TextIO,

)

from .. import cliutil

PROGNAME = 'extract-odf-links'

logger = logging.getLogger('conservancy_beancount.tools.extract_odf_links')

def parse_delimiter(arg: str) -> str:

    try:

        retval = eval('"{}"'.format(arg.replace('"', r'\"')), {})

    except SyntaxError:

        retval = None

    if isinstance(retval, str):

        return retval

    else:

        raise ValueError(f"not a valid string: {arg!r}")

def parse_arguments(arglist: Optional[Sequence[str]]=None) -> argparse.Namespace:

    parser = argparse.ArgumentParser(prog=PROGNAME)

    cliutil.add_version_argument(parser)

    cliutil.add_loglevel_argument(parser)

    parser.add_argument(

        '--delimiter', '-d',

        type=parse_delimiter,

        default='\\n',

        help="""String to output between links. Accepts all backslash escapes

""")

    parser.add_argument(

        '--zero', '--null', '-z', '-0',

        action='store_const',

        dest='delimiter',

        const='\0',

        help="""Shortcut for --delimiter=\\0

""")

    parser.add_argument(

        'odf_paths',

        metavar='ODF_PATH',

        type=Path,

        nargs=argparse.ONE_OR_MORE,

        help="""ODF file(s) to extract links from. Note that %(prog)s cannot

read from stdin because it needs to know document paths to resolve links.

""")

    return parser.parse_args(arglist)

def extract_links(odf_path: Path) -> Iterator[Path]:

    odf_root = odf_path.parent.resolve()

    with odf_path.open('rb') as odf_file:

        odf_doc = odf.opendocument.load(odf_file)

    for a_elem in odf_doc.getElementsByType(odf.text.A):

        parts = urllib.parse.urlparse(a_elem.getAttribute('href') or '')