Module org.elasticsearch.server
Class SparseVectorFieldMapper.SparseVectorFieldType
java.lang.Object
org.elasticsearch.index.mapper.MappedFieldType
org.elasticsearch.index.mapper.vectors.SparseVectorFieldMapper.SparseVectorFieldType
- Enclosing class:
SparseVectorFieldMapper
-
Nested Class Summary
Nested classes/interfaces inherited from class org.elasticsearch.index.mapper.MappedFieldType
MappedFieldType.BlockLoaderContext, MappedFieldType.CollapseType, MappedFieldType.FielddataOperation, MappedFieldType.FieldExtractPreference, MappedFieldType.Relation -
Constructor Summary
ConstructorsConstructorDescriptionSparseVectorFieldType(IndexVersion indexVersionCreated, String name, boolean isStored, Map<String, String> meta) SparseVectorFieldType(IndexVersion indexVersionCreated, String name, boolean isStored, Map<String, String> meta, SparseVectorFieldMapper.SparseVectorIndexOptions indexOptions) -
Method Summary
Modifier and TypeMethodDescriptionorg.apache.lucene.search.QueryexistsQuery(SearchExecutionContext context) fielddataBuilder(FieldDataContext fieldDataContext) Return a fielddata builder for this fieldorg.apache.lucene.search.QueryfinalizeSparseVectorQuery(SearchExecutionContext context, String fieldName, List<WeightedToken> queryVectors, Boolean shouldPruneTokensFromQuery, TokenPruningConfig tokenPruningConfigFromQuery) booleanVector embeddings are typically large and not intended for human consumption, so such fields may be excluded from responses.org.apache.lucene.search.QuerytermQuery(Object value, SearchExecutionContext context) Generates a query that will only match documents that contain the given value.typeName()Returns the name of this type, as would be specified in mapping propertiesvalueFetcher(SearchExecutionContext context, String format) Create a helper class to fetch field values during theFetchFieldsPhase.Methods inherited from class org.elasticsearch.index.mapper.MappedFieldType
automatonQuery, blockLoader, checkNoFormat, checkNoTimeZone, collapseType, dimensions, distanceFeatureQuery, docValueFormat, eagerGlobalOrdinals, failIfNoDocValues, failIfNotIndexed, failIfNotIndexedNorDocValuesFallback, familyTypeName, fieldHasValue, fuzzyIntervals, fuzzyQuery, fuzzyQuery, getDefaultHighlighter, getMetricType, getTerms, getTextSearchInfo, hasDocValues, hasScriptValues, isAggregatable, isDimension, isFieldWithinQuery, isIndexed, isSearchable, isStored, mayExistInIndex, meta, multiPhraseQuery, name, normalizedWildcardQuery, phrasePrefixQuery, phraseQuery, pointReaderIfPossible, prefixIntervals, prefixQuery, prefixQuery, rangeIntervals, rangeQuery, regexpIntervals, regexpQuery, spanPrefixQuery, termIntervals, termQueryCaseInsensitive, termsQuery, validateMatchedRoutingPath, valueForDisplay, wildcardIntervals, wildcardLikeQuery, wildcardQuery, wildcardQuery
-
Constructor Details
-
SparseVectorFieldType
public SparseVectorFieldType(IndexVersion indexVersionCreated, String name, boolean isStored, Map<String, String> meta) -
SparseVectorFieldType
public SparseVectorFieldType(IndexVersion indexVersionCreated, String name, boolean isStored, Map<String, String> meta, @Nullable SparseVectorFieldMapper.SparseVectorIndexOptions indexOptions)
-
-
Method Details
-
getIndexOptions
-
typeName
Description copied from class:MappedFieldTypeReturns the name of this type, as would be specified in mapping properties- Specified by:
typeNamein classMappedFieldType
-
isVectorEmbedding
public boolean isVectorEmbedding()Description copied from class:MappedFieldTypeVector embeddings are typically large and not intended for human consumption, so such fields may be excluded from responses.- Overrides:
isVectorEmbeddingin classMappedFieldType- Returns:
- true if this field contains vector embeddings.
-
fielddataBuilder
Description copied from class:MappedFieldTypeReturn a fielddata builder for this field- Overrides:
fielddataBuilderin classMappedFieldType- Parameters:
fieldDataContext- the context for the fielddata
-
valueFetcher
Description copied from class:MappedFieldTypeCreate a helper class to fetch field values during theFetchFieldsPhase. New field types must implement this method in order to support the search 'fields' option. Except for metadata fields, field types should not throwUnsupportedOperationExceptionsince this could cause a search retrieving multiple fields (like "fields": ["*"]) to fail.- Specified by:
valueFetcherin classMappedFieldType
-
termQuery
Description copied from class:MappedFieldTypeGenerates a query that will only match documents that contain the given value. The default implementation returns aTermQueryover the value bytes- Specified by:
termQueryin classMappedFieldType
-
existsQuery
- Overrides:
existsQueryin classMappedFieldType
-
finalizeSparseVectorQuery
public org.apache.lucene.search.Query finalizeSparseVectorQuery(SearchExecutionContext context, String fieldName, List<WeightedToken> queryVectors, Boolean shouldPruneTokensFromQuery, TokenPruningConfig tokenPruningConfigFromQuery) throws IOException - Throws:
IOException
-