Back to LanguageTool Homepage - Privacy - Imprint

[PT] Help with a rule

Hello guys!
I am doing a research at uni using languagetool and for that I need to explain in detail some rules. However there is one I don’t get to understand and also didn’t find other examples like that. Would someone please help me reading it?
I am having trouble to understand specifically why this token <exception postag='(?:[NZ].|A..)F.+' postag_regexp='yes' negate_pos='yes'/> is negated and as exception, given it’s the exact same token than the previous one.
Here it goes:

<rule id='ERRO_DE_CONCORDNCIA_DO_GÊNERO_MASCULINO_O' name="Concordância de gênero: O + Feminino">
              <token postag='(?:[NZ].|A..)F.+' postag_regexp='yes'>
                <exception postag='(?:[NZ].|A..)F.+' postag_regexp='yes' negate_pos='yes'/>
                <exception regexp='yes'>umas?|e|ser|nome|termo</exception></token>
              <token negate='yes' regexp='yes'>&hifen;</token>
            <suggestion>a \2</suggestion>
            <suggestion>o <match no='2' postag='(A..|N.|Z.)FS(.+)' postag_replace='$1MS$2' postag_regexp='yes'/></suggestion>
          <example correction='A cadela|O cão'><marker>O cadela</marker> está no pasto.</example>
          <example type='correct'>A cadela está no pasto.</example>

Maybe it’s this construct:

Thank you a lot! It’s exactly this, maybe I haven’t checked this page before =]
Also, do you know if LT has a chunker for Brazilian Portuguese? I’m no sure how to check this information.

You’d need to look in the code to find out (class BrazilianPortuguese or Portuguese), but the answer is “no”.