How to define a dictionary for new language?

users
#1

Hi,

a friend of mine would like to build a dictionary for a local dialect
and he asked for my help.

First I build the aff and dic files and I tested them with hunspell. It
works.

Next I was able to construct an oxt which install correctly. However the
new language do not appear in the list of the installed ones so I was
not able to use it for checking the open document.

As a quick hack, I chose an existing language, replaced the aff and dic
file and installed it. It works.

So the step I am missing to accomplish my task, it is how to "register"
a new language in order to be seen in the combo box in the spelling
checker dialog.

Can someone help me with instruction or a pointer to a guide?

TIA

Sergio Martino

#2

So the step I am missing to accomplish my task, it is how to "register"
a new language in order to be seen in the combo box in the spelling
checker dialog.

Unlike Microsoft Office, it cannot be done. The code changes must be made.
So you will have to stick to dummy language names.

If that is a real language, you can submit a bug to add it.

#3

Hi Sergio

Have you tried to add it under "Tools - Options - Language settings" and then the sub-options of "Languages" and "Writing aids". This might guide you to get the language into your workspace.

Regards

Andrew Brown

#4

Hi :) 
1.  Have you been in contact with the translators mailing list?  The one international one brings together people with a lot of skills from most of the various languages on offer.  They might be able to help with this sort of thing.

2.  Which language?

3.  Congrats on an amazing amount of good work so far! :D  Nicely done! :))

Regards from
Tom :slight_smile:

#5

Hi Sergio

Ok, went back to look at this to try and help further, as I was able to load my countries two of the 11 languages I wanted to use, this way.

In "Tools - Options - Language settings - Languages" in the section "Language of" and field "Locale setting" can you see your created language/dialect you are wanting to use. And then further down under the section "Default languages for documents" and field "Western" or "Asian" or "CTL", can you see your created language/dialect there.

Likewise under "Tools - Options - Language settings - Writing aids", in the "User-defined dictionaries", you can click on the "New" radio button and add the name you gave the dictionary, then in the next drop-down menu "language" see if you can again locate your created language/dialect".

Hope this helps you to get further. If not then I would follow Tom's proposal in his reply.

Regards

Andrew Brown

#6

Can you help me to do this, is there any tutorial or something? How to build aff and dic so it can be tested?

Kruno

#7

https://bugs.freedesktop.org

Product: LibreOffice
Component: Linguistic
Severity: enhancement

#8

Can you help me to do this, is there any tutorial or something? How to
build aff and dic so it can be tested?

This one is a good start: http://linux.die.net/man/4/hunspell

To test it with LO, download some dictionary extension, rename it, throw away unneeded files and make corresponding changes in description.xml and Dictionaries.xcu

#9

I have made variations on the en_US dictionaries, but I used existing .aff files. I never have found any references to creating one. Never really found one for the .dic file and the .oxt file[s] as well.

So I experimented wit the .dic and .oxt files till I got them to work correctly.

You never listed whatthe local dialect of what language you were creating the .oxt dictionary for, or at least I have not seen it listed.

I am a little behind the "game" right now, but I tried to make a list of all of the different .oxt language files [dictionaries, thesaurus, etc.] on my dictionary page listing.

http://libreoffice-na.us/English-4.0-installs/dictionary.html

I hope to have some time in the next month to check for updates to the listed files. As I update the files, I try to remember to listthe last change of the word list or the thesaurus files. With 180 +/- listings, it can take a lot of time to work on finding the updates. But hopefully I will be able to take time in August to do some of them.

#10

Hi :slight_smile:
It seems to be for a fairly specific area of Italy.  I'm not even sure if the dialect has a separate name.

In England we have a lot of different types of 'English' some of which are completely incomprehensible to an outsider living as far as 30miles away.  None of my family ever understood my Gran for example, but she was always there offering cups of tea with a rock-hard scone or porridge only slightly less runny than cement (actually it was all good stuff really but don't tell her that).  In the case of cockney that was a deliberate attempt to avoid passing anything onto "the old bill" by accident.  Liverpudlian and Geordie are perhaps due to different peoples having invaded us at different times and different kingdoms all over the place or different tribes claiming different parts.  I'm sure it's much the same in any other country.

Regards from
Tom :slight_smile:

#11

Hi :slight_smile:

/snip/

In England we have a lot of different types of 'English' some of which are completely incomprehensible to an outsider living as far as 30miles away. None of my family ever understood my Gran for example, but she was always there offering cups of tea with a rock-hard scone or porridge only slightly less runny than cement (actually it was all good stuff really but don't tell her that). In the case of cockney that was a deliberate attempt to avoid passing anything onto "the old bill" by accident. Liverpudlian and Geordie are perhaps due to different peoples having invaded us at different times and different kingdoms all over the place or different tribes claiming different parts. I'm sure it's much the same in any other country.

Regards from
Tom

Is this still true? I am aware that it was true in the past, but I would
have thought that with radio, TV and movies, that the local
dialects would have mostly disappeared. But what do I know. sitting
here on the other side of the pond, where dialects really have pretty
much disappeared.

(55 years ago, when I was in the Air Force here, I ran into some boys
from the backwoods of Kentucky, and they spoke a dialect that was
reminiscent of what you read in Shakespeare. I'm pretty sure that's
all gone, now. We get news reports with interviews of the locals from
all over the US, and there's very little "drawl" even. Probably those
of us in New York or Boston have more of a unique accent today. Altho
there is a woman reading commercials on KSEY-FM, in Seymore, TX, who
really sounds hillbilly! [KSEY is accessible by the Net, and plays
classic country music.])

I ask this OT question because I have been interested in language
all my life, and I notice accents. And of course, if _you_ can't
understand some folks in Merry Olde, surely I couldn't!

--doug

#12

Hi :slight_smile:
I still hear a lot of differences in even very common phrases used in different areas of the US.  I think it's inevitable whenever people group together in any way.  The media seems to average things out a bit but it's more like a trading language that doesn't really belong to anywhere and isn't really anyone's "native" language but is added to be all sorts and then made instantly bland.  Baltimore sounds different from other places, even phrases are different. 
Regards from
Tom :slight_smile:

#13

There are "official" dialects and then there are those that just pronounce the words differently or uses different phrase styles.

As for a language dialect, say using different words, like the various non-English languages spoken in the UK, or ones that are based in English variants from 500 year ago and never really changed much over the years, these may need to be part of some install of LO.

Look at all the different language dialects that India has. Only a few are part of the "official" list of languages that LO supports.

Regional languages or dialects could be a real "mess" to try and get LO to support all of them.

How many languages/dialects are supported right now? 100, 150?

How many thousands are out there that may never be supported due to the small market? I could only guess.

I lady I met was overjoyed when she heard that OOo [before LO ever came out] supported Hebrew. She worked in a tourist center that needed to use her English, French, and Hebrew skills to answer questions and create documents.

If you want to create a dialect of a language, that is close to the "root" language that LO supports, you could create a dictionary that uses the "root" language and add the dialect elements. Look at Spanish. There are about 20 regional dialects that have come out in dictionaries for LO based on the country that it is spoken in. It is still Spanish, but with a regional name included as part of the dictionary's name. Spanish - Mexico, Spanish - Columbia, etc..

The hardest part of the whole process it finding the needed documentation to create the .aff and .dic files. When I last searched, I came up empty. So I just used some common sense and used the "default" en_US .aff file and then just used a word list of all the words with their various prefixes and suffics included. I ended up with over 700,000 of them. I did have a en_US dictionary that included over 2 million words and their proper variations, but that seemed to go too far. Some of those really "rare spellings" might be not so popular with certain English professors or bosses, even though they were correctly spelled.

As for the UK, I saw a person's list of "proper" Shakespearian era words. Someone told me that in that era, they spelled their words whichever way they felt like. So I decided not to make a dictionary for that.

But there are other dialects that could be added. All it takes is some willing souls to make it a working project.