Kobo Japanese Dictionary Enhancer 1.1
Lots of releases in quick succession – the new Kobo Japanese Dictionary Enhancer brings multi-dictionary support and merged translation support. Using the Wadoku project’s edict2 database we can now add also German translations.
Looking at the numbers, we have now 326064 translated entries when using the English edict2, and 368943 translated entries when using the German Wadoku edict version. And more than that, as an extra feature it is now also possible to have merged translations, so to have both German and English translations added.
Please head over to the main page of the project for details and download instructions. If you need my help in creating the updated dictionary, please feel free to contact me.
Enjoy.
Out of curiosity, is the furigana being added dynamically or is that in the book contents?
That is in the book contents.
Thank you very much for providing an alternative to the regular Ja-Ja-Dictionary by offering a script which adds the entries from the Edict- and Wadoku-Project to the dictionary.
I am using Mac OS 10.10.5, installed p7zip via Macports but still encounter a problem when using the original dichthtml_jaxxds.zip (~93mb). Apparently the cp command in the create_dict function runs into an error because the “Argument list too long”.
$ perl enhance-dictionary.pl –dict=wadokudict2_20150705/wadokudict2 –dict=edict2 –merge –debug
Using the following dictionaries as source for translations: wadokudict2_20150705/wadokudict2 edict2
loading edict2 type from wadokudict2_20150705/wadokudict2 … Didn’t get description, skipping. Line nr 251299, =Interkontinentalflug m/Use of uninitialized value $fields[1] in pattern match (m//) at enhance-dictionary.pl line 279, line 253208.
Use of uninitialized value in split at enhance-dictionary.pl line 282, line 253208.
Didn’t get description, skipping. Line nr 253208, =領ろしめす;知ろしめす;done
loading edict2 type from edict2 … done
unpacking original dictionary … Y
done
reading file αa (.html)t files … 0%
find_translation: searching for alpha blogger
reading file ‥a (.html)t files … 0.01%
…
find_translation: searching for ZIPコード
find_translation: searching for ZIP+4
loading and merging dict files … done
total entries: 922379
entries used from wadokudict2_20150705/wadokudict2: 367736
entries used from edict2: 324937
creating output html … done
creating update dictionary in dicthtml-jaxxdjs-201508311538.zip … sh: /opt/local/bin/7z: Argument list too long
done
I am not too familiar with advanced Perl-scripting nor with Unix shell commands which is why I had little success in altering your script using the the find command (which supposedly creates an action per file instead of working with stacks)
Also I didn’t found a way to increase the maximal stack size.
$ getconf ARG_MAX
262144
This might be an MacOS/BSD problem but I would like to hear your opinion.
Thank you very much in advance.
Thanks, indeed. I have fixed this in the script, so either get the new script from the github page, or replace in the script line 496 with
Mind only the * is changed to a ..
That should fix it.
Thanks
I forgot: Concerning the error around “Interkontinentalflug”: That is an error in the wadoku edict file. If you edit it and search for “Interkontinentalflug” you will find about 10 consecutive wrong line breaks. Just remove them and move the definition which is on a line by itself up to the line where “Intercontinentalflug” is defined.
Thank you very much for the quick reply.
The updated script fixed the problem. I will later edit the wadoku-database-file so that “Interkontinentalflug” will be displayed correctly.
Thank you again for developing ways to customise Kobo eReaders and making them public. It helps me a lot since I have to read lots japanese papers for my major.
Hi Norbert,
I found your blog a while back and I am very impressed. You have a lot of cool stuff going on!
Since you seem to be one of the few experts on this topic, please allow me to ask a question here. I have been studying Japanese for some years now and are really getting into reading Japanese ebooks, as the dictionary function just beats having to look up unknown words manually.
Recently I have been trying hard to improve my Japanese intonation and am studying pitch accent. As the Wadoku online dictionary shows pitch accent, I was wondering if you knew any way of including the pitch accent (just a number would also be superb) into one of the Kobo dictionaries? Have you experimented with that yourself?
Would love to hear your thoughts on this.
Best wishes,
Jakob
追加：I also found the pitch accent database file “pitch_accents.sqlite” from the EDICT dictionary. If this could be added somehow.
Hi Jakob,
thanks for your comments. Where did you find the pitch_accents.sqlite file? I can take a look at it, but not sure how this can be included.
Hi Norbert,
thanks for your quick reply.
The pitch_accents.sqlite file is included in the download of “JGlossator”, a sort of offline desktop dictionary.
http://jglossator.sourceforge.net/
After downloading the zip, it is located in under
JGlossator_v5.0.zip\JGlossator\Dic\PitchAccent