Add support for correlated subquery

[mustafasrepo]

Is your feature request related to a problem or challenge?

When I run the query below

SELECT s.amount * (
         SELECT e.amount
         FROM sales_us AS e
         WHERE s.currency = e.currency AND
  		         s.ts >= e.ts
         ORDER BY e.ts DESC
         LIMIT 1
       ) AS amount_usd
FROM sales_global AS s
ORDER BY s.sn;

Datafusion returns

Error: NotImplemented("Physical plan does not support logical expression (<subquery>)")
Same query successfully runs on Postgre.

Describe the solution you'd like

I would like to have this feature.

Describe alternatives you've considered

No response

Additional context

To reproduce you can use test below

#[tokio::test]
async fn test_subquery() -> Result<()> {
    let config = SessionConfig::new()
        .with_target_partitions(1);
    let ctx = SessionContext::with_config(config);
    ctx.sql("CREATE TABLE sales_us (
          ts TIMESTAMP,
          currency VARCHAR(3),
          amount INT
        ) as VALUES
              ('2022-01-01 10:00:00'::timestamp, 'USD', 100.00),
              ('2022-01-01 11:00:00'::timestamp, 'USD', 200.00),
              ('2022-01-02 09:00:00'::timestamp, 'USD', 300.00),
              ('2022-01-02 10:00:00'::timestamp, 'USD', 150.00)").await?;
    ctx.sql("CREATE TABLE sales_global (
          ts TIMESTAMP,
          currency VARCHAR(3),
          amount INT
        ) as VALUES
          ('2022-01-01 08:00:00'::timestamp, 'EUR', 50.00),
          ('2022-01-01 11:30:00'::timestamp, 'EUR', 75.00),
          ('2022-01-02 12:00:00'::timestamp, 'EUR', 200.00),
          ('2022-01-03 10:00:00'::timestamp, 'EUR', 100.00)").await?;
    let sql = "SELECT s.amount * (
                 SELECT e.amount
                 FROM sales_us AS e
                 WHERE s.currency = e.currency AND
                         s.ts >= e.ts
                 ORDER BY e.ts DESC
                 LIMIT 1
               ) AS amount_usd
        FROM sales_global AS s
        ORDER BY s.sn
        ";

    let msg = format!("Creating logical plan for '{sql}'");
    let dataframe: DataFrame = ctx.sql(sql).await.expect(&msg);
    let physical_plan = dataframe.create_physical_plan().await?;
    let batches = collect(physical_plan, ctx.task_ctx()).await?;
    print_batches(&batches)?;
    Ok(())
}

[mingmwang]

I will work on this soon.

[ozankabak]

@mingmwang any thoughts on how to make progress here?

[aalexandrov]

@alamb can you please assign me here? I will take a stab at this next week. If I succeed, you might be able to also close some other issues listed in the enclosing epic as obsolete because there won't be any correlated queries left for the physical planning process.

[alamb]

Thanks @aalexandrov -- I have done so.

BTW you can also make a single word comment take and we have a bot that will assign you.

[aalexandrov]

BTW you can also make a single word comment take and we have a bot that will assign you.

Oh, good to know—I will take a look at the available commands next time.

[Lordworms]

Interested in this one

[alamb]

I believe @eejbyfeldt and @Dandandan have been working a bit on correlated subqueries in the context of getting all TPCS-DS running