Alternative dictionaries

Search This thread

saizenn

New member
Feb 15, 2012
4
1
Wow that was GREAT and exactly what i was looking for... it made almost 600.000 inflected entries, that after deleting duplicates and filtering became 480.000 that is not bad at all!
At the moment I am double crossing it with my tblwords.term spanish word list (around 100.000 words) and we will see how many are left... I will post the results.
Thank you ApokrifX for the inflected list, great help!
 
  • Like
Reactions: ApokrifX

saizenn

New member
Feb 15, 2012
4
1
This is finished and I got almost 400.000 valid items for the inflected table.
The final files are around 23 MB for the inflected table and over 50 for the main words and the balance aprox 100.000 main and 400.000 inflected. Now I will test it and see in real how many words I have from a book and all that so I can check if it is enough
 

saizenn

New member
Feb 15, 2012
4
1
I finished a working version, that is a bit slow yet cuz of the many words I used but I will improve it making descriptions shorter.
If someone wants to try can be downloaded here:

saizenpro .com/nook
(remove space)

At the moment are just english to spanish and spanish definitions.
 

danielgessuti

New member
Apr 22, 2012
1
0
unzip field "description" of table "tblWords" in java

Hi!
Hello, I'm trying to write a small java program to unzip the field "description" of table "tblWords".
Someone has already done? Thanks in advance

Daniel
 

Renate

Recognized Contributor / Inactive Recognized Dev
empty inflectedwords.db?

Code:
sqlite3 inflectedwords.db

CREATE TABLE android_metadata
(locale TEXT);

INSERT INTO android_metadata VALUES('en_US');

CREATE TABLE tblInflections
(_id INTEGER PRIMARY KEY AUTOINCREMENT,
 inflected TEXT,
 headword TEXT);

CREATE INDEX inflected_index on tblInflections (inflected ASC);

@danielgessuti: The stuff I did, I did in C on a desktop.
In Java use java.util.zip.ZipFile and java.util.zip.ZipEntry
 
  • Like
Reactions: funvn

ApokrifX

Senior Member
May 23, 2011
468
50
Renate,
I saw, you started fiddling with injecting smali code, so I was wondering…
How difficult it’ll be to replace native reader dictionary browser, with a market one (ColorDict, Fora Dictionary, you name it)
I guess, it’s gotta be some (standard?) API to call already installed dictionary.
I can post some other “I guess” things, but I think you know all those things better than me anyway. :eek:
 

fufutos

Senior Member
Oct 23, 2010
71
4
This seems very interesting... If there is a method it would be perfect for languages spoken by few (like Greek)...
 

yov9

Member
May 1, 2012
10
0
hebrew\arabic?

hey renate
changing the dictionary is a great idea!
i could use an english-hebrew dictionary, and i think there are quite a few that would be glad having an english-arabic around here.

i know RTL languages make everything harder, but if you think this is possible and worth the effort i will be happy to try and help. (as soon as ill finish rooting my device, new around here)
 

Renate

Recognized Contributor / Inactive Recognized Dev
Ok, here's a French to English dictionary.

There is also an English to Arabic dictionary.
I checked what I could.
The squiggles seem to look the same as Google translator.

In either case, the file gets extracted and copied to /system/media/reference/basewords.db
 

Attachments

  • fra-eng.zip
    532 KB · Views: 140
  • eng-ara.zip
    6.2 MB · Views: 49
  • Like
Reactions: rhhd and qft1967

rhhd

Senior Member
Jan 19, 2011
94
10
MUC
Google Pixel 6a
Thanks a lot for the dictionary! :)

Installation was without problems.

I expected the dictionary not to find conjugated forms of verbs. But even finding plural nouns almost always fails, despite the fact that in French the plural is almost always simply marked by an "-s", like in English. I found exactly one noun where the plural was recognized: "femmes", don't know why it works there.

But many nouns are not even recognized when they are in singular, e.g., "comportement" or "neutralité" which are very frequently used words. Not clear whether these words are missing from the dictionary or whether it may be a problem with my epub file?
 

Renate

Recognized Contributor / Inactive Recognized Dev
Of those two dictionaries, the French-English one is pretty small.
I did a look at the source, neither comportement nor neutralité are in there.

Normally inflectedwords.db supplies mapping for plurals.

The English-Arabic dictionary is a fairly large one and directly includes entries for inflected words.
 

rhhd

Senior Member
Jan 19, 2011
94
10
MUC
Google Pixel 6a
Okay, thanks.

Maybe "femme" is so common a french word that it is even contained in the english dictionary and therefore the mapping is contained in the original inflectedwords.db.
 

Renate

Recognized Contributor / Inactive Recognized Dev
Sorry, I must have missed that, ApokrifX.

The stock Reader.apk has both the dictionary provider and a dictionary client in it.
There's a lot of things that could be done with this.
I'm not sure which way you want to go with this.

You could switch the stock Reader to use an entirely different dictionary provider
or switch another reader to use the stock Reader dictionary.

The English dictionary that comes with the Reader is actually not bad, although the formatting was munched during the export.

Maybe I will write a little app to just lookup the dictionary using typed-in words.
 

Top Liked Posts

  • There are no posts matching your filters.
  • 3
    hey @nmyshkin ... i thought you had deleted and exited out of xda ... surprised & glad to find you back active !!!

    any new tweaks and modd for our device ?

    :)
    Not to hijack this thread, but as Mark Twain is reported to have said, "Reports of my death have been greatly exaggerated." I have however been dealing with some serious health issues, the pain meds for which don't improve my concentration or analytical abilities. I'm slowly dropping down on the pain meds but as for the rest, we'll have to see how I respond to the rest of the therapy.

    I have been doing some work--you've probably already seen it--a new version of the Nook Settings app, and a true weather widget. Actually, I am sort of running out of ideas. I do these things for my own enjoyment of my device and then share on the off chance that someone else may find them useful. I now have a Glowlight which is completetly B&N free (see post) and someone has gotten me vaguely interested in a library app that would show covers (and I'd like access to the summary meta data such as you can see in Calibre). I think I understand how to handle epubs, but mobi files are heavily encrypted and extracting even a cover image is seemingly impossible. PDFs are just as bad. So this one is going to require a lot more thought.

    But I'm not dead yet.
    2
    Oh, great, there is some interest!

    I haven't posted the converter or any of the databases yet. As a newbie here I wasn't sure about posting links or 15 Meg attachments.

    The dictionaries from freedict.org are pretty basic word translations. There are some multi-word phrases, but of course the lookup on the Nook only works on single words.

    The original dictionary also has a database of inflected words, so that a lookup of "cows" will go to the headword "cow". I know of no source for inflected words in other languages.

    For an English/French dictionary you could keep the original English inflected word database and replace the headword database with an English/French database.

    I also want to make a companion EPUB book dictionary so that you could search (if necessary) arbitrarily for words.
    2
    Hi,

    I have just managed to convert some free Babylon dictionaries (.bgl) to Renate's dictionary format. It uses the stardict dictionary format (see http://code.google.com/p/stardict-3/) as an intermediate format. Moreover, one of the scripts converts from the stardict format to Renate's, so this method could be used also for another formats (not only .bgl) as long as 1) there is one tool to convert from that particular format to stardict's and 2) the entries are formatted in simple HTML (which is the case of the babylon dictionaries that I have tested).

    The instructions are included in the attachment. It first uses the PyGlossary project (see https://github.com/ilius/pyglossary) to convert from Babylon format to stardict format. Then I have slightly modified the PyStarDict project (see http://sourceforge.net/projects/pystardict/files/pystardict/v0.6/) and created one script in order to generate the database file according to Renate's format. So all the credits to the authors of these open-source projects and to Renate.

    I do not support piracy, so please do use this script only with Babylon free dictionaries or if you have the rights to use it with other Babylon dictionaries.

    ---

    EDIT: file updated so now it takes the parameters from the command line (the source code has not to be changed).
    2
    Sorry, Nook reader versions are very incompatible.
    I hadn't looked into this recently, the patches for Lookup only work on 1.2.1
    (The latest Nook, the white Glow 2 is also a big mix of old and new.)

    I think that I'd update to 1.2.1
    That's probably the last update we'll ever see.

    sqlite3 for the Nook is below (it's a very thin program).
    Make system read/write:
    Code:
    mount -o rw,remount /dev/block/mmcblk0p5 /system
    adb push it over to /system/bin, then:
    Code:
    chmod 755 /system/bin/sqlite3