[8.19] Don't build can_match queries we can't push to data nodes (#130210) (#130386)

* Don't build can_match queries we can't push to data nodes (#130210)

Passes the minimum transport version down to expressions when
we convert them into queries that we'll use for can_match.
Right now all this is used for is skipping the can_match from
the wildcard like queries. The queries we make there aren't
serializable. We'll fix that - but this should give us the
levers that we need to do it in a backwards incompatible way.

* Fix for backport

* don't be backwards

Author: nik9000

server/src/main/java/org/elasticsearch/index/query/AutomatonQueryBuilder.java

@@ -48,6 +48,10 @@ public String fieldName() {
         return fieldName;
     }
 
+    public String description() {
+        return description;
+    }
+
     @Override
     public String getWriteableName() {
         throw new UnsupportedOperationException("AutomatonQueryBuilder does not support getWriteableName");

x-pack/plugin/esql/qa/server/multi-clusters/src/javaRestTest/java/org/elasticsearch/xpack/esql/ccq/MultiClustersIT.java

@@ -29,6 +29,7 @@
 import org.junit.rules.TestRule;
 
 import java.io.IOException;
+import java.util.ArrayList;
 import java.util.List;
 import java.util.Map;
 import java.util.Set;
@@ -37,11 +38,15 @@
 import java.util.stream.Stream;
 
 import static org.elasticsearch.test.MapMatcher.assertMap;
+import static org.elasticsearch.test.MapMatcher.matchesMap;
 import static org.elasticsearch.xpack.esql.ccq.Clusters.REMOTE_CLUSTER_NAME;
 import static org.hamcrest.Matchers.any;
+import static org.hamcrest.Matchers.anyOf;
 import static org.hamcrest.Matchers.equalTo;
 import static org.hamcrest.Matchers.greaterThanOrEqualTo;
 import static org.hamcrest.Matchers.hasKey;
+import static org.hamcrest.Matchers.is;
+import static org.hamcrest.Matchers.nullValue;
 
 @ThreadLeakFilters(filters = TestClustersThreadFilter.class)
 public class MultiClustersIT extends ESRestTestCase {
@@ -129,7 +134,7 @@ void indexDocs(RestClient client, String index, List<Doc> docs) throws IOExcepti
     }
 
     private Map<String, Object> run(String query, boolean includeCCSMetadata) throws IOException {
-        var queryBuilder = new RestEsqlTestCase.RequestObjectBuilder().query(query);
+        var queryBuilder = new RestEsqlTestCase.RequestObjectBuilder().query(query).profile(true);
         if (includeCCSMetadata) {
             queryBuilder.includeCCSMetadata(true);
         }
@@ -158,12 +163,51 @@ private Map<String, Object> runEsql(RestEsqlTestCase.RequestObjectBuilder reques
         }
     }
 
+    private <C, V> void assertResultMapForLike(
+        boolean includeCCSMetadata,
+        Map<String, Object> result,
+        C columns,
+        V values,
+        boolean remoteOnly,
+        boolean requireLikeListCapability
+    ) throws IOException {
+        List<String> requiredCapabilities = new ArrayList<>(List.of("like_on_index_fields"));
+        if (requireLikeListCapability) {
+            requiredCapabilities.add("like_list_on_index_fields");
+        }
+        // the feature is completely supported if both local and remote clusters support it
+        boolean isSupported = clusterHasCapability("POST", "/_query", List.of(), requiredCapabilities).orElse(false);
+        try (RestClient remoteClient = remoteClusterClient()) {
+            isSupported = isSupported
+                && clusterHasCapability(remoteClient, "POST", "/_query", List.of(), requiredCapabilities).orElse(false);
+        }
+
+        if (isSupported) {
+            assertResultMap(includeCCSMetadata, result, columns, values, remoteOnly);
+        } else {
+            logger.info("-->  skipping data check for like index test, cluster does not support like index feature");
+            // just verify that we did not get a partial result
+            var clusters = result.get("_clusters");
+            var reason = "unexpected partial results" + (clusters != null ? ": _clusters=" + clusters : "");
+            assertThat(reason, result.get("is_partial"), anyOf(nullValue(), is(false)));
+        }
+    }
+
+    private boolean capabilitiesSupportedNewAndOld(List<String> requiredCapabilities) throws IOException {
+        boolean isSupported = clusterHasCapability("POST", "/_query", List.of(), requiredCapabilities).orElse(false);
+        try (RestClient remoteClient = remoteClusterClient()) {
+            isSupported = isSupported
+                && clusterHasCapability(remoteClient, "POST", "/_query", List.of(), requiredCapabilities).orElse(false);
+        }
+        return isSupported;
+    }
+
     private <C, V> void assertResultMap(boolean includeCCSMetadata, Map<String, Object> result, C columns, V values, boolean remoteOnly) {
         MapMatcher mapMatcher = getResultMatcher(
             ccsMetadataAvailable(),
             result.containsKey("is_partial"),
             result.containsKey("documents_found")
-        );
+        ).extraOk();
         if (includeCCSMetadata) {
             mapMatcher = mapMatcher.entry("_clusters", any(Map.class));
         }
@@ -251,11 +295,13 @@ private void assertClusterDetailsMap(Map<String, Object> result, boolean remoteO
 
         @SuppressWarnings("unchecked")
         Map<String, Object> remoteClusterShards = (Map<String, Object>) remoteCluster.get("_shards");
-        assertThat(remoteClusterShards.keySet(), equalTo(Set.of("total", "successful", "skipped", "failed")));
-        assertThat((Integer) remoteClusterShards.get("total"), greaterThanOrEqualTo(0));
-        assertThat((Integer) remoteClusterShards.get("successful"), equalTo((Integer) remoteClusterShards.get("total")));
-        assertThat((Integer) remoteClusterShards.get("skipped"), equalTo(0));
-        assertThat((Integer) remoteClusterShards.get("failed"), equalTo(0));
+        assertThat(
+            remoteClusterShards,
+            matchesMap().entry("total", greaterThanOrEqualTo(0))
+                .entry("successful", remoteClusterShards.get("total"))
+                .entry("skipped", greaterThanOrEqualTo(0))
+                .entry("failed", 0)
+        );
 
         if (remoteOnly == false) {
             @SuppressWarnings("unchecked")
@@ -267,11 +313,13 @@ private void assertClusterDetailsMap(Map<String, Object> result, boolean remoteO
 
             @SuppressWarnings("unchecked")
             Map<String, Object> localClusterShards = (Map<String, Object>) localCluster.get("_shards");
-            assertThat(localClusterShards.keySet(), equalTo(Set.of("total", "successful", "skipped", "failed")));
-            assertThat((Integer) localClusterShards.get("total"), greaterThanOrEqualTo(0));
-            assertThat((Integer) localClusterShards.get("successful"), equalTo((Integer) localClusterShards.get("total")));
-            assertThat((Integer) localClusterShards.get("skipped"), equalTo(0));
-            assertThat((Integer) localClusterShards.get("failed"), equalTo(0));
+            assertThat(
+                localClusterShards,
+                matchesMap().entry("total", greaterThanOrEqualTo(0))
+                    .entry("successful", localClusterShards.get("total"))
+                    .entry("skipped", greaterThanOrEqualTo(0))
+                    .entry("failed", 0)
+            );
         }
     }
 
@@ -371,6 +419,116 @@ public void testStats() throws IOException {
         assertThat(clusterData, hasKey("took"));
     }
 
+    public void testLikeIndex() throws Exception {
+
+        boolean includeCCSMetadata = includeCCSMetadata();
+        Map<String, Object> result = run("""
+            FROM test-local-index,*:test-remote-index METADATA _index
+            | WHERE _index LIKE "*remote*"
+            | STATS c = COUNT(*) BY _index
+            | SORT _index ASC
+            """, includeCCSMetadata);
+        var columns = List.of(Map.of("name", "c", "type", "long"), Map.of("name", "_index", "type", "keyword"));
+        var values = List.of(List.of(remoteDocs.size(), REMOTE_CLUSTER_NAME + ":" + remoteIndex));
+        assertResultMapForLike(includeCCSMetadata, result, columns, values, false, false);
+    }
+
+    public void testNotLikeIndex() throws Exception {
+        boolean includeCCSMetadata = includeCCSMetadata();
+        Map<String, Object> result = run("""
+            FROM test-local-index,*:test-remote-index METADATA _index
+            | WHERE _index NOT LIKE "*remote*"
+            | STATS c = COUNT(*) BY _index
+            | SORT _index ASC
+            """, includeCCSMetadata);
+        var columns = List.of(Map.of("name", "c", "type", "long"), Map.of("name", "_index", "type", "keyword"));
+        var values = List.of(List.of(localDocs.size(), localIndex));
+        assertResultMapForLike(includeCCSMetadata, result, columns, values, false, false);
+    }
+
+    public void testLikeListIndex() throws Exception {
+        List<String> requiredCapabilities = new ArrayList<>(List.of("like_list_on_index_fields"));
+        // the feature is completely supported if both local and remote clusters support it
+        if (capabilitiesSupportedNewAndOld(requiredCapabilities) == false) {
+            logger.info("-->  skipping testNotLikeListIndex, due to missing capability");
+            return;
+        }
+        boolean includeCCSMetadata = includeCCSMetadata();
+        Map<String, Object> result = run("""
+            FROM test-local-index,*:test-remote-index METADATA _index
+            | WHERE _index LIKE ("*remote*", "not-exist*")
+            | STATS c = COUNT(*) BY _index
+            | SORT _index ASC
+            """, includeCCSMetadata);
+        var columns = List.of(Map.of("name", "c", "type", "long"), Map.of("name", "_index", "type", "keyword"));
+        var values = List.of(List.of(remoteDocs.size(), REMOTE_CLUSTER_NAME + ":" + remoteIndex));
+        assertResultMapForLike(includeCCSMetadata, result, columns, values, false, true);
+    }
+
+    public void testNotLikeListIndex() throws Exception {
+        List<String> requiredCapabilities = new ArrayList<>(List.of("like_list_on_index_fields"));
+        // the feature is completely supported if both local and remote clusters support it
+        if (capabilitiesSupportedNewAndOld(requiredCapabilities) == false) {
+            logger.info("-->  skipping testNotLikeListIndex, due to missing capability");
+            return;
+        }
+        boolean includeCCSMetadata = includeCCSMetadata();
+        Map<String, Object> result = run("""
+            FROM test-local-index,*:test-remote-index METADATA _index
+            | WHERE _index NOT LIKE ("*remote*", "not-exist*")
+            | STATS c = COUNT(*) BY _index
+            | SORT _index ASC
+            """, includeCCSMetadata);
+        var columns = List.of(Map.of("name", "c", "type", "long"), Map.of("name", "_index", "type", "keyword"));
+        var values = List.of(List.of(localDocs.size(), localIndex));
+        assertResultMapForLike(includeCCSMetadata, result, columns, values, false, true);
+    }
+
+    public void testNotLikeListKeyWord() throws Exception {
+        List<String> requiredCapabilities = new ArrayList<>(List.of("like_list_on_index_fields"));
+        // the feature is completely supported if both local and remote clusters support it
+        if (capabilitiesSupportedNewAndOld(requiredCapabilities) == false) {
+            logger.info("-->  skipping testNotLikeListIndex, due to missing capability");
+            return;
+        }
+        boolean includeCCSMetadata = includeCCSMetadata();
+        Map<String, Object> result = run("""
+            FROM test-local-index,*:test-remote-index METADATA _index
+            | WHERE color NOT LIKE ("*blue*", "*red*")
+            | STATS c = COUNT(*) BY _index
+            | SORT _index ASC
+            """, includeCCSMetadata);
+        var columns = List.of(Map.of("name", "c", "type", "long"), Map.of("name", "_index", "type", "keyword"));
+        var values = List.of(List.of(localDocs.size(), localIndex));
+        assertResultMapForLike(includeCCSMetadata, result, columns, values, false, true);
+    }
+
+    public void testRLikeIndex() throws Exception {
+        boolean includeCCSMetadata = includeCCSMetadata();
+        Map<String, Object> result = run("""
+            FROM test-local-index,*:test-remote-index METADATA _index
+            | WHERE _index RLIKE ".*remote.*"
+            | STATS c = COUNT(*) BY _index
+            | SORT _index ASC
+            """, includeCCSMetadata);
+        var columns = List.of(Map.of("name", "c", "type", "long"), Map.of("name", "_index", "type", "keyword"));
+        var values = List.of(List.of(remoteDocs.size(), REMOTE_CLUSTER_NAME + ":" + remoteIndex));
+        assertResultMapForLike(includeCCSMetadata, result, columns, values, false, false);
+    }
+
+    public void testNotRLikeIndex() throws Exception {
+        boolean includeCCSMetadata = includeCCSMetadata();
+        Map<String, Object> result = run("""
+            FROM test-local-index,*:test-remote-index METADATA _index
+            | WHERE _index NOT RLIKE ".*remote.*"
+            | STATS c = COUNT(*) BY _index
+            | SORT _index ASC
+            """, includeCCSMetadata);
+        var columns = List.of(Map.of("name", "c", "type", "long"), Map.of("name", "_index", "type", "keyword"));
+        var values = List.of(List.of(localDocs.size(), localIndex));
+        assertResultMapForLike(includeCCSMetadata, result, columns, values, false, false);
+    }
+
     private RestClient remoteClusterClient() throws IOException {
         var clusterHosts = parseClusterHosts(remoteCluster.getHttpAddresses());
         return buildClient(restClientSettings(), clusterHosts.toArray(new HttpHost[0]));

x-pack/plugin/esql/src/main/java/org/elasticsearch/xpack/esql/expression/function/scalar/string/regex/WildcardLikeList.java

@@ -92,8 +92,12 @@ protected WildcardLikeList replaceChild(Expression newLeft) {
      */
     @Override
     public Translatable translatable(LucenePushdownPredicates pushdownPredicates) {
-        return pushdownPredicates.isPushableAttribute(field()) ? Translatable.YES : Translatable.NO;
-
+        if (pushdownPredicates.minTransportVersion() == null) {
+            return pushdownPredicates.isPushableAttribute(field()) ? Translatable.YES : Translatable.NO;
+        } else {
+            // The AutomatonQuery that we use right now isn't serializable.
+            return Translatable.NO;
+        }
     }
 
     /**

x-pack/plugin/esql/src/main/java/org/elasticsearch/xpack/esql/optimizer/rules/physical/local/LucenePushdownPredicates.java

@@ -7,6 +7,9 @@
 
 package org.elasticsearch.xpack.esql.optimizer.rules.physical.local;
 
+import org.elasticsearch.TransportVersion;
+import org.elasticsearch.core.Nullable;
+import org.elasticsearch.index.query.QueryBuilder;
 import org.elasticsearch.xpack.esql.core.expression.Expression;
 import org.elasticsearch.xpack.esql.core.expression.FieldAttribute;
 import org.elasticsearch.xpack.esql.core.expression.MetadataAttribute;
@@ -30,6 +33,24 @@
  * </ol>
  */
 public interface LucenePushdownPredicates {
+    /**
+     * If we're extracting a query for {@code can_match} then this is the
+     * minimum transport version in the cluster. Otherwise, this is {@code null}.
+     * <p>
+     *     If this is not null {@link Expression}s should not claim to be
+     *     serializable unless their {@link QueryBuilder}
+     *     {@link QueryBuilder#getMinimalSupportedVersion supports} the version.
+     * </p>
+     * <p>
+     *     This is done on the coordinating node <strong>and</strong>. And for
+     *     cross cluster search this is done on the coordinating node on the
+     *     remote cluster. So! We actually <strong>have</strong> the minimum
+     *     cluster transport version.
+     * </p>
+     */
+    @Nullable
+    TransportVersion minTransportVersion();
+
     /**
      * For TEXT fields, we need to check if the field has a subfield of type KEYWORD that can be used instead.
      */
@@ -101,29 +122,41 @@ static String pushableAttributeName(TypedAttribute attribute) {
      * In particular, it assumes TEXT fields have no exact subfields (underlying keyword field),
      * and that isAggregatable means indexed and has hasDocValues.
      */
-    LucenePushdownPredicates DEFAULT = new LucenePushdownPredicates() {
-        @Override
-        public boolean hasExactSubfield(FieldAttribute attr) {
-            return false;
-        }
+    LucenePushdownPredicates DEFAULT = forCanMatch(null);
 
-        @Override
-        public boolean isIndexedAndHasDocValues(FieldAttribute attr) {
-            // Is the FieldType.isAggregatable() check correct here? In FieldType isAggregatable usually only means hasDocValues
-            return attr.field().isAggregatable();
-        }
+    /**
+     * A {@link LucenePushdownPredicates} for use with the {@code can_match} phase.
+     */
+    static LucenePushdownPredicates forCanMatch(TransportVersion minTransportVersion) {
+        return new LucenePushdownPredicates() {
+            @Override
+            public TransportVersion minTransportVersion() {
+                return minTransportVersion;
+            }
 
-        @Override
-        public boolean isIndexed(FieldAttribute attr) {
-            // TODO: This is the original behaviour, but is it correct? In FieldType isAggregatable usually only means hasDocValues
-            return attr.field().isAggregatable();
-        }
+            @Override
+            public boolean hasExactSubfield(FieldAttribute attr) {
+                return false;
+            }
 
-        @Override
-        public boolean canUseEqualityOnSyntheticSourceDelegate(FieldAttribute attr, String value) {
-            return false;
-        }
-    };
+            @Override
+            public boolean isIndexedAndHasDocValues(FieldAttribute attr) {
+                // Is the FieldType.isAggregatable() check correct here? In FieldType isAggregatable usually only means hasDocValues
+                return attr.field().isAggregatable();
+            }
+
+            @Override
+            public boolean isIndexed(FieldAttribute attr) {
+                // TODO: This is the original behaviour, but is it correct? In FieldType isAggregatable usually only means hasDocValues
+                return attr.field().isAggregatable();
+            }
+
+            @Override
+            public boolean canUseEqualityOnSyntheticSourceDelegate(FieldAttribute attr, String value) {
+                return false;
+            }
+        };
+    }
 
     /**
      * If we have access to {@link SearchStats} over a collection of shards, we can make more fine-grained decisions about what can be
@@ -133,6 +166,11 @@ static LucenePushdownPredicates from(SearchStats stats) {
         // TODO: use FieldAttribute#fieldName, otherwise this doesn't apply to field attributes used for union types.
         // C.f. https://github.com/elastic/elasticsearch/issues/128905
         return new LucenePushdownPredicates() {
+            @Override
+            public TransportVersion minTransportVersion() {
+                return null;
+            }
+
             @Override
             public boolean hasExactSubfield(FieldAttribute attr) {
                 return stats.hasExactSubfield(new FieldAttribute.FieldName(attr.name()));