Function _ensure_padded_year

xarray/coding/times.py:171–202

(ref_date: str)

def _ensure_padded_year(ref_date: str) -> str:
    # Reference dates without a padded year (e.g. since 1-1-1 or since 2-3-4)
    # are ambiguous (is it YMD or DMY?). This can lead to some very odd
    # behaviour e.g. pandas (via dateutil) passes '1-1-1 00:00:0.0' as
    # '2001-01-01 00:00:00' (because it assumes a) DMY and b) that year 1 is
    # shorthand for 2001 (like 02 would be shorthand for year 2002)).

    # Here we ensure that there is always a four-digit year, with the
    # assumption being that year comes first if we get something ambiguous.
    matches_year = re.match(r".*\d{4}.*", ref_date)
    if matches_year:
        # all good, return
        return ref_date

    # No four-digit strings, assume the first digits are the year and pad
    # appropriately
    matches_start_digits = re.match(r"(\d+)(.*)", ref_date)
    if not matches_start_digits:
        raise ValueError(f"invalid reference date for time units: {ref_date}")
    ref_year, everything_else = (s for s in matches_start_digits.groups())
    ref_date_padded = f"{int(ref_year):04d}{everything_else}"

    warning_msg = (
        f"Ambiguous reference date string: {ref_date}. The first value is "
        "assumed to be the year hence will be padded with zeros to remove "
        f"the ambiguity (the padded reference date string is: {ref_date_padded}). "
        "To remove this message, remove the ambiguity by padding your reference "
        "date strings with zeros."
    )
    warnings.warn(warning_msg, SerializationWarning, stacklevel=2)

    return ref_date_padded


def _unpack_netcdf_time_units(units: str) -> tuple[str, str]: