Transforms complex documents like PDFs and Office docs into LLM-ready markdown/JSON for your Agentic workflows.
Python ai4sciencedocument-analysisdocxextract-datalayout-analysisocrparserpdfpdf-converterpdf-extractor-llmpdf-extractor-pretrainpdf-extractor-ragpdf-parserpptxpythonxlsx
from __future__ import annotations
import tomllib
from pathlib import Path
import pytest
import mineru.utils.stdio as stdio
def test_configure_standard_streams_uses_utf8(monkeypatch: pytest.MonkeyPatch) -> None:
class _Stream:
def __init__(self) -> None:
self.calls: list[dict[str, str]] = []
def reconfigure(self, **kwargs: str) -> None:
self.calls.append(kwargs)
stdout = _Stream()
stderr = _Stream()
monkeypatch.setattr(stdio.sys, "stdout", stdout)
monkeypatch.setattr(stdio.sys, "stderr", stderr)
stdio.configure_standard_streams()
expected = [{"encoding": "utf-8", "errors": "backslashreplace"}]
assert stdout.calls == expected
assert stderr.calls == expected
def test_utf8_subprocess_env_overrides_inherited_stdio_encoding() -> None:
env = stdio.utf8_subprocess_env(
{
"PYTHONIOENCODING": "gbk",
"MINERU_TEST_VALUE": "preserved",
}
)
assert env["PYTHONIOENCODING"] == "utf-8:backslashreplace"
assert env["MINERU_TEST_VALUE"] == "preserved"
def test_console_scripts_use_standard_stream_configuring_entrypoints() -> None:
pyproject_path = Path(__file__).resolve().parents[2] / "pyproject.toml"
scripts = tomllib.loads(pyproject_path.read_text(encoding="utf-8"))["project"]["scripts"]
assert scripts["mineru"] == "mineru.cli.main:main"
assert scripts["mineru-kit"] == "mineru.kit.main:main"