Function match_column_identifier

csvkit/cli.py:487–512

Determine what column a single column id (name or index) matches in a series of column names. Note that integer values are *always* treated as positional identifiers. If you happen to have column names which are also integers, you must specify them using a positional index.

(column_names, c, column_offset=1)

def match_column_identifier(column_names, c, column_offset=1):
    """
    Determine what column a single column id (name or index) matches in a series of column names.
    Note that integer values are *always* treated as positional identifiers. If you happen to have
    column names which are also integers, you must specify them using a positional index.
    """
    if isinstance(c, str) and not c.isdigit() and c in column_names:
        return column_names.index(c)

    try:
        c = int(c) - column_offset
    # Fail out if neither a column name nor an integer
    except ValueError:
        raise ColumnIdentifierError("Column '%s' is invalid. It is neither an integer nor a column name. "
                                    "Column names are: %s" % (c, repr(column_names)[1:-1]))

    # Fail out if index is 0-based
    if c < 0:
        raise ColumnIdentifierError("Column %i is invalid. Columns are 1-based." % (c + column_offset))

    # Fail out if index is out of range
    if c >= len(column_names):
        raise ColumnIdentifierError("Column %i is invalid. The last column is '%s' at index %i." % (
            c + column_offset, column_names[-1], len(column_names) - 1 + column_offset))

    return c


def parse_column_identifiers(ids, column_names, column_offset=1, excluded_columns=None):