Struct VelesCollection 

pub struct VelesCollection { /* private fields */ }

A collection of vectors with associated metadata.

inner is a detached core collection leaf with no observer reference, so every governed read is routed back through the owning database’s control-plane gate (db.gated_search / db.authorize_read) rather than hitting the leaf directly — restoring observer governance for the mobile direct-search API (audit F-5.4, #1392). Reads with no observer registered take a single Option check and then the same leaf call as before (zero-overhead contract).

Implementations

impl VelesCollection

pub fn search( &self, vector: Vec<f32>, limit: u32, ) -> Result<Vec<SearchResult>, VelesError>

Searches for the k nearest neighbors to the query vector.

Arguments

• vector - Query vector
• limit - Maximum number of results to return

Returns

Vector of search results sorted by similarity.

pub fn search_with_quality( &self, vector: Vec<f32>, limit: u32, quality: SearchQuality, ) -> Result<Vec<SearchResult>, VelesError>

Searches with a specific quality profile controlling recall/latency.

Arguments

• vector - Query vector
• limit - Maximum number of results to return
• quality - Search quality profile (Fast, Balanced, Accurate, etc.)

Returns

Vector of search results sorted by similarity.

pub fn upsert(&self, point: VelesPoint) -> Result<(), VelesError>

Inserts or updates a single point.

Arguments

• point - The point to upsert

pub fn upsert_batch(&self, points: Vec<VelesPoint>) -> Result<(), VelesError>

Inserts or updates multiple points in batch.

Arguments

• points - Points to upsert

pub fn delete(&self, id: u64) -> Result<(), VelesError>

Deletes a point by ID.

pub fn count(&self) -> u64

Returns the number of points in the collection.

pub fn dimension(&self) -> u32

Returns the vector dimension.

pub fn get(&self, ids: Vec<u64>) -> Vec<VelesPoint>

Gets points by their IDs.

Arguments

• ids - List of point IDs to retrieve

Returns

Vector of points found. Missing IDs are silently skipped.

pub fn get_by_id(&self, id: u64) -> Option<VelesPoint>

Gets a single point by ID.

Arguments

• id - Point ID to retrieve

Returns

The point if found, None otherwise.

pub fn is_metadata_only(&self) -> bool

Checks if this is a metadata-only collection.

pub fn text_search( &self, query: String, limit: u32, ) -> Result<Vec<SearchResult>, VelesError>

Performs full-text search using BM25.

Arguments

• query - Text query to search for
• limit - Maximum number of results to return

Returns

Vector of search results sorted by BM25 score.

pub fn hybrid_search( &self, vector: Vec<f32>, text_query: String, limit: u32, vector_weight: f32, ) -> Result<Vec<SearchResult>, VelesError>

Performs hybrid search combining vector similarity and BM25 text search.

Arguments

• vector - Query vector for similarity search
• text_query - Text query for BM25 search
• limit - Maximum number of results
• vector_weight - Weight for vector similarity (0.0-1.0)

Returns

Vector of search results sorted by fused score.

pub fn search_with_filter( &self, vector: Vec<f32>, limit: u32, filter_json: String, ) -> Result<Vec<SearchResult>, VelesError>

Searches with metadata filtering.

Arguments

• vector - Query vector
• limit - Maximum number of results
• filter_json - JSON filter string (e.g., {"condition": {"type": "eq", "field": "category", "value": "tech"}})

Returns

Vector of search results matching the filter.

pub fn batch_search( &self, searches: Vec<IndividualSearchRequest>, ) -> Result<Vec<Vec<SearchResult>>, VelesError>

Performs batch search for multiple query vectors in parallel.

Arguments

• searches - List of search requests

Returns

List of result lists (one per query vector).

pub fn text_search_with_filter( &self, query: String, limit: u32, filter_json: String, ) -> Result<Vec<SearchResult>, VelesError>

Performs text search with metadata filtering.

Arguments

• query - Text query
• limit - Maximum number of results
• filter_json - JSON filter string

pub fn hybrid_search_with_filter( &self, vector: Vec<f32>, text_query: String, limit: u32, vector_weight: f32, filter_json: String, ) -> Result<Vec<SearchResult>, VelesError>

Performs hybrid search with metadata filtering.

Arguments

• vector - Query vector
• text_query - Text query
• limit - Maximum number of results
• vector_weight - Weight for vector similarity (0.0-1.0)
• filter_json - JSON filter string

pub fn query( &self, query_str: String, params_json: Option<String>, ) -> Result<Vec<SearchResult>, VelesError>

Executes a VelesQL query.

Arguments

• query_str - VelesQL query string
• params_json - Optional JSON object with query parameters

Returns

Vector of search results.

Example

let results = try collection.query(
    "SELECT * FROM vectors WHERE category = 'tech' LIMIT 10",
    nil
)

pub fn enable_streaming( &self, config: Option<MobileStreamingConfig>, ) -> Result<(), VelesError>

Enables streaming ingestion on this collection.

Must be called before stream_insert; otherwise stream inserts fail with “not configured”. Calling it again replaces the existing ingester.

Arguments

• config - Optional MobileStreamingConfig. None uses the engine defaults (buffer_size=10000, batch_size=128, flush_interval_ms=50).

pub fn stream_insert(&self, points: Vec<VelesPoint>) -> Result<u64, VelesError>

Queues a batch of points for streaming ingestion.

Requires enable_streaming to have been called first. Returns the number of points successfully queued.

Arguments

• points - Points to queue for ingestion

pub fn flush(&self) -> Result<(), VelesError>

Flushes collection data to durable storage.

pub fn compact_storage(&self) -> Result<u64, VelesError>

Compacts on-disk storage, reclaiming space left by deleted vectors.

Returns the number of bytes reclaimed.

pub fn guard_rails(&self) -> MobileQueryLimits

Returns the current query guardrail limits for this collection.

pub fn apply_advanced_config( &self, config: MobileAdvancedConfig, ) -> Result<(), VelesError>

Applies post-creation overrides to advanced configuration fields (pq_rescore_oversampling, deferred_indexing, async_index_builder) and persists the updated config. Each Some field is applied; each None field is left unchanged.

pub fn all_ids(&self) -> Vec<u64>

Returns all point IDs currently present in the collection.

pub fn create_index(&self, field_name: String) -> Result<(), VelesError>

Creates a secondary metadata index for a payload field.

pub fn has_secondary_index(&self, field_name: String) -> bool

Checks whether a secondary metadata index exists for a field.

pub fn create_property_index( &self, label: String, property: String, ) -> Result<(), VelesError>

Creates a graph/property index for equality lookups.

pub fn create_range_index( &self, label: String, property: String, ) -> Result<(), VelesError>

Creates a graph/range index for range queries.

pub fn has_property_index(&self, label: String, property: String) -> bool

Checks if a property index exists.

pub fn has_range_index(&self, label: String, property: String) -> bool

Checks if a range index exists.

pub fn list_indexes(&self) -> Vec<MobileIndexInfo>

Lists all index definitions on this collection.

pub fn drop_index( &self, label: String, property: String, ) -> Result<bool, VelesError>

Drops an index and returns true when something was removed.

pub fn indexes_memory_usage(&self) -> u64

Returns total memory usage used by indexes.

pub fn analyze(&self) -> Result<MobileCollectionStats, VelesError>

Runs ANALYZE and returns fresh statistics for this collection.

pub fn get_stats(&self) -> MobileCollectionStats

Returns the latest known collection statistics snapshot.

pub fn diagnostics(&self) -> MobileCollectionDiagnostics

Returns a health/readiness diagnostics snapshot for this collection.

impl VelesCollection

pub fn sparse_search( &self, sparse_vector: VelesSparseVector, limit: u32, index_name: Option<String>, ) -> Result<Vec<SearchResult>, VelesError>

Performs sparse-only search using an inverted index.

Arguments

• sparse_vector - Query sparse vector (parallel arrays of indices/values)
• limit - Maximum number of results
• index_name - Name of the sparse index (empty string for default)

Returns

Vector of search results sorted by sparse similarity.

pub fn hybrid_sparse_search( &self, vector: Vec<f32>, sparse_vector: VelesSparseVector, limit: u32, index_name: Option<String>, ) -> Result<Vec<SearchResult>, VelesError>

Performs hybrid dense+sparse search with RRF fusion.

Combines vector similarity search with sparse (keyword) search using Reciprocal Rank Fusion.

Arguments

• vector - Dense query vector
• sparse_vector - Sparse query vector (parallel arrays)
• limit - Maximum number of results
• index_name - Name of the sparse index (empty string or None for default)

Returns

Vector of fused search results.

pub fn multi_query_search( &self, vectors: Vec<Vec<f32>>, limit: u32, strategy: FusionStrategy, ) -> Result<Vec<SearchResult>, VelesError>

Performs multi-query search with result fusion.

pub fn multi_query_search_ids( &self, vectors: Vec<Vec<f32>>, limit: u32, strategy: FusionStrategy, ) -> Result<Vec<SearchResult>, VelesError>

Performs multi-query search returning IDs and scores only.

Id-only twin of Self::multi_query_search: reuses the same fusion path but strips payloads, avoiding payload materialization.

pub fn multi_query_search_with_filter( &self, vectors: Vec<Vec<f32>>, limit: u32, strategy: FusionStrategy, filter_json: String, ) -> Result<Vec<SearchResult>, VelesError>

Performs multi-query search with metadata filtering.

pub fn upsert_with_sparse( &self, point: VelesPoint, sparse_vector: VelesSparseVector, ) -> Result<(), VelesError>

Inserts or updates a point with an associated sparse vector.

Arguments

• point - The point to upsert (dense vector + payload)
• sparse_vector - Sparse vector to associate with this point

Trait Implementations

impl<UT> LiftRef<UT> for VelesCollection

type LiftType = Arc<VelesCollection>

impl<UT> LowerError<UT> for VelesCollection

fn lower_error(obj: Self) -> RustBuffer

Lower this value for scaffolding function return

impl<UT> LowerReturn<UT> for VelesCollection

type ReturnType = <Arc<VelesCollection> as LowerReturn<UniFfiTag>>::ReturnType

The type that should be returned by scaffolding functions for this type.

fn lower_return(obj: Self) -> Result<Self::ReturnType, RustCallError>

Lower the return value from an scaffolding call

fn handle_failed_lift( error: LiftArgsError, ) -> Result<Self::ReturnType, RustCallError>

Lower the return value for failed argument lifts

impl<UT> TypeId<UT> for VelesCollection

const TYPE_ID_META: MetadataBuffer