
}
}
}
catch (IOException e) {
e.printStackTrace();
return;
}
// Calculate the number of bytes each thread has to process
long numBytesToProcessPerThread = numTotalBytes / actualNumThreads;
long remainingNumBytesToProcess = numTotalBytes % actualNumThreads;
var exec = Executors.newFixedThreadPool(actualNumThreads);
var futures = new ArrayList<Future<RecordCollectorResult>>();
for (int i = 0; i < actualNumThreads; i++) {
RecordCollector recordCollector;
if (i == actualNumThreads - 1) {
recordCollector = new RecordCollector(
INPUT_FILEPATH,
i * numBytesToProcessPerThread,
numBytesToProcessPerThread + remainingNumBytesToProcess,
DELIMITER,
BYTE_BUFFER_SIZE,
BIS_BUFFER_SIZE);
}
else {
recordCollector = new RecordCollector(
INPUT_FILEPATH,
i * numBytesToProcessPerThread,
numBytesToProcessPerThread,
DELIMITER,
BYTE_BUFFER_SIZE,
BIS_BUFFER_SIZE);
}
Future<RecordCollectorResult> future = exec.submit(recordCollector);
futures.add(future);
}
// Consolidate results of each thread
var mins = new HashMap<String, Double>();
var maxes = new HashMap<String, Double>();
var sums = new HashMap<String, Double>();
var counts = new HashMap<String, Integer>();
try {
for (var future : futures) {
RecordCollectorResult result = future.get();
result.mins.forEach((k, v) -> {
if (!mins.containsKey(k)) {
mins.put(k, v);
}
else {
mins.put(k, Double.min(v, mins.get(k)));
}
});
result.maxes.forEach((k, v) -> {
if (!maxes.containsKey(k)) {
maxes.put(k, v);
}
else {
maxes.put(k, Double.max(v, maxes.get(k)));
}
});
result.sums.forEach((k, v) -> {
if (!sums.containsKey(k)) {
sums.put(k, v);
}
else {
sums.put(k, Double.sum(v, sums.get(k)));
}
});
result.counts.forEach((k, v) -> {
if (!counts.containsKey(k)) {
counts.put(k, v);
}
else {
counts.put(k, Integer.sum(v, counts.get(k)));
}
});
}
}
catch (InterruptedException | ExecutionException e) {
e.printStackTrace();
return;
}
finally {
exec.shutdown();
}
// Calculate means
var means = new HashMap<String, Double>();
sums.forEach((k, v) -> means.put(k, v / counts.get(k)));
// Sort station names
List<String> sortedStationNames = means
.keySet()
.stream()
.sorted()
.toList();
// Create output string
var sb = new StringBuilder();
sb.append("{");
sortedStationNames.forEach(stationName -> {
sb
.append(stationName)
.append("=")
.append(round(mins.get(stationName)))
.append("/")
.append(round(means.get(stationName)))
.append("/")
.append(round(maxes.get(stationName)))
.append(", ");
});
sb.delete(sb.length() - 2, sb.length());
sb.append("}");
// Print result string
System.out.println(sb);
}
}
|