SearchSpec

const val EMBEDDING_SEARCH_METRIC_TYPE_COSINE = 1: Int

Cosine similarity as metric for embedding search and ranking.

const val EMBEDDING_SEARCH_METRIC_TYPE_DOT_PRODUCT = 2: Int

Dot product similarity as metric for embedding search and ranking.

const val EMBEDDING_SEARCH_METRIC_TYPE_EUCLIDEAN = 3: Int

Euclidean distance as metric for embedding search and ranking.

const val GROUPING_TYPE_PER_NAMESPACE = 2: Int

Results should be grouped together by namespace for the purpose of enforcing a limit on the number of results returned per namespace.

const val GROUPING_TYPE_PER_PACKAGE = 1: Int

Results should be grouped together by package for the purpose of enforcing a limit on the number of results returned per package.

@RequiresFeature(enforcement = "androidx.appsearch.app.Features#isFeatureSupported", name = Features.SEARCH_SPEC_GROUPING_TYPE_PER_SCHEMA)
const val GROUPING_TYPE_PER_SCHEMA = 4: Int

Results should be grouped together by schema type for the purpose of enforcing a limit on the number of results returned per schema type.

const val ORDER_ASCENDING = 1: Int

Search results will be returned in an ascending order.

const val ORDER_DESCENDING = 0: Int

Search results will be returned in a descending order.

const val PROJECTION_SCHEMA_TYPE_WILDCARD = "*": String!

Schema type to be used in addProjection to apply property paths to all results, excepting any types that have had their own, specific property paths set.

const val RANKING_STRATEGY_ADVANCED_RANKING_EXPRESSION = 9: Int

Ranked by the advanced ranking expression provided.

const val RANKING_STRATEGY_CREATION_TIMESTAMP = 2: Int

Ranked by document creation timestamps.

const val RANKING_STRATEGY_DOCUMENT_SCORE = 1: Int

Ranked by app-provided document scores.

const val RANKING_STRATEGY_JOIN_AGGREGATE_SCORE = 8: Int

Ranked by the aggregated ranking signal of the joined documents.

Which aggregation strategy is used to determine a ranking signal is specified in the JoinSpec set by setJoinSpec. This ranking strategy may not be used if no JoinSpec is provided.

const val RANKING_STRATEGY_NONE = 0: Int

No Ranking, results are returned in arbitrary order.

const val RANKING_STRATEGY_RELEVANCE_SCORE = 3: Int

Ranked by document relevance score.

const val RANKING_STRATEGY_SYSTEM_USAGE_COUNT = 6: Int

Ranked by number of usages from a system UI surface.

const val RANKING_STRATEGY_SYSTEM_USAGE_LAST_USED_TIMESTAMP = 7: Int

Ranked by timestamp of last usage from a system UI surface.

const val RANKING_STRATEGY_USAGE_COUNT = 4: Int

Ranked by number of usages, as reported by the app.

const val RANKING_STRATEGY_USAGE_LAST_USED_TIMESTAMP = 5: Int

Ranked by timestamp of last usage, as reported by the app.

const val SCHEMA_TYPE_WILDCARD = "*": String!

Schema type to be used in addFilterProperties and addProjection to apply property paths to all results, excepting any types that have had their own, specific property paths set.

const val TERM_MATCH_EXACT_ONLY = 1: Int

Query terms will only match exact tokens in the index.

For example, a query term "foo" will only match indexed token "foo", and not "foot" or "football".

const val TERM_MATCH_PREFIX = 2: Int

Query terms will match indexed tokens when the query term is a prefix of the token.

For example, a query term "foo" will match indexed tokens like "foo", "foot", and "football".

fun getAdvancedRankingExpression(): String

Get the advanced ranking expression, or "" if setRankingStrategy was not called.

fun getDefaultEmbeddingSearchMetricType(): Int

Returns the default embedding metric type used for embedding search (see search) and ranking (see setRankingStrategy).

fun getEmbeddingParameters(): (Mutable)List<EmbeddingVector!>

Returns the list of EmbeddingVector that can be referenced in the query through the "getEmbeddingParameter({index})" function.

fun getFilterNamespaces(): (Mutable)List<String!>

Returns the list of namespaces to search over.

If empty, the query will search over all namespaces.

fun getFilterPackageNames(): (Mutable)List<String!>

Returns the list of package name filters to search over.

If empty, the query will search over all packages that the caller has access to. If package names are specified which caller doesn't have access to, then those package names will be ignored.

fun getFilterProperties(): (Mutable)Map<String!, (Mutable)List<String!>!>

Returns the map of schema and target properties to search over.

If empty, will search over all schema and properties.

Calling this function repeatedly is inefficient. Prefer to retain the Map returned by this function, rather than calling it multiple times.

fun getFilterSchemas(): (Mutable)List<String!>

Returns the list of schema types to search for.

If empty, the query will search over all schema types.

fun getInformationalRankingExpressions(): (Mutable)List<String!>

Returns the informational ranking expressions.

fun getJoinSpec(): JoinSpec?

Returns specification on which documents need to be joined.

fun getMaxSnippetSize(): Int

Returns the maximum size of a snippet in characters.

fun getOrder(): Int

Returns the order of returned search results (descending or ascending).

fun getProjectionPaths(): (Mutable)Map<String!, (Mutable)List<PropertyPath!>!>

Returns a map from schema type to property paths to be used for projection.

If the map is empty, then all properties will be retrieved for all results.

Calling this function repeatedly is inefficient. Prefer to retain the Map returned by this function, rather than calling it multiple times.

fun getProjections(): (Mutable)Map<String!, (Mutable)List<String!>!>

Returns a map from schema type to property paths to be used for projection.

If the map is empty, then all properties will be retrieved for all results.

Calling this function repeatedly is inefficient. Prefer to retain the Map returned by this function, rather than calling it multiple times.

fun getPropertyWeightPaths(): (Mutable)Map<String!, (Mutable)Map<PropertyPath!, Double!>!>

Returns properties weights to be used for scoring.

Calling this function repeatedly is inefficient. Prefer to retain the Map returned by this function, rather than calling it multiple times.

fun getPropertyWeights(): (Mutable)Map<String!, (Mutable)Map<String!, Double!>!>

Returns properties weights to be used for scoring.

Calling this function repeatedly is inefficient. Prefer to retain the Map returned by this function, rather than calling it multiple times.

fun getRankingStrategy(): Int

Returns the ranking strategy.

fun getResultCountPerPage(): Int

Returns the number of results per page in the result set.

fun getResultGroupingLimit(): Int

Get the maximum number of results to return for each group.

fun getResultGroupingTypeFlags(): Int

Get the type of grouping limit to apply, or 0 if setResultGrouping was not called.

fun getSearchSourceLogTag(): String?

Gets a tag to indicate the source of this search, or null if setSearchSourceLogTag was not called.

Some AppSearch implementations may log a hash of this tag using statsd. This tag may be used for tracing performance issues and crashes to a component of an app.

Call setSearchSourceLogTag and give a unique value if you want to distinguish this search scenario with other search scenarios during performance analysis.

Under no circumstances will AppSearch log the raw String value using statsd, but it will be provided as-is to custom AppSearchLogger implementations you have registered in your app.

fun getSearchStringParameters(): (Mutable)List<String!>

Returns the list of String parameters that can be referenced in the query through the "getSearchStringParameter({index})" function.

fun getSnippetCount(): Int

Returns how many documents to generate snippets for.

fun getSnippetCountPerProperty(): Int

Returns how many matches for each property of a matching document to generate snippets for.

fun getTermMatch(): Int

Returns how the query terms should match terms in the index.

fun isListFilterHasPropertyFunctionEnabled(): Boolean

Returns whether the LIST_FILTER_HAS_PROPERTY_FUNCTION feature is enabled.

fun isListFilterQueryLanguageEnabled(): Boolean

Returns whether the LIST_FILTER_QUERY_LANGUAGE feature is enabled.

fun isNumericSearchEnabled(): Boolean

Returns whether the NUMERIC_SEARCH feature is enabled.

fun isVerbatimSearchEnabled(): Boolean

Returns whether the VERBATIM_SEARCH feature is enabled.