2026-08-17 13:40:09 +03:00

94 lines
3.1 KiB
Python

"""Tests for the CSV variant adapter against small in-memory payloads."""
from __future__ import annotations
import sys
from pathlib import Path
import pytest
sys.path.insert(0, str(Path(__file__).resolve().parents[1] / "src"))
from csv_variant_adapter.plugin import CsvOptions, CsvVariantAdapter, parse_csv
def test_sniffs_comma_delimiter_and_types_numbers() -> None:
payload = b"time,x,y\n0,1.25,2\n1,3.5,4\n"
rows = parse_csv(payload)
assert rows == [
{"time": 0, "x": 1.25, "y": 2},
{"time": 1, "x": 3.5, "y": 4},
]
# int stays int, decimals become float.
assert isinstance(rows[0]["time"], int)
assert isinstance(rows[0]["x"], float)
def test_european_dialect_with_semicolon_comma_decimal_and_thousands() -> None:
payload = "time;pressure\n0;1.013,25\n60;995,40\n".encode("utf-8")
options = CsvOptions(delimiter=";", decimal=",", thousands=".")
rows = parse_csv(payload, options)
assert rows == [
{"time": 0, "pressure": 1013.25},
{"time": 60, "pressure": 995.40},
]
def test_utf8_bom_is_stripped_by_default_encoding() -> None:
payload = "time,value\n0,10\n".encode("utf-8")
rows = parse_csv(payload)
assert list(rows[0]) == ["time", "value"]
def test_blank_cell_becomes_none() -> None:
rows = parse_csv(b"time,x\n0,\n1,2\n")
assert rows[0]["x"] is None
assert rows[1]["x"] == 2
def test_invoke_returns_envelope_with_columns_and_count() -> None:
adapter = CsvVariantAdapter()
result = adapter.invoke({"payload": "time;value\n0;1,5\n1;2,5\n",
"options": {"delimiter": ";", "decimal": ","}})
assert result["row_count"] == 2
assert result["columns"] == ["time", "value"]
assert result["records"][1] == {"time": 1, "value": 2.5}
def test_records_convert_to_canonical_dataset_shape() -> None:
adapter = CsvVariantAdapter()
records = adapter.invoke({"payload": b"time,x\n0,1.0\n1,2.0\n2,3.0\n"})["records"]
time = tuple(float(row["time"]) for row in records)
columns = {"x": tuple(float(row["x"]) for row in records)}
# Canonical Dataset requires strictly increasing time and aligned columns.
assert all(b > a for a, b in zip(time, time[1:]))
assert len(columns["x"]) == len(time)
def test_duplicate_header_names_are_rejected() -> None:
with pytest.raises(ValueError, match="unique"):
parse_csv(b"time,x,x\n0,1,2\n")
def test_empty_header_is_rejected() -> None:
with pytest.raises(ValueError, match="header"):
parse_csv(b"time,,y\n0,1,2\n")
def test_max_bytes_limit_is_enforced() -> None:
options = CsvOptions(delimiter=",", max_bytes=8)
with pytest.raises(ValueError, match="max_bytes"):
parse_csv(b"time,x\n0,1\n1,2\n", options)
def test_invalid_options_are_rejected_before_parsing() -> None:
with pytest.raises(ValueError, match="decimal"):
CsvOptions(decimal="x")
with pytest.raises(ValueError, match="delimiter"):
CsvOptions(delimiter=";;")
def test_invoke_rejects_non_text_non_bytes_payload() -> None:
with pytest.raises(TypeError):
CsvVariantAdapter().invoke({"payload": 123})