Bool query (must, should, must_not, filter) in Elasticsearch - Step-by-Step Execution

Concept Flow - Bool query (must, should, must_not, filter)

Start Bool Query

↓

Combine with OR logic

↓

Exclude matching docs

↓

Filter without scoring

↓

Return matched documents

Bool query combines multiple conditions: must (all required), should (optional), must_not (exclude), and filter (no scoring, just filter).

Execution Sample

Elasticsearch

{
  "bool": {
    "must": [{"match": {"title": "apple"}}],
    "should": [{"term": {"color": "red"}}],
    "must_not": [{"term": {"status": "sold"}}],
    "filter": [{"range": {"price": {"lte": 100}}}]
  }
}

This query finds documents with title containing 'apple', optionally red color, excludes sold items, and filters price less or equal 100.

Execution Table

Step	Clause	Condition	Evaluation	Effect on Results
1	must	title contains 'apple'	Check if document title has 'apple'	Include only if true
2	should	color is 'red'	Check if document color is 'red'	Boost score if true, but not required
3	must_not	status is 'sold'	Check if document status is 'sold'	Exclude document if true
4	filter	price <= 100	Check if price is less or equal 100	Include only if true, no score change
5	final	Combine all clauses	Documents must pass must and filter, must_not excludes, should boosts score	Return matching documents

💡 Documents not matching must or filter clauses or matching must_not are excluded; should clauses influence scoring but do not exclude.

Variable Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	After Step 4	Final
Documents	All documents	Only with title 'apple'	Same set, score boosted if color 'red'	Exclude those with status 'sold'	Only with price <= 100	Final filtered and scored documents

Key Moments - 3 Insights

Why do documents that don't match 'should' clauses still appear in results?

Because 'should' clauses boost score but are not required; only 'must' and 'filter' clauses strictly include documents (see execution_table step 2).

Does 'filter' affect document scoring?

'Filter' clauses only include or exclude documents without changing their score, unlike 'must' or 'should' (see execution_table step 4).

What happens if a document matches a 'must_not' clause?

It is excluded from results regardless of other matches (see execution_table step 3).

Visual Quiz - 3 Questions

According to the execution_table, what happens at step 3?

A 'must_not' boosts document score

B Documents matching 'must_not' are included

C Documents matching 'must_not' are excluded

D 'must_not' filters documents without exclusion

💡 Hint

Check execution_table row with Step 3 describing 'must_not' clause effect

Looking at variable_tracker, what is the state of documents after step 4?

A Documents filtered to price <= 100

B Documents boosted by 'should' clause

C Documents with price > 100 included

D Documents excluded by 'must_not' clause

💡 Hint

See variable_tracker column 'After Step 4' describing price filtering

If we remove the 'must' clause, how does the execution_table change?

A 'must_not' clause becomes required

B 'must' step is skipped, documents no longer required to have title 'apple'

C 'should' clause excludes documents

D 'filter' clause boosts scores

💡 Hint

Refer to execution_table step 1 about 'must' clause requirement

Concept Snapshot

Bool query syntax:
{
  "bool": {
    "must": [...],
    "should": [...],
    "must_not": [...],
    "filter": [...]
  }
}

Behavior:
- must: required conditions
- should: optional, boosts score
- must_not: excludes matches
- filter: filters without scoring

Use to combine multiple query parts logically.

Full Transcript

A Bool query in Elasticsearch combines multiple conditions to find documents. The 'must' clause requires documents to match all its conditions. The 'should' clause adds optional conditions that boost document scores but do not exclude documents if not matched. The 'must_not' clause excludes documents that match its conditions. The 'filter' clause filters documents without affecting their score. Execution proceeds by checking each clause in order: first 'must', then 'should', then 'must_not', and finally 'filter'. Documents must pass 'must' and 'filter' clauses, must not match 'must_not', and may match 'should' to improve ranking. This combination allows flexible and powerful searches.