Skip to content

Schema & Indexing

Uni lets you define schema, constraints, and indexes to enforce correctness and speed up queries. You can start schemaless and evolve toward typed properties.

What It Provides

  • Typed properties and constraints (unique, exists, check).
  • Scalar indexes, vector indexes, and full-text indexes.
  • Fast predicate pushdown and index usage in the planner.

Example

use uni_db::{DataType, IndexType, ScalarType, Uni};

# async fn demo() -> Result<(), uni_db::UniError> {
let db = Uni::open("./my_db").build().await?;

db.schema()
    .label("User")
        .property("email", DataType::String)
        .property("age", DataType::Int64)
        .index("email", IndexType::Scalar(ScalarType::BTree))
    .apply()
    .await?;

# Ok(())
# }

import uni_db

db = uni_db.Database("./my_db")

db.schema() \
    .label("User") \
        .property("email", "string") \
        .property("age", "int64") \
        .index("email", "btree") \
        .done() \
    .apply()

Use Cases

  • Enforcing data consistency in production.
  • Speeding up high-selectivity filters and joins.
  • Adding indexes incrementally as query patterns stabilize.

When To Use

Start with a minimal schema for agility, then add indexes and constraints once your access patterns are clear.