Update sparse_vector field mapping to include default setting for token pruning #126739

markjhoy · 2025-04-12T01:05:31Z

Updates the SparseVectorFieldMapper type to include index options for pruning tokens and associated configuration values.

Before this update, token pruning for sparse vector types is only available via the query (see parameters for the sparse vector query ).

With this PR, by default, any new indices with a sparse_vector field type will by default have token pruning turned on.

Example:

{
  "properties": {
    "example_field": {
       "type": "sparse_vector",
        "index_options": {
          "prune": (boolean, default is `true`),
          "pruning_config": {
            "tokens_freq_ratio_threshold": (integer, range 1-100, default is 5),
            "tokens_weight_threshold": (double, range 0.0-1.0, default if 0.4)
          }
        }
     }
  }
}

elasticsearchmachine · 2025-05-06T17:19:36Z

Pinging @elastic/search-eng (Team:SearchOrg)

elasticsearchmachine · 2025-05-06T17:19:37Z

Pinging @elastic/search-relevance (Team:Search - Relevance)

Mikep86

Partial review, nice start! Will take me a bit to work through this as it's a hefty PR!

Mikep86 · 2025-05-06T20:36:04Z

rest-api-spec/src/yamlRestTest/resources/rest-api-spec/test/cluster.stats/10_basic.yml

+  - do:
+      indices.delete:
+        index: ["test1", "test2"]
+        ignore: 404


Apologies if this has been covered already, but why do we need to explicitly delete the indices? I thought this was handled automatically between YAML test suites.

It should - but there's a number of tests that try to create indices with the same names, and we've found that it can be flaky sometimes so, best to ensure their removal with each individual test.

Mikep86 · 2025-05-06T20:44:56Z

server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

+    public static final IndexVersion SPARSE_VECTOR_PRUNING_INDEX_OPTIONS_VERSION =
+        IndexVersions.SPARSE_VECTOR_PRUNING_INDEX_OPTIONS_SUPPORT;


Nit: can we make this package private like the other index versions?

It's used in the SparseVectorQueryBuilder to ensure we use the same index version... we could add the same variable there, but, I think this would be more consistent.

server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

Mikep86 · 2025-05-06T21:10:37Z

server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

+        }
+
+        public static Boolean parseIndexOptionsPruneValue(Map<String, Object> indexOptionsMap) {
+            Object shouldPrune = indexOptionsMap.remove(IndexOptions.PRUNE_FIELD_NAME);


In general I'm not a fan of mutating the passed-in map. We accept it elsewhere because it's baked in at this point, but IMO we shouldn't do it for new implementations.

Mikep86 · 2025-05-06T21:12:03Z

server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

+        }
+
+        public static TokenPruningConfig parseIndexOptionsPruningConfig(Boolean prune, Map<String, Object> indexOptionsMap) {
+            Object pruningConfiguration = indexOptionsMap.remove(IndexOptions.PRUNING_CONFIG_FIELD_NAME);


server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

elasticsearchmachine · 2025-05-07T13:00:30Z

Hi @markjhoy, I've created a changelog YAML for you.

Mikep86

Got further in my partial review. Found some bugs and some areas where we can simplify.

Mikep86 · 2025-05-07T13:16:35Z

docs/changelog/126739.yaml

+pr: 126739
+summary: Update `sparse_vector` field mapping to include default setting for token
+  pruning
+area: Relevance


Mapping may be a better area for this PR

Mikep86 · 2025-05-07T13:24:34Z

server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

+            }
+
+            if (prune == null || prune == false) {
+                throw new MapperParsingException("[index_options] field [pruning_config] should only be set if [prune] is set to true");


Nit: better to use variables to build the string ( i.e PRUNE_FIELD_NAME, PRUNING_CONFIG_FIELD_NAME)

Mikep86 · 2025-05-07T13:29:59Z

server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

+                (b, n, v) -> {
+                    if (v != null) {
+                        b.field(n, v);
+                    }
+                },


Nit: I think we can simplify this to XContentBuilder::field

Mikep86 · 2025-05-07T13:41:20Z

server/src/main/java/org/elasticsearch/index/mapper/vectors/TokenPruningConfig.java

+    public static TokenPruningConfig parseFromMap(Map<String, Object> pruningConfigMap) {
+        Object mappedTokensFreqRatioThreshold = pruningConfigMap.remove(TOKENS_FREQ_RATIO_THRESHOLD.getPreferredName());
+        Object mappedTokensWeightThreshold = pruningConfigMap.remove(TOKENS_WEIGHT_THRESHOLD.getPreferredName());
+        Object mappedOnlyScorePrunedTokens = pruningConfigMap.remove(ONLY_SCORE_PRUNED_TOKENS_FIELD.getPreferredName());
+
+        if (pruningConfigMap.isEmpty() == false) {
+            throw new MapperParsingException("[" + PRUNING_CONFIG_FIELD + "] has unknown fields");
+        }
+
+        Float tokensFreqRatioThreshold = parseTokensFreqRatioThreshold(mappedTokensFreqRatioThreshold);
+        Float tokensWeightThreshold = parseTokensWeightThreshold(mappedTokensWeightThreshold);
+        boolean onlyScorePrunedTokens = mappedOnlyScorePrunedTokens != null ? parseScorePrunedTokens(mappedOnlyScorePrunedTokens) : false;
+
+        if (tokensFreqRatioThreshold != null || tokensWeightThreshold != null) {
+            return new TokenPruningConfig(tokensFreqRatioThreshold, tokensWeightThreshold, onlyScorePrunedTokens);
+        }
+
+        return null;
+    }
+
+    private static Float parseFloatNumberFromObject(Object numberObject, String fieldName, String exceptionDetails) {
+        if (numberObject instanceof Integer intValue) {
+            return (float) intValue;
+        } else if (numberObject instanceof Float floatValue) {
+            return floatValue;
+        } else if (numberObject instanceof Double doubleValue) {
+            return ((Double) numberObject).floatValue();
+        }
+
+        throw new MapperParsingException("[" + PRUNING_CONFIG_FIELD + "] field [" + fieldName + "]" + exceptionDetails);
+    }
+
+    private static Float parseTokensWeightThreshold(Object mappedTokensWeightThreshold) {
+        if (mappedTokensWeightThreshold == null) {
+            return DEFAULT_TOKENS_WEIGHT_THRESHOLD;
+        }
+
+        Float tokensWeightThreshold = parseFloatNumberFromObject(
+            mappedTokensWeightThreshold,
+            TOKENS_WEIGHT_THRESHOLD.getPreferredName(),
+            "field should be a number between 0.0 and 1.0"
+        );
+
+        if (tokensWeightThreshold < MIN_TOKENS_WEIGHT_THRESHOLD || tokensWeightThreshold > MAX_TOKENS_WEIGHT_THRESHOLD) {
+            throw new MapperParsingException(
+                "["
+                    + PRUNING_CONFIG_FIELD
+                    + "] field ["
+                    + TOKENS_WEIGHT_THRESHOLD.getPreferredName()
+                    + "] field should be a number between 0.0 and 1.0"
+            );
+        }
+        return tokensWeightThreshold;
+    }
+
+    private static Float parseTokensFreqRatioThreshold(Object mappedTokensFreqRatioThreshold) {
+        if (mappedTokensFreqRatioThreshold == null) {
+            return DEFAULT_TOKENS_FREQ_RATIO_THRESHOLD;
+        }
+
+        Float tokensFreqRatioThreshold = parseFloatNumberFromObject(
+            mappedTokensFreqRatioThreshold,
+            TOKENS_FREQ_RATIO_THRESHOLD.getPreferredName(),
+            "field should be a number between 1 and 100"
+        );
+
+        if (tokensFreqRatioThreshold < MIN_TOKENS_FREQ_RATIO_THRESHOLD || tokensFreqRatioThreshold > MAX_TOKENS_FREQ_RATIO_THRESHOLD) {
+            throw new MapperParsingException(
+                "["
+                    + PRUNING_CONFIG_FIELD
+                    + "] field ["
+                    + TOKENS_FREQ_RATIO_THRESHOLD.getPreferredName()
+                    + "] field should be a number between 1 and 100"
+            );
+        }
+
+        return tokensFreqRatioThreshold;
+    }
+
+    private static boolean parseScorePrunedTokens(Object mappedScorePrunedTokens) {
+        if (mappedScorePrunedTokens == null) {
+            return false;
+        }
+
+        if (mappedScorePrunedTokens instanceof Boolean boolValue) {
+            return boolValue;
+        }
+
+        throw new MapperParsingException(
+            "[" + PRUNING_CONFIG_FIELD + "] field [" + ONLY_SCORE_PRUNED_TOKENS_FIELD.getPreferredName() + "] field should be true or false"
+        );
+    }


A lot of this logic seems to be duplicating that provided by ConstructingObjectParser. I suggest we refactor this to use that instead. See here for an example:

elasticsearch/x-pack/plugin/inference/src/main/java/org/elasticsearch/xpack/inference/mapper/SemanticTextField.java

Lines 113 to 129 in e7d4a28

static MinimalServiceSettings parseModelSettingsFromMap(Object node) {

if (node == null) {

return null;

}

try {

Map<String, Object> map = XContentMapValues.nodeMapValue(node, MODEL_SETTINGS_FIELD);

XContentParser parser = new MapXContentParser(

NamedXContentRegistry.EMPTY,

DeprecationHandler.IGNORE_DEPRECATIONS,

map,

XContentType.JSON

);

return MinimalServiceSettings.parse(parser);

} catch (Exception exc) {

throw new ElasticsearchException(exc);

}

}

Mikep86 · 2025-05-07T13:48:50Z

server/src/test/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapperTests.java

+        assertEquals(2, fields.size());
+        assertThat(fields.get(0), Matchers.instanceOf(XFeatureField.class));
+        XFeatureField featureField1 = null;
+        XFeatureField featureField2 = null;
+        for (IndexableField field : fields) {
+            if (field.stringValue().equals("ten")) {
+                featureField1 = (XFeatureField) field;
+            } else if (field.stringValue().equals("twenty")) {
+                featureField2 = (XFeatureField) field;
+            } else {
+                throw new UnsupportedOperationException();
+            }
+        }
+
+        int freq1 = getFrequency(featureField1.tokenStream(null, null));
+        int freq2 = getFrequency(featureField2.tokenStream(null, null));
+        assertTrue(freq1 < freq2);


Could we factor this check into a helper method instead duplicating the code?

Mikep86 · 2025-05-07T14:42:22Z

...ugin/core/src/main/java/org/elasticsearch/xpack/core/ml/search/SparseVectorQueryBuilder.java

+            if (shouldPruneTokens == null && indexPruningSettings.prune != null && indexPruningSettings.prune) {
+                doPruneTokens = true;
+            }


There's a bug here: what if the query sets shouldPruneTokens to true without setting the pruning config? By my reading of this method, doPruneTokens will end up false in this case.

Mikep86 · 2025-05-07T14:46:39Z

...ugin/core/src/main/java/org/elasticsearch/xpack/core/ml/search/SparseVectorQueryBuilder.java

+    private TokenPruningSet setPruningConfigFromIndexIfNeeded(
+        Boolean queryPruneTokens,
+        TokenPruningConfig queryPruningConfig,
+        SparseVectorFieldMapper fieldMapper
+    ) {
+        boolean doPruneTokens = false;
+        TokenPruningConfig setTokenPruningConfig = queryPruningConfig;
+        if (queryPruneTokens == null || queryPruningConfig == null) {
+            IndexFieldPruningSettings indexPruningSettings = getIndexFieldPruningSettings(fieldMapper);
+            if (shouldPruneTokens == null && indexPruningSettings.prune != null && indexPruningSettings.prune) {
+                doPruneTokens = true;
+            }
+            if (setTokenPruningConfig == null && indexPruningSettings.pruningConfig != null) {
+                setTokenPruningConfig = indexPruningSettings.pruningConfig;
+            }
+        } else {
+            doPruneTokens = queryPruneTokens;
+        }
+
+        return new TokenPruningSet(doPruneTokens, setTokenPruningConfig);
+    }


This method is hard to read and contains at least one bug (see #126739 (comment)). For example, queryPruneTokens is always set to shouldPruneTokens, yet shouldPruneTokens is also referenced in this method.

IMO we should refactor it to just return the pruning settings from the mapping and then handle query-time overrides where it is called.

Mikep86 · 2025-05-07T14:57:19Z

server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

+    public static final IndexVersion SPARSE_VECTOR_PRUNING_INDEX_OPTIONS_VERSION =
+        IndexVersions.SPARSE_VECTOR_PRUNING_INDEX_OPTIONS_SUPPORT;
+
+    private final SparseVectorFieldMapper.IndexOptions indexOptions;


There's no need to store indexOptions in the mapper, we're already storing it in the field type

Mikep86 · 2025-05-07T14:57:57Z

server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

                builderParams(this, context)
            );
        }
    }

+    public IndexOptions getIndexOptions() {
+        return this.indexOptions;


You can get the index options by calling fieldType().getIndexOptions() instead

Mikep86 · 2025-05-07T15:04:47Z

...ugin/core/src/main/java/org/elasticsearch/xpack/core/ml/search/SparseVectorQueryBuilder.java

+            // if the query options for pruning are not set,
+            // we need to check the index options for this field
+            // and use those if set.
+            SparseVectorFieldMapper sparseVectorFieldMapper = getSparseVectorFieldMapperForQueryRewrite(fieldName, queryRewriteContext);
+            TokenPruningSet pruningOptions = setPruningConfigFromIndexIfNeeded(
+                shouldPruneTokens,
+                tokenPruningConfig,
+                sparseVectorFieldMapper
+            );


Why are we parsing the mapping pruning options here as well as in doToQuery? Unless I'm missing something, it makes sense to parse the mapping pruning options as late as possible, which is in doToQuery.

Also, this logic is not going to act the way you think it will in a multi-node environment. Part of the query rewrite cycle happens on the coordinator node, which doesn't have access to the mappings. You're going to get an NPE here in that case. Another reason to remove this logic and do it only in doToQuery, where you are guaranteed to have access to mappings.

Mikep86

Got through everything except the documentation, nice work! Some bug fixes, simplification, and more testing coverage is required, but the high-level functionality looks good 👍

Mikep86 · 2025-05-07T15:30:29Z

...core/src/test/java/org/elasticsearch/xpack/core/ml/search/SparseVectorQueryBuilderTests.java

+    public void testWeCorrectlyRewriteQueryIntoVectorsWithIndexOptions() {
+        wrapTestSparseVectorIndexOptions((c) -> {
+            SearchExecutionContext searchExecutionContext = createSearchExecutionContext();
+
+            TokenPruningConfig defaultTokenPruningConfig = new TokenPruningConfig(12, 0.6f, false);
+
+            SparseVectorQueryBuilder queryBuilder = createTestQueryBuilder(null);
+            QueryBuilder rewrittenQueryBuilder = rewriteAndFetch(queryBuilder, searchExecutionContext);
+            assertTrue(rewrittenQueryBuilder instanceof SparseVectorQueryBuilder);
+            assertEquals(queryBuilder.shouldPruneTokens(), ((SparseVectorQueryBuilder) rewrittenQueryBuilder).shouldPruneTokens());
+            assertNotNull(((SparseVectorQueryBuilder) rewrittenQueryBuilder).getQueryVectors());
+        });
+    }
+
+    private void wrapTestSparseVectorIndexOptions(Consumer<Boolean> testMethod) {
+        testWithSparseVectorFieldIndexOptions = true;
+        try {
+            testMethod.accept(true);
+        } finally {
+            testWithSparseVectorFieldIndexOptions = false;
+        }
+    }


Rather than doing this test wrapping approach, I recommend randomizing the value of testWithSparseVectorFieldIndexOptions and using doAssertLuceneQuery to validate the results. See how we randomize inferenceResultType in SemanticQueryBuilderTests as an example.

Mikep86 · 2025-05-07T15:31:16Z

...core/src/test/java/org/elasticsearch/xpack/core/ml/search/SparseVectorQueryBuilderTests.java

+    private void addSparseVectorIndexOptionsMapping(MapperService mapperService) throws IOException {
+        String addIndexOptionsTemplate = "{\"properties\":{\""
+            + SPARSE_VECTOR_FIELD
+            + "\":{\"type\":\"sparse_vector\",\"index_options\""
+            + ":{\"prune\":true,\"pruning_config\":{\"tokens_freq_ratio_threshold\""
+            + ":12,\"tokens_weight_threshold\":0.6}}}}}";
+        mapperService.merge("_doc", new CompressedXContent(addIndexOptionsTemplate), MapperService.MergeReason.MAPPING_UPDATE);
+    }


Can we modify the string passed to mapperService.merge in initializeAdditionalMappings instead of duplicating the logic?

Mikep86 · 2025-05-07T15:39:41Z

...core/src/test/java/org/elasticsearch/xpack/core/ml/search/SparseVectorQueryBuilderTests.java

We need more tests here to cover query-time overrides of token pruning behavior, including partial overrides (like if only prune: true without a pruning config is specified in the query). We shouldn't use randomization to test this, however we need to implement it with the randomization of the test setup in mind. In other words, we should deterministically set the query-time overrides, but the expected result will be dependent on the random value of testWithSparseVectorFieldIndexOptions.

Mikep86 · 2025-05-07T15:44:19Z

server/src/main/java/org/elasticsearch/index/mapper/vectors/SparseVectorFieldMapper.java

+
+    private final SparseVectorFieldMapper.IndexOptions indexOptions;
+
+    public static final NodeFeature SPARSE_VECTOR_INDEX_OPTIONS_FEATURE = new NodeFeature("sparse_vector_index_options_supported");


Tiniest of nits here, but we normally pseudo-namespace node features using .. In this case, I think sparse_vector.index_options_supported would work well, where sparse_vector. is the pseudo-namespace.

Mikep86 · 2025-05-07T15:48:47Z

...tests-with-security/src/test/resources/rest-api-spec/test/multi_cluster/50_sparse_vector.yml

+              embeddings:
+                type: sparse_vector


What's the purpose of the embeddings field in these tests?

elasticsearchmachine · 2025-05-07T16:42:41Z

Pinging @elastic/es-search-foundations (Team:Search Foundations)

Initial checkin - needs tests

e02cd3a

elasticsearchmachine added the v9.1.0 label Apr 12, 2025

markjhoy added 2 commits April 11, 2025 21:34

Missing s in IndexVersions

e24ab76

add changelog and docs for index_options

f39b78a

github-actions bot deployed to docs-preview April 15, 2025 00:50 View deployment

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

eeebfd8

github-actions bot deployed to docs-preview April 15, 2025 00:56 View deployment

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

51aab0c

github-actions bot deployed to docs-preview April 21, 2025 13:29 View deployment

correct index version

983ddf1

github-actions bot deployed to docs-preview April 21, 2025 14:06 View deployment

update tests

9545a0c

github-actions bot deployed to docs-preview April 21, 2025 17:53 View deployment

Complete tests for SparseVectorFieldMapper

19fe72d

github-actions bot deployed to docs-preview April 22, 2025 14:22 View deployment

[CI] Auto commit changes from spotless

58f9909

github-actions bot deployed to docs-preview April 22, 2025 14:32 View deployment

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

5f8e7b9

github-actions bot deployed to docs-preview April 25, 2025 12:34 View deployment

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

d7d27ba

github-actions bot deployed to docs-preview April 25, 2025 12:47 View deployment

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

d342656

github-actions bot deployed to docs-preview April 25, 2025 18:51 View deployment

fix docs

96096ba

github-actions bot deployed to docs-preview April 25, 2025 19:01 View deployment

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

eed88c6

github-actions bot deployed to docs-preview April 25, 2025 20:11 View deployment

fix lint

6a6052a

github-actions bot deployed to docs-preview April 25, 2025 21:03 View deployment

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

f977ea8

fix docs; add backport 8.x index version

5bb6561

markjhoy requested a review from Samiul-TheSoccerFan May 6, 2025 15:12

github-actions bot deployed to docs-preview May 6, 2025 15:13 View deployment

markjhoy added Team:SearchOrg Meta label for the Search Org (Enterprise Search) :SearchOrg/Relevance Label for the Search (solution/org) Relevance team labels May 6, 2025

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

e4d547a

github-actions bot deployed to docs-preview May 6, 2025 17:21 View deployment

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

6c5e253

github-actions bot deployed to docs-preview May 6, 2025 18:27 View deployment

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

29dae8a

github-actions bot deployed to docs-preview May 6, 2025 20:43 View deployment

Mikep86 reviewed May 6, 2025

View reviewed changes

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

31f9e6d

github-actions bot deployed to docs-preview May 6, 2025 21:40 View deployment

cleanups/optimizations

0b1c1d2

github-actions bot deployed to docs-preview May 6, 2025 23:46 View deployment

markjhoy requested a review from Mikep86 May 6, 2025 23:46

markjhoy added the >enhancement label May 7, 2025

Merge branch 'main' into markjhoy/default_token_pruning_sparse_vector

b48deea

Update docs/changelog/126739.yaml

7f60eca

github-actions bot deployed to docs-preview May 7, 2025 13:01 View deployment

refactor equals for IndexOptions

d8f3c63

github-actions bot deployed to docs-preview May 7, 2025 14:32 View deployment

Mikep86 reviewed May 7, 2025

View reviewed changes

markjhoy added the :Search Foundations/Mapping Index mappings, including merging and defining field types label May 7, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Update sparse_vector field mapping to include default setting for token pruning #126739

Update sparse_vector field mapping to include default setting for token pruning #126739

markjhoy commented Apr 12, 2025 •

edited

Loading

elasticsearchmachine commented May 6, 2025

elasticsearchmachine commented May 6, 2025

Mikep86 left a comment

Mikep86 May 6, 2025

markjhoy May 6, 2025

Mikep86 May 6, 2025

markjhoy May 6, 2025

Mikep86 May 6, 2025

Mikep86 May 6, 2025

elasticsearchmachine commented May 7, 2025

Mikep86 left a comment

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 May 7, 2025 •

edited

Loading

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 left a comment

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 May 7, 2025

Mikep86 May 7, 2025

elasticsearchmachine commented May 7, 2025

		public static final IndexVersion SPARSE_VECTOR_PRUNING_INDEX_OPTIONS_VERSION =
		IndexVersions.SPARSE_VECTOR_PRUNING_INDEX_OPTIONS_SUPPORT;

	static MinimalServiceSettings parseModelSettingsFromMap(Object node) {
	if (node == null) {
	return null;
	}
	try {
	Map<String, Object> map = XContentMapValues.nodeMapValue(node, MODEL_SETTINGS_FIELD);
	XContentParser parser = new MapXContentParser(
	NamedXContentRegistry.EMPTY,
	DeprecationHandler.IGNORE_DEPRECATIONS,
	map,
	XContentType.JSON
	);
	return MinimalServiceSettings.parse(parser);
	} catch (Exception exc) {
	throw new ElasticsearchException(exc);
	}
	}


		private final SparseVectorFieldMapper.IndexOptions indexOptions;

		public static final NodeFeature SPARSE_VECTOR_INDEX_OPTIONS_FEATURE = new NodeFeature("sparse_vector_index_options_supported");

Update sparse_vector field mapping to include default setting for token pruning #126739

Are you sure you want to change the base?

Update sparse_vector field mapping to include default setting for token pruning #126739

Conversation

markjhoy commented Apr 12, 2025 • edited Loading

elasticsearchmachine commented May 6, 2025

elasticsearchmachine commented May 6, 2025

Mikep86 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

elasticsearchmachine commented May 7, 2025

Mikep86 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Mikep86 May 7, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Mikep86 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

elasticsearchmachine commented May 7, 2025

markjhoy commented Apr 12, 2025 •

edited

Loading

Mikep86 May 7, 2025 •

edited

Loading