Dictionary converter and irregular inflections

[ninpuukamui]

Hi,

Does anyone know how to contact whoever wrote the dictionary converter? I've been creating XDXF files out of some dictionaries, and so far I've failed to implement verb conjugations. In Spanish there are about 200 irregular verbs with about 50 conjugations each, so it's a bit of a problem.

My goal is that when you search for a verb form other than infinitive, it will get you the definition for the infinitive verb. This can be done with the morphems file for regular verbs, but irregular verbs are a bit trickier. I tried adding all irregular forms to the morphems file, but the converter.exe ignores entries after certain point fi the file is too large.

I also tried using the logical format of XDXF, which deals with this kind of things, but it seems that converter.exe supports only very basic XDXF, so that didn't work either. I have attached a manual for logical XDXF, as found in http://xdxf.revdanica.com/drafts/logical/download/

The other option is include one entry per each non regular form of each irregular verb. I'm do have a list of irregular verbs conjugations, but I'm not going to start separating the irregular forms from the regular ones, and at about 50 conjugations per verb, it means the dictionary would have an extra 10,000 entries just for that. Not the best solution really ^_^

Anyhow, if anyone has any idea on how to feed this kind of information (any kind of irregular inflection) to the converter, please let me know.

Thank you very much!

[Schnitzelbrain]

Here are some readymade spanish dic files:
http://translate.googleusercontent.c...3FZrckUhHA#top

Page 15 last post

[ninpuukamui]

That you very much, but the es-en dictionary has the same problem as the one I'm trying to make, i.e. can't find irregular inflections. If you look for "tuvo", you'll see you don't get "tener" back.

[ninpuukamui]

I sent an email to info@pocketbookreader.com and got no answer either. I'm going to send an email in English to service@pocketbook.com.ua and see what happens.

[Ulli]

If you will get no reply, you can just try service@pocketbook.de please refer to this Thread in your E-Mail.

[ninpuukamui]

I haven't received any answer from any of the support emails (English, Russian or German), not even to acknowledge the email.

I am pretty disappointed

[Schnitzelbrain]

Originally Posted by ninpuukamui

I haven't received any answer from any of the support emails (English, Russian or German), not even to acknowledge the email.

I am pretty disappointed

Welcome to the club, they really do not care about the dict thing.
There are many posts to get some more languages done, so we need the morphems. But nobody cares, in the russian forum nobody cares. U can not access (Download) 99% of bookfast.org (lot of dicts there) nobody cares.

[Pol Import]

Only first key term in XDXF is indexed by converter.exe

Dear members,

I've made a dictionary Spanish-Dutch (50,194 entries, from on old Windows dictionary program), converted it to XDXF and made a dic file of it for my pocketbook touch lux 3. It worked fine.
To make it more complete I added 27,000 articles from a Spanish dictionary (Diccionario Vox) for entries missing in the first version. It suited me even better than it's predecessor. Although with quite a lot of words one tap is sufficient to make pop up the article, this is not the case with a lot of inflections and conjugations. In those cases, after reading translation not found, one has to tap the pencil button, type a few characters before the desired entry appears in the search list, tap on the word to make it the first item in the list, and finally tap on it to let pop up the meaning.

To solve this problem I extracted from another Spanish dictionary nearly half a million conjugations and inflections and integrated them into the XDXF-file as extra Key terms in the corresponding articles (See attachment)
Unfortunately only the first key term is indexed and as a result the problem still exists. All key terms appear now in the definition popup.

A babylon version of the file works perfectly in Goldendict on windows and smartphone. Strangely enough the bgl file could only be successfully made In Babyblon Glossary Builder from a gls file (in ANSI) and not from an Excel workbook.

My questions:
Is this a wrong approach (adding extra key terms) or a bug in the converter or dictionary program on my e-reader?

Would adding the half million inflections to the morphems.txt be a solution ?

Thanks in advance.

Pol

[rkomar]

I think the solution would be to create a proper morphems.txt file for Spanish. That file contains rules for inflections in the language as a whole, and not on a word-by-word basis. You need to come up with the rule for generating the root word from each type of inflection and creating the proper line in the morphems.txt file for that. I expect that one or two dozen rules will cover more than 90% of your cases.

I google-translated the instructions for creating a dictionary from Russian to English, and here is the part for the morphemes. It should give you an idea of how to create the rules. You can look at the english morphems.txt file provided with the converter to see some simple examples of rules.

Edit: I did not look above to see that the original poster ran into problems with the size of the morphems.txt file when adding rules for the irregular verbs. Perhaps you should add rules for the regular verbs first, and then worry about tackling the irregular verbs after.

Morphems.txt file may contain the following elements:

1) The classes of characters:

% = Number charlist

For example, the character classes "Russian vowels," "Russian consonants" described lines:

%1=аеиоу
%2=бвгджзклмнпрстфхцчшщ

2) Morphemes:


ˆsuffix1/suffix2/...=variant1/variant2/...

If the end of the selected word matches one of the endings in the left side, just look at all the options on the right side of endings (the priority is the original version of the word, then the options left to right).
On the left side you can use the symbol "?" (Any character) and 0-9 (class number of characters). In the right part of the "dot" denotes the original character in the replacement end.


Examples of morphemes:

^2е/2и/2у/2ы/2ой/2ам/2ами=.а

The words "книги", "книгой" "книгам" will be replaced by "книга".

^2ой/2ей/2ому/2ему/2ых/2их/2ым/2им=.ый/.ой/.ий

For the word "быстрому" options are checked, and the replacement of the "быстрый".