Struct Query 

pub struct Query;

Query builder entry point. Start all queries with Q::.

Examples

// Vector similarity search
Q::similar_to(&embedding, 10)

// Start from a graph node
Q::from_node("node-label")

// Filter table rows
Q::table("hosts").where_("status").equals("active")

// Full collection scan
Q::all_in("vulnerabilities")

Implementations

impl Q

pub fn similar_to(vector: &[f32], k: usize) -> VectorQueryBuilder

Start a vector similarity query

pub fn from_node(label: impl Into<String>) -> GraphQueryBuilder

Start a query from a specific graph node

pub fn from_id(id: EntityId) -> GraphQueryBuilder

Start a query from a specific entity by ID

pub fn table(name: impl Into<String>) -> TableQueryBuilder

Query a specific table/collection

pub fn collection(name: impl Into<String>) -> TableQueryBuilder

Shorthand for collection query

pub fn all_in(collection: impl Into<String>) -> ScanQueryBuilder

Query all entities in a collection

pub fn refs_from(id: EntityId, ref_type: RefType) -> RefQueryBuilder

Find entities by cross-reference

pub fn text_search(query: impl Into<String>) -> TextSearchBuilder

Text search across all indexed content

pub fn document(collection: impl Into<String>) -> TableQueryBuilder

Query documents in a collection

Documents are stored as enriched table rows, so this is a convenience alias that delegates to TableQueryBuilder. The flattened top-level document fields support path-based filtering through the table query DSL.

Example

let results = Q::document("articles")
    .where_("title").equals("Hello World")
    .limit(10)
    .execute(&store)?;

pub fn kv(collection: impl Into<String>) -> KvQueryBuilder

Key-Value operations on a collection

Examples

// Get a value
let result = Q::kv("config").get("theme").execute(&store)?;

// Set a value (upsert)
let result = Q::kv("config").set("theme", Value::text("dark")).execute(&store)?;

// Delete a key
let result = Q::kv("config").delete("theme").execute(&store)?;

// List all KV pairs
let result = Q::kv("config").list().execute(&store)?;

pub fn hybrid() -> HybridQueryBuilder

Hybrid query combining multiple modes

pub fn cross_modal() -> ThreeWayJoinBuilder

Three-way cross-modal JOIN

Efficiently chains queries across Vector → Graph → Table (or any order).

Example

// Find CVEs similar to vector, traverse to affected hosts, get host records
let results = Q::cross_modal()
    .start_vector(&cve_embedding, 10)
    .traverse("AFFECTS", 2)
    .join_table("hosts")
    .execute(&store)?;