Elasticsearch Query DSL Overview: Queries, Filters, and Advanced Types

After understanding mapping and analysis in Elasticsearch, users often focus on constructing query statements to retrieve the desired data; this article introduces the Query DSL that serves this purpose.

Query DSL (Query Domain Specific Language) is a language dedicated to querying Elasticsearch, with two main categories: leaf query clauses that match specific field values, and compound query clauses that combine multiple simple or compound clauses logically.

The behavior of a query depends on its context: in filter context, a document either matches or not, and results are cached for high efficiency; in query context, relevance scores (_score) are calculated based on the proportion of matched terms.

Basic match‑all queries include match_all (matches all documents) and match_none (matches none).

Full‑text queries analyze the query string before execution and include match, match_phrase, match_phrase_prefix, multi_match, common_terms, query_string, and simple_query_string.

Term‑level queries operate directly on exact terms stored in the inverted index and cover term, terms, terms_set, range, exists, prefix, wildcard, regexp, fuzzy, type, and ids.

Compound queries combine other queries and can modify scoring or switch contexts; examples are constant_score, bool, dis_max, function_score, and boosting.

Joining queries, which avoid costly SQL‑style joins in a distributed system, include nested, has_child / has_parent, and parent_id.

Geo queries support two data types— geo_point and geo_shape —and provide geo_shape, geo_bounding_box, geo_distance, and geo_polygon queries.

Specialized queries not covered elsewhere are more_like_this, script, percolate, and wrapper.

Span queries offer low‑level control over term order and proximity, useful for legal or patent documents; they include span_term, span_multi, span_first, span_near, span_or, span_not, span_containing, span_within, and field_masking_span.

The article concludes that this overview gives a general impression of Elasticsearch's query capabilities, and readers should refer to the official documentation for detailed specifics.