In need of a second opinion.
I created an analyzer that generates the tokens by reducing the term to its radical.
Ex (I'm using a Brazilian analyzer, my language is Portuguese):
confeitaria -> confeit
conflito -> conflit
So far so good, my problem is when I start using fuzzy in my search query.
If I search for "confeitaria" I will retrieve both the term "confeitar" (ok) and "confeitaria" (ok) but it also comes "conflito" (which I want to eliminate)
Fuzzy in this case makes it possible to return the term "conflito" because he fixes confeit to conflit,
I don't want to give up fuzzy but in certain situations it gets in the way.
Does anyone have any tips?