Fix query builder rewriting when using lookup join (#123498)

Author: ioanatia

x-pack/plugin/esql/qa/testFixtures/src/main/resources/data/semantic_text.csv

@@ -1,4 +1,4 @@
-_id:keyword,semantic_text_field:semantic_text,st_bool:semantic_text,st_cartesian_point:semantic_text,st_cartesian_shape:semantic_text,st_datetime:semantic_text,st_double:semantic_text,st_geopoint:semantic_text,st_geoshape:semantic_text,st_integer:semantic_text,st_ip:semantic_text,st_long:semantic_text,st_unsigned_long:semantic_text,st_version:semantic_text,st_multi_value:semantic_text,st_unicode:semantic_text,host:keyword,description:text,value:long,st_base64:semantic_text,st_logs:semantic_text
-1,live long and prosper,false,"POINT(4297.11 -1475.53)",,1953-09-02T00:00:00.000Z,5.20128E11,"POINT(42.97109630194 14.7552534413725)","POLYGON ((30 10\, 40 40\, 20 40\, 10 20\, 30 10))",23,1.1.1.1,2147483648,2147483648,1.2.3,["Hello there!", "This is a random value", "for testing purposes"],你吃饭了吗,"host1","some description1",1001,ZWxhc3RpYw==,"2024-12-23T12:15:00.000Z 1.2.3.4 [email protected] 4553"
-2,all we have to decide is what to do with the time that is given to us,true,"POINT(7580.93 2272.77)",,2023-09-24T15:57:00.000Z,4541.11,"POINT(37.97109630194 21.7552534413725)","POLYGON ((30 10\, 40 40\, 20 40\, 10 20\, 30 10))",122,1.1.2.1,123,2147483648.2,9.0.0,["nice to meet you", "bye bye!"],["谢谢", "对不起我的中文不好"],"host2","some description2",1002,aGVsbG8=,"2024-01-23T12:15:00.000Z 1.2.3.4 [email protected] 42"
-3,be excellent to each other,,,,,,,,,,,,,,,"host3","some description3",1003,,"2023-01-23T12:15:00.000Z 127.0.0.1 [email protected] 42"
+_id:keyword,semantic_text_field:semantic_text,st_bool:semantic_text,st_cartesian_point:semantic_text,st_cartesian_shape:semantic_text,st_datetime:semantic_text,st_double:semantic_text,st_geopoint:semantic_text,st_geoshape:semantic_text,st_integer:semantic_text,st_ip:semantic_text,st_long:semantic_text,st_unsigned_long:semantic_text,st_version:semantic_text,st_multi_value:semantic_text,st_unicode:semantic_text,host:keyword,description:text,value:long,st_base64:semantic_text,st_logs:semantic_text,language_name:keyword
+1,live long and prosper,false,"POINT(4297.11 -1475.53)",,1953-09-02T00:00:00.000Z,5.20128E11,"POINT(42.97109630194 14.7552534413725)","POLYGON ((30 10\, 40 40\, 20 40\, 10 20\, 30 10))",23,1.1.1.1,2147483648,2147483648,1.2.3,["Hello there!", "This is a random value", "for testing purposes"],你吃饭了吗,"host1","some description1",1001,ZWxhc3RpYw==,"2024-12-23T12:15:00.000Z 1.2.3.4 [email protected] 4553",English
+2,all we have to decide is what to do with the time that is given to us,true,"POINT(7580.93 2272.77)",,2023-09-24T15:57:00.000Z,4541.11,"POINT(37.97109630194 21.7552534413725)","POLYGON ((30 10\, 40 40\, 20 40\, 10 20\, 30 10))",122,1.1.2.1,123,2147483648.2,9.0.0,["nice to meet you", "bye bye!"],["谢谢", "对不起我的中文不好"],"host2","some description2",1002,aGVsbG8=,"2024-01-23T12:15:00.000Z 1.2.3.4 [email protected] 42",French
+3,be excellent to each other,,,,,,,,,,,,,,,"host3","some description3",1003,,"2023-01-23T12:15:00.000Z 127.0.0.1 [email protected] 42",Spanish

x-pack/plugin/esql/qa/testFixtures/src/main/resources/mapping-semantic_text.json

@@ -76,6 +76,9 @@
     "st_logs": {
       "type": "semantic_text",
       "inference_id": "test_sparse_inference"
+    },
+    "language_name": {
+      "type": "keyword"
     }
   }
 }

x-pack/plugin/esql/qa/testFixtures/src/main/resources/match-operator.csv-spec

@@ -721,3 +721,39 @@ book_no:keyword
 7140
 2714
 ;
+
+testMatchWithSemanticTextAndLookupJoin
+required_capability: match_operator_colon
+required_capability: semantic_text_field_caps
+required_capability: join_lookup_v12
+
+from semantic_text METADATA _id
+| where semantic_text_field:"something"
+| LOOKUP JOIN languages_lookup ON language_name
+| KEEP _id, language_name, language_code
+;
+
+ignoreOrder: true
+
+_id:keyword | language_name:keyword | language_code:integer
+1           | English               | 1
+2           | French                | 2
+3           | Spanish               | 3
+;
+
+testMatchWithSemanticTextKqlAndLookupJoin
+required_capability: match_operator_colon
+required_capability: semantic_text_type
+required_capability: kql_function
+required_capability: semantic_text_field_caps
+required_capability: join_lookup_v12
+
+from semantic_text
+| where kql("host:host1") AND semantic_text_field:"something"
+| LOOKUP JOIN languages_lookup ON language_name
+| KEEP host, semantic_text_field, language_name, language_code
+;
+
+host:keyword | semantic_text_field:text | language_name:keyword | language_code:integer
+"host1"      | live long and prosper    | English               | 1
+;

x-pack/plugin/esql/src/main/java/org/elasticsearch/xpack/esql/session/QueryBuilderResolver.java

@@ -16,7 +16,6 @@
 import org.elasticsearch.index.query.Rewriteable;
 import org.elasticsearch.search.SearchService;
 import org.elasticsearch.transport.TransportService;
-import org.elasticsearch.xpack.esql.core.util.Holder;
 import org.elasticsearch.xpack.esql.expression.function.fulltext.FullTextFunction;
 import org.elasticsearch.xpack.esql.plan.logical.EsRelation;
 import org.elasticsearch.xpack.esql.plan.logical.LogicalPlan;
@@ -98,9 +97,9 @@ private Set<FullTextFunction> fullTextFunctions(LogicalPlan plan) {
     }
 
     public Set<String> indexNames(LogicalPlan plan) {
-        Holder<Set<String>> indexNames = new Holder<>();
-        plan.forEachDown(EsRelation.class, esRelation -> indexNames.set(esRelation.concreteIndices()));
-        return indexNames.get();
+        Set<String> indexNames = new HashSet<>();
+        plan.forEachDown(EsRelation.class, esRelation -> indexNames.addAll(esRelation.concreteIndices()));
+        return indexNames;
     }
 
     public LogicalPlan planWithResolvedQueryBuilders(LogicalPlan plan, Map<FullTextFunction, QueryBuilder> newQueryBuilders) {