HunSpell, remove words from .dic

Terry R

@Alan-Kilborn said in HunSpell, remove words from .dic:

I know we’re being annoyed.

AFAIK Thomas-2020 WAS Pan-Jan. If you look at the first posting under Thomas the replies all referenced Pan-Jan. So is it possible to change one’s forum ID? I then wonder how the forum allows for a new ID to be made from one that “already existed” before albeit a renamed one.

I was formulating a response to Pan-Jan. I am willing to give him the benefit of the doubt, however I have been watching the responses from other forum members, lots of them seasoned ones whom I trust judgement from. It would seem we are all thinking along the same lines.

Terry

PeterJones

Oddly coming to @Thomas-2020 / @Pan-Jan 's defense:

The original Hunspell thread died when T2/PJ said that he couldn’t edit the dictionary and no one replied
The Hunspell/Aspell tangent of “Add the icon to the toolbar” devolved into a … discussion … on native language and translators.

So it may be that there’s still something that T2/PJ doesn’t understand about the process.

@Thomas-2020: Guy went to a lot of effort to explain to you how to edit a Hunspell dictionary in the post at https://community.notepad-plus-plus.org/post/56134 . And then pointed you to the Aspell dictionary spec in the post at https://community.notepad-plus-plus.org/post/56145 .

In what way did those not answer your question? What steps are you using to try to edit the dictionary? And what happens to tell you that “it didn’t work”? Please give us details as to how you are having problems. If you show effort, then we will try to help you. But if you keep trying to hide behind name changes, and blaming translation for your lack of explanation, we won’t be able to or willing to help you. Help us to help you.

Pan Jan

Where in this post is the answer to removing words from .dic or Convert?
https://community.notepad-plus-plus.org/topic/19650/hunspell?_=1596716637917

What’s wrong with asking again?
Maybe someone who knows something about it will look here now?

Why can’t I change the login to my real name?
“Pan. Jan” was stupid, but I had to register
and I had no choice.

I opened a topic for outside help.
It turns out I’m a troll.

“Ekopalypse” knows Polish and instead of helping with explanation
ambiguities, this exacerbates the conflict.

Wouldn’t it be easier to ask why I changed my login?
Why does he think I’m hiding under a new login.

What else should I write?
If you don’t like what I write here, I’ll stop.
What’s the problem?

PeterJones

@Pan-Jan ,

You are allowed to change your username; I did (once).

If you thought the “Pan Jan” name was stupid, why did you create a new account with that name, rather than continuing to use @Thomas-2020? And why are you using the new @Pan-Jan account rather than the @Thomas-2020 account?

Where in this post is the answer to removing words from .dic or Convert?

Nowhere. I said you didn’t get the answer there.

But I did point you to the other thread, with two obvious links, where Guy explained exactly how to edit the dictionary files:

edit hunspell: https://community.notepad-plus-plus.org/post/56134
aspell spec: https://community.notepad-plus-plus.org/post/56145

To delete the word from the Hunspell dictionary, did you try to edit the file and remove the word? After saving the file, then exiting and re-starting Notepad++, was the word still in the dictionary, or was it removed?

What’s the problem?

You fight with us. You don’t listen to what we say. You make excuses.

If you want help, show effort. Try the things we recommend. Read the posts that we make. Reply to direct questions.

Pan Jan

why did you create a new account with that name

The profile itself changed to “Pan Jan” (or maybe someone helped). I didn’t rummage there.

It has been activated “associate with Google”
There is Pan Jan in “Googlach”

My questions are not answered here.

I did exactly what Guy wrote about Hunspell
and still not working

Aspell writes to pl.pws

and it doesn’t work either.

That’s why I opened a new post, maybe someone new here will know how to do it.
I have been trying to solve this problem for several years.
I tried on Polish and German forums,
and unfortunately nothing came of it

Reply to direct questions.

I give my best thanks to “Google Translator”
and that’s how they understand me on this forum You fight with us and You make excuses

Believe me, I have no evil intentions and do not want to offend anyone.

I want to learn to use RegEx,
and this forum has the best specialists.

PeterJones

@Pan-Jan ,

I asked previously,

To delete the word from the Hunspell dictionary, did you try to edit the file and remove the word? After saving the file, then exiting and re-starting Notepad++, was the word still in the dictionary, or was it removed?

Please answer those questions.

Pan Jan

I removed it abaka/MnN

I also changed 249884 to 249883
it’s still there abak after restarting Npp

What if I ever want to retype a word abaka
So how do I know these signs have to be too /MnN?

Someone changed something that the profiles no longer work.
I will have to delete my account.

PeterJones

I installed DSpellCheck and downloaded the standard US dictionary.

Version:

I am using the Hunspell library, in case that makes a difference:

I scrolled to the end of the en_US.dic, and saw zygotic as a good candidate.

By default, it says the word zygotic is correct:

If I delete line 49524, save, exit Notepad++ completely, and come back in, it now says that zygotic is misspelled:

So I tried the Polish dictionary (mine has more words than yours listed, by the way… it may have been updated):
Before editing: abaka is okay =

Delete line 31. Exit Notepad++. Reload Notepad++. Hmm, it still shows abaka as okay =

When I look down a few lines, there is what’s now my line 37 abak/NOsT, which is a shorter word with the same first letters. I am guessing the stuff after the / is defining something about prefixes and suffixes, so I wonder if one of those settings allows abaka to be a valid suffix version of abak (I know nothing linguistically about Polish, so it’s a wild guess).

I removed lines 37-38 (abak and abakus), save, exit, reload =

Now it shows abaka as misspelled. So DSpellCheck seems to be accepting abaka as a result of having abak/NOsT in the dictionary. Since I don’t know Polish, I don’t know if that’s right.

Unfortunately, I don’t know what the characters after the / specifically mean, so I don’t know how to interpret what kind of prefixes and suffixes it will allow. It might be possible to edit that to not accept abaka but still accept abak and other real versions of that word. But I don’t know how. If someone else is a DSpellCheck power user, I encourage them to chime in.

Otherwise, I am at-mentioning @Predelnik, the author of DSpellCheck, but his profile says he hasn’t logged on here in a year. If you have a question specifically about DSpellCheck and what the meaning of the characters after the / are in the dictionary file, you might want to open an issue at https://github.com/Predelnik/DSpellCheck/issues .

Ekopalypse

@PeterJones - DSpellCheck is just the tool which uses Hunspell or Aspell under the hood. Hunspell format is explained here.

@Pan-Jan

“Ekopalypse” knows Polish

I don’t know anything about Polish - my translator does.

instead of helping …

Since I was probably the one who fueled the “you’ve already received the answer” thread,
I also feel responsible for clarifying the circumstances.

I don’t see you adding value in any way to this forum.
On the contrary, I think you are trolling.

Every forum in which I am a part of does not like it if a user posts several threads to a question.
Because every new thread makes it more difficult for subsequent users
to gather all the information about this question.
But it’s no problem at all to continue discussing the topic in the already opened thread.
Also for the 100th time, if it must be.

To the current question @PeterJones has already posted the links
where they tried to provide you with the information.
My opinion on this - that should have been known to you as the
questioner,
why did another forum member have to post this again?
Sure you are not a troll?
It has been explained to you several times already that if something
does not work, you need to provide additional information.
Information that helps to clarify what you did, what you expected and
what the result was. This is rarely or never the case with you.
Only after re-asking and re-asking you come with information but sparsely.
Another thing you like to do, you change the direction of questions within an open discussion.
In my opinion, these are tactics that are only used by trolls,
because a questioner who is interested in a solution will do everything
to support the helpers and not to confuse them further and further.

My opinion - you are a troll, but you are free to prove that this is not the case.

Alan Kilborn

@Ekopalypse said in HunSpell, remove words from .dic:

I don’t see you adding value in any way to this forum.

LOL – and well-said.

I like the fact that on the flip side of things, the last post of @PeterJones was packed with good information for others interested in learning some things.

Pan Jan

Thanks, the deletion worked, but it only works one way.

What if I ever want to re-enter abaka into my dictionary.
So how do I know that such signs abaka/MnN? must also be?

To enter this, a new (additional dictionary - UserDic.dic) is created.
I don’t have pl_PL.dic full control over it.

That’s why I switched to Aspell,
because I have full control over ** pl.rws **.
But this time there is another problem

does not work in Aspell.

If this could be resolved, the topic would be closed to me.

PeterJones

@Pan-Jan said in HunSpell, remove words from .dic:

Thanks, the deletion worked, but it only works one way.
What if I ever want to re-enter abaka into my dictionary.
So how do I know that such signs abaka/MnN? must also be?

Deletion is a one-way operation. If you want to save the information you deleted (liked that abaka was paired with /MnN), instead of just deleting it, cut it out of the main dictionary and store it in some “archive” file.

To enter this, a new (additional dictionary - UserDic.dic) is created.
I don’t have pl_PL.dic full control over it.

That is the design of that tool, yes.

[add to dictionary] does not work in Aspell.
If this could be resolved, the topic would be closed to me.

I don’t have the Aspell library, so I cannot test that variant.

Maybe someone else here has the Aspell library, and can help you with that portion.

However, at this point, I’d recommend opening two issues at the DSpellCheck plugin’s github (https://github.com/Predelnik/DSpellCheck/issues)

Ask if there is documentation for what the /MnN means in a dictionary entry
Report the bug you experienced in Add [word] to Dictionary when using Aspell.

I suggest keeping the two issues separate, because they are two unrelated topics.

If you create those issues, I recommend you come back here and paste the URL’s to the two issues, so that anyone else reading this topic in the future will be able to find them.

Pan Jan

Thanks,
https://github.com/Predelnik/DSpellCheck/issues
I found nothing there
like abaka/MnN?
I’m sure I need special DOS commands for this.

I also believe that this post was needed.
The forum is visited by people with different knowledge.

PeterJones

@Pan-Jan said in HunSpell, remove words from .dic:

I found nothing there

I said you needed to create the two issues there.

Go to the issues page. (https://github.com/Predelnik/DSpellCheck/issues)

Przejdź do strony z problemami.
Create a new issue. Add the text (in English), “In the Hunspell dictionary files, what do the symbols after the /, like /MnM mean? Is it documented somewhere”. Submit this issue.

Stwórzcie nowy numer. Dodaj tekst (w języku angielskim): “In the Hunspell dictionary files, what do the symbols after the /, like /MnM mean? Czy jest to gdzieś udokumentowane”. Prześlij ten numer.

By “numer”, I mean “issue” or “problem report” or “request for help”

przez “numer” mam na myśli „problem”, „zgłoszenie problemu” lub „prośba o pomoc”
Create a second issue. Add the text (in English), “When I use the Aspell library, and I try to add a word to the dictionary using the right click ‘Add … to Dictionary’, the word does not get added to the dictionary.”

Stwórz drugi numer. Dodaj tekst (w języku angielskim), “Kiedy używam biblioteki Aspell i próbuję dodać słowo do słownika używając prawego przycisku myszy ‘Add … to Dictionary’, słowo to nie zostanie dodane do słownika”.

like abaka/MnN?
I’m sure I need special DOS commands for this.

I’m not sure why you think that a line of text in the dictionary file would need a special DOS command.

Nie jestem pewien, dlaczego uważasz, że linia tekstu w pliku słownika wymagałaby specjalnego polecenia DOS.

----
I have used DeepL.com to translate to Polish. When I take the resulting Polish and translate it back to English, it mostly makes sense to me, except it used “numer” in Polish (“number” in English) when I used “issue” in English. Other than that, I have to assume that the Polish itself matches my intent.

Użyłem DeepL.com do tłumaczenia na język polski. Kiedy zabieram wynik po polsku i tłumaczę go z powrotem na język angielski, ma to dla mnie głównie sens, z wyjątkiem tego, że użyłem “numer” w języku polskim (“numer” w języku angielskim), kiedy użyłem “issue” w języku angielskim, więc dodałem tekst wyjaśnienia powyżej. Poza tym, muszę założyć, że sam język polski jest zgodny z moimi intencjami.

Pan Jan

I’m not sure why you think that a line of text in the dictionary file would need a special DOS command
I use the analogy to ASpell … there I use DOS.

It is a pity that there is nobody here who knows Polish or German.
I can’t explain it correctly.
I propose to finish the topic.

I have used DeepL.com to translate to Polish.
However, Google Translator does it better.

PeterJones

@Pan-Jan said in HunSpell, remove words from .dic:

I can’t explain it correctly.
I propose to finish the topic.

Maybe @Ekopalypse can chime in with some German on this one. I am assuming that he understands what I’ve tried to tell you. Aber meine Schule Deutsch von vor Jahrzehnte is nicht gut genau. (I used google translator for “from decades ago” )

PeterJones

@Ekopalypse ,

Specifically, I don’t think he understood that I was suggesting it would be best if he added two issues to DSpellCheck (as listed above).

I know he claimed this topic was done, but I actually hoped he’d get an answer publicly for this one, because it’s actually useful to others to know what the extra characters in the dictionary mean. And if he is right and Add To Dictionary doesn’t work with Aspell, then that’s a definite bug that needs to be reported.

edit Or, since you’ve apparently used DSpellCheck more than I have (at least, you submitted a bug report), maybe you could verify the Aspell bug, and if so, report it yourself./edit

PeterJones

So I did a bit of searching, and found that the Hunspell library has a man 4 hunspell manpage describing the dictionary file. You can see it online on sites like https://linux.die.net/man/4/hunspell

Apparently, if I understand correctly, the stuff after the / can either be literal text, or it can be characters representing prefixes and/or suffixes defined in an associated file. In DSpellCheck, the *.aff files have that prefix/suffix-defining syntax.

Looking at the pl_PL.aff file to try to figure out /MnN syntax. The SFX M section of that file seems to be one set of rules for when it’s okay to add a to the end of the word, and how to modify it. SFX n is five rules for a/ja/ia, and SFX N … I cannot tell what it’s trying to do.

So my suggestion for the first issue is no longer relevant: hunspell.4 defines this syntax, and someone with time and inclination could study that man page.

Ekopalypse

@Pan-Jan, ich bitte Dir an das wir dein Problem nochmal ausführlich diskutieren und bin bereit Dir dies in Deutsch zu erläutern.
Ich erwarte aber, dass Du da mitarbeitest, soll heißen ich werde auf konkrete Frage antworten.
Ich werde aber nicht deine Hand halten und dich Schritt für Schritt zur Lösung begleiten.

Meine erste Frage hierzu wäre, was ist eigentlich das Problem.
Hinweis: ich erwarte keine Antwort ala “ich will ein Wort aus einem Wörterbuch entfernen” sondern eher "mein Ziel ist es mit Applikation A
das und jenes zu machen und muß hierbei das polnische Wörtbuch ändern. Die muß ich während der Laufzeit oder bei jedem Start … was auch immer - weißt Du worauf ich hinaus ziele? Nur wenn wir verstehen was das eigentliche Ziel ist können wir auch vernünftige Lösungen zu einem Problem geben.

Nun bist Du an der Reihe.

@Pan-Jan, please let us discuss your problem in detail and I am ready to explain it to you in German.
But I expect you to cooperate, that means I will work on answer a specific question.
But I will not hold your hand and accompany you step by step to the solution.

My first question would be, what is actually the problem.
Note: I don’t expect an answer like “I want to remove a word from a dictionary” but rather "my goal is to use application A
to do this and that and must change the Polish dictionary. I have to do that during runtime or at every start … whatever - you know what I’m aiming at? Only if we understand what the real goal is we can give reasonable solutions to a problem.

Now it is your turn.

Ekopalypse

@PeterJones said in HunSpell, remove words from .dic:

So I did a bit of searching, and found that the Hunspell library has a man 4 hunspell manpage describing the dictionary file. You can see it online on sites like https://linux.die.net/man/4/hunspell

:-(
You missed my link, didn’t you?