Distinguish between simple matches with and without the terms index

modules/mapper-extras/src/main/java/org/elasticsearch/index/mapper/ScaledFloatFieldMapper.java

@@ -136,7 +136,7 @@ public static final class ScaledFloatFieldType extends SimpleMappedFieldType {
 
         public ScaledFloatFieldType(String name, boolean indexed, boolean stored, boolean hasDocValues,
                                     Map<String, String> meta, double scalingFactor, Double nullValue) {
-            super(name, indexed, stored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_ONLY, meta);
+            super(name, indexed, stored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, meta);
             this.scalingFactor = scalingFactor;
             this.nullValue = nullValue;
         }

server/src/internalClusterTest/java/org/elasticsearch/search/aggregations/bucket/SignificantTermsSignificanceScoreIT.java

@@ -556,15 +556,9 @@ public void testScriptCaching() throws Exception {
         ScriptHeuristic scriptHeuristic = new ScriptHeuristic(
             new Script(ScriptType.INLINE, "mockscript", "Math.random()", Collections.emptyMap())
         );
-        boolean useSigText = randomBoolean();
         SearchResponse r;
-        if (useSigText) {
-            r = client().prepareSearch("cache_test_idx").setSize(0)
-                    .addAggregation(significantText("foo", "s").significanceHeuristic(scriptHeuristic)).get();
-        } else {
-            r = client().prepareSearch("cache_test_idx").setSize(0)
-                    .addAggregation(significantTerms("foo").field("s").significanceHeuristic(scriptHeuristic)).get();
-        }
+        r = client().prepareSearch("cache_test_idx").setSize(0)
+                .addAggregation(significantTerms("foo").field("s").significanceHeuristic(scriptHeuristic)).get();
         assertSearchResponse(r);
 
         assertThat(client().admin().indices().prepareStats("cache_test_idx").setRequestCache(true).get().getTotal().getRequestCache()
@@ -574,14 +568,8 @@ public void testScriptCaching() throws Exception {
 
         // Test that a request using a deterministic script gets cached
         scriptHeuristic = getScriptSignificanceHeuristic();
-        useSigText = randomBoolean();
-        if (useSigText) {
-            r = client().prepareSearch("cache_test_idx").setSize(0)
-                    .addAggregation(significantText("foo", "s").significanceHeuristic(scriptHeuristic)).get();
-        } else {
-            r = client().prepareSearch("cache_test_idx").setSize(0)
-                    .addAggregation(significantTerms("foo").field("s").significanceHeuristic(scriptHeuristic)).get();
-        }
+        r = client().prepareSearch("cache_test_idx").setSize(0)
+                .addAggregation(significantTerms("foo").field("s").significanceHeuristic(scriptHeuristic)).get();
         assertSearchResponse(r);
 
         assertThat(client().admin().indices().prepareStats("cache_test_idx").setRequestCache(true).get().getTotal().getRequestCache()
@@ -590,11 +578,7 @@ public void testScriptCaching() throws Exception {
                 .getMissCount(), equalTo(1L));
 
         // Ensure that non-scripted requests are cached as normal
-        if (useSigText) {
-            r = client().prepareSearch("cache_test_idx").setSize(0).addAggregation(significantText("foo", "s")).get();
-        } else {
-            r = client().prepareSearch("cache_test_idx").setSize(0).addAggregation(significantTerms("foo").field("s")).get();
-        }
+        r = client().prepareSearch("cache_test_idx").setSize(0).addAggregation(significantTerms("foo").field("s")).get();
         assertSearchResponse(r);
 
         assertThat(client().admin().indices().prepareStats("cache_test_idx").setRequestCache(true).get().getTotal().getRequestCache()

server/src/main/java/org/elasticsearch/index/mapper/ConstantFieldType.java

@@ -42,7 +42,7 @@
 public abstract class ConstantFieldType extends MappedFieldType {
 
     public ConstantFieldType(String name, Map<String, String> meta) {
-        super(name, true, false, true, TextSearchInfo.SIMPLE_MATCH_ONLY, meta);
+        super(name, true, false, true, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, meta);
     }
 
     @Override

server/src/main/java/org/elasticsearch/index/mapper/DateFieldMapper.java

@@ -273,7 +273,7 @@ public static final class DateFieldType extends MappedFieldType {
         public DateFieldType(String name, boolean isSearchable, boolean isStored, boolean hasDocValues,
                              DateFormatter dateTimeFormatter, Resolution resolution, String nullValue,
                              Map<String, String> meta) {
-            super(name, isSearchable, isStored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_ONLY, meta);
+            super(name, isSearchable, isStored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, meta);
             this.dateTimeFormatter = dateTimeFormatter;
             this.dateMathParser = dateTimeFormatter.toDateMathParser();
             this.resolution = resolution;

server/src/main/java/org/elasticsearch/index/mapper/IpFieldMapper.java

@@ -133,7 +133,7 @@ public static final class IpFieldType extends SimpleMappedFieldType {
 
         public IpFieldType(String name, boolean indexed, boolean stored, boolean hasDocValues,
                            InetAddress nullValue, Map<String, String> meta) {
-            super(name, indexed, stored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_ONLY, meta);
+            super(name, indexed, stored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, meta);
             this.nullValue = nullValue;
         }
 

server/src/main/java/org/elasticsearch/index/mapper/MappedFieldType.java

@@ -406,9 +406,11 @@ public Map<String, String> meta() {
      * Returns information on how any text in this field is indexed
      *
      * Fields that do not support any text-based queries should return
-     * {@link TextSearchInfo#NONE}.  Some fields (eg numeric) may support
+     * {@link TextSearchInfo#NONE}.  Some fields (eg keyword) may support
      * only simple match queries, and can return
-     * {@link TextSearchInfo#SIMPLE_MATCH_ONLY}
+     * {@link TextSearchInfo#SIMPLE_MATCH_ONLY}; other fields may support
+     * simple match queries without using the terms index, and can return
+     * {@link TextSearchInfo#SIMPLE_MATCH_WITHOUT_TERMS}
      */
     public TextSearchInfo getTextSearchInfo() {
         return textSearchInfo;

server/src/main/java/org/elasticsearch/index/mapper/NumberFieldMapper.java

@@ -894,7 +894,7 @@ public static class NumberFieldType extends SimpleMappedFieldType {
 
         public NumberFieldType(String name, NumberType type, boolean isSearchable, boolean isStored,
                                boolean hasDocValues, boolean coerce, Number nullValue, Map<String, String> meta) {
-            super(name, isSearchable, isStored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_ONLY, meta);
+            super(name, isSearchable, isStored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, meta);
             this.type = Objects.requireNonNull(type);
             this.coerce = coerce;
             this.nullValue = nullValue;

server/src/main/java/org/elasticsearch/index/mapper/RangeFieldMapper.java

@@ -153,7 +153,7 @@ public static final class RangeFieldType extends MappedFieldType {
 
         public RangeFieldType(String name, RangeType type, boolean indexed, boolean stored,
                               boolean hasDocValues, boolean coerce, Map<String, String> meta) {
-            super(name, indexed, stored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_ONLY, meta);
+            super(name, indexed, stored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, meta);
             assert type != RangeType.DATE;
             this.rangeType = Objects.requireNonNull(type);
             dateTimeFormatter = null;
@@ -168,7 +168,7 @@ public RangeFieldType(String name, RangeType type) {
 
         public RangeFieldType(String name, boolean indexed, boolean stored,  boolean hasDocValues, DateFormatter formatter,
                               boolean coerce, Map<String, String> meta) {
-            super(name, indexed, stored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_ONLY, meta);
+            super(name, indexed, stored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, meta);
             this.rangeType = RangeType.DATE;
             this.dateTimeFormatter = Objects.requireNonNull(formatter);
             this.dateMathParser = dateTimeFormatter.toDateMathParser();

server/src/main/java/org/elasticsearch/index/mapper/SeqNoFieldMapper.java

@@ -96,7 +96,7 @@ static final class SeqNoFieldType extends SimpleMappedFieldType {
         private static final SeqNoFieldType INSTANCE = new SeqNoFieldType();
 
         private SeqNoFieldType() {
-            super(NAME, true, false, true, TextSearchInfo.SIMPLE_MATCH_ONLY, Collections.emptyMap());
+            super(NAME, true, false, true, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, Collections.emptyMap());
         }
 
         @Override

server/src/main/java/org/elasticsearch/index/mapper/TextSearchInfo.java

@@ -54,6 +54,13 @@ public class TextSearchInfo {
     public static final TextSearchInfo WHITESPACE_MATCH_ONLY
         = new TextSearchInfo(SIMPLE_MATCH_ONLY_FIELD_TYPE, null, Lucene.WHITESPACE_ANALYZER, Lucene.WHITESPACE_ANALYZER);
 
+    /**
+     * Defines indexing information for fields that support simple match text queries
+     * without using the terms index
+     */
+    public static final TextSearchInfo SIMPLE_MATCH_WITHOUT_TERMS
+        = new TextSearchInfo(SIMPLE_MATCH_ONLY_FIELD_TYPE, null, Lucene.KEYWORD_ANALYZER, Lucene.KEYWORD_ANALYZER);
+
     private static final NamedAnalyzer FORBIDDEN_ANALYZER = new NamedAnalyzer("", AnalyzerScope.GLOBAL,
         new Analyzer() {
             @Override

server/src/main/java/org/elasticsearch/index/mapper/VersionFieldMapper.java

@@ -42,7 +42,7 @@ static final class VersionFieldType extends MappedFieldType {
         public static final VersionFieldType INSTANCE = new VersionFieldType();
 
         private VersionFieldType() {
-            super(NAME, false, false, true, TextSearchInfo.SIMPLE_MATCH_ONLY, Collections.emptyMap());
+            super(NAME, false, false, true, TextSearchInfo.NONE, Collections.emptyMap());
         }
 
         @Override

server/src/main/java/org/elasticsearch/search/aggregations/bucket/terms/SignificantTextAggregatorFactory.java

@@ -32,6 +32,7 @@
 import org.elasticsearch.common.util.BytesRefHash;
 import org.elasticsearch.common.util.ObjectArray;
 import org.elasticsearch.index.mapper.MappedFieldType;
+import org.elasticsearch.index.mapper.TextSearchInfo;
 import org.elasticsearch.index.query.QueryBuilder;
 import org.elasticsearch.search.DocValueFormat;
 import org.elasticsearch.search.aggregations.Aggregator;
@@ -84,9 +85,9 @@ public SignificantTextAggregatorFactory(String name,
         // Note that if the field is unmapped (its field type is null), we don't fail,
         // and just use the given field name as a placeholder.
         this.fieldType = context.getFieldType(fieldName);
-        if (fieldType != null && fieldType.indexAnalyzer() == null) {
-            throw new IllegalArgumentException("Field [" + fieldType.name() + "] has no analyzer, but SignificantText " +
-                "requires an analyzed field");
+        if (fieldType != null && supportsAgg(fieldType) == false) {
+            throw new IllegalArgumentException("Field [" + fieldType.name() + "] of type ["
+                + fieldType.typeName() + "] does not support significant text aggregations");
         }
         this.indexedFieldName = fieldType != null ? fieldType.name() : fieldName;
         this.sourceFieldNames = sourceFieldNames == null ? new String[] { indexedFieldName } : sourceFieldNames;
@@ -98,6 +99,13 @@ public SignificantTextAggregatorFactory(String name,
         this.significanceHeuristic = significanceHeuristic;
     }
 
+    private static boolean supportsAgg(MappedFieldType ft) {
+        if (ft.getTextSearchInfo() == TextSearchInfo.NONE) {
+            return false;
+        }
+        return ft.getTextSearchInfo() != TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS;
+    }
+
     @Override
     protected Aggregator createInternal(SearchContext searchContext, Aggregator parent, CardinalityUpperBound cardinality,
                                         Map<String, Object> metadata) throws IOException {

server/src/test/java/org/elasticsearch/search/aggregations/bucket/terms/SignificantTextAggregatorTests.java

@@ -78,11 +78,7 @@ protected AggregationBuilder createAggBuilderForTypeTest(MappedFieldType fieldTy
 
     @Override
     protected List<ValuesSourceType> getSupportedValuesSourceTypes() {
-        // TODO it is likely accidental that SigText supports anything other than Bytes, and then only text fields
-        return List.of(CoreValuesSourceType.NUMERIC,
-            CoreValuesSourceType.BYTES,
-            CoreValuesSourceType.RANGE,
-            CoreValuesSourceType.GEOPOINT);
+        return List.of(CoreValuesSourceType.BYTES);
     }
 
     @Override

x-pack/plugin/mapper-unsigned-long/src/main/java/org/elasticsearch/xpack/unsignedlong/UnsignedLongFieldMapper.java

@@ -147,7 +147,7 @@ public UnsignedLongFieldType(
             Number nullValueFormatted,
             Map<String, String> meta
         ) {
-            super(name, indexed, isStored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_ONLY, meta);
+            super(name, indexed, isStored, hasDocValues, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, meta);
             this.nullValueFormatted = nullValueFormatted;
         }
 

x-pack/plugin/runtime-fields/src/main/java/org/elasticsearch/xpack/runtimefields/mapper/AbstractScriptFieldType.java

@@ -44,7 +44,7 @@ abstract class AbstractScriptFieldType<LeafFactory> extends MappedFieldType {
         TriFunction<String, Map<String, Object>, SearchLookup, LeafFactory> factory,
         Map<String, String> meta
     ) {
-        super(name, false, false, false, TextSearchInfo.SIMPLE_MATCH_ONLY, meta);
+        super(name, false, false, false, TextSearchInfo.SIMPLE_MATCH_WITHOUT_TERMS, meta);
         this.script = script;
         this.factory = factory;
     }