Hunspell dictionaries

[Igor Kmitowski]

Hello again,

The hunspell dictionaries are probably already installed in your system (with the other software I mentioned in the update announcement). You can also download them from here: http://extensions.services.openoffice.org/dictionary. They are packed as extensions (.oxt files). After downloading, rename the .oxt file to the .zip file and unpack. Next, copy the two files for your target languages .dic and .aff (e.g. en_US.dic and en_US.aff) to Cafetran's resources/spellchecker/ folder inside CT installation. After running CafeTran you should see the dictionaries visible in the menu Edit | Spell checker.

Enjoy your mistakes ;)
Igor

[Hans list]

Igor Kmitowski wrote:
>
> After running CafeTran you should see the dictionaries visible in the menu Edit | Spell checker.
>

Hi Igor,

Thanks again for this great improvement. I've deactivated the OpenOffice connection:

> Edit | Options | Connection to: - select the second (empty) choice and restart.

I've placed nl_NL.dic and nl_nl.aff in the /Applications/CafeTran.app/Contents/Resources/Java/resources/spellchecker/ folder: the spell checker works fine.

Is it possible to see/change Hunspell's checking modus (e.g. Ignore words with numbers)?

Is it possible to quickly add a word to Hunspell's dictionary of unknown words?

Thanks,

Hans

[Selcuk Akyuz]

Unfortunately Turkish is not included in this list. It is surprising that I can use some other Turkish resources in MemoQ but it does not work in CT.  I cannot attach that file here, it is 4 MB. But what can be the problem?

I tested the English spell checker, I can see the red underlined words and the proposed words when I right-click. Though the menu key does not work (yes I use Windows).

Selcuk

[Hans list]

Selcuk Akyuz wrote:
>
> I can see the red underlined words and the proposed words when I right-click

thanks for the hint!

an option, add to spelling dictionary here would be greatly appreciated

saying this although I almost never make a spellig error

[Hans van den Broek]

On 3 Jul 2012, at 23:31, Selcuk Akyuz wrote:

Unfortunately Turkish is not included in this list.

That's silly. Turkish is a pretty important language, both because of the number of native speakers, and its cultural and economic value. Does the spell checker that comes with OOo still work? You can choose Turkish in OOo, and therefore I suppose it's included in the LanguageTool. Turkish is also available in Aspell (http://cocoaspell.leuski.net), so a (free) solution seems to be possible.

BTW HansL, you may want to have a look at http://extensions.services.openoffice.org/en/project/Linguist

Cheers,

Hans

-- 

Hans van den Broek

Schrijf-, vertaal- en redigeerwerk

Peleman Rejowinangun KG1/513

RT029 RW009

Yogyakarta 55171

Indonesia

T  +62 (0)85868607333

SKYPE: hanstranslations

transl...@hansvandenbroek.eu

http://cafetran4mac.blogspot.com

[Selcuk Akyuz]

Hi Hans,

Igor solved the problem :) After renaming both files from tr to tr_TR, now spell check works. Turkish users can download Hunspell dictionaries from http://code.google.com/p/tr-spell/downloads/list I use the Firefox version.

Also tested Zemberek Turkish spell checker extension in LibreOffice. It works but I don't want an instance of LibreOffice running in the background.

Selcuk

[Hans van den Broek]

On 4 Jul 2012, at 08:32, Selcuk Akyuz wrote:

Igor solved the problem :) After renaming both files from tr to tr_TR, now spell check works. Turkish users can download Hunspell dictionaries from http://code.google.com/p/tr-spell/downloads/list I use the Firefox version.

I'm just writing about it for my blog. Will add this information. TRIMS! (that's SMS Speak for plural "terima kasih" in Indonesian, or just plain Thank you!)

[Hans list]

Hans van den Broek wrote:
>
> you may want to have a look at http://extensions.services.openoffice.org/en/project/Linguist

The extension didn't work for me. The macro listed on the extension site () doesn't work either. This modified version works: [see below]


However: I'm still looking for the file and location as to where CafeTran's Hunspell adds words with unknown spelling!!!!!

If I can edit that file manually, I can easily trhansmigrate my old Transit list of unknown words. What they are, I cannot say – since they are, er, unknown.

====

Sub WrongWordsList
oDocModel = ThisComponent
If IsNull(oDocModel) Then
MsgBox("There's no active document.")
Exit Sub
End If

If Not HasUnoInterfaces (oDocModel, "com.sun.star.text.XTextDocument") Then
MsgBox("This document doesn't support the ‘XTextDocument' interface.")
Exit Sub
End If

oTextCursor = oDocModel.Text.createTextCursor()
oTextCursor.gotoStart(False)

oLinguSvcMgr = createUnoService("com.sun.star.linguistic2.LinguServiceManager")
If Not IsNull(oLinguSvcMgr) Then
oSpellChk = oLinguSvcMgr.getSpellChecker()
End If
If IsNull (oSpellChk) Then
MsgBox("It's not possible to access to the spellcheck.")
Exit Sub
End If

Do
If oTextCursor.isStartOfWord() Then
oTextCursor.gotoEndOfWord(True)
oCharLoc = oTextCursor.getPropertyValue("CharLocale")
If Not isEmpty (oCharLoc) Then
If Not oSpellChk.com_sun_star_linguistic2_XSpellChecker_isValid(oTextCursor.getString(),oCharLoc, Array()) Then
sListaPalabras = sListaPalabras + oTextCursor.getString() + Chr(13)
End If
End If
oTextCursor.collapseToEnd()
End If
Loop While oTextCursor.gotoNextWord(False)

If Len(sListaPalabras) = 0 Then
MsgBox("There are no errors in the document.")
Exit Sub
End If

oListDocModel = StarDesktop.loadComponentFromURL("private:factory/swriter", "_default", 0, Array())
oListDocModel.Text.String = sListaPalabras
oListDocModel.CurrentController.Frame.activate()

End Sub

However, AFAIR the macro crashes on extremely large documents, containing thousands of unknown words (at least that was what I noticed when I tried to easily add my CafeTransit list).

[Igor Kmitowski]

Hi Hans,

> Is it possible to see/change Hunspell's checking modus (e.g. Ignore words with numbers)? 

I don't know. You may ask some Hunspell guys if this is possible.

> Is it possible to quickly add a word to Hunspell's dictionary of unknown words? 

The update has a user.dic file located in the resources/spellchecker folder. You may add your unknown words there. I don't know if there is any limit to the number of words that you can add. The first line in this file is number which informs about the number of entries. You may increase it accordingly but do not delete it by mistake.

Igor 

[Hans list]

Hi Igor,

Thanks for your quick answer!

> Is it possible to see/change Hunspell's checking modus (e.g. Ignore words with numbers)? 

I don't know. You may ask some Hunspell guys if this is possible.

I guess this is possible, since e.g. OpenOffice allows for quite a bunch of useful settings:

> Is it possible to quickly add a word to Hunspell's dictionary of unknown words? 

The update has a user.dic file located in the resources/spellchecker folder. You may add your unknown words there. I don't know if there is any limit to the number of words that you can add. 

For OO it is 64 K – not sure whether this is the total number of lines or the file size.

Okay, I'll do some experimenting.

Cheers,

Hans

(enjoying a day off with the kids in the garden)

[Hans list]

Hans list wrote:
>
>> The update has a user.dic file located in the resources/spellchecker folder. You may add your unknown words there. I don't know if there is any limit to the number of words that you can add.
>

Igor,

Is it by design that the words in the User.dic aren't considered when correcting typos with Ctrl+space (or is it Cmd+space)?

Hans
(always asking the impossible)

[Hans list]

>>>
>>>
>>> I don't know if there is any limit to the number of words that you can add.
>>
>

I've found this:

Main features of Hunspell spell checker and morphological analyzer:

- Unicode support (affix rules work only with the first 65535 Unicode characters)

- Morphological analysis (in custom item and arrangement style) and stemming

- Max. 65535 affix classes and twofold affix stripping (for agglutinative
languages, like Azeri, Basque, Estonian, Finnish, Hungarian, Turkish, etc.)

Does this mean that the User.dic can have a size of 65535 KB?

Hans

[Igor Kmitowski]

Yes, the typos are autocorrected (CTRL + Space) with words form the main
dictionary. I am not sure if autocorrecting with two dictionaries at the
same time is possible.

> Is it by design that the words in the User.dic aren't considered when
> correcting typos with Ctrl+space (or is it Cmd+space)?
>
> Hans
> (always asking the impossible)

--
Igor Kmitowski
Translator and Java developer
CafeTran website: http://www.cafetran.com
CafeTran support: cafetran...@gmail.com

[Igor Kmitowski]

> I've found this:
>
> Main features of Hunspell spell checker and morphological analyzer:
>
> - Unicode support (affix rules work only with the first 65535 Unicode
> characters)
>
> - Morphological analysis (in custom item and arrangement style) and
> stemming
>
> - Max. 65535 affix classes and twofold affix stripping (for agglutinative
> languages, like Azeri, Basque, Estonian, Finnish, Hungarian, Turkish,
> etc.)
>
> Does this mean that the User.dic can have a size of 65535 KB?

The info says about the limit for affixes. It does not say anything about
the max. number of words. However, this may imply the number of words as
well.

Igor

[Wolfgang Schoene]

Hi
why is there no resources/spellchecker/ folder inside my CT installation?
Do I have to create it?

Wolfgang


Am 03/07/12 17:14 schrieb "Igor Kmitowski" unter <cafetran...@gmail.com>:

[Hans van den Broek]

On 4 Jul 2012, at 16:58, Wolfgang Schoene wrote:

why is there no resources/spellchecker/ folder inside my CT installation?

There is, but you should go a few steps further. The folder isn't in "Resources," but in "resources." See my blog...

[Hans list]

Igor Kmitowski wrote:
>
> The info says about the limit for affixes. It does not say anything about the max. number of words. However, this may imply the number of words as well.

Well, I'll use my OOw macro to extract unknown words from the Dutch column of my big papa. First I'll have to split it in chunks. Will see how many words are not in Hunspell's list already :). It is pretty good, I noticed.

[Hans list]

Igor Kmitowski wrote:
>
> I am not sure if autocorrecting with two dictionaries at the same time is possible.

Hi Igor,

I guess, when OO can do this (just checked) CafeTran surely can do it ;). Is there a Java forum for Hunspell where one can ask questions like this (and about ignoring words with numbers/allow lower case only)?

I'll write an KB article about combined use of user.dic by OOw and CafeTran, since there are some complications there.

Cheers,

Hans

[Igor Kmitowski]

Hi Hans,

> I guess, when OO can do this (just checked) CafeTran surely can do it ;).

Autocorrection with the user dictionary will be added in the next update.

> Is there a Java forum for Hunspell where one can ask questions like this
> (and about ignoring words with numbers/allow lower case only)?

I don't know. But I think it is a developer task to implement ignoring
words with numbers and allow lower case only. Let it be RFE.

Cheers,
Igor

[Wolfgang Schoene]

...sometimes drives me insane.

Hi All
having played around with windows layouts and not having found one that
meets my requirements, the question is, how to go back to the original
supposedly default layout. I won't take "Go to Window -- Window layout 1
(default)" for an answer ;-) as by no means it returns to the very first
layout nor do the other options. Hence my first remark. ;-)

Best regards
Wolfgang

[Hans list]

Too late for you this time, Wolfgang, but for next time before you start experimenting: make a backup of the Java prefs file: com.apple.java.util.prefs.plist

[Wolfgang Schoene]

... et maintenant... que vais je faire ...?

Thanks Hans,
so I'm left with reinstalling CT?

Wolfgang

Am 09/07/12 10:44 schrieb "Hans list" unter <hans...@gmail.com>:

[Igor Kmitowski]

Hi Wolfgang,

To come back to the default layout:

1. Window layout 1 (default)

Based on this default layout revert the other options that you may have
changed.

1. Horizontal/Vertical segment windows.
2. Reverse window layout.
2. Flip segment windows.

If that doesn't help, send a screenshot of your current layout to help you.

Igor

[Hans list]

On Jul 9, 2012, at 10:49 AM, Wolfgang Schoene wrote:

> ... et maintenant... que vais je faire …?

I think you'd take Igor's advice best …

That being said, it would be interesting to know where CafeTran stores its windows settings in the plist file (which can be edited with TextWrangler).

BTW: If you delete the existing plist file, CafeTran will create a new one (that of course has default settings).

[Wolfgang Schoene]

Hi Igor
all good and well, but how can I tell what I may have selected, there are no
check signs anywhere.
I'd say somewhat ironically that one shouldn't underestimate the fun factor
contained within CT, given one has enough time and patience.
Sorry but there's no way to come back to the original.
Will send you a screenshot of one of the layout proposals and use another
tool in the meantime.

Wolfgang

Am 09/07/12 11:18 schrieb "Igor Kmitowski" unter
<cafetran...@gmail.com>:

[Wolfgang Schoene]

Hi Igor
attached hereto the screenshot of one of the random layouts.

Wolfgang



____________________________
Wolfgang Schoene
138 Bd de Provence
06700 Saint-Laurent-du-Var
Cell: +33687923918
Technical translator
Voice over speaker
Adaption-Subtitling
DE-IT-FR-EN
in...@soundandword.com
www.soundandword.com
Member of: SFT/ITI/DTT/ADAP/ATAA
Skype: soundandword
Twitter: @WSchoene




Am 07/07/12 11:16 schrieb "Igor Kmitowski" unter
<cafetran...@gmail.com>:

[Wolfgang Schoene]

Sorry for this one, it was meant for support.

Wolfgang


Am 09/07/12 14:44 schrieb "Wolfgang Schoene" unter <soun...@gmail.com>:

[Wolfgang Schoene]

Hi Hans,
can you give me a hint where I can find the Java prefs file. Since nothing
helped to change the CT windows settings back to default, I'm afraid I will
have to use heavy weapons and start from scratch with the prefs file.

Thank you

Wolfgang


Am 09/07/12 10:44 schrieb "Hans list" unter <hans...@gmail.com>:

[Wolfgang Schoene]

Hi again
I've found it, it was hidden.

Wolfgang


Am 10/07/12 08:53 schrieb "Wolfgang Schoene" unter <wsch...@gmail.com>:

[Igor Kmitowski]

Hi Wolfgang,

To restore your layout as seen on the screenshot to default:

1. Go to Edit | Window | Flip Segment Windows.
2. Drag down and left/right the segments windows to resize them.
3. If you wish to undock any previously docked tabs, click the docking
icons (the last ones to the right at the window toolbars)

I think, the next CT update should include restoring to default all window
layout options.

Cheerio,
Igor

[Wolfgang Schoene]

Thanks Igor,
but nothing helps.
Best is I reinstall the last release and save my java preference file before
playing around.
I suppose you will have to send me a new .jar file after reinstalling?

Wolfgang


Am 10/07/12 09:08 schrieb "Igor Kmitowski" unter
<cafetran...@gmail.com>:

[Igor Kmitowski]

You don't have to reinstall. Just remove this java preferences file and
restart the program.

Igor

[Wolfgang Schoene]

Done! Thanks a lot and yes, you really should create an Undo to default
button.

Best regards
Wolfgang



Am 10/07/12 09:28 schrieb "Igor Kmitowski" unter

[Hans list]

On Jul 10, 2012, at 8:53 AM, Wolfgang Schoene wrote:

> I'm afraid I will
> have to use heavy weapons and start from scratch with the prefs file.

1. Make screen shots first

OR

2. Open the prefs file with TextWrangler en copy the info manually.

BTW: If you find which fields in the pref are relevant, please let us know.

[Wolfgang Schoene]

Hi Hans
well, it's working now and I won't touch it again. Will save the java
preferences file somewhere for future needs.

Cheerio
Wolfgang


Am 10/07/12 11:27 schrieb "Hans list" unter <hans...@gmail.com>:

[MikeTrans]

Just a simple question; I checked this jungle, but didn't find it:

Can you add a term in the user dictionnary and how?

Thanks,
Mike

[Hans list]

Mike,

On Jul 10, 2012, at 1:22 PM, MikeTrans wrote:

>
> Just a simple question; I checked this jungle, but didn't find it:
>
> Can you add a term in the user dictionnary and how?
>

Just a fast answer: you have to edit User.dic manually and restart CafeTran.

Cheers,

Hans

PS Adding words from within CafeTran is planned.

[Hans van den Broek]

On 10 Jul 2012, at 16:49, Wolfgang Schoene wrote:

> Will save the java preferences file somewhere for future needs.

But where is it located? This morning, I tried my brilliant Time Machine solution I mentioned yesterday. And brilliant it was. It didn't work. I first restored CT to yesterday. Now that didn't work, so I suppose the .plist file isn't located there, but at least I could "undo" it. I then restored the Java app in Utilities to yesterday's settings. That didn't work either, but it took me more than half an hour to undo it. The Java app was crossed out, and I wasn't allowed to undo the "restoration." Bad luck. I finally "restored" Java to an external disk, and then copied it to the Utilities folder, but by then I had swallowed several aspirins. Pretty scary.
So I still don't know where to find the relevant .plist...

[Hans list]

On Jul 10, 2012, at 2:25 PM, Hans van den Broek wrote:

On 10 Jul 2012, at 16:49, Wolfgang Schoene wrote:

Will save the java preferences file somewhere for future needs.

But where is it located? This morning, I tried my brilliant Time Machine solution I mentioned yesterday. And brilliant it was. It didn't work. I first restored CT to yesterday. Now that didn't work, so I suppose the .plist file isn't located there, but at least I could "undo" it. I then restored the Java app in Utilities to yesterday's settings. That didn't work either, but it took me more than half an hour to undo it. The Java app was crossed out, and I wasn't allowed to undo the "restoration." Bad luck. I finally "restored" Java to an external disk, and then copied it to the Utilities folder, but by then I had swallowed several aspirins. Pretty scary.

So I still don't know where to find the relevant .plist…

I thought I had written a KB article about this, but couldn't find it yesterday.

Anyway. For future: download Find any File for finding files.

For now:

[Hans van den Broek]

On 10 Jul 2012, at 19:30, Hans list wrote:

I thought I had written a KB article about this, but couldn't find it yesterday.

You did mention it yesterday, but not on your KB, methinks. Anyway, thanks. No wonder Time Machine didn't do the trick. 

[Hans list]

On Jul 10, 2012, at 2:43 PM, Hans van den Broek wrote:

> No wonder Time Machine didn't do the trick.

We'll have to discuss putting all CafeTran related stuff in your Dropbox after my holidays.

[Hans van den Broek]

On 10 Jul 2012, at 19:44, Hans list wrote:

> We'll have to discuss putting all CafeTran related stuff in your Dropbox after my holidays.

Why in MY Dropbox and after YOUR holidays? And why in the Dropbox anyway?

I can't even afford going on holidays. Food is a problem...

[Hans list]

On Jul 10, 2012, at 2:48 PM, Hans van den Broek wrote:

On 10 Jul 2012, at 19:44, Hans list wrote:

We'll have to discuss putting all CafeTran related stuff in your Dropbox after my holidays.

Why in MY Dropbox and after YOUR holidays? And why in the Dropbox anyway?

Going back to the original concept of CafeTran. First I wanted to put it on a nice LaCie USB key:

Then I realized that nearly every Irish pub and Biergarten has wifi access nowadays.

So with an excellent Bruce Eckel's book "Thinking in Java" in hand I  
started writing some text manipulation program which later became a core  
of CT translation engine. As soon as the program managed to parse simple  
text files matching segments with the translation memories I put it on the  
web around 2005. Then, there wasn't any versioning scheme (no pre-1.0 or  
so). The first name was TexTran but I quickly changed it to CafeTran  
coming from my idea of perfect translator's job traveling around the world  
and translating in cafes for a living. By the way, do such translators  
exist? I guess, with a current technology it is possible.

[Hans van den Broek]

On 10 Jul 2012, at 19:52, Hans list wrote:

Then I realized that nearly every Irish pub and Biergarten has wifi access nowadays.

So why not use your KB for that purpose? It's a bit silly that all CT related stuff is spread over the Help, this group, your KB, ProZ, and my blog. But Your KB would be the perfect solution for this problem, especially if you provide links to the other sources. I don't mind contributing to it, nothing much to do anyway (I write this down on purpose, so starting tomorrow I won't have a minute without work).

[Wolfgang Schoene]

Hi Hans
I hope I got you right ... the java settings file is in your Home folder --- Library --- Java.
Was that your problem?

Wolfgang


Am 10/07/12 14:30 schrieb "Hans list" unter <hans...@gmail.com>:

[Hans van den Broek]

On 10 Jul 2012, at 20:58, Wolfgang Schoene wrote:

I hope I got you right ...

Well, I didn't have problems restoring the lay-out using CT. I was just curious. And stupid.

the java settings file is in your Home folder --- Library --- Java.

Users/[user name]/Library/Preferences, com.apple.java.util.prefs.plist (no wonder I couldn't find it)

Was that your problem?

Yep. The other Hans solved it.

[Tim Thompson]

I copied the two target-language files from my Firefox installation into resources/spellchecker and restarted CT, but it's not working. The option for "en_US" does appear in the Spell checker menu, but no red underlining for mistakes. Any suggestions?

I also wanted to clarify regarding the user.dic file. Do I need to increment the number on the first line for each new word I add? Is there a reason why the number 16 is the default number?

Thanks again!

Tim

Em terça-feira, 3 de julho de 2012 11h14min24s UTC-4, Igor Kmitowski escreveu:

Hello again,

The hunspell dictionaries are probably already installed in your system (with the other software I mentioned in the update announcement). You can also download them from here: http://extensions.services.openoffice.org/dictionary. They are packed as extensions (.oxt files). After downloading, rename the .oxt file to the .zip file and unpack. Next, copy the two files for your target languages .dic and .aff (e.g. en_US.dic and en_US.aff) to Cafetran's resources/spellchecker/ folder inside CT installation. After running CafeTran you should see the dictionaries visible in the menu Edit | Spell checker.

Enjoy your mistakes ;)
Igor

[Tim Thompson]

My second question still stands, but never mind the first one. For some reason, the Firefox Hunspell files were named "en-US" instead of "en_US." The spellchecker worked after renaming the files.

Cheers,

Tim

[Igor Kmitowski]

The Hunspell .dic file requires the first line to be the number. This
number informs about the number of entries in the dictionary. It seems you
don't have to increase this number every time you add a new entry. Perhaps
once a 100th entry will do. CT will increase this number automatically in
one of the future updates.

Igor

> I also wanted to clarify regarding the user.dic file. Do I need to
> increment the number on the first line for each new word I add? Is there
> a reason why the number 16 is the default number?