ActiveRecord.php 21 KB
Newer Older
1 2 3
<?php
/**
 * @link http://www.yiiframework.com/
4
 * @copyright Copyright (c) 2008 Yii Software LLC
5 6 7 8 9 10 11
 * @license http://www.yiiframework.com/license/
 */

namespace yii\elasticsearch;

use yii\base\InvalidCallException;
use yii\base\InvalidConfigException;
12
use yii\db\BaseActiveRecord;
13
use yii\helpers\ArrayHelper;
14 15
use yii\helpers\Inflector;
use yii\helpers\Json;
16
use yii\helpers\StringHelper;
17 18 19 20

/**
 * ActiveRecord is the base class for classes representing relational data in terms of objects.
 *
Carsten Brandt committed
21 22 23 24 25 26
 * This class implements the ActiveRecord pattern for the fulltext search and data storage
 * [elasticsearch](http://www.elasticsearch.org/).
 *
 * For defining a record a subclass should at least implement the [[attributes()]] method to define
 * attributes.
 * The primary key (the `_id` field in elasticsearch terms) is represented by `getId()` and `setId()`.
27
 * The primary key is not part of the attributes.
Carsten Brandt committed
28 29 30 31 32 33 34 35 36 37 38 39
 *
 * The following is an example model called `Customer`:
 *
 * ```php
 * class Customer extends \yii\elasticsearch\ActiveRecord
 * {
 *     public function attributes()
 *     {
 *         return ['id', 'name', 'address', 'registration_date'];
 *     }
 * }
 * ```
40
 *
Carsten Brandt committed
41
 * You may override [[index()]] and [[type()]] to define the index and type this record represents.
42
 *
43 44 45
 * @property float $score Returns the score of this record when it was retrieved via a [[find()]] query. This
 * property is read-only.
 *
46 47 48
 * @author Carsten Brandt <mail@cebe.cc>
 * @since 2.0
 */
49
class ActiveRecord extends BaseActiveRecord
50
{
51 52 53 54 55 56 57 58 59 60 61 62
    private $_id;
    private $_score;
    private $_version;

    /**
     * Returns the database connection used by this AR class.
     * By default, the "elasticsearch" application component is used as the database connection.
     * You may override this method if you want to use a different database connection.
     * @return Connection the database connection used by this AR class.
     */
    public static function getDb()
    {
63
        return \Yii::$app->get('elasticsearch');
64 65 66 67 68
    }

    /**
     * @inheritdoc
     */
Alexander Makarov committed
69
    public static function find()
70
    {
Alexander Makarov committed
71 72
        return new ActiveQuery(get_called_class());
    }
73

Alexander Makarov committed
74 75 76
    /**
     * @inheritdoc
     */
77
    public static function findOne($condition)
Alexander Makarov committed
78 79
    {
        $query = static::find();
80 81
        if (is_array($condition)) {
            return $query->andWhere($condition)->one();
82
        } else {
83 84 85 86 87 88 89 90 91 92 93 94 95 96
            return static::get($condition);
        }
    }

    /**
     * @inheritdoc
     */
    public static function findAll($condition)
    {
        $query = static::find();
        if (ArrayHelper::isAssociative($condition)) {
            return $query->andWhere($condition)->all();
        } else {
            return static::mget((array) $condition);
97 98 99 100 101 102
        }
    }

    /**
     * Gets a record by its primary key.
     *
103 104 105 106 107
     * @param mixed $primaryKey the primaryKey value
     * @param array $options options given in this parameter are passed to elasticsearch
     * as request URI parameters.
     * Please refer to the [elasticsearch documentation](http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/docs-get.html)
     * for more details on these options.
108 109 110 111 112 113 114 115 116
     * @return static|null The record instance or null if it was not found.
     */
    public static function get($primaryKey, $options = [])
    {
        if ($primaryKey === null) {
            return null;
        }
        $command = static::getDb()->createCommand();
        $result = $command->get(static::index(), static::type(), $primaryKey, $options);
117
        if ($result['found']) {
118 119 120 121 122 123 124 125 126 127 128 129 130 131
            $model = static::instantiate($result);
            static::populateRecord($model, $result);
            $model->afterFind();

            return $model;
        }

        return null;
    }

    /**
     * Gets a list of records by its primary keys.
     *
     * @param array $primaryKeys an array of primaryKey values
132 133
     * @param array $options options given in this parameter are passed to elasticsearch
     * as request URI parameters.
134 135 136
     *
     * Please refer to the [elasticsearch documentation](http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/docs-get.html)
     * for more details on these options.
137
     * @return array The record instances, or empty array if nothing was found
138
     */
139
    public static function mget(array $primaryKeys, $options = [])
140 141 142 143
    {
        if (empty($primaryKeys)) {
            return [];
        }
144 145 146 147 148
        if (count($primaryKeys) === 1) {
            $model = static::get(reset($primaryKeys));
            return $model === null ? [] : [$model];
        }

149 150 151 152
        $command = static::getDb()->createCommand();
        $result = $command->mget(static::index(), static::type(), $primaryKeys, $options);
        $models = [];
        foreach ($result['docs'] as $doc) {
153
            if ($doc['found']) {
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
                $model = static::instantiate($doc);
                static::populateRecord($model, $doc);
                $model->afterFind();
                $models[] = $model;
            }
        }

        return $models;
    }

    // TODO add more like this feature http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-more-like-this.html

    // TODO add percolate functionality http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-percolate.html

    // TODO implement copy and move as pk change is not possible

    /**
     * @return float returns the score of this record when it was retrieved via a [[find()]] query.
     */
    public function getScore()
    {
        return $this->_score;
    }

    /**
     * Sets the primary key
180
     * @param mixed $value
181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284
     * @throws \yii\base\InvalidCallException when record is not new
     */
    public function setPrimaryKey($value)
    {
        $pk = static::primaryKey()[0];
        if ($this->getIsNewRecord() || $pk != '_id') {
            $this->$pk = $value;
        } else {
            throw new InvalidCallException('Changing the primaryKey of an already saved record is not allowed.');
        }
    }

    /**
     * @inheritdoc
     */
    public function getPrimaryKey($asArray = false)
    {
        $pk = static::primaryKey()[0];
        if ($asArray) {
            return [$pk => $this->$pk];
        } else {
            return $this->$pk;
        }
    }

    /**
     * @inheritdoc
     */
    public function getOldPrimaryKey($asArray = false)
    {
        $pk = static::primaryKey()[0];
        if ($this->getIsNewRecord()) {
            $id = null;
        } elseif ($pk == '_id') {
            $id = $this->_id;
        } else {
            $id = $this->getOldAttribute($pk);
        }
        if ($asArray) {
            return [$pk => $id];
        } else {
            return $id;
        }
    }

    /**
     * This method defines the attribute that uniquely identifies a record.
     *
     * The primaryKey for elasticsearch documents is the `_id` field by default. This field is not part of the
     * ActiveRecord attributes so you should never add `_id` to the list of [[attributes()|attributes]].
     *
     * You may overide this method to define the primary key name when you have defined
     * [path mapping](http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/mapping-id-field.html)
     * for the `_id` field so that it is part of the `_source` and thus part of the [[attributes()|attributes]].
     *
     * Note that elasticsearch only supports _one_ attribute to be the primary key. However to match the signature
     * of the [[\yii\db\ActiveRecordInterface|ActiveRecordInterface]] this methods returns an array instead of a
     * single string.
     *
     * @return string[] array of primary key attributes. Only the first element of the array will be used.
     */
    public static function primaryKey()
    {
        return ['_id'];
    }

    /**
     * Returns the list of all attribute names of the model.
     *
     * This method must be overridden by child classes to define available attributes.
     *
     * Attributes are names of fields of the corresponding elasticsearch document.
     * The primaryKey for elasticsearch documents is the `_id` field by default which is not part of the attributes.
     * You may define [path mapping](http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/mapping-id-field.html)
     * for the `_id` field so that it is part of the `_source` fields and thus becomes part of the attributes.
     *
     * @return string[] list of attribute names.
     */
    public function attributes()
    {
        throw new InvalidConfigException('The attributes() method of elasticsearch ActiveRecord has to be implemented by child classes.');
    }

    /**
     * @return string the name of the index this record is stored in.
     */
    public static function index()
    {
        return Inflector::pluralize(Inflector::camel2id(StringHelper::basename(get_called_class()), '-'));
    }

    /**
     * @return string the name of the type of this record.
     */
    public static function type()
    {
        return Inflector::camel2id(StringHelper::basename(get_called_class()), '-');
    }

    /**
     * @inheritdoc
     */
    public static function populateRecord($record, $row)
    {
285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301
        $attributes = [];
        if (isset($row['_source'])) {
            $attributes = $row['_source'];
        }
        if (isset($row['fields'])) {
            // reset fields in case it is scalar value TODO use field metadata for this
            foreach($row['fields'] as $key => $value) {
                if (count($value) == 1) {
                    $row['fields'][$key] = reset($value);
                }
            }
            $attributes = array_merge($attributes, $row['fields']);
        }

        parent::populateRecord($record, $attributes);

        $pk = static::primaryKey()[0];//TODO should always set ID in case of fields are not returned
302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
        if ($pk === '_id') {
            $record->_id = $row['_id'];
        }
        $record->_score = isset($row['_score']) ? $row['_score'] : null;
        $record->_version = isset($row['_version']) ? $row['_version'] : null; // TODO version should always be available...
    }

    /**
     * Creates an active record instance.
     *
     * This method is called together with [[populateRecord()]] by [[ActiveQuery]].
     * It is not meant to be used for creating new records directly.
     *
     * You may override this method if the instance being created
     * depends on the row data to be populated into the record.
     * For example, by creating a record based on the value of a column,
     * you may implement the so-called single-table inheritance mapping.
319 320 321 322 323
     * @param array $row row data to be populated into the record.
     * This array consists of the following keys:
     *  - `_source`: refers to the attributes of the record.
     *  - `_type`: the type this record is stored in.
     *  - `_index`: the index this record is stored in.
324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364
     * @return static the newly created active record
     */
    public static function instantiate($row)
    {
        return new static;
    }

    /**
     * Inserts a document into the associated index using the attribute values of this record.
     *
     * This method performs the following steps in order:
     *
     * 1. call [[beforeValidate()]] when `$runValidation` is true. If validation
     *    fails, it will skip the rest of the steps;
     * 2. call [[afterValidate()]] when `$runValidation` is true.
     * 3. call [[beforeSave()]]. If the method returns false, it will skip the
     *    rest of the steps;
     * 4. insert the record into database. If this fails, it will skip the rest of the steps;
     * 5. call [[afterSave()]];
     *
     * In the above step 1, 2, 3 and 5, events [[EVENT_BEFORE_VALIDATE]],
     * [[EVENT_BEFORE_INSERT]], [[EVENT_AFTER_INSERT]] and [[EVENT_AFTER_VALIDATE]]
     * will be raised by the corresponding methods.
     *
     * Only the [[dirtyAttributes|changed attribute values]] will be inserted into database.
     *
     * If the [[primaryKey|primary key]] is not set (null) during insertion,
     * it will be populated with a
     * [randomly generated value](http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/docs-index_.html#_automatic_id_generation)
     * after insertion.
     *
     * For example, to insert a customer record:
     *
     * ~~~
     * $customer = new Customer;
     * $customer->name = $name;
     * $customer->email = $email;
     * $customer->insert();
     * ~~~
     *
     * @param boolean $runValidation whether to perform validation before saving the record.
365 366 367 368 369
     * If the validation fails, the record will not be inserted into the database.
     * @param array $attributes list of attributes that need to be saved. Defaults to null,
     * meaning all attributes will be saved.
     * @param array $options options given in this parameter are passed to elasticsearch
     * as request URI parameters. These are among others:
370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396
     *
     * - `routing` define shard placement of this record.
     * - `parent` by giving the primaryKey of another record this defines a parent-child relation
     * - `timestamp` specifies the timestamp to store along with the document. Default is indexing time.
     *
     * Please refer to the [elasticsearch documentation](http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/docs-index_.html)
     * for more details on these options.
     *
     * By default the `op_type` is set to `create`.
     * @return boolean whether the attributes are valid and the record is inserted successfully.
     */
    public function insert($runValidation = true, $attributes = null, $options = ['op_type' => 'create'])
    {
        if ($runValidation && !$this->validate($attributes)) {
            return false;
        }
        if ($this->beforeSave(true)) {
            $values = $this->getDirtyAttributes($attributes);

            $response = static::getDb()->createCommand()->insert(
                static::index(),
                static::type(),
                $values,
                $this->getPrimaryKey(),
                $options
            );

397 398 399
//            if (!isset($response['ok'])) {
//                return false;
//            }
400 401 402 403 404 405 406
            $pk = static::primaryKey()[0];
            $this->$pk = $response['_id'];
            if ($pk != '_id') {
                $values[$pk] = $response['_id'];
            }
            $this->_version = $response['_version'];
            $this->_score = null;
407

408
            $this->afterSave(true);
409
            $this->setOldAttributes($values);
410 411 412 413 414 415 416 417 418 419 420 421 422 423 424

            return true;
        }

        return false;
    }

    /**
     * Updates all records whos primary keys are given.
     * For example, to change the status to be 1 for all customers whose status is 2:
     *
     * ~~~
     * Customer::updateAll(['status' => 1], [2, 3, 4]);
     * ~~~
     *
425 426 427
     * @param array $attributes attribute values (name-value pairs) to be saved into the table
     * @param array $condition the conditions that will be put in the WHERE part of the UPDATE SQL.
     * Please refer to [[ActiveQuery::where()]] on how to specify this parameter.
428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461
     * @return integer the number of rows updated
     */
    public static function updateAll($attributes, $condition = [])
    {
        $pkName = static::primaryKey()[0];
        if (count($condition) == 1 && isset($condition[$pkName])) {
            $primaryKeys = is_array($condition[$pkName]) ? $condition[$pkName] : [$condition[$pkName]];
        } else {
            $primaryKeys = static::find()->where($condition)->column($pkName); // TODO check whether this works with default pk _id
        }
        if (empty($primaryKeys)) {
            return 0;
        }
        $bulk = '';
        foreach ($primaryKeys as $pk) {
            $action = Json::encode([
                "update" => [
                    "_id" => $pk,
                    "_type" => static::type(),
                    "_index" => static::index(),
                ],
            ]);
            $data = Json::encode([
                "doc" => $attributes
            ]);
            $bulk .= $action . "\n" . $data . "\n";
        }

        // TODO do this via command
        $url = [static::index(), static::type(), '_bulk'];
        $response = static::getDb()->post($url, [], $bulk);
        $n = 0;
        $errors = [];
        foreach ($response['items'] as $item) {
462
            if (isset($item['update']['status']) && $item['update']['status'] == 200) {
463
                $n++;
464 465
            } else {
                $errors[] = $item['update'];
466 467
            }
        }
468
        if (!empty($errors) || isset($response['errors']) && $response['errors']) {
469 470 471 472 473 474 475 476 477 478 479 480 481 482
            throw new Exception(__METHOD__ . ' failed updating records.', $errors);
        }

        return $n;
    }

    /**
     * Updates all matching records using the provided counter changes and conditions.
     * For example, to increment all customers' age by 1,
     *
     * ~~~
     * Customer::updateAllCounters(['age' => 1]);
     * ~~~
     *
483 484 485 486 487
     * @param array $counters the counters to be updated (attribute name => increment value).
     * Use negative values if you want to decrement the counters.
     * @param string|array $condition the conditions that will be put in the WHERE part of the UPDATE SQL.
     * Please refer to [[Query::where()]] on how to specify this parameter.
     * @return integer the number of rows updated
488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525
     */
    public static function updateAllCounters($counters, $condition = [])
    {
        $pkName = static::primaryKey()[0];
        if (count($condition) == 1 && isset($condition[$pkName])) {
            $primaryKeys = is_array($condition[$pkName]) ? $condition[$pkName] : [$condition[$pkName]];
        } else {
            $primaryKeys = static::find()->where($condition)->column($pkName); // TODO check whether this works with default pk _id
        }
        if (empty($primaryKeys) || empty($counters)) {
            return 0;
        }
        $bulk = '';
        foreach ($primaryKeys as $pk) {
            $action = Json::encode([
                "update" => [
                    "_id" => $pk,
                    "_type" => static::type(),
                    "_index" => static::index(),
                ],
            ]);
            $script = '';
            foreach ($counters as $counter => $value) {
                $script .= "ctx._source.$counter += $counter;\n";
            }
            $data = Json::encode([
                "script" => $script,
                "params" => $counters
            ]);
            $bulk .= $action . "\n" . $data . "\n";
        }

        // TODO do this via command
        $url = [static::index(), static::type(), '_bulk'];
        $response = static::getDb()->post($url, [], $bulk);
        $n = 0;
        $errors = [];
        foreach ($response['items'] as $item) {
526
            if (isset($item['update']['status']) && $item['update']['status'] == 200) {
527
                $n++;
528 529
            } else {
                $errors[] = $item['update'];
530 531
            }
        }
532
        if (!empty($errors) || isset($response['errors']) && $response['errors']) {
533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548
            throw new Exception(__METHOD__ . ' failed updating records counters.', $errors);
        }

        return $n;
    }

    /**
     * Deletes rows in the table using the provided conditions.
     * WARNING: If you do not specify any condition, this method will delete ALL rows in the table.
     *
     * For example, to delete all customers whose status is 3:
     *
     * ~~~
     * Customer::deleteAll('status = 3');
     * ~~~
     *
549 550
     * @param array $condition the conditions that will be put in the WHERE part of the DELETE SQL.
     * Please refer to [[ActiveQuery::where()]] on how to specify this parameter.
551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580
     * @return integer the number of rows deleted
     */
    public static function deleteAll($condition = [])
    {
        $pkName = static::primaryKey()[0];
        if (count($condition) == 1 && isset($condition[$pkName])) {
            $primaryKeys = is_array($condition[$pkName]) ? $condition[$pkName] : [$condition[$pkName]];
        } else {
            $primaryKeys = static::find()->where($condition)->column($pkName); // TODO check whether this works with default pk _id
        }
        if (empty($primaryKeys)) {
            return 0;
        }
        $bulk = '';
        foreach ($primaryKeys as $pk) {
            $bulk .= Json::encode([
                "delete" => [
                    "_id" => $pk,
                    "_type" => static::type(),
                    "_index" => static::index(),
                ],
            ]) . "\n";
        }

        // TODO do this via command
        $url = [static::index(), static::type(), '_bulk'];
        $response = static::getDb()->post($url, [], $bulk);
        $n = 0;
        $errors = [];
        foreach ($response['items'] as $item) {
581 582 583 584 585
            if (isset($item['delete']['status']) && $item['delete']['status'] == 200) {
                if (isset($item['delete']['found']) && $item['delete']['found']) {
                    $n++;
                }
            } else {
586 587 588
                $errors[] = $item['delete'];
            }
        }
589
        if (!empty($errors) || isset($response['errors']) && $response['errors']) {
590 591 592 593 594
            throw new Exception(__METHOD__ . ' failed deleting records.', $errors);
        }

        return $n;
    }
595
}