Function get_unicode_xml_data

scripts/test_helpers.py:28–59

Download and parse Unicode UCD XML, caching in temp directory. Args: version: Unicode version string (e.g., "17.0.0") Returns: Root Element of the parsed XML tree

(version: str = UNICODE_VERSION)

def get_unicode_xml_data(version: str = UNICODE_VERSION) -> ET.Element:
    """Download and parse Unicode UCD XML, caching in temp directory.

    Args:
        version: Unicode version string (e.g., "17.0.0")

    Returns:
        Root Element of the parsed XML tree
    """
    cache_path = os.path.join(tempfile.gettempdir(), f"ucd-{version}.all.flat.xml")

    if not os.path.exists(cache_path):
        url = f"https://www.unicode.org/Public/{version}/ucdxml/ucd.all.flat.zip"
        print(f"Downloading Unicode {version} UCD XML from {url}...")
        try:
            with urllib.request.urlopen(url) as response:
                zip_data = response.read()
            zip_bytes = io.BytesIO(zip_data)
            with zipfile.ZipFile(zip_bytes) as zf:
                xml_filename = zf.namelist()[0]
                with zf.open(xml_filename) as xml_file:
                    xml_content = xml_file.read()
                    with open(cache_path, "wb") as f:
                        f.write(xml_content)
            print(f"Cached to {cache_path}")
        except Exception as e:
            raise UnicodeDataDownloadError(f"Could not download UCD XML from {url}: {e}")
    else:
        print(f"Using cached Unicode {version} UCD XML: {cache_path}")

    tree = ET.parse(cache_path)
    return tree.getroot()

def get_all_codepoints(version: str = UNICODE_VERSION) -> List[int]:
    """Return all assigned/defined codepoints in Unicode."""