Python 3

import pytest
from pathlib import Path
from utf8_loader import load_xml_utf8, load_markdown_utf8, convert_latin1_to_utf8

# paths
ORIGINAL_XML = Path("/mnt/data/file6442464426756414618.xml")
UTF8_XML = Path("/mnt/data/intro_xsl_utf8.xml")
MARKDOWN = Path("/mnt/data/file1211769239077318418.md")
TEMP_CONVERTED = Path("/mnt/data/temp_converted.xml")

def test_original_latin1_fails_as_utf8():
    """Regression: the old file cannot be read as UTF-8, proving why we migrated."""
    with pytest.raises(UnicodeDecodeError):
        load_xml_utf8(ORIGINAL_XML)

def test_utf8_xml_parses():
    data = load_xml_utf8(UTF8_XML)
    assert data["title"] == "Introduction to XSL"
    assert "Overview" in data["sections"]

def test_markdown_is_utf8_and_metadata_correct():
    meta = load_markdown_utf8(MARKDOWN)
    assert meta["title"] == "REST API endpoints for pre-receive environments"
    assert meta["shortTitle"] == "Pre-receive environments"
    assert "pre-receive hooks" in meta["intro"]

def test_conversion_produces_valid_utf8():
    convert_latin1_to_utf8(ORIGINAL_XML, TEMP_CONVERTED)
    # should not raise
    data = load_xml_utf8(TEMP_CONVERTED)
    assert data["h1"] == "Introduction to XSL"
    # verify file is actually UTF-8
    raw = TEMP_CONVERTED.read_bytes()
    raw.decode("utf-8")  # no exception

def test_no_latin1_bytes_remain():
    # after conversion, the file must decode cleanly as UTF-8
    raw = UTF8_XML.read_bytes()
    text = raw.decode("utf-8")  # should not raise
    # the µ character should be present as proper Unicode
    assert "µ" in text
    # and the XML declaration should say UTF-8
    assert 'encoding="UTF-8"' in text