How do you format your reading material to work with LingQ?

[JeremyBrunette]

My 어린 왕자 has words on different lines and even on different pages.
For example:

나는 이렇게 진심어린 대화를 나눌 사람 하나 없이 고독하
게 살았다. 육 년 전 사하라 사막에서 비행기가 고장을 일으
킬 때까지도. 엔진 내부 어딘가가 파손되어 기사도 승객도 없
이 혼자서 어려운 수리에 나선 참이었다.

and

신비로움이 너무 강렬하면 순순히 따르게 마련이다. 사람
12
13
내 친구는 너그럽고 상
사는 마을에서 수천 마일 떨어진 곳에서 죽음과 대면한 처지
냥하게 미소 지었다.

Obviously this doesn’t work with LingQ. How do y’all deal with this?
I’ve tried reformatting with AI but the AI I’ve used (ChatGPT/Google AI Studio) aren’t reliable or consistent.

[vet8t6z79pc4]

The AI tools you mentioned earlier are the best way so far. With proper prompting, you could achieve good results.

Why "LingQ import" modify my content?

If LingQ uses some LLM with a prompt to make a non-formatted transcript neat version that is easier to read, consider trying this prompt that I am using personally, It works better than LingQ’s original importer. You are an expert in restoring punctuation. Your task is to transform the following text into complete sentences while preserving every original word and phrase. Please follow these steps to complete the task: 1. Convert the input text into sentences according to these guidelines: …

This is a prompt I used to reformat the YouTube subtitles into complete sentences. It will help you.

[JeremyBrunette]

Yeah I started using your prompt a while back. It isn’t working with these separated words but maybe that’s because I have been modifying it based on my needs. I’ll return to the original that you provide and give it one more try.

[JeremyBrunette]

Your prompt doesn’t work for what I need. It doesn’t break the text up into sentences. I’m getting my text from an actual published story so the grammar is already fine. I just need each sentence to start a new line and no words to be cut by a line break (obviously that would be two “new” words with LingQ). I try editing the prompt but the AI, after 2 or 3 uses (or after a large paragraph), stops being accurate.

[vet8t6z79pc4]

Then, extract the EPUB from Google Books. That’s a raw and clean text, so it will not have that issue.

[Obsttorte]

In regards to word splitting. When I import E-books I firstly modify them using Calibre. If you go to the modify book section, you can usually find a style sheet document that handles the default look of the text. It is often named default.css. In the body or root section, depending on what you find, add the line
word-break: keep-all;
(mind the semicolon at the end). It needs to be in the main section, so to speak, for it to affect the whole document. The naming might be root, body or something similar. If no such section exists, you may have to add it (use root then). You may have to experiment. You can see the effect immediately.

도와드렸으면 좋겠어요. 안녕히 가세요.