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
486
487
488
489
490
491
|
/**
* Copyright (c) 2011 10gen Inc.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License, version 3,
* as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
* As a special exception, the copyright holders give permission to link the
* code of portions of this program with the OpenSSL library under certain
* conditions as described in each individual source file and distribute
* linked combinations including the program with the OpenSSL library. You
* must comply with the GNU Affero General Public License in all respects for
* all of the code used other than as permitted herein. If you modify file(s)
* with this exception, you may extend this exception to your version of the
* file(s), but you are not obligated to do so. If you do not wish to do so,
* delete this exception statement from your version. If you delete this
* exception statement from all source files in the program, then also delete
* it in the license file.
*/
#include "mongo/platform/basic.h"
#include "mongo/db/pipeline/document.h"
#include <boost/functional/hash.hpp>
#include "mongo/bson/bson_depth.h"
#include "mongo/db/jsobj.h"
#include "mongo/db/pipeline/field_path.h"
#include "mongo/util/mongoutils/str.h"
namespace mongo {
using namespace mongoutils;
using boost::intrusive_ptr;
using std::string;
using std::vector;
const DocumentStorage DocumentStorage::kEmptyDoc;
Position DocumentStorage::findField(StringData requested) const {
int reqSize = requested.size(); // get size calculation out of the way if needed
if (_numFields >= HASH_TAB_MIN) { // hash lookup
const unsigned bucket = bucketForKey(requested);
Position pos = _hashTab[bucket];
while (pos.found()) {
const ValueElement& elem = getField(pos);
if (elem.nameLen == reqSize && memcmp(requested.rawData(), elem._name, reqSize) == 0) {
return pos;
}
// possible collision
pos = elem.nextCollision;
}
} else { // linear scan
for (DocumentStorageIterator it = iteratorAll(); !it.atEnd(); it.advance()) {
if (it->nameLen == reqSize && memcmp(requested.rawData(), it->_name, reqSize) == 0) {
return it.position();
}
}
}
// if we got here, there's no such field
return Position();
}
Value& DocumentStorage::appendField(StringData name) {
Position pos = getNextPosition();
const int nameSize = name.size();
// these are the same for everyone
const Position nextCollision;
const Value value;
// Make room for new field (and padding at end for alignment)
const unsigned newUsed = ValueElement::align(_usedBytes + sizeof(ValueElement) + nameSize);
if (_buffer + newUsed > _bufferEnd)
alloc(newUsed);
_usedBytes = newUsed;
// Append structure of a ValueElement
char* dest = _buffer + pos.index; // must be after alloc since it changes _buffer
#define append(x) \
memcpy(dest, &(x), sizeof(x)); \
dest += sizeof(x)
append(value);
append(nextCollision);
append(nameSize);
name.copyTo(dest, true);
// Padding for alignment handled above
#undef append
// Make sure next field starts where we expect it
fassert(16486, getField(pos).next()->ptr() == _buffer + _usedBytes);
_numFields++;
if (_numFields > HASH_TAB_MIN) {
addFieldToHashTable(pos);
} else if (_numFields == HASH_TAB_MIN) {
// adds all fields to hash table (including the one we just added)
rehash();
}
return getField(pos).val;
}
// Call after adding field to _fields and increasing _numFields
void DocumentStorage::addFieldToHashTable(Position pos) {
ValueElement& elem = getField(pos);
elem.nextCollision = Position();
const unsigned bucket = bucketForKey(elem.nameSD());
Position* posPtr = &_hashTab[bucket];
while (posPtr->found()) {
// collision: walk links and add new to end
posPtr = &getField(*posPtr).nextCollision;
}
*posPtr = Position(pos.index);
}
void DocumentStorage::alloc(unsigned newSize) {
const bool firstAlloc = !_buffer;
const bool doingRehash = needRehash();
const size_t oldCapacity = _bufferEnd - _buffer;
// make new bucket count big enough
while (needRehash() || hashTabBuckets() < HASH_TAB_INIT_SIZE)
_hashTabMask = hashTabBuckets() * 2 - 1;
// only allocate power-of-two sized space > 128 bytes
size_t capacity = 128;
while (capacity < newSize + hashTabBytes())
capacity *= 2;
uassert(16490, "Tried to make oversized document", capacity <= size_t(BufferMaxSize));
std::unique_ptr<char[]> oldBuf(_buffer);
_buffer = new char[capacity];
_bufferEnd = _buffer + capacity - hashTabBytes();
if (!firstAlloc) {
// This just copies the elements
memcpy(_buffer, oldBuf.get(), _usedBytes);
if (_numFields >= HASH_TAB_MIN) {
// if we were hashing, deal with the hash table
if (doingRehash) {
rehash();
} else {
// no rehash needed so just slide table down to new position
memcpy(_hashTab, oldBuf.get() + oldCapacity, hashTabBytes());
}
}
}
}
void DocumentStorage::reserveFields(size_t expectedFields) {
fassert(16487, !_buffer);
unsigned buckets = HASH_TAB_INIT_SIZE;
while (buckets < expectedFields)
buckets *= 2;
_hashTabMask = buckets - 1;
// Using expectedFields+1 to allow space for long field names
const size_t newSize = (expectedFields + 1) * ValueElement::align(sizeof(ValueElement));
uassert(16491, "Tried to make oversized document", newSize <= size_t(BufferMaxSize));
_buffer = new char[newSize + hashTabBytes()];
_bufferEnd = _buffer + newSize;
}
intrusive_ptr<DocumentStorage> DocumentStorage::clone() const {
intrusive_ptr<DocumentStorage> out(new DocumentStorage());
// Make a copy of the buffer.
// It is very important that the positions of each field are the same after cloning.
const size_t bufferBytes = allocatedBytes();
out->_buffer = new char[bufferBytes];
out->_bufferEnd = out->_buffer + (_bufferEnd - _buffer);
if (bufferBytes > 0) {
memcpy(out->_buffer, _buffer, bufferBytes);
}
// Copy remaining fields
out->_usedBytes = _usedBytes;
out->_numFields = _numFields;
out->_hashTabMask = _hashTabMask;
out->_metaFields = _metaFields;
out->_textScore = _textScore;
out->_randVal = _randVal;
// Tell values that they have been memcpyed (updates ref counts)
for (DocumentStorageIterator it = out->iteratorAll(); !it.atEnd(); it.advance()) {
it->val.memcpyed();
}
return out;
}
DocumentStorage::~DocumentStorage() {
std::unique_ptr<char[]> deleteBufferAtScopeEnd(_buffer);
for (DocumentStorageIterator it = iteratorAll(); !it.atEnd(); it.advance()) {
it->val.~Value(); // explicit destructor call
}
}
Document::Document(const BSONObj& bson) {
MutableDocument md(bson.nFields());
BSONObjIterator it(bson);
while (it.more()) {
BSONElement bsonElement(it.next());
md.addField(bsonElement.fieldNameStringData(), Value(bsonElement));
}
*this = md.freeze();
}
Document::Document(std::initializer_list<std::pair<StringData, ImplicitValue>> initializerList) {
MutableDocument mutableDoc(initializerList.size());
for (auto&& pair : initializerList) {
mutableDoc.addField(pair.first, pair.second);
}
*this = mutableDoc.freeze();
}
BSONObjBuilder& operator<<(BSONObjBuilderValueStream& builder, const Document& doc) {
BSONObjBuilder subobj(builder.subobjStart());
doc.toBson(&subobj);
subobj.doneFast();
return builder.builder();
}
void Document::toBson(BSONObjBuilder* builder, size_t recursionLevel) const {
uassert(ErrorCodes::Overflow,
str::stream() << "cannot convert document to BSON because it exceeds the limit of "
<< BSONDepth::getMaxAllowableDepth()
<< " levels of nesting",
recursionLevel <= BSONDepth::getMaxAllowableDepth());
for (DocumentStorageIterator it = storage().iterator(); !it.atEnd(); it.advance()) {
it->val.addToBsonObj(builder, it->nameSD(), recursionLevel);
}
}
BSONObj Document::toBson() const {
BSONObjBuilder bb;
toBson(&bb);
return bb.obj();
}
const StringData Document::metaFieldTextScore("$textScore"_sd);
const StringData Document::metaFieldRandVal("$randVal"_sd);
BSONObj Document::toBsonWithMetaData() const {
BSONObjBuilder bb;
toBson(&bb);
if (hasTextScore())
bb.append(metaFieldTextScore, getTextScore());
if (hasRandMetaField())
bb.append(metaFieldRandVal, getRandMetaField());
return bb.obj();
}
Document Document::fromBsonWithMetaData(const BSONObj& bson) {
MutableDocument md;
BSONObjIterator it(bson);
while (it.more()) {
BSONElement elem(it.next());
auto fieldName = elem.fieldNameStringData();
if (fieldName[0] == '$') {
if (fieldName == metaFieldTextScore) {
md.setTextScore(elem.Double());
continue;
} else if (fieldName == metaFieldRandVal) {
md.setRandMetaField(elem.Double());
continue;
}
}
// Note: this will not parse out metadata in embedded documents.
md.addField(fieldName, Value(elem));
}
return md.freeze();
}
MutableDocument::MutableDocument(size_t expectedFields)
: _storageHolder(NULL), _storage(_storageHolder) {
if (expectedFields) {
storage().reserveFields(expectedFields);
}
}
MutableValue MutableDocument::getNestedFieldHelper(const FieldPath& dottedField, size_t level) {
if (level == dottedField.getPathLength() - 1) {
return getField(dottedField.getFieldName(level));
} else {
MutableDocument nested(getField(dottedField.getFieldName(level)));
return nested.getNestedFieldHelper(dottedField, level + 1);
}
}
MutableValue MutableDocument::getNestedField(const FieldPath& dottedField) {
fassert(16601, dottedField.getPathLength());
return getNestedFieldHelper(dottedField, 0);
}
MutableValue MutableDocument::getNestedFieldHelper(const vector<Position>& positions,
size_t level) {
if (level == positions.size() - 1) {
return getField(positions[level]);
} else {
MutableDocument nested(getField(positions[level]));
return nested.getNestedFieldHelper(positions, level + 1);
}
}
MutableValue MutableDocument::getNestedField(const vector<Position>& positions) {
fassert(16488, !positions.empty());
return getNestedFieldHelper(positions, 0);
}
static Value getNestedFieldHelper(const Document& doc,
const FieldPath& fieldNames,
vector<Position>* positions,
size_t level) {
const auto fieldName = fieldNames.getFieldName(level);
const Position pos = doc.positionOf(fieldName);
if (!pos.found())
return Value();
if (positions)
positions->push_back(pos);
if (level == fieldNames.getPathLength() - 1)
return doc.getField(pos);
Value val = doc.getField(pos);
if (val.getType() != Object)
return Value();
return getNestedFieldHelper(val.getDocument(), fieldNames, positions, level + 1);
}
const Value Document::getNestedField(const FieldPath& path, vector<Position>* positions) const {
fassert(16489, path.getPathLength());
return getNestedFieldHelper(*this, path, positions, 0);
}
size_t Document::getApproximateSize() const {
if (!_storage)
return 0; // we've allocated no memory
size_t size = sizeof(DocumentStorage);
size += storage().allocatedBytes();
for (DocumentStorageIterator it = storage().iterator(); !it.atEnd(); it.advance()) {
size += it->val.getApproximateSize();
size -= sizeof(Value); // already accounted for above
}
return size;
}
void Document::hash_combine(size_t& seed,
const StringData::ComparatorInterface* stringComparator) const {
for (DocumentStorageIterator it = storage().iterator(); !it.atEnd(); it.advance()) {
StringData name = it->nameSD();
boost::hash_range(seed, name.rawData(), name.rawData() + name.size());
it->val.hash_combine(seed, stringComparator);
}
}
int Document::compare(const Document& rL,
const Document& rR,
const StringData::ComparatorInterface* stringComparator) {
DocumentStorageIterator lIt = rL.storage().iterator();
DocumentStorageIterator rIt = rR.storage().iterator();
while (true) {
if (lIt.atEnd()) {
if (rIt.atEnd())
return 0; // documents are the same length
return -1; // left document is shorter
}
if (rIt.atEnd())
return 1; // right document is shorter
const ValueElement& rField = rIt.get();
const ValueElement& lField = lIt.get();
// For compatibility with BSONObj::woCompare() consider the canonical type of values
// before considerting their names.
if (lField.val.getType() != rField.val.getType()) {
const int rCType = canonicalizeBSONType(rField.val.getType());
const int lCType = canonicalizeBSONType(lField.val.getType());
if (lCType != rCType)
return lCType < rCType ? -1 : 1;
}
const int nameCmp = lField.nameSD().compare(rField.nameSD());
if (nameCmp)
return nameCmp; // field names are unequal
const int valueCmp = Value::compare(lField.val, rField.val, stringComparator);
if (valueCmp)
return valueCmp; // fields are unequal
rIt.advance();
lIt.advance();
}
}
string Document::toString() const {
if (empty())
return "{}";
StringBuilder out;
const char* prefix = "{";
for (DocumentStorageIterator it = storage().iterator(); !it.atEnd(); it.advance()) {
out << prefix << it->nameSD() << ": " << it->val.toString();
prefix = ", ";
}
out << '}';
return out.str();
}
void Document::serializeForSorter(BufBuilder& buf) const {
const int numElems = size();
buf.appendNum(numElems);
for (DocumentStorageIterator it = storage().iterator(); !it.atEnd(); it.advance()) {
buf.appendStr(it->nameSD(), /*NUL byte*/ true);
it->val.serializeForSorter(buf);
}
if (hasTextScore()) {
buf.appendNum(char(DocumentStorage::MetaType::TEXT_SCORE + 1));
buf.appendNum(getTextScore());
}
if (hasRandMetaField()) {
buf.appendNum(char(DocumentStorage::MetaType::RAND_VAL + 1));
buf.appendNum(getRandMetaField());
}
buf.appendNum(char(0));
}
Document Document::deserializeForSorter(BufReader& buf, const SorterDeserializeSettings&) {
const int numElems = buf.read<LittleEndian<int>>();
MutableDocument doc(numElems);
for (int i = 0; i < numElems; i++) {
StringData name = buf.readCStr();
doc.addField(name, Value::deserializeForSorter(buf, Value::SorterDeserializeSettings()));
}
while (char marker = buf.read<char>()) {
if (marker == char(DocumentStorage::MetaType::TEXT_SCORE) + 1) {
doc.setTextScore(buf.read<LittleEndian<double>>());
} else if (marker == char(DocumentStorage::MetaType::RAND_VAL) + 1) {
doc.setRandMetaField(buf.read<LittleEndian<double>>());
} else {
uasserted(28744, "Unrecognized marker, unable to deserialize buffer");
}
}
return doc.freeze();
}
}
|