How to use the bashlex.parsesingle function in bashlex

def evaluate(value: str, environment: Dict[str, str], executor: Optional[EnvironmentExecutor] = None) -> str:
    if not value:
        # empty string evaluates to empty string
        # (but trips up bashlex)
        return ''

    command_node = bashlex.parsesingle(value)

    if len(command_node.parts) != 1:
        raise ValueError(f'"{value}" has too many parts')

    value_word_node = command_node.parts[0]

    return evaluate_node(
        value_word_node,
        context=NodeExecutionContext(environment=environment, input=value, executor=executor or local_environment_executor)
    )

['VAR=abc']
    >>> split_env_items('VAR="a string" THING=3')
    ['VAR="a string"', 'THING=3']
    >>> split_env_items('VAR="a string" THING=\\'single "quotes"\\'')
    ['VAR="a string"', 'THING=\\'single "quotes"\\'']
    >>> split_env_items('VAR="dont \\\\"forget\\\\" about backslashes"')
    ['VAR="dont \\\\"forget\\\\" about backslashes"']
    >>> split_env_items('PATH="$PATH;/opt/cibw_test_path"')
    ['PATH="$PATH;/opt/cibw_test_path"']
    >>> split_env_items('PATH2="something with spaces"')
    ['PATH2="something with spaces"']
    '''
    if not env_string:
        return []

    command_node = bashlex.parsesingle(env_string)
    result = []

    for word_node in command_node.parts:
        part_string = env_string[word_node.pos[0]:word_node.pos[1]]
        result.append(part_string)

    return result