Largest palindrome in a string? C ++

The code below searches a phrase for the sequence of characters that form a palindrome, as long as the phrase does not use a comma or accent, for this case you simply include a check for each round of for in the string sub_sentence .

#include <iostream>

using namespace std;

bool check_palindrome(string sentence) {

    string sentence_without_space;

    //Isto é necessário caso queira vê palidromes de várias palavras em uma frase
    for (int i = 0; i < sentence.size(); i++) {

        if (sentence[i] == ' ') {

            continue;
        }

        //Removendo os espaços em branco
        sentence_without_space.push_back(sentence[i]);
    }

    //Verificando se é palindrome
    for (int i = 0, j = sentence_without_space.size() - 1; i <= j; i++, j--) {

        if (sentence_without_space[i] != sentence_without_space[j]) {

            return false;
        }
    }

    return true;
}


string analyze_sentence(string sentence) {

    string greater_palindrome;
    string sub_sentence;
    string sub_sentence_palindrome;
    //---

    //Normaliza a entrada de dados
    for (int i = 0; i < sentence.size(); i++) {

        sentence[i] = tolower(sentence[i]);
    }

    //Percorre toda a sequencia de caracteres da frase, comparando se há um palindrome a partir do caracter atual
    for (int i = 0; i < sentence.size(); i++) {

        //Caso o caracter inicial seja fazio ignora e não verifica
        if (sentence[i] == ' ')
            continue;

        //Selecionando a senquancia de busca atual
        sub_sentence = sentence.substr(i);

        //Percorre do final para o inicio verificando se o conjunto de caracteres entre os indices i e j formam um palindrome
        for (int j = sub_sentence.size(); j > 0; j--) {

            //Caso o caracter final seja fazio ignora e não verifica
            if (sub_sentence[j] == ' ')
                continue;

            //Extrai o possível palindrome
            sub_sentence_palindrome = sub_sentence.substr(0, j);

            //Verifica se é um palindrome
            if (check_palindrome(sub_sentence_palindrome)) {

                //Sendo um palindrome verifica e salva o de maior tamanho
                if (sub_sentence_palindrome.size() > greater_palindrome.size()) {

                    greater_palindrome = sub_sentence_palindrome;
                }
                break; //Aqui você pode encerrar a busca pois este for só irá decrementar o J e procurar por coisa menores ao palindrome atual
            }
        }
    }

    return greater_palindrome;
}

int main() {

    string greater_palindrome = analyze_sentence("Enquanto ele estava seco de raiva coloco no colo caviar e doces!");

    printf("%s\n", greater_palindrome.c_str());

    system("pause");
    return 0;
}

The code is very commented I think it will be easy to understand. In general I analyze all the sequences of characters from the first index of the phrase and always decrementing the size of the sequence investigated. For example, in the phrase 'The blue macaw is beautiful!' start analyzing everything then decrement the size of the string in 1 and then my search is in the string 'The blue macaw is beautiful', then 'The blue macaw is lind', and so on. After doing this, I based on the beginning of the string being the first position, increment the index of beginning of the sequence of characters and make the search from 'blue macaw is beautiful!', And continue decreasing the size of this analysis in 'blue macaw is beautiful'. Whenever a new sequence is formed to be evaluated whether it is a palindrome or not, I check the size of the palindrome to save only the larger one.

Note: Mario's commentary on Manacher's algorithm should be interesting because it should be something more optimized, I did not read it.