[8.x] ESQL: Properly skip datasets with lookup indices (elastic#118682)

Do not rely on specific naming, read the actual settings.
davidkyle · Dec 16, 2024 · 52a178b · 52a178b
1 parent fd20e28
commit 52a178b
Showing 1 changed file with 30 additions and 13 deletions.
diff --git a/...n/esql/qa/testFixtures/src/main/java/org/elasticsearch/xpack/esql/CsvTestsDataLoader.java b/...n/esql/qa/testFixtures/src/main/java/org/elasticsearch/xpack/esql/CsvTestsDataLoader.java
@@ -41,7 +41,6 @@
 import java.util.List;
 import java.util.Map;
 import java.util.Set;
-import java.util.stream.Collectors;
 
 import static org.elasticsearch.common.logging.LoggerMessageFormat.format;
 import static org.elasticsearch.xpack.esql.CsvTestUtils.COMMA_ESCAPING_REGEX;
@@ -258,11 +257,22 @@ public static void main(String[] args) throws IOException {
     public static Set<TestsDataset> availableDatasetsForEs(RestClient client, boolean supportsIndexModeLookup) throws IOException {
         boolean inferenceEnabled = clusterHasInferenceEndpoint(client);
 
-        return CSV_DATASET_MAP.values()
-            .stream()
-            .filter(d -> d.requiresInferenceEndpoint == false || inferenceEnabled)
-            .filter(d -> supportsIndexModeLookup || d.indexName.endsWith("_lookup") == false) // TODO: use actual index settings
-            .collect(Collectors.toCollection(HashSet::new));
+        Set<TestsDataset> testDataSets = new HashSet<>();
+
+        for (TestsDataset dataset : CSV_DATASET_MAP.values()) {
+            if ((inferenceEnabled || dataset.requiresInferenceEndpoint == false)
+                && (supportsIndexModeLookup || isLookupDataset(dataset) == false)) {
+                testDataSets.add(dataset);
+            }
+        }
+
+        return testDataSets;
+    }
+
+    public static boolean isLookupDataset(TestsDataset dataset) throws IOException {
+        Settings settings = dataset.readSettingsFile();
+        String mode = settings.get("index.mode");
+        return (mode != null && mode.equalsIgnoreCase("lookup"));
     }
 
     public static void loadDataSetIntoEs(RestClient client, boolean supportsIndexModeLookup) throws IOException {
@@ -354,13 +364,8 @@ private static void load(RestClient client, TestsDataset dataset, Logger logger,
         if (data == null) {
             throw new IllegalArgumentException("Cannot find resource " + dataName);
         }
-        Settings indexSettings = Settings.EMPTY;
-        final String settingName = dataset.settingFileName != null ? "/" + dataset.settingFileName : null;
-        if (settingName != null) {
-            indexSettings = Settings.builder()
-                .loadFromStream(settingName, CsvTestsDataLoader.class.getResourceAsStream(settingName), false)
-                .build();
-        }
+
+        Settings indexSettings = dataset.readSettingsFile();
         indexCreator.createIndex(client, dataset.indexName, readMappingFile(mapping, dataset.typeMapping), indexSettings);
         loadCsvData(client, dataset.indexName, data, dataset.allowSubFields, logger);
     }
@@ -669,6 +674,18 @@ public TestsDataset withTypeMapping(Map<String, String> typeMapping) {
         public TestsDataset withInferenceEndpoint(boolean needsInference) {
             return new TestsDataset(indexName, mappingFileName, dataFileName, settingFileName, allowSubFields, typeMapping, needsInference);
         }
+
+        private Settings readSettingsFile() throws IOException {
+            Settings indexSettings = Settings.EMPTY;
+            final String settingName = settingFileName != null ? "/" + settingFileName : null;
+            if (settingName != null) {
+                indexSettings = Settings.builder()
+                    .loadFromStream(settingName, CsvTestsDataLoader.class.getResourceAsStream(settingName), false)
+                    .build();
+            }
+
+            return indexSettings;
+        }
     }
 
     public record EnrichConfig(String policyName, String policyFileName) {}