robin-sparkless 4.4.0

PySpark-like DataFrame API in Rust on Polars; no JVM.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

"""Tests for #375: flatMap behavior (PySpark)."""

from __future__ import annotations


def test_flat_map(spark) -> None:
    """PySpark: use RDD.flatMap for row expansion."""
    df = spark.createDataFrame([("a b",), ("c d e",)], ["word"])
    words = df.rdd.flatMap(lambda row: row["word"].split()).collect()
    assert words == ["a", "b", "c", "d", "e"]


def test_flat_map_empty(spark) -> None:
    """PySpark: flatMap can return empty output."""
    df = spark.createDataFrame([(1,)], ["x"])
    out = df.rdd.flatMap(lambda row: []).collect()
    assert out == []