Add an option to replace _source field values with synthetic references

benchmarks/src/main/java/org/elasticsearch/benchmark/script/ScriptScoreBenchmark.java

@@ -30,8 +30,10 @@
 import org.elasticsearch.index.fielddata.IndexFieldDataCache;
 import org.elasticsearch.index.fielddata.IndexNumericFieldData;
 import org.elasticsearch.index.mapper.MappedFieldType;
+import org.elasticsearch.index.mapper.MappingLookup;
 import org.elasticsearch.index.mapper.NumberFieldMapper.NumberFieldType;
 import org.elasticsearch.index.mapper.NumberFieldMapper.NumberType;
+import org.elasticsearch.index.mapper.SourceFieldMetrics;
 import org.elasticsearch.indices.breaker.CircuitBreakerService;
 import org.elasticsearch.indices.breaker.NoneCircuitBreakerService;
 import org.elasticsearch.plugins.PluginsService;
@@ -89,7 +91,7 @@ public class ScriptScoreBenchmark {
     private final SearchLookup lookup = new SearchLookup(
         fieldTypes::get,
         (mft, lookup, fdo) -> mft.fielddataBuilder(FieldDataContext.noRuntimeFields("benchmark")).build(fieldDataCache, breakerService),
-        SourceProvider.fromStoredFields()
+        SourceProvider.fromLookup(MappingLookup.EMPTY, SourceFieldMetrics.NOOP)
     );
 
     @Param({ "expression", "metal", "painless_cast", "painless_def" })

plugins/mapper-murmur3/src/test/java/org/elasticsearch/index/mapper/murmur3/Murmur3FieldMapperTests.java

@@ -107,7 +107,8 @@ protected void assertFetch(MapperService mapperService, String field, Object val
         ValueFetcher nativeFetcher = ft.valueFetcher(searchExecutionContext, format);
         ParsedDocument doc = mapperService.documentMapper().parse(source);
         withLuceneIndex(mapperService, iw -> iw.addDocuments(doc.docs()), ir -> {
-            Source s = SourceProvider.fromStoredFields().getSource(ir.leaves().get(0), 0);
+            Source s = SourceProvider.fromLookup(mapperService.mappingLookup(), mapperService.getMapperMetrics().sourceFieldMetrics())
+                .getSource(ir.leaves().get(0), 0);
             docValueFetcher.setNextReader(ir.leaves().get(0));
             nativeFetcher.setNextReader(ir.leaves().get(0));
             List<Object> fromDocValues = docValueFetcher.fetchValues(s, 0, new ArrayList<>());

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

@@ -17,8 +17,10 @@
 import org.elasticsearch.index.IndexSettings;
 import org.elasticsearch.index.analysis.IndexAnalyzers;
 import org.elasticsearch.index.mapper.MapperService.MergeReason;
+import org.elasticsearch.search.builder.SearchSourceBuilder;
 import org.elasticsearch.xcontent.FilterXContentParserWrapper;
 import org.elasticsearch.xcontent.FlatteningXContentParser;
+import org.elasticsearch.xcontent.XContentLocation;
 import org.elasticsearch.xcontent.XContentParser;
 
 import java.io.IOException;
@@ -36,6 +38,8 @@
  * the lucene data structures and mappings to be dynamically created as the outcome of parsing a document.
  */
 public abstract class DocumentParserContext {
+    public record XContentPatch(String fullPath, XContentLocation location, int id) {}
+
     /**
      * Wraps a given context while allowing to override some of its behaviour by re-implementing some of the non final methods
      */
@@ -124,6 +128,7 @@ public int get() {
      * in this document and therefore is not present in mapping yet.
      */
     private final Set<String> copyToFields;
+    private final Map<XContentLocation, XContentPatch> sourcePatches;
 
     // Indicates if the source for this context has been cloned and gets parsed multiple times.
     private boolean clonedSource;
@@ -146,6 +151,7 @@ private DocumentParserContext(
         Set<String> fieldsAppliedFromTemplates,
         Set<String> copyToFields,
         DynamicMapperSize dynamicMapperSize,
+        Map<XContentLocation, XContentPatch> sourcePatches,
         boolean clonedSource
     ) {
         this.mappingLookup = mappingLookup;
@@ -165,6 +171,7 @@ private DocumentParserContext(
         this.fieldsAppliedFromTemplates = fieldsAppliedFromTemplates;
         this.copyToFields = copyToFields;
         this.dynamicMappersSize = dynamicMapperSize;
+        this.sourcePatches = sourcePatches;
         this.clonedSource = clonedSource;
     }
 
@@ -187,6 +194,7 @@ private DocumentParserContext(ObjectMapper parent, ObjectMapper.Dynamic dynamic,
             in.fieldsAppliedFromTemplates,
             in.copyToFields,
             in.dynamicMappersSize,
+            in.sourcePatches,
             in.clonedSource
         );
     }
@@ -216,6 +224,7 @@ protected DocumentParserContext(
             new HashSet<>(),
             new HashSet<>(mappingLookup.fieldTypesLookup().getCopyToDestinationFields()),
             new DynamicMapperSize(),
+            new HashMap<>(),
             false
         );
     }
@@ -300,6 +309,36 @@ public final void addToFieldNames(String field) {
         }
     }
 
+    /**
+     * Registers a patch for a specific field at a given location in the original source.
+     * The patch modifies the field value in the original `_source` by replacing it with a numerical ID,
+     * which is then stored in a document field accessible via {@link SearchSourceBuilder#equals(Object)}.
+     *
+     * The original field value must still be retrievable when the original `_source` is requested.
+     * This responsibility lies with the {@link FieldMapper}, ensuring that the patched field can provide
+     * the original value during retrieval.
+     *
+     * Note: Only one patch per field is permitted for each document.
+     *
+     * @param fieldMapper The {@link FieldMapper} responsible for applying the patch to the field.
+     * @param xContentLocation The {@link XContentLocation} representing the location of the field in the original source.
+     */
+    public void addSourceFieldPatch(FieldMapper fieldMapper, XContentLocation xContentLocation) {
+        var sourceFieldMapper = (SourceFieldMapper) getMetadataMapper(SourceFieldMapper.NAME);
+        if (sourceFieldMapper == null) {
+            return;
+        }
+        sourceFieldMapper.indexFieldPatch(doc(), fieldMapper, xContentLocation, getSourcePatches());
+    }
+
+    /**
+     * Returns all {@link XContentPatch} currently registered in this context,
+     * mapped by their {@link XContentLocation} in the original `_source`.
+     */
+    public Map<XContentLocation, XContentPatch> getSourcePatches() {
+        return sourcePatches;
+    }
+
     public final Field version() {
         return this.version;
     }

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

@@ -18,6 +18,7 @@
 import org.elasticsearch.index.IndexVersions;
 import org.elasticsearch.xcontent.ToXContentFragment;
 import org.elasticsearch.xcontent.XContentBuilder;
+import org.elasticsearch.xcontent.XContentLocation;
 
 import java.io.IOException;
 import java.util.Arrays;
@@ -175,6 +176,16 @@ public SourceLoader.SyntheticFieldLoader syntheticFieldLoader() {
         throw new IllegalArgumentException("field [" + fullPath() + "] of type [" + typeName() + "] doesn't support synthetic source");
     }
 
+    /**
+     * Creates a {@link SourceLoader.PatchFieldLoader} to load patches that were previously indexed
+     * with {@link DocumentParserContext#addSourceFieldPatch(FieldMapper, XContentLocation)}.
+     *
+     * Returns {@code null} if the field doesn't allow patching.
+     */
+    protected SourceLoader.PatchFieldLoader patchFieldLoader() {
+        return null;
+    }
+
     @Override
     public String toString() {
         return Strings.toString(this);

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

@@ -131,6 +131,10 @@ public SourceLoader.SyntheticFieldLoader syntheticFieldLoader() {
         return root.syntheticFieldLoader(stream);
     }
 
+    protected SourceLoader.PatchFieldLoader patchFieldLoader() {
+        return root.patchFieldLoader();
+    }
+
     /**
      * Merges a new mapping into the existing one.
      *

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

@@ -10,6 +10,7 @@
 package org.elasticsearch.index.mapper;
 
 import org.apache.lucene.index.LeafReader;
+import org.apache.lucene.search.CheckedIntConsumer;
 import org.apache.lucene.search.DocIdSetIterator;
 import org.apache.lucene.search.IndexSearcher;
 import org.apache.lucene.search.Query;
@@ -391,6 +392,27 @@ protected MapperMergeContext createChildContext(MapperMergeContext mapperMergeCo
         );
     }
 
+    @Override
+    protected SourceLoader.PatchFieldLoader patchFieldLoader() {
+        var patchLoader = super.patchFieldLoader();
+        if (patchLoader == null) {
+            return null;
+        }
+        return context -> {
+            var patchFieldLoader = patchLoader.leaf(context);
+            if (patchFieldLoader == null) {
+                return null;
+            }
+            IndexSearcher searcher = new IndexSearcher(context.reader());
+            searcher.setQueryCache(null);
+            var childScorer = searcher.createWeight(nestedTypeFilter, ScoreMode.COMPLETE_NO_SCORES, 1f).scorer(context);
+            var parentsDocs = bitsetProducer.apply(parentTypeFilter).getBitSet(context);
+            return (doc, acc) -> {
+                collectChildren(nestedTypePath, doc, parentsDocs, childScorer.iterator(), child -> patchFieldLoader.load(child, acc));
+            };
+        };
+    }
+
     @Override
     public SourceLoader.SyntheticFieldLoader syntheticFieldLoader() {
         if (storeArraySource()) {
@@ -444,29 +466,15 @@ public DocValuesLoader docValuesLoader(LeafReader leafReader, int[] docIdsInLeaf
             if (childScorer != null) {
                 var parentDocs = parentBitSetProducer.get().getBitSet(leafReader.getContext());
                 return parentDoc -> {
-                    collectChildren(parentDoc, parentDocs, childScorer.iterator());
+                    children.clear();
+                    collectChildren(nestedTypePath, parentDoc, parentDocs, childScorer.iterator(), child -> children.add(child));
                     return children.size() > 0;
                 };
             } else {
                 return parentDoc -> false;
             }
         }
 
-        private List<Integer> collectChildren(int parentDoc, BitSet parentDocs, DocIdSetIterator childIt) throws IOException {
-            assert parentDoc < 0 || parentDocs.get(parentDoc) : "wrong context, doc " + parentDoc + " is not a parent of " + nestedTypePath;
-            final int prevParentDoc = parentDoc > 0 ? parentDocs.prevSetBit(parentDoc - 1) : -1;
-            int childDocId = childIt.docID();
-            if (childDocId <= prevParentDoc) {
-                childDocId = childIt.advance(prevParentDoc + 1);
-            }
-
-            children.clear();
-            for (; childDocId < parentDoc; childDocId = childIt.nextDoc()) {
-                children.add(childDocId);
-            }
-            return children;
-        }
-
         @Override
         public boolean hasValue() {
             return children.size() > 0;
@@ -494,4 +502,24 @@ public String fieldName() {
             return NestedObjectMapper.this.fullPath();
         }
     }
+
+    private static void collectChildren(
+        String nestedTypePath,
+        int parentDoc,
+        BitSet parentDocs,
+        DocIdSetIterator childIt,
+        CheckedIntConsumer<IOException> childConsumer
+    ) throws IOException {
+        assert parentDoc < 0 || parentDocs.get(parentDoc) : "wrong context, doc " + parentDoc + " is not a parent of " + nestedTypePath;
+        final int prevParentDoc = parentDoc > 0 ? parentDocs.prevSetBit(parentDoc - 1) : -1;
+        int childDocId = childIt.docID();
+        if (childDocId <= prevParentDoc) {
+            childDocId = childIt.advance(prevParentDoc + 1);
+        }
+
+        for (; childDocId < parentDoc; childDocId = childIt.nextDoc()) {
+            childConsumer.accept(childDocId);
+        }
+    }
+
 }

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

@@ -37,6 +37,7 @@
 import java.util.Objects;
 import java.util.Optional;
 import java.util.TreeMap;
+import java.util.stream.Collectors;
 import java.util.stream.Stream;
 
 public class ObjectMapper extends Mapper {
@@ -808,6 +809,31 @@ protected void doXContent(XContentBuilder builder, Params params) throws IOExcep
 
     }
 
+    @Override
+    protected SourceLoader.PatchFieldLoader patchFieldLoader() {
+        var loaders = mappers.values().stream().map(m -> m.patchFieldLoader()).filter(l -> l != null).collect(Collectors.toList());
+        if (loaders.isEmpty()) {
+            return null;
+        }
+        return context -> {
+            final List<SourceLoader.PatchFieldLoader.Leaf> leaves = new ArrayList<>();
+            for (var loader : loaders) {
+                var leaf = loader.leaf(context);
+                if (leaf != null) {
+                    leaves.add(leaf);
+                }
+            }
+            if (leaves.isEmpty()) {
+                return null;
+            }
+            return (doc, acc) -> {
+                for (var leaf : leaves) {
+                    leaf.load(doc, acc);
+                }
+            };
+        };
+    }
+
     protected SourceLoader.SyntheticFieldLoader syntheticFieldLoader(Stream<Mapper> mappers, boolean isFragment) {
         var fields = mappers.sorted(Comparator.comparing(Mapper::fullPath))
             .map(Mapper::syntheticFieldLoader)

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

@@ -0,0 +1,88 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the "Elastic License
+ * 2.0", the "GNU Affero General Public License v3.0 only", and the "Server Side
+ * Public License v 1"; you may not use this file except in compliance with, at
+ * your election, the "Elastic License 2.0", the "GNU Affero General Public
+ * License v3.0 only", or the "Server Side Public License, v 1".
+ */
+
+package org.elasticsearch.index.mapper;
+
+import org.elasticsearch.common.bytes.BytesReference;
+import org.elasticsearch.common.io.stream.BytesStreamOutput;
+import org.elasticsearch.common.xcontent.XContentHelper;
+import org.elasticsearch.xcontent.XContent;
+import org.elasticsearch.xcontent.XContentBuilder;
+import org.elasticsearch.xcontent.XContentGenerator;
+import org.elasticsearch.xcontent.XContentParser;
+import org.elasticsearch.xcontent.XContentParserConfiguration;
+
+import java.io.IOException;
+import java.util.Set;
+
+public class PatchSourceUtils {
+    public interface CheckedTriConsumer<S, T, U> {
+        void apply(S s, T t, U u) throws IOException;
+    }
+
+    /**
+     * Parses the given {@code source} and returns a new version with all values referenced by the
+     * provided {@code patchFullPaths} replaced using the {@code patchApply} consumer.
+     */
+    public static BytesReference patchSource(
+        BytesReference source,
+        XContent xContent,
+        Set<String> patchFullPaths,
+        CheckedTriConsumer<String, XContentParser, XContentGenerator> patchApply
+    ) throws IOException {
+        BytesStreamOutput streamOutput = new BytesStreamOutput();
+        XContentBuilder builder = new XContentBuilder(xContent, streamOutput);
+        try (XContentParser parser = XContentHelper.createParserNotCompressed(XContentParserConfiguration.EMPTY, source, xContent.type())) {
+            if ((parser.currentToken() == null) && (parser.nextToken() == null)) {
+                return source;
+            }
+            parseAndPatchSource(builder.generator(), parser, "", patchFullPaths, patchApply);
+            return BytesReference.bytes(builder);
+        }
+    }
+
+    private static void parseAndPatchSource(
+        XContentGenerator destination,
+        XContentParser parser,
+        String fullPath,
+        Set<String> patchFullPaths,
+        CheckedTriConsumer<String, XContentParser, XContentGenerator> patchApply
+    ) throws IOException {
+        XContentParser.Token token = parser.currentToken();
+        if (token == XContentParser.Token.FIELD_NAME) {
+            String fieldName = parser.currentName();
+            destination.writeFieldName(fieldName);
+            token = parser.nextToken();
+            fullPath = fullPath + (fullPath.isEmpty() ? "" : ".") + fieldName;
+            if (patchFullPaths.contains(fullPath)) {
+                patchApply.apply(fullPath, parser, destination);
+                return;
+            }
+        }
+
+        switch (token) {
+            case START_ARRAY -> {
+                destination.writeStartArray();
+                while (parser.nextToken() != XContentParser.Token.END_ARRAY) {
+                    parseAndPatchSource(destination, parser, fullPath, patchFullPaths, patchApply);
+                }
+                destination.writeEndArray();
+            }
+            case START_OBJECT -> {
+                destination.writeStartObject();
+                while (parser.nextToken() != XContentParser.Token.END_OBJECT) {
+                    parseAndPatchSource(destination, parser, fullPath, patchFullPaths, patchApply);
+                }
+                destination.writeEndObject();
+            }
+            default -> // others are simple:
+                destination.copyCurrentEvent(parser);
+        }
+    }
+}