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
|
/**
* 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_source.h"
#include "mongo/db/jsobj.h"
#include "mongo/db/pipeline/document.h"
#include "mongo/db/pipeline/expression.h"
#include "mongo/db/pipeline/expression_context.h"
#include "mongo/db/pipeline/value.h"
namespace mongo {
using boost::intrusive_ptr;
using std::unique_ptr;
using std::make_pair;
using std::string;
using std::vector;
DocumentSourceSort::DocumentSourceSort(const intrusive_ptr<ExpressionContext>& pExpCtx)
: DocumentSource(pExpCtx), populated(false), _mergingPresorted(false) {}
REGISTER_DOCUMENT_SOURCE(sort, DocumentSourceSort::createFromBson);
const char* DocumentSourceSort::getSourceName() const {
return "$sort";
}
boost::optional<Document> DocumentSourceSort::getNext() {
pExpCtx->checkForInterrupt();
if (!populated)
populate();
if (!_output || !_output->more()) {
// Need to be sure connections are marked as done so they can be returned to the connection
// pool. This only needs to happen in the _mergingPresorted case, but it doesn't hurt to
// always do it.
dispose();
return boost::none;
}
return _output->next().second;
}
void DocumentSourceSort::serializeToArray(vector<Value>& array, bool explain) const {
if (explain) { // always one Value for combined $sort + $limit
array.push_back(
Value(DOC(getSourceName()
<< DOC("sortKey" << serializeSortKey(explain) << "mergePresorted"
<< (_mergingPresorted ? Value(true) : Value()) << "limit"
<< (limitSrc ? Value(limitSrc->getLimit()) : Value())))));
} else { // one Value for $sort and maybe a Value for $limit
MutableDocument inner(serializeSortKey(explain));
if (_mergingPresorted)
inner["$mergePresorted"] = Value(true);
array.push_back(Value(DOC(getSourceName() << inner.freeze())));
if (limitSrc) {
limitSrc->serializeToArray(array);
}
}
}
void DocumentSourceSort::dispose() {
_output.reset();
if (pSource) {
pSource->dispose();
}
}
long long DocumentSourceSort::getLimit() const {
return limitSrc ? limitSrc->getLimit() : -1;
}
bool DocumentSourceSort::coalesce(const intrusive_ptr<DocumentSource>& pNextSource) {
if (!limitSrc) {
limitSrc = dynamic_cast<DocumentSourceLimit*>(pNextSource.get());
return limitSrc.get(); // false if next is not a $limit
} else {
return limitSrc->coalesce(pNextSource);
}
}
void DocumentSourceSort::addKey(const string& fieldPath, bool ascending) {
VariablesIdGenerator idGenerator;
VariablesParseState vps(&idGenerator);
vSortKey.push_back(ExpressionFieldPath::parse("$$ROOT." + fieldPath, vps));
vAscending.push_back(ascending);
}
Document DocumentSourceSort::serializeSortKey(bool explain) const {
MutableDocument keyObj;
// add the key fields
const size_t n = vSortKey.size();
for (size_t i = 0; i < n; ++i) {
if (ExpressionFieldPath* efp = dynamic_cast<ExpressionFieldPath*>(vSortKey[i].get())) {
// ExpressionFieldPath gets special syntax that includes direction
const FieldPath& withVariable = efp->getFieldPath();
verify(withVariable.getPathLength() > 1);
verify(withVariable.getFieldName(0) == "ROOT");
const string fieldPath = withVariable.tail().getPath(false);
// append a named integer based on the sort order
keyObj.setField(fieldPath, Value(vAscending[i] ? 1 : -1));
} else {
// other expressions use a made-up field name
keyObj[string(str::stream() << "$computed" << i)] = vSortKey[i]->serialize(explain);
}
}
return keyObj.freeze();
}
DocumentSource::GetDepsReturn DocumentSourceSort::getDependencies(DepsTracker* deps) const {
for (size_t i = 0; i < vSortKey.size(); ++i) {
vSortKey[i]->addDependencies(deps);
}
return SEE_NEXT;
}
intrusive_ptr<DocumentSource> DocumentSourceSort::createFromBson(
BSONElement elem, const intrusive_ptr<ExpressionContext>& pExpCtx) {
uassert(15973, "the $sort key specification must be an object", elem.type() == Object);
return create(pExpCtx, elem.embeddedObject());
}
intrusive_ptr<DocumentSourceSort> DocumentSourceSort::create(
const intrusive_ptr<ExpressionContext>& pExpCtx, BSONObj sortOrder, long long limit) {
intrusive_ptr<DocumentSourceSort> pSort = new DocumentSourceSort(pExpCtx);
/* check for then iterate over the sort object */
BSONForEach(keyField, sortOrder) {
const char* fieldName = keyField.fieldName();
if (str::equals(fieldName, "$mergePresorted")) {
verify(keyField.Bool());
pSort->_mergingPresorted = true;
continue;
}
if (keyField.type() == Object) {
BSONObj metaDoc = keyField.Obj();
// this restriction is due to needing to figure out sort direction
uassert(17312,
"$meta is the only expression supported by $sort right now",
metaDoc.firstElement().fieldNameStringData() == "$meta");
VariablesIdGenerator idGen;
VariablesParseState vps(&idGen);
pSort->vSortKey.push_back(ExpressionMeta::parse(metaDoc.firstElement(), vps));
// If sorting by textScore, sort highest scores first. If sorting by randVal, order
// doesn't matter, so just always use descending.
pSort->vAscending.push_back(false);
continue;
}
uassert(15974,
"$sort key ordering must be specified using a number or {$meta: 'textScore'}",
keyField.isNumber());
int sortOrder = keyField.numberInt();
uassert(15975,
"$sort key ordering must be 1 (for ascending) or -1 (for descending)",
((sortOrder == 1) || (sortOrder == -1)));
pSort->addKey(fieldName, (sortOrder > 0));
}
uassert(15976, "$sort stage must have at least one sort key", !pSort->vSortKey.empty());
if (limit > 0) {
bool coalesced = pSort->coalesce(DocumentSourceLimit::create(pExpCtx, limit));
verify(coalesced); // should always coalesce
verify(pSort->getLimit() == limit);
}
return pSort;
}
SortOptions DocumentSourceSort::makeSortOptions() const {
/* make sure we've got a sort key */
verify(vSortKey.size());
SortOptions opts;
if (limitSrc)
opts.limit = limitSrc->getLimit();
opts.maxMemoryUsageBytes = 100 * 1024 * 1024;
if (pExpCtx->extSortAllowed && !pExpCtx->inRouter) {
opts.extSortAllowed = true;
opts.tempDir = pExpCtx->tempDir;
}
return opts;
}
void DocumentSourceSort::populate() {
if (_mergingPresorted) {
typedef DocumentSourceMergeCursors DSCursors;
if (DSCursors* castedSource = dynamic_cast<DSCursors*>(pSource)) {
populateFromCursors(castedSource->getCursors());
} else {
msgasserted(17196, "can only mergePresorted from MergeCursors");
}
} else {
while (boost::optional<Document> next = pSource->getNext()) {
loadDocument(std::move(*next));
}
loadingDone();
}
}
void DocumentSourceSort::loadDocument(const Document& doc) {
invariant(!populated);
if (!_sorter) {
_sorter.reset(MySorter::make(makeSortOptions(), Comparator(*this)));
}
_sorter->add(extractKey(doc), doc);
}
void DocumentSourceSort::loadingDone() {
if (!_sorter) {
_sorter.reset(MySorter::make(makeSortOptions(), Comparator(*this)));
}
_output.reset(_sorter->done());
_sorter.reset();
populated = true;
}
class DocumentSourceSort::IteratorFromCursor : public MySorter::Iterator {
public:
IteratorFromCursor(DocumentSourceSort* sorter, DBClientCursor* cursor)
: _sorter(sorter), _cursor(cursor) {}
bool more() {
return _cursor->more();
}
Data next() {
const Document doc = DocumentSourceMergeCursors::nextSafeFrom(_cursor);
return make_pair(_sorter->extractKey(doc), doc);
}
private:
DocumentSourceSort* _sorter;
DBClientCursor* _cursor;
};
void DocumentSourceSort::populateFromCursors(const vector<DBClientCursor*>& cursors) {
vector<std::shared_ptr<MySorter::Iterator>> iterators;
for (size_t i = 0; i < cursors.size(); i++) {
iterators.push_back(std::make_shared<IteratorFromCursor>(this, cursors[i]));
}
_output.reset(MySorter::Iterator::merge(iterators, makeSortOptions(), Comparator(*this)));
populated = true;
}
Value DocumentSourceSort::extractKey(const Document& d) const {
Variables vars(0, d);
if (vSortKey.size() == 1) {
return vSortKey[0]->evaluate(&vars);
}
vector<Value> keys;
keys.reserve(vSortKey.size());
for (size_t i = 0; i < vSortKey.size(); i++) {
keys.push_back(vSortKey[i]->evaluate(&vars));
}
return Value(std::move(keys));
}
int DocumentSourceSort::compare(const Value& lhs, const Value& rhs) const {
/*
populate() already checked that there is a non-empty sort key,
so we shouldn't have to worry about that here.
However, the tricky part is what to do is none of the sort keys are
present. In this case, consider the document less.
*/
const size_t n = vSortKey.size();
if (n == 1) { // simple fast case
if (vAscending[0])
return Value::compare(lhs, rhs);
else
return -Value::compare(lhs, rhs);
}
// compound sort
for (size_t i = 0; i < n; i++) {
int cmp = Value::compare(lhs[i], rhs[i]);
if (cmp) {
/* if necessary, adjust the return value by the key ordering */
if (!vAscending[i])
cmp = -cmp;
return cmp;
}
}
/*
If we got here, everything matched (or didn't exist), so we'll
consider the documents equal for purposes of this sort.
*/
return 0;
}
intrusive_ptr<DocumentSource> DocumentSourceSort::getShardSource() {
verify(!_mergingPresorted);
return this;
}
intrusive_ptr<DocumentSource> DocumentSourceSort::getMergeSource() {
verify(!_mergingPresorted);
intrusive_ptr<DocumentSourceSort> other = new DocumentSourceSort(pExpCtx);
other->vAscending = vAscending;
other->vSortKey = vSortKey;
other->limitSrc = limitSrc;
other->_mergingPresorted = true;
return other;
}
}
#include "mongo/db/sorter/sorter.cpp"
// Explicit instantiation unneeded since we aren't exposing Sorter outside of this file.
|