Flatten parsing result

[BrunoGugli]

This issue is to ask if there is an expression in grammar syntax to turn a capture into a flatten string; here is an example to explain my question:

import tatsu

grammar = r'''
    @@grammar::Base64Command
    @@whitespace :: /[ \t]+/

    start = base64_decode_command;

    target =  'Taking' code:code

    code = { let_dig | special_char }+ ;

    let_dig = letter | digit  ;

    special_char = '+' | '/' | '=' ;

    letter = /[a-zA-Z]+/ ;

    digit = /\d+/ ;
'''
    
    text = 'Taking VGhpcyBpcyBhIGJhc2U2NCBjb2RlIGZvciBhbiBleGFtcGxlCg=='
    
    parser = tatsu.compile(grammar)
    
    parse_result = parser.parse(text)
    
    print("Result: ", parse_result)

The output of this script is:

Result:  {'code': ['VGhpcyBpcyBhIGJhc', '2', 'U', '2', 'NCBjb', '2', 'RlIGZvciBhbiBleGFtcGxlCg', '=', '=']}

What I'm looking for it's an expression that allows to specify in the grammar, that "code" must be flatten, so with that, the output should be this:

Result:  {'code': ['VGhpcyBpcyBhIGJhc2U2NCBjb2RlIGZvciBhbiBleGFtcGxlCg==']}

Obviously there is a way to take "code" as a flatten string, and it's defining code like this:

code = /[a-zA-Z0-9+/=]+/ ;

But my target is to maintain the structure and clarity of the grammar.