|
1
|
|
|
<?php |
|
2
|
|
|
/** |
|
3
|
|
|
* Efficiently run operations on batches of results for any function |
|
4
|
|
|
* that supports an options array. |
|
5
|
|
|
* |
|
6
|
|
|
* This is usually used with elgg_get_entities() and friends, |
|
7
|
|
|
* elgg_get_annotations(), and elgg_get_metadata(). |
|
8
|
|
|
* |
|
9
|
|
|
* If you pass a valid PHP callback, all results will be run through that |
|
10
|
|
|
* callback. You can still foreach() through the result set after. Valid |
|
11
|
|
|
* PHP callbacks can be a string, an array, or a closure. |
|
12
|
|
|
* {@link http://php.net/manual/en/language.pseudo-types.php} |
|
13
|
|
|
* |
|
14
|
|
|
* The callback function must accept 3 arguments: an entity, the getter |
|
15
|
|
|
* used, and the options used. |
|
16
|
|
|
* |
|
17
|
|
|
* Results from the callback are stored in callbackResult. If the callback |
|
18
|
|
|
* returns only booleans, callbackResults will be the combined result of |
|
19
|
|
|
* all calls. If no entities are processed, callbackResults will be null. |
|
20
|
|
|
* |
|
21
|
|
|
* If the callback returns anything else, callbackresult will be an indexed |
|
22
|
|
|
* array of whatever the callback returns. If returning error handling |
|
23
|
|
|
* information, you should include enough information to determine which |
|
24
|
|
|
* result you're referring to. |
|
25
|
|
|
* |
|
26
|
|
|
* Don't combine returning bools and returning something else. |
|
27
|
|
|
* |
|
28
|
|
|
* Note that returning false will not stop the foreach. |
|
29
|
|
|
* |
|
30
|
|
|
* @warning If your callback or foreach loop deletes or disable entities |
|
31
|
|
|
* you MUST call setIncrementOffset(false) or set that when instantiating. |
|
32
|
|
|
* This forces the offset to stay what it was in the $options array. |
|
33
|
|
|
* |
|
34
|
|
|
* @example |
|
35
|
|
|
* <code> |
|
36
|
|
|
* // using foreach |
|
37
|
|
|
* $batch = new \ElggBatch('elgg_get_entities', array()); |
|
38
|
|
|
* $batch->setIncrementOffset(false); |
|
39
|
|
|
* |
|
40
|
|
|
* foreach ($batch as $entity) { |
|
41
|
|
|
* $entity->disable(); |
|
42
|
|
|
* } |
|
43
|
|
|
* |
|
44
|
|
|
* // using both a callback |
|
45
|
|
|
* $callback = function($result, $getter, $options) { |
|
46
|
|
|
* var_dump("Looking at annotation id: $result->id"); |
|
47
|
|
|
* return true; |
|
48
|
|
|
* } |
|
49
|
|
|
* |
|
50
|
|
|
* $batch = new \ElggBatch('elgg_get_annotations', array('guid' => 2), $callback); |
|
51
|
|
|
* </code> |
|
52
|
|
|
* |
|
53
|
|
|
* @package Elgg.Core |
|
54
|
|
|
* @subpackage DataModel |
|
55
|
|
|
* @since 1.8 |
|
56
|
|
|
*/ |
|
57
|
|
|
class ElggBatch |
|
58
|
|
|
implements \Iterator { |
|
59
|
|
|
|
|
60
|
|
|
/** |
|
61
|
|
|
* The objects to interator over. |
|
62
|
|
|
* |
|
63
|
|
|
* @var array |
|
64
|
|
|
*/ |
|
65
|
|
|
private $results = array(); |
|
66
|
|
|
|
|
67
|
|
|
/** |
|
68
|
|
|
* The function used to get results. |
|
69
|
|
|
* |
|
70
|
|
|
* @var mixed A string, array, or closure, or lamda function |
|
71
|
|
|
*/ |
|
72
|
|
|
private $getter = null; |
|
73
|
|
|
|
|
74
|
|
|
/** |
|
75
|
|
|
* The number of results to grab at a time. |
|
76
|
|
|
* |
|
77
|
|
|
* @var int |
|
78
|
|
|
*/ |
|
79
|
|
|
private $chunkSize = 25; |
|
80
|
|
|
|
|
81
|
|
|
/** |
|
82
|
|
|
* A callback function to pass results through. |
|
83
|
|
|
* |
|
84
|
|
|
* @var mixed A string, array, or closure, or lamda function |
|
85
|
|
|
*/ |
|
86
|
|
|
private $callback = null; |
|
87
|
|
|
|
|
88
|
|
|
/** |
|
89
|
|
|
* Start after this many results. |
|
90
|
|
|
* |
|
91
|
|
|
* @var int |
|
92
|
|
|
*/ |
|
93
|
|
|
private $offset = 0; |
|
94
|
|
|
|
|
95
|
|
|
/** |
|
96
|
|
|
* Stop after this many results. |
|
97
|
|
|
* |
|
98
|
|
|
* @var int |
|
99
|
|
|
*/ |
|
100
|
|
|
private $limit = 0; |
|
101
|
|
|
|
|
102
|
|
|
/** |
|
103
|
|
|
* Number of processed results. |
|
104
|
|
|
* |
|
105
|
|
|
* @var int |
|
106
|
|
|
*/ |
|
107
|
|
|
private $retrievedResults = 0; |
|
108
|
|
|
|
|
109
|
|
|
/** |
|
110
|
|
|
* The index of the current result within the current chunk |
|
111
|
|
|
* |
|
112
|
|
|
* @var int |
|
113
|
|
|
*/ |
|
114
|
|
|
private $resultIndex = 0; |
|
115
|
|
|
|
|
116
|
|
|
/** |
|
117
|
|
|
* The index of the current chunk |
|
118
|
|
|
* |
|
119
|
|
|
* @var int |
|
120
|
|
|
*/ |
|
121
|
|
|
private $chunkIndex = 0; |
|
122
|
|
|
|
|
123
|
|
|
/** |
|
124
|
|
|
* The number of results iterated through |
|
125
|
|
|
* |
|
126
|
|
|
* @var int |
|
127
|
|
|
*/ |
|
128
|
|
|
private $processedResults = 0; |
|
129
|
|
|
|
|
130
|
|
|
/** |
|
131
|
|
|
* Is the getter a valid callback |
|
132
|
|
|
* |
|
133
|
|
|
* @var bool |
|
134
|
|
|
*/ |
|
135
|
|
|
private $validGetter = null; |
|
136
|
|
|
|
|
137
|
|
|
/** |
|
138
|
|
|
* The result of running all entities through the callback function. |
|
139
|
|
|
* |
|
140
|
|
|
* @var mixed |
|
141
|
|
|
*/ |
|
142
|
|
|
public $callbackResult = null; |
|
143
|
|
|
|
|
144
|
|
|
/** |
|
145
|
|
|
* If false, offset will not be incremented. This is used for callbacks/loops that delete. |
|
146
|
|
|
* |
|
147
|
|
|
* @var bool |
|
148
|
|
|
*/ |
|
149
|
|
|
private $incrementOffset = true; |
|
150
|
|
|
|
|
151
|
|
|
/** |
|
152
|
|
|
* Entities that could not be instantiated during a fetch |
|
153
|
|
|
* |
|
154
|
|
|
* @var \stdClass[] |
|
155
|
|
|
*/ |
|
156
|
|
|
private $incompleteEntities = array(); |
|
157
|
|
|
|
|
158
|
|
|
/** |
|
159
|
|
|
* Total number of incomplete entities fetched |
|
160
|
|
|
* |
|
161
|
|
|
* @var int |
|
162
|
|
|
*/ |
|
163
|
|
|
private $totalIncompletes = 0; |
|
164
|
|
|
|
|
165
|
|
|
/** |
|
166
|
|
|
* Batches operations on any elgg_get_*() or compatible function that supports |
|
167
|
|
|
* an options array. |
|
168
|
|
|
* |
|
169
|
|
|
* Instead of returning all objects in memory, it goes through $chunk_size |
|
170
|
|
|
* objects, then requests more from the server. This avoids OOM errors. |
|
171
|
|
|
* |
|
172
|
|
|
* @param string $getter The function used to get objects. Usually |
|
173
|
|
|
* an elgg_get_*() function, but can be any valid PHP callback. |
|
174
|
|
|
* @param array $options The options array to pass to the getter function. If limit is |
|
175
|
|
|
* not set, 10 is used as the default. In most cases that is not |
|
176
|
|
|
* what you want. |
|
177
|
|
|
* @param mixed $callback An optional callback function that all results will be passed |
|
178
|
|
|
* to upon load. The callback needs to accept $result, $getter, |
|
179
|
|
|
* $options. |
|
180
|
|
|
* @param int $chunk_size The number of entities to pull in before requesting more. |
|
181
|
|
|
* You have to balance this between running out of memory in PHP |
|
182
|
|
|
* and hitting the db server too often. |
|
183
|
|
|
* @param bool $inc_offset Increment the offset on each fetch. This must be false for |
|
184
|
|
|
* callbacks that delete rows. You can set this after the |
|
185
|
|
|
* object is created with {@link \ElggBatch::setIncrementOffset()}. |
|
186
|
|
|
*/ |
|
187
|
|
|
public function __construct($getter, $options, $callback = null, $chunk_size = 25, |
|
188
|
|
|
$inc_offset = true) { |
|
189
|
|
|
|
|
190
|
|
|
$this->getter = $getter; |
|
191
|
|
|
$this->options = $options; |
|
|
|
|
|
|
192
|
|
|
$this->callback = $callback; |
|
193
|
|
|
$this->chunkSize = $chunk_size; |
|
194
|
|
|
$this->setIncrementOffset($inc_offset); |
|
195
|
|
|
|
|
196
|
|
|
if ($this->chunkSize <= 0) { |
|
197
|
|
|
$this->chunkSize = 25; |
|
198
|
|
|
} |
|
199
|
|
|
|
|
200
|
|
|
// store these so we can compare later |
|
201
|
|
|
$this->offset = elgg_extract('offset', $options, 0); |
|
202
|
|
|
$this->limit = elgg_extract('limit', $options, elgg_get_config('default_limit')); |
|
203
|
|
|
|
|
204
|
|
|
// if passed a callback, create a new \ElggBatch with the same options |
|
205
|
|
|
// and pass each to the callback. |
|
206
|
|
|
if ($callback && is_callable($callback)) { |
|
207
|
|
|
$batch = new \ElggBatch($getter, $options, null, $chunk_size, $inc_offset); |
|
208
|
|
|
|
|
209
|
|
|
$all_results = null; |
|
210
|
|
|
|
|
211
|
|
|
foreach ($batch as $result) { |
|
212
|
|
|
$result = call_user_func($callback, $result, $getter, $options); |
|
213
|
|
|
|
|
214
|
|
|
if (!isset($all_results)) { |
|
215
|
|
|
if ($result === true || $result === false || $result === null) { |
|
216
|
|
|
$all_results = $result; |
|
217
|
|
|
} else { |
|
218
|
|
|
$all_results = array(); |
|
219
|
|
|
} |
|
220
|
|
|
} |
|
221
|
|
|
|
|
222
|
|
|
if (($result === true || $result === false || $result === null) && !is_array($all_results)) { |
|
223
|
|
|
$all_results = $result && $all_results; |
|
224
|
|
|
} else { |
|
225
|
|
|
$all_results[] = $result; |
|
226
|
|
|
} |
|
227
|
|
|
} |
|
228
|
|
|
|
|
229
|
|
|
$this->callbackResult = $all_results; |
|
230
|
|
|
} |
|
231
|
|
|
} |
|
232
|
|
|
|
|
233
|
|
|
/** |
|
234
|
|
|
* Tell the process that an entity was incomplete during a fetch |
|
235
|
|
|
* |
|
236
|
|
|
* @param \stdClass $row |
|
237
|
|
|
* |
|
238
|
|
|
* @access private |
|
239
|
|
|
*/ |
|
240
|
|
|
public function reportIncompleteEntity(\stdClass $row) { |
|
241
|
|
|
$this->incompleteEntities[] = $row; |
|
242
|
|
|
} |
|
243
|
|
|
|
|
244
|
|
|
/** |
|
245
|
|
|
* Fetches the next chunk of results |
|
246
|
|
|
* |
|
247
|
|
|
* @return bool |
|
248
|
|
|
*/ |
|
249
|
|
|
private function getNextResultsChunk() { |
|
250
|
|
|
|
|
251
|
|
|
// always reset results. |
|
252
|
|
|
$this->results = array(); |
|
253
|
|
|
|
|
254
|
|
|
if (!isset($this->validGetter)) { |
|
255
|
|
|
$this->validGetter = is_callable($this->getter); |
|
256
|
|
|
} |
|
257
|
|
|
|
|
258
|
|
|
if (!$this->validGetter) { |
|
259
|
|
|
return false; |
|
260
|
|
|
} |
|
261
|
|
|
|
|
262
|
|
|
$limit = $this->chunkSize; |
|
263
|
|
|
|
|
264
|
|
|
// if someone passed limit = 0 they want everything. |
|
265
|
|
|
if ($this->limit != 0) { |
|
266
|
|
|
if ($this->retrievedResults >= $this->limit) { |
|
267
|
|
|
return false; |
|
268
|
|
|
} |
|
269
|
|
|
|
|
270
|
|
|
// if original limit < chunk size, set limit to original limit |
|
271
|
|
|
// else if the number of results we'll fetch if greater than the original limit |
|
272
|
|
|
if ($this->limit < $this->chunkSize) { |
|
273
|
|
|
$limit = $this->limit; |
|
274
|
|
|
} elseif ($this->retrievedResults + $this->chunkSize > $this->limit) { |
|
275
|
|
|
// set the limit to the number of results remaining in the original limit |
|
276
|
|
|
$limit = $this->limit - $this->retrievedResults; |
|
277
|
|
|
} |
|
278
|
|
|
} |
|
279
|
|
|
|
|
280
|
|
|
if ($this->incrementOffset) { |
|
281
|
|
|
$offset = $this->offset + $this->retrievedResults; |
|
282
|
|
|
} else { |
|
283
|
|
|
$offset = $this->offset + $this->totalIncompletes; |
|
284
|
|
|
} |
|
285
|
|
|
|
|
286
|
|
|
$current_options = array( |
|
287
|
|
|
'limit' => $limit, |
|
288
|
|
|
'offset' => $offset, |
|
289
|
|
|
'__ElggBatch' => $this, |
|
290
|
|
|
); |
|
291
|
|
|
|
|
292
|
|
|
$options = array_merge($this->options, $current_options); |
|
293
|
|
|
|
|
294
|
|
|
$this->incompleteEntities = array(); |
|
295
|
|
|
$this->results = call_user_func($this->getter, $options); |
|
|
|
|
|
|
296
|
|
|
|
|
297
|
|
|
// batch result sets tend to be large; we don't want to cache these. |
|
298
|
|
|
_elgg_services()->db->disableQueryCache(); |
|
299
|
|
|
|
|
300
|
|
|
$num_results = count($this->results); |
|
301
|
|
|
$num_incomplete = count($this->incompleteEntities); |
|
302
|
|
|
|
|
303
|
|
|
$this->totalIncompletes += $num_incomplete; |
|
304
|
|
|
|
|
305
|
|
|
if ($this->incompleteEntities) { |
|
|
|
|
|
|
306
|
|
|
// pad the front of the results with nulls representing the incompletes |
|
307
|
|
|
array_splice($this->results, 0, 0, array_pad(array(), $num_incomplete, null)); |
|
308
|
|
|
// ...and skip past them |
|
309
|
|
|
reset($this->results); |
|
310
|
|
|
for ($i = 0; $i < $num_incomplete; $i++) { |
|
311
|
|
|
next($this->results); |
|
312
|
|
|
} |
|
313
|
|
|
} |
|
314
|
|
|
|
|
315
|
|
|
if ($this->results) { |
|
316
|
|
|
$this->chunkIndex++; |
|
317
|
|
|
|
|
318
|
|
|
// let the system know we've jumped past the nulls |
|
319
|
|
|
$this->resultIndex = $num_incomplete; |
|
320
|
|
|
|
|
321
|
|
|
$this->retrievedResults += ($num_results + $num_incomplete); |
|
322
|
|
|
if ($num_results == 0) { |
|
323
|
|
|
// This fetch was *all* incompletes! We need to fetch until we can either |
|
324
|
|
|
// offer at least one row to iterate over, or give up. |
|
325
|
|
|
return $this->getNextResultsChunk(); |
|
326
|
|
|
} |
|
327
|
|
|
_elgg_services()->db->enableQueryCache(); |
|
328
|
|
|
return true; |
|
329
|
|
|
} else { |
|
330
|
|
|
_elgg_services()->db->enableQueryCache(); |
|
331
|
|
|
return false; |
|
332
|
|
|
} |
|
333
|
|
|
} |
|
334
|
|
|
|
|
335
|
|
|
/** |
|
336
|
|
|
* Increment the offset from the original options array? Setting to |
|
337
|
|
|
* false is required for callbacks that delete rows. |
|
338
|
|
|
* |
|
339
|
|
|
* @param bool $increment Set to false when deleting data |
|
340
|
|
|
* @return void |
|
341
|
|
|
*/ |
|
342
|
|
|
public function setIncrementOffset($increment = true) { |
|
343
|
|
|
$this->incrementOffset = (bool) $increment; |
|
344
|
|
|
} |
|
345
|
|
|
|
|
346
|
|
|
/** |
|
347
|
|
|
* Implements Iterator |
|
348
|
|
|
*/ |
|
349
|
|
|
|
|
350
|
|
|
/** |
|
351
|
|
|
* PHP Iterator Interface |
|
352
|
|
|
* |
|
353
|
|
|
* @see Iterator::rewind() |
|
354
|
|
|
* @return void |
|
355
|
|
|
*/ |
|
356
|
|
|
public function rewind() { |
|
357
|
|
|
$this->resultIndex = 0; |
|
358
|
|
|
$this->retrievedResults = 0; |
|
359
|
|
|
$this->processedResults = 0; |
|
360
|
|
|
|
|
361
|
|
|
// only grab results if we haven't yet or we're crossing chunks |
|
362
|
|
|
if ($this->chunkIndex == 0 || $this->limit > $this->chunkSize) { |
|
363
|
|
|
$this->chunkIndex = 0; |
|
364
|
|
|
$this->getNextResultsChunk(); |
|
365
|
|
|
} |
|
366
|
|
|
} |
|
367
|
|
|
|
|
368
|
|
|
/** |
|
369
|
|
|
* PHP Iterator Interface |
|
370
|
|
|
* |
|
371
|
|
|
* @see Iterator::current() |
|
372
|
|
|
* @return mixed |
|
373
|
|
|
*/ |
|
374
|
|
|
public function current() { |
|
375
|
|
|
return current($this->results); |
|
376
|
|
|
} |
|
377
|
|
|
|
|
378
|
|
|
/** |
|
379
|
|
|
* PHP Iterator Interface |
|
380
|
|
|
* |
|
381
|
|
|
* @see Iterator::key() |
|
382
|
|
|
* @return int |
|
383
|
|
|
*/ |
|
384
|
|
|
public function key() { |
|
385
|
|
|
return $this->processedResults; |
|
386
|
|
|
} |
|
387
|
|
|
|
|
388
|
|
|
/** |
|
389
|
|
|
* PHP Iterator Interface |
|
390
|
|
|
* |
|
391
|
|
|
* @see Iterator::next() |
|
392
|
|
|
* @return mixed |
|
393
|
|
|
*/ |
|
394
|
|
|
public function next() { |
|
395
|
|
|
// if we'll be at the end. |
|
396
|
|
|
if (($this->processedResults + 1) >= $this->limit && $this->limit > 0) { |
|
397
|
|
|
$this->results = array(); |
|
398
|
|
|
return false; |
|
399
|
|
|
} |
|
400
|
|
|
|
|
401
|
|
|
// if we'll need new results. |
|
402
|
|
|
if (($this->resultIndex + 1) >= $this->chunkSize) { |
|
403
|
|
|
if (!$this->getNextResultsChunk()) { |
|
404
|
|
|
$this->results = array(); |
|
405
|
|
|
return false; |
|
406
|
|
|
} |
|
407
|
|
|
|
|
408
|
|
|
$result = current($this->results); |
|
409
|
|
|
} else { |
|
410
|
|
|
// the function above resets the indexes, so only inc if not |
|
411
|
|
|
// getting new set |
|
412
|
|
|
$this->resultIndex++; |
|
413
|
|
|
$result = next($this->results); |
|
414
|
|
|
} |
|
415
|
|
|
|
|
416
|
|
|
$this->processedResults++; |
|
417
|
|
|
return $result; |
|
418
|
|
|
} |
|
419
|
|
|
|
|
420
|
|
|
/** |
|
421
|
|
|
* PHP Iterator Interface |
|
422
|
|
|
* |
|
423
|
|
|
* @see Iterator::valid() |
|
424
|
|
|
* @return bool |
|
425
|
|
|
*/ |
|
426
|
|
|
public function valid() { |
|
427
|
|
|
if (!is_array($this->results)) { |
|
428
|
|
|
return false; |
|
429
|
|
|
} |
|
430
|
|
|
$key = key($this->results); |
|
431
|
|
|
return ($key !== null && $key !== false); |
|
432
|
|
|
} |
|
433
|
|
|
} |
|
434
|
|
|
|
In PHP it is possible to write to properties without declaring them. For example, the following is perfectly valid PHP code:
Generally, it is a good practice to explictly declare properties to avoid accidental typos and provide IDE auto-completion: