How to filter items that do not contain a word in a list?

Question

I have a: List<Grupos> lista = new List<Grupos>();

And I have this code

var txtFiltro = "noivas,unhas";
var palavrasFiltro = txtFiltro.ToLower().Split(',');
matches = lista.Where(x => !palavrasFiltro.Contains(x.Nome_Grupo.ToLower().ToString())).ToList();

This code filters everything that is different from the text I am informing, but only works if you have one word, if you have phrases does not work, as could do the implementation of this filter?

Here is a code that the user @jbueno helped me.

Answer 1

The algorithm presented in link is wrong. You’re looking for the phrase within words, it’s not going to work at all. You can’t look for a larger text inside a smaller one, the larger one will never be inside the smaller one. If this is what you wanted, it’s mathematically impossible and it doesn’t make any sense.

If you are looking for sentences, just use the phrase. The space between words will not disturb anything.

If you want to use a few loose words as a filter, you need to create a more complex filter mechanism. Although you can do 100% on LINQ I find it interesting to create an extension method that solves this (you can even use LINQ on it, if you like). It doesn’t even need to be an extension method, but it is more convenient to use it.

This method has to verify that each of the strings used in the filter is contained within each sentence.

It has more performative ways of doing this, but it’s more complex, I don’t know if it pays.

public static bool ContainsAny(this string haystack, params string[] needles) {
    foreach (var needle in needles) {
        if (haystack.Contains(needle)) return true;
    }
    return false;
}

Behold working in the ideone. And in the .NET Fiddle. Also put on the Github for future reference.

Form with LINQ:

public static bool ContainsAny(this string haystack, params string[] needles) => needles.Any(x => haystack.Contains(x));

In fact, if you want to make sure it’s just exact words, it complicates things. This example (based on the previous one applied, so if it does not do what it wants, the other will not do it either) if you look for "bride", you will find "betrothal", even if you do not want this. This search is not for words, it is for text snippets, it does not observe the syntax of the text.

If I wanted to solve this naively I would have to separate the phrase into words and verify the equality of each one.

Behold working in the ideone. And in the .NET Fiddle. Also put on the Github for future reference.

This does not solve all cases. A Parsing more complete is necessary to meet all cases, then becomes complicated.

Then you must think: if I want "bride" and "bride"? You have to treat it as if it were two words. The same goes for verb conjugation, gender change, etc.

Answer 2

I think the problem is that you are looking for the larger string inside the smaller one, i.e.: you are seeing if the filter contains the phrase and not the other way around.

I suggest:

string[] palavrasFiltro; // mantive array porque é o que você usa...
List<string> lista; // ... mas eu prefiro trabalhar com listas. Suas frases ficam aqui.
List<string> remocao = new List<string>();

foreach (string elemento in lista)
{
    foreach (string filtro in palavrasFiltro)
    {
        if (elemento.Contains(filtro))
        {
            remocao.Add(elemento);
            break;
        }
    }
}

foreach (string s in remocao)
{
    lista.Remove(s);
}