[pt] Portuguese rule contribution/discussion

@tiagosantos

Hello Tiago,

I was translating this sentence for DALnet:
"O líder actual da Equipa de Documentação desde 2002, tradutor Alemão oficial"

LT flagged it saying it has no verb.

But, it seems to be well written.

:slight_smile:

Could you take a look?

Thanks!

This precise sentence can not trigger that rule, since it lacks any type of final punctuation.
If it indeed triggered the rule, you saw it with a final stop in the end of it, making it a sentence. It has indeed no verbs that I can recognize.
As such the sentence could be corrected as follows:

“O líder actual da Equipa de Documentação desde 2002 é o tradutor Alemão oficial”
or
“[Ele] é o líder actual da Equipa de Documentação desde 2002, tradutor Alemão oficial”

Oh… I see what you mean. The sentence with the quotes! Nice catch. I will try to add an exception to handle that situation, though it may be tricky.

Many thanks.

Well, I made a few scenarios:

Coelho Tavares é “O líder atual da Equipa de Documentação desde 2002, tradutor Alemão oficial”.
Coelho Tavares “O líder atual da Equipa de Documentação desde 2002, tradutor Alemão oficial” disse que não.
“O líder atual da Equipa de Documentação desde 2002, tradutor Alemão oficial” disse que não.
“O líder atual da Equipa de Documentação desde 2002, tradutor Alemão oficial” — disse Maria.

None of the above trigger the rule. Can you give a realistic example, where the rule should be skipped?

"O líder atual da Equipa de Documentação desde 2002, tradutor Alemão oficial" que não à queixa.

The above triggers but with a good reason.

@tiagosantos

Sorry… I guess you are right, because it was something like this in DALnet:

4.5. FredFred
O líder actual blah blah

It has numbered bullets for each leader, so I didn’t use “É”.

Now I am not sure.

Thank you for analysing it.

Kind regards,

I see. That is indeed another bug. A new line should be always considered as a SENT_END. Changing that may have a lot of unintended conseuqences, so I guess that is why it was not solved so far.
Thank you for your report.
Best regards.

@tiagosantos

Hello!

I have added a new rule:
já mais > jamais.

I added exceptions and done three sentences to check it:

  1. Eu queria já mais documentação. VERB+JÁ MAIS (EXCEPTION)
  2. Há já mais pessoas na sala. HÁ + JÁ MAIS (EXCEPTION)
  3. Eu já mais faria mal a um animal.

This rule only flags 3).

Feel free to enhance it if you find more exceptions.

Thanks!

Good stuff! That and the one that you pushed moment ago (Quanto muito) are great. If possible also add urls to those rules, like:

Note that Wikicheck behaviour was partially reverted yesterday, and it is likely that today’s regression tests will be extremely verbose. To easily verify the rules, search the rule ids in the test result.

@tiagosantos

Hello!

I wanted to add this rule:
“bolsa de estudo” -> “bolsa de estudos

To be sure I used Google translator and typed there:
“I want to get a scholarship”.

The problem is that when I tried this sentence in Portuguese with both “estudo” and “estudos”, it accepted both and did the correct translation (scholarship).

Are both correct?

Thanks!

@tiagosantos

Hello!

I believe there is a false positive with Portuguese postal codes.

For example, LT flags the dash of:
2746–501 Queluz

But I believe I am using the correct type.

Ahhhh…

It also flags subchapters:
A.1. Primeiro Anexo
9.7.1. Subcapítulo

In this case, it says there isn’t a verb in the sentence.

Tiago, I know you are very busy and I am sorry to bring so many issues at the same time.

I believe they are. I use to use and hear “bolsa de estudo”, but I haven’t found any solid reference on this.
In my opinion, if there are many doubts there is no solid rule.

www.ensino.eu
Regulamento de Atribuição de Bolsas de Estudo a Estudantes do …
Bolsa de estudo – Wikipédia, a enciclopédia livre.

I could not reproduce this issue. In which tool is this occuring (LT, standalone, FF)?

Tiago, in LibreOffice.

Ahhh… with last night’s nightly, now it only flags A.1. Primeiro Anexo

:slight_smile:

I will try it, but I can’t reproduce in the standalone, nor in the website.

On the other day, I mentioned the URL, but you already had it on the other rule. My usual hastiness. Really sorry.
However, regarding the rule you pushed today. I was searching for references on that. I couldn’t find any. Can you add an URL to a style book/grammarian/“Ciberdúvidas”? Maybe, that rule should be also moved to ‘Formal speech’.

I took the rule from the usual book, a “Prontuário de Português” which is full of grammar rules and such.

I couldn’t find any URL that I could link to.

The other day rule was also taken from this book but I could find a URL that covered it.

Excellent. This one, I am not familiar with and I haven’t looked it up in my paper grammars. Not sure these cases would be covered anyway.

I was not able to reproduce the issue with 0108 daily build, not even in LibreOffice. Anyway, it would be odd, since the Rule only works for sentences with 3 or more words, 4 or more if they are exclamations. May be related with sentence end recognition, but again, I can’t reproduce the issue on my end.

LO_Tiago_Santos_subchapters_20170802.zip (6.6 KB)

Tiago, here is the sample ODT which I zipped and uploaded.

Even with your file, it doesn’t show any underline there. Both with LO 5.1.6 and LO DEV snap version. Maybe it is Windows related. May be worth posting the file and the bug on GitHub, because I can not reproduce that behaviour.

However I found another bug that may be related. My version is not showing some rule issues. Legibility (less than 45 words) is no longer being detected in the Portuguese language, but it is detected for the same text if is it tagged as English.

For a test example:
“Os dois amigos, de 35 e 34 anos, estavam tão perto das vítimas que assistiram de perto ao impacto do embate e foram dos primeiros a tentar socorrê-los, embora pouco já houvesse a fazer “Foi enquanto a avioneta estava a aterrar que bateu nas pessoas que morreram, ainda não tinha aterrado”, conta Joana Sentada na areia, uns metros acima do local onde a CS-AVA ainda espera ser recolhida, Joana e Filipe têm ainda o chapéu de sol amarelo montados. Marcada pelas imagens que viu, faltam forças a esta psicóloga para agarrar nas suas coisas e pegar no carro.” (from observador.pt)

Can you confirm?

I figure this out. This happens due to rule priority confict in the pt-PT file. I must refactor the specific pt-PT priorities or remove them.

And now it also trigger the "NO_VERB’ rule only in Libreoffice. You were detecting it first since you are using pt-pre-AO (pt-AO), right?
This may be related with issue #669, since it doesn’t happen in the other tools. Not sure if I that can be solved so easily as the other problem.