Shared highlights - Githubissues

To show shared highlights of the same passage between different versions of the same book.

Create postgres database with pg_trgm extension.
- Create the table Book(id, name, author, publisher) with an indexed column for book name.
For each book, fetch book metadata
- Book Name
- Author
- Publisher
Preprocess book name and author name
- Remove spaces, punctuation, brackets[](){}, as well as content inside brackets. Keep alphanumerical characters and Japanese characters.
- Execute similarity match on the Book table book column with the book name.
- If similarity exceeds threshold, check the similarity of author name.
  - If both book and author name exceeds respective thresholds, or if author name in either book is empty, match success and grab the book id.
- If no matches, create new book row with the preprocessed metadata.

Create the table Highlights(id, bookId, content, chapter, page, progress_percentage, prefix, suffix) with indexes for content and progress_percentage.
On load book, query highlight count with the bookId
- If highlight count == 0, do nothing
- If highlight count > 0 && <= 100, grab all highlights
- if highlight count > 100, grab all highlights limit 100 where highlights progress_percentage < (current_percentage +- 20%)
Given current page number n, grab all text between n-2 to n+2.
- For each highlight, preprocess it (remove spaces, punctuation, brackets and content within brackets).
- Preprocess the grabbed text as well and do an exact match to see if any highlights exist in the text.
- If exists, check if prefix (sentence) and suffix (sentence) matches.
  - If match, select text node on the reader and display the highlight.
On creating a new highlight, save the content, prefix, suffix, chapter and page. (Chapters and pages are for reference purposes.) Progress percentage is calculated on first creation.

Notes: It might be helpful to update the progress percentage with a Gaussian Bayes model.

mathewthe2 / immersion_reader