Rust API Overview

Overview

The krishiv-api crate is the user-facing Rust API. It re-exports every public type at the crate root. Add it to your Cargo.toml:

[dependencies]
krishiv = { path = "../krishiv", features = ["embedded"] }

Or use the api crate directly:

[dependencies]
krishiv-api = { path = "../crates/krishiv-api" }

Key Top-Level Exports

Type / Function	Description
Session, SessionBuilder	Entry point for all Krishiv workloads.
DataFrame, GroupedDataFrame	Lazy query plan builder.
Stream, KeyedStream	Streaming data pipeline builder.
IncrementalFlow	Incremental view maintenance.
WindowedStream, SessionWindowedStream, SlidingWindowedStream	Windowed streaming aggregation.
DataStreamReader, DataStreamWriter, StreamingQuery	Structured streaming API (Spark-style).
QueryHandle, QueryResult	Async query execution and result collection.
PreparedStatement	Parameterised SQL statements.
col, lit, expr, avg, count, sum, min, max	Expression builder functions.
Pipeline, PipelineBuilder	Source-to-sink pipeline execution.
ValueState, MapState, ListState	Keyed operator state.
KrishivError, Result	Error type and alias.
RecordBatch, Schema, Field, DataType	Re-exported Arrow types.

Quick Example

use krishiv_api::{Session, Result, col, lit};

#[tokio::main]
async fn main() -> Result<()> {
    let session = Session::embedded().await?;

    // SQL path
    let result = session.sql("SELECT 1 + 1 AS two").await?.collect().await?;

    // DataFrame path
    let df = session.read_parquet("data/sales.parquet").await?
        .filter(col("amount").gt(lit(100)))?
        .select(&["customer_id", "amount"])?;
    df.show().await?;

    Ok(())
}