robin-sparkless 0.16.0

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
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64

"""
Tests for #420: createDataFrame verify_schema strict per-row validation (PySpark parity).
Rust: create_dataframe_from_rows(..., verify_schema=true) is implemented. Python binding must
pass verify_schema through when createDataFrame(..., verify_schema=True) is called.
"""

from __future__ import annotations

import pytest

import robin_sparkless as rs


def _spark() -> rs.SparkSession:
    return rs.SparkSession.builder().app_name("issue_420").get_or_create()


def _supports_verify_schema_kw() -> bool:
    """True if createDataFrame(..., schema=..., verify_schema=True) is accepted."""
    try:
        spark = _spark()
        spark.createDataFrame(
            [{"name": "a", "age": 1}],
            schema=[("name", "string"), ("age", "bigint")],
            verify_schema=False,
        )
        return True
    except TypeError:
        return False


@pytest.mark.skipif(
    not _supports_verify_schema_kw(),
    reason="Python binding does not yet accept verify_schema= keyword; Rust API create_dataframe_from_rows(..., verify_schema) is ready.",
)
def test_verify_schema_true_raises_on_type_mismatch() -> None:
    """When verify_schema=True, wrong type raises with Row/column message."""
    spark = _spark()
    # Schema says age is bigint; second row has age as string.
    data = [{"name": "Alice", "age": 25}, {"name": "Bob", "age": "thirty"}]
    with pytest.raises(TypeError) as exc_info:
        spark.createDataFrame(
            data, schema=[("name", "string"), ("age", "bigint")], verify_schema=True
        )
    msg = str(exc_info.value)
    assert "Row 1" in msg or "row 1" in msg.lower()
    assert "age" in msg or "column" in msg.lower()
    assert "bigint" in msg or "number" in msg.lower()


@pytest.mark.skipif(
    not _supports_verify_schema_kw(),
    reason="Python binding does not yet accept verify_schema= keyword; Rust API is ready.",
)
def test_verify_schema_false_allows_mismatch_or_fails_later() -> None:
    """When verify_schema=False, creation may succeed or fail later."""
    spark = _spark()
    data = [{"name": "Alice", "age": 25}, {"name": "Bob", "age": 30}]
    df = spark.createDataFrame(
        data, schema=[("name", "string"), ("age", "bigint")], verify_schema=False
    )
    rows = df.collect()
    assert len(rows) == 2
    assert rows[0]["age"] == 25 and rows[1]["age"] == 30