add aggregation, stats and suggester to elasticsearch

fixes #4452, close #4452
10 years ago · 3fb680964c
3 changed files with 68 additions and 121 deletions
--- a/extensions/elasticsearch/CHANGELOG.md
+++ b/extensions/elasticsearch/CHANGELOG.md
@ -11,6 +11,9 @@ Yii Framework 2 elasticsearch extension Change Log
 - Enh #4048: Added `init` event to `ActiveQuery` classes (qiangxue)
 - Enh #4086: changedAttributes of afterSave Event now contain old values (dizews)
 - Enh: Make error messages more readable in HTML output (cebe)
+- Enh: Added support for query stats (cebe)
+- Enh: Added support for query suggesters (cebe)
+- Chg #4451: Removed support for facets and replaced them with aggregations (cebe, tadaszelvys)
 - Chg: asArray in ActiveQuery is now equal to using the normal Query. This means, that the output structure has changed and `with` is supported anymore. (cebe)
 - Chg: Deletion of a record is now also considered successful if the record did not exist. (cebe)
 - Chg: Requirement changes: Yii now requires elasticsearch version 1.0 or higher (cebe)
--- a/extensions/elasticsearch/Query.php
+++ b/extensions/elasticsearch/Query.php
@ -135,11 +135,29 @@ class Query extends Component implements QueryInterface
    /**
     * @var array The highlight part of this search query. This is an array that allows to highlight search results
     * on one or more fields.
+     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/1.x/search-request-highlighting.html
     */
    public $highlight;
-    public $facets = [];
+    /**
+     * @var array List of aggregations to add to this query.
+     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/1.x/search-aggregations.html
+     */
+    public $aggregations = [];
+    /**
+     * @var array the 'stats' part of the query. An array of groups to maintain a statistics aggregation for.
+     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search.html#stats-groups
+     */
+    public $stats = [];
+    /**
+     * @var array list of suggesters to add to this query.
+     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-suggesters.html
+     */
+    public $suggest = [];


+    /**
+     * @inheritdoc
+     */
    public function init()
    {
        parent::init();
@ -244,8 +262,6 @@ class Query extends Component implements QueryInterface
        return $result;
    }

-    // TODO add query stats http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search.html#stats-groups
-
    // TODO add scroll/scan http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-request-search-type.html#scan

    /**
@ -342,137 +358,72 @@ class Query extends Component implements QueryInterface
    }

    /**
-     * Adds a facet search to this query.
-     * @param string $name the name of this facet
-     * @param string $type the facet type. e.g. `terms`, `range`, `histogram`...
-     * @param string|array $options the configuration options for this facet. Can be an array or a json string.
+     * Adds a 'stats' part to the query.
+     * @param array $groups an array of groups to maintain a statistics aggregation for.
     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-query-facet.html
+     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search.html#stats-groups
     */
-    public function addFacet($name, $type, $options)
+    public function stats($groups)
    {
-        $this->facets[$name] = [$type => $options];
+        $this->stats = $groups;
        return $this;
    }

    /**
-     * The `terms facet` allow to specify field facets that return the N most frequent terms.
-     * @param string $name the name of this facet
-     * @param array $options additional option. Please refer to the elasticsearch documentation for details.
-     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-terms-facet.html
-     */
-    public function addTermFacet($name, $options)
-    {
-        return $this->addFacet($name, 'terms', $options);
-    }
-
-    /**
-     * Range facet allows to specify a set of ranges and get both the number of docs (count) that fall
-     * within each range, and aggregated data either based on the field, or using another field.
-     * @param string $name the name of this facet
-     * @param array $options additional option. Please refer to the elasticsearch documentation for details.
-     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-range-facet.html
-     */
-    public function addRangeFacet($name, $options)
-    {
-        return $this->addFacet($name, 'range', $options);
-    }
-
-    /**
-     * The histogram facet works with numeric data by building a histogram across intervals of the field values.
-     * Each value is "rounded" into an interval (or placed in a bucket), and statistics are provided per
-     * interval/bucket (count and total).
-     * @param string $name the name of this facet
-     * @param array $options additional option. Please refer to the elasticsearch documentation for details.
-     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-histogram-facet.html
-     */
-    public function addHistogramFacet($name, $options)
-    {
-        return $this->addFacet($name, 'histogram', $options);
-    }
-
-    /**
-     * A specific histogram facet that can work with date field types enhancing it over the regular histogram facet.
-     * @param string $name the name of this facet
-     * @param array $options additional option. Please refer to the elasticsearch documentation for details.
-     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-date-histogram-facet.html
-     */
-    public function addDateHistogramFacet($name, $options)
-    {
-        return $this->addFacet($name, 'date_histogram', $options);
-    }
-
-    /**
-     * A filter facet (not to be confused with a facet filter) allows you to return a count of the hits matching the filter.
-     * The filter itself can be expressed using the Query DSL.
-     * @param string $name the name of this facet
-     * @param string $filter the query in Query DSL
-     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-filter-facet.html
-     */
-    public function addFilterFacet($name, $filter)
-    {
-        return $this->addFacet($name, 'filter', $filter);
-    }
-
-    /**
-     * A facet query allows to return a count of the hits matching the facet query.
-     * The query itself can be expressed using the Query DSL.
-     * @param string $name the name of this facet
-     * @param string $query the query in Query DSL
+     * Sets a highlight parameters to retrieve from the documents.
+     * @param array $highlight array of parameters to highlight results.
     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-query-facet.html
+     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-request-highlighting.html
     */
-    public function addQueryFacet($name, $query)
+    public function highlight($highlight)
    {
-        return $this->addFacet($name, 'query', $query);
+        $this->highlight = $highlight;
+        return $this;
    }

    /**
-     * Statistical facet allows to compute statistical data on a numeric fields. The statistical data include count,
-     * total, sum of squares, mean (average), minimum, maximum, variance, and standard deviation.
-     * @param string $name the name of this facet
-     * @param array $options additional option. Please refer to the elasticsearch documentation for details.
+     * Adds an aggregation to this query.
+     * @param string $name the name of the aggregation
+     * @param string $type the aggregation type. e.g. `terms`, `range`, `histogram`...
+     * @param string|array $options the configuration options for this aggregation. Can be an array or a json string.
     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-statistical-facet.html
+     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/1.x/search-aggregations.html
     */
-    public function addStatisticalFacet($name, $options)
+    public function addAggregation($name, $type, $options)
    {
-        return $this->addFacet($name, 'statistical', $options);
+        $this->aggregations[$name] = [$type => $options];
+        return $this;
    }

    /**
-     * The `terms_stats` facet combines both the terms and statistical allowing to compute stats computed on a field,
-     * per term value driven by another field.
-     * @param string $name the name of this facet
-     * @param array $options additional option. Please refer to the elasticsearch documentation for details.
+     * Adds an aggregation to this query.
+     *
+     * This is an alias for [[addAggregation]].
+     *
+     * @param string $name the name of the aggregation
+     * @param string $type the aggregation type. e.g. `terms`, `range`, `histogram`...
+     * @param string|array $options the configuration options for this aggregation. Can be an array or a json string.
     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-terms-stats-facet.html
+     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/1.x/search-aggregations.html
     */
-    public function addTermsStatsFacet($name, $options)
+    public function addAgg($name, $type, $options)
    {
-        return $this->addFacet($name, 'terms_stats', $options);
+        return $this->addAggregation($name, $type, $options);
    }

    /**
-     * The `geo_distance` facet is a facet providing information for ranges of distances from a provided `geo_point`
-     * including count of the number of hits that fall within each range, and aggregation information (like `total`).
-     * @param string $name the name of this facet
-     * @param array $options additional option. Please refer to the elasticsearch documentation for details.
+     * Adds a suggester to this query.
+     * @param string $name the name of the suggester
+     * @param string|array $definition the configuration options for this suggester. Can be an array or a json string.
     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-facets-geo-distance-facet.html
+     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-suggesters.html
     */
-    public function addGeoDistanceFacet($name, $options)
+    public function addSuggester($name, $definition)
    {
-        return $this->addFacet($name, 'geo_distance', $options);
+        $this->suggest[$name] = $definition;
+        return $this;
    }

-    // TODO add suggesters http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-suggesters.html
-
    // TODO add validate query http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-validate.html

    // TODO support multi query via static method http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-multi-search.html
@ -532,18 +483,6 @@ class Query extends Component implements QueryInterface
    }

    /**
-     * Sets a highlight parameters to retrieve from the documents.
-     * @param array $highlight array of parameters to highlight results.
-     * @return static the query object itself
-     * @see http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-request-highlighting.html
-     */
-    public function highlight($highlight)
-    {
-        $this->highlight = $highlight;
-        return $this;
-    }
-
-    /**
     * Sets the source filtering, specifying how the `_source` field of the document should be returned.
     * @param array $source the source patterns to be selected.
     * @return static the query object itself
--- a/extensions/elasticsearch/QueryBuilder.php
+++ b/extensions/elasticsearch/QueryBuilder.php
@ -98,19 +98,24 @@ class QueryBuilder extends \yii\base\Object
            $parts['filter'] = $whereFilter;
        }

-        if($query->highlight) {
+        if (!empty($query->highlight)) {
            $parts['highlight'] = $query->highlight;
        }
+        if (!empty($query->aggregations)) {
+            $parts['aggregations'] = $query->aggregations;
+        }
+        if (!empty($query->stats)) {
+            $parts['stats'] = $query->stats;
+        }
+        if (!empty($query->suggest)) {
+            $parts['suggest'] = $query->suggest;
+        }

        $sort = $this->buildOrderBy($query->orderBy);
        if (!empty($sort)) {
            $parts['sort'] = $sort;
        }

-        if (!empty($query->facets)) {
-            $parts['facets'] = $query->facets;
-        }
-
        $options = [];
        if ($query->timeout !== null) {
            $options['timeout'] = $query->timeout;