You probably don't need to execute this code. If you are only looking for the dictionaries, download them here
This repo does two things:
- The first is creating a performant database containing words and their definitions, including all inflections and proper linkages, using English Wiktionary data (and for Russian OpenRussian as well).
- Based on this data, it creates high-quality dictionaries compatible with recent ebook readers. It uses an algorithm to fix bugs in the Kindle lookup engine for this that prevent inflections from being found - even official dictionaries suffer from it.
If you want another language or format, you can open an issue and I will take a look if I have time. I apologize that the code right now is somewhat messy.
Contributions and feedback about words that are not displayed correctly or unhelpful definitions (I tried to remove them all/link them to parent definitions) are always welcome. Additionally, you can tell me if the dictionary for your language is missing essential information.
You can install the package with pip install git+https://github.com/Vuizur/ebook_dictionary_creator
And then use it like this:
from ebook_dictionary_creator import DictionaryCreator
dict_creator = DictionaryCreator("Czech", "English")
dict_creator.download_data_from_kaikki()
dict_creator.create_database()
dict_creator.export_to_tabfile()
dict_creator.export_to_kindle(
kindlegen_path="C:/Users/hanne/AppData/Local/Amazon/Kindle Previewer 3/lib/fc/bin/kindlegen.exe",
try_to_fix_failed_inflections=False, # This tries to fix kindle's problems with inflections.
# For example, if oso is in the dictionary as a headword, it will only display this headword,
# - if oso is the inflection of osar, osar will not be found. This tries to fix it.
# Currently only works for latin scripts.
author="Vuizur",
title="Czech to English dictionary",
mobi_output_file_path="Czech-English.mobi",
)
dict_creator.export_to_stardict("Vuizur", "Czech to English dictionary", "Czech-English-dict.ifo")
This project would not have been possible without the https://kaikki.org/ data provided by Tatu Ylonen, the OpenRussian data and the Pyglossary library for the creation of the Kindle dictionary
This project looks very good, but focuses more on DBnary translations (and doesn't seem to have inflections): https://github.com/karlb/wikdict-gen
https://github.com/nyg/wiktionary-to-kindle parses the Wiktionary itself.
https://github.com/efskap/kindlewick apparently works very similarly, also supports inflections, as of now it only supports smaller languages though (has been tested for Finnish).
https://github.com/BoboTiG/ebook-reader-dict is a program that also parses the Wiktionary dump itself and outputs Kobo compatible files