/
UnfilteredPartitionIterators.java
420 lines (366 loc) · 15.1 KB
/
UnfilteredPartitionIterators.java
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
/*
* 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.cassandra.db.partitions;
import java.io.IOError;
import java.io.IOException;
import java.util.*;
import org.apache.cassandra.db.*;
import org.apache.cassandra.db.filter.ColumnFilter;
import org.apache.cassandra.db.rows.*;
import org.apache.cassandra.db.transform.FilteredPartitions;
import org.apache.cassandra.db.transform.MorePartitions;
import org.apache.cassandra.db.transform.Transformation;
import org.apache.cassandra.io.util.DataInputPlus;
import org.apache.cassandra.io.util.DataOutputPlus;
import org.apache.cassandra.schema.TableMetadata;
import org.apache.cassandra.utils.MergeIterator;
/**
* Static methods to work with partition iterators.
*/
public abstract class UnfilteredPartitionIterators
{
private static final Serializer serializer = new Serializer();
private static final Comparator<UnfilteredRowIterator> partitionComparator = (p1, p2) -> p1.partitionKey().compareTo(p2.partitionKey());
private UnfilteredPartitionIterators() {}
public interface MergeListener
{
/**
* Returns true if the merger needs to preserve the position of sources within the merge when passing data to
* the listener. If false, the merger can avoid creating empty sources for non-present partitions and
* significantly speed up processing.
*
* @return True to preserve position of source iterators.
*/
public default boolean preserveOrder() { return true; }
public UnfilteredRowIterators.MergeListener getRowMergeListener(DecoratedKey partitionKey, List<UnfilteredRowIterator> versions);
public default void close() {}
public static MergeListener NOOP = new MergeListener()
{
@Override
public boolean preserveOrder()
{
return false;
}
public UnfilteredRowIterators.MergeListener getRowMergeListener(DecoratedKey partitionKey, List<UnfilteredRowIterator> versions)
{
return UnfilteredRowIterators.MergeListener.NOOP;
}
};
}
public static UnfilteredRowIterator getOnlyElement(final UnfilteredPartitionIterator iter, SinglePartitionReadCommand command)
{
// If the query has no results, we'll get an empty iterator, but we still
// want a RowIterator out of this method, so we return an empty one.
UnfilteredRowIterator toReturn = iter.hasNext()
? iter.next()
: EmptyIterators.unfilteredRow(command.metadata(),
command.partitionKey(),
command.clusteringIndexFilter().isReversed());
// Note that in general, we should wrap the result so that it's close method actually
// close the whole UnfilteredPartitionIterator.
class Close extends Transformation
{
public void onPartitionClose()
{
// asserting this only now because it bothers Serializer if hasNext() is called before
// the previously returned iterator hasn't been fully consumed.
boolean hadNext = iter.hasNext();
iter.close();
assert !hadNext;
}
}
return Transformation.apply(toReturn, new Close());
}
public static UnfilteredPartitionIterator concat(final List<UnfilteredPartitionIterator> iterators)
{
if (iterators.size() == 1)
return iterators.get(0);
class Extend implements MorePartitions<UnfilteredPartitionIterator>
{
int i = 1;
public UnfilteredPartitionIterator moreContents()
{
if (i >= iterators.size())
return null;
return iterators.get(i++);
}
}
return MorePartitions.extend(iterators.get(0), new Extend());
}
public static PartitionIterator filter(final UnfilteredPartitionIterator iterator, final long nowInSec)
{
return FilteredPartitions.filter(iterator, nowInSec);
}
public static UnfilteredPartitionIterator merge(final List<? extends UnfilteredPartitionIterator> iterators, final MergeListener listener)
{
assert !iterators.isEmpty();
final TableMetadata metadata = iterators.get(0).metadata();
final boolean preserveOrder = listener != null && listener.preserveOrder();
final MergeIterator<UnfilteredRowIterator, UnfilteredRowIterator> merged = MergeIterator.get(iterators, partitionComparator, new MergeIterator.Reducer<UnfilteredRowIterator, UnfilteredRowIterator>()
{
private final List<UnfilteredRowIterator> toMerge = new ArrayList<>(iterators.size());
private DecoratedKey partitionKey;
private boolean isReverseOrder;
public void reduce(int idx, UnfilteredRowIterator current)
{
partitionKey = current.partitionKey();
isReverseOrder = current.isReverseOrder();
if (preserveOrder)
{
// Note that because the MergeListener cares about it, we want to preserve the index of the iterator.
// Non-present iterator will thus be set to empty in getReduced.
toMerge.set(idx, current);
}
else
{
toMerge.add(current);
}
}
protected UnfilteredRowIterator getReduced()
{
UnfilteredRowIterators.MergeListener rowListener = listener == null
? null
: listener.getRowMergeListener(partitionKey, toMerge);
if (preserveOrder)
{
// Make a single empty iterator object to merge, we don't need toMerge.size() copiess
UnfilteredRowIterator empty = null;
// Replace nulls by empty iterators
for (int i = 0; i < toMerge.size(); i++)
{
if (toMerge.get(i) == null)
{
if (null == empty)
empty = EmptyIterators.unfilteredRow(metadata, partitionKey, isReverseOrder);
toMerge.set(i, empty);
}
}
}
return UnfilteredRowIterators.merge(toMerge, rowListener);
}
protected void onKeyChange()
{
toMerge.clear();
if (preserveOrder)
{
for (int i = 0; i < iterators.size(); i++)
toMerge.add(null);
}
}
});
return new AbstractUnfilteredPartitionIterator()
{
public TableMetadata metadata()
{
return metadata;
}
public boolean hasNext()
{
return merged.hasNext();
}
public UnfilteredRowIterator next()
{
return merged.next();
}
@Override
public void close()
{
merged.close();
if (listener != null)
listener.close();
}
};
}
public static UnfilteredPartitionIterator mergeLazily(final List<? extends UnfilteredPartitionIterator> iterators)
{
assert !iterators.isEmpty();
if (iterators.size() == 1)
return iterators.get(0);
final TableMetadata metadata = iterators.get(0).metadata();
final MergeIterator<UnfilteredRowIterator, UnfilteredRowIterator> merged = MergeIterator.get(iterators, partitionComparator, new MergeIterator.Reducer<UnfilteredRowIterator, UnfilteredRowIterator>()
{
private final List<UnfilteredRowIterator> toMerge = new ArrayList<>(iterators.size());
public void reduce(int idx, UnfilteredRowIterator current)
{
toMerge.add(current);
}
protected UnfilteredRowIterator getReduced()
{
return new LazilyInitializedUnfilteredRowIterator(toMerge.get(0).partitionKey())
{
protected UnfilteredRowIterator initializeIterator()
{
return UnfilteredRowIterators.merge(toMerge);
}
};
}
protected void onKeyChange()
{
toMerge.clear();
}
});
return new AbstractUnfilteredPartitionIterator()
{
public TableMetadata metadata()
{
return metadata;
}
public boolean hasNext()
{
return merged.hasNext();
}
public UnfilteredRowIterator next()
{
return merged.next();
}
@Override
public void close()
{
merged.close();
}
};
}
/**
* Digests the the provided iterator.
*
* Caller must close the provided iterator.
*
* @param iterator the iterator to digest.
* @param digest the {@link Digest} to use.
* @param version the messaging protocol to use when producing the digest.
*/
public static void digest(UnfilteredPartitionIterator iterator, Digest digest, int version)
{
while (iterator.hasNext())
{
try (UnfilteredRowIterator partition = iterator.next())
{
UnfilteredRowIterators.digest(partition, digest, version);
}
}
}
public static Serializer serializerForIntraNode()
{
return serializer;
}
/**
* Wraps the provided iterator so it logs the returned rows/RT for debugging purposes.
* <p>
* Note that this is only meant for debugging as this can log a very large amount of
* logging at INFO.
*/
public static UnfilteredPartitionIterator loggingIterator(UnfilteredPartitionIterator iterator, final String id, final boolean fullDetails)
{
class Logging extends Transformation<UnfilteredRowIterator>
{
public UnfilteredRowIterator applyToPartition(UnfilteredRowIterator partition)
{
return UnfilteredRowIterators.loggingIterator(partition, id, fullDetails);
}
}
return Transformation.apply(iterator, new Logging());
}
/**
* Serialize each UnfilteredSerializer one after the other, with an initial byte that indicates whether
* we're done or not.
*/
public static class Serializer
{
public void serialize(UnfilteredPartitionIterator iter, ColumnFilter selection, DataOutputPlus out, int version) throws IOException
{
// Previously, a boolean indicating if this was for a thrift query.
// Unused since 4.0 but kept on wire for compatibility.
out.writeBoolean(false);
while (iter.hasNext())
{
out.writeBoolean(true);
try (UnfilteredRowIterator partition = iter.next())
{
UnfilteredRowIteratorSerializer.serializer.serialize(partition, selection, out, version);
}
}
out.writeBoolean(false);
}
public UnfilteredPartitionIterator deserialize(final DataInputPlus in, final int version, final TableMetadata metadata, final ColumnFilter selection, final DeserializationHelper.Flag flag) throws IOException
{
// Skip now unused isForThrift boolean
in.readBoolean();
return new AbstractUnfilteredPartitionIterator()
{
private UnfilteredRowIterator next;
private boolean hasNext;
private boolean nextReturned = true;
public TableMetadata metadata()
{
return metadata;
}
public boolean hasNext()
{
if (!nextReturned)
return hasNext;
/*
* We must consume the previous iterator before we start deserializing the next partition, so
* that we start from the right position in the byte stream.
*
* It's possible however that it hasn't been fully consumed by upstream consumers - for example,
* if a per partition limit caused merge iterator to stop early (see CASSANDRA-13911).
*
* In that case we must drain the unconsumed iterator fully ourselves, here.
*
* NOTE: transformations of the upstream BaseRows won't be applied for these consumed elements,
* so, for exmaple, they won't be counted.
*/
if (null != next)
while (next.hasNext())
next.next();
try
{
hasNext = in.readBoolean();
nextReturned = false;
return hasNext;
}
catch (IOException e)
{
throw new IOError(e);
}
}
public UnfilteredRowIterator next()
{
if (nextReturned && !hasNext())
throw new NoSuchElementException();
try
{
nextReturned = true;
next = UnfilteredRowIteratorSerializer.serializer.deserialize(in, version, metadata, selection, flag);
return next;
}
catch (IOException e)
{
throw new IOError(e);
}
}
@Override
public void close()
{
if (next != null)
next.close();
}
};
}
}
}