vuizur / add-stress-to-epub Goto Github PK

View Code? Open in Web Editor NEW

19.0 4.0 2.0 580.72 MB

A program that sets the stress and the letter ё of Russian text and ebooks using Wiktionary data and grammar analysis.

License: GNU Affero General Public License v3.0

Python 100.00%

epub python dictionary russian russian-stress language-learning russian-accent

add-stress-to-epub's People

Contributors

Stargazers

Watchers

Forkers

freelanguagetools equwal

add-stress-to-epub's Issues

Offical™ embedded font

Hello,

I'm very impressed by the project and found it a few minutes ago.

I've noticed that the display of accented characters has a lot of variety depending on whatever fonts are embedded in the book or whatever. It would be nice for users to be able to add a font known to be good with calibre (maybe a wiki page)?

(edit: looks like this is a bug in KOReader, where it calls up a "fallback font" for the accented characters)

Set stress for words that have two different possibilities

For example удалось. There should be a new row in the database for this to save the other version

Parse comparative forms from Wiktionary starting with paranthesis

Like (по)умнее

One-syllable words from OpenRussian are accented

Use a CSS hack to add accents to preserve selected words

Hello
This software seems to physically add the accents on the words themselves. This requires that the user has special dictionary programs or files which can handle the words in their accented forms. I propose a possible way to add accents without needing special dictionaries.
We use a CSS ::before property to provide the accented character in a zero width inline-block span. The text generated by CSS is not selectable, at least in all the browsers I tested. It also works in Calibre and Foliate, partially in KOReader (it does not select as a whole word, but you can just drag a bit). It might work with all reader software, but it still seems useful.
Example html:

<html>
  <head>
    <style>
    [data-content]::before {
      content: attr(data-content); 
    }
    </style>
  </head>
  <body>
    нали<span data-content='&#x301;'></span>чный
  </body>
</html>

This should display нали́чный, but if you try to select it, it will select наличный (without the accent)
It would be great if this is an option for this tool!

Publishing a PyPI module

Hello!
I am the maintainer for VocabSieve. It would be great if you you can publish this to PyPI for programmatic use. All the existing ones (like russtress) does not consider context and make mistakes quite often.

On another note, is it really necessary to have spacy for this? It is a rather large dependency and in my experience works somewhat slowly. Have you tried pymorphy2? It seems to be able to tag words too.

Любом is ambiguous for some reason

Add links to proper dictionaries

Finish benchmarks

Put everything together and get the basic results
Perform detailed benchmark
Benchmark of old versions

Error when importing

Hi there!
Cool concept and exactly what I've been looking for to add stress marks to sentences from the SMARTool database to make Anki decks.

I installed on my Linux Mint 20 machine using
pip3 install git+https://github.com/Vuizur/add-stress-to-epub

When I run your example in my Python script, I get the following error at the import line:

File "/home/user_name/.local/lib/python3.8/site-packages/russian_text_stresser/russian_dictionary.py", line 45, in <module>
    class RussianDictionary:
  File "/home/user_name/.local/lib/python3.8/site-packages/russian_text_stresser/russian_dictionary.py", line 46, in RussianDictionary
    def __init__(self, db_file: str, simple_cases_file: str | None) -> None:
TypeError: unsupported operand type(s) for |: 'type' and 'NoneType'