Legal_Books / README.md
QomSSLab's picture
Update README
a8666f3 verified
|
Raw History Blame Contribute Delete
1.14 kB
metadata
language:
  - fa
pretty_name: Persian Legal Books
tags:
  - legal
  - persian
  - books
configs:
  - config_name: default
    data_files:
      - split: train
        path: books.jsonl

Persian Legal Books

A collection of Persian law books. One row = one book, with the text kept page by page.

Structure

Field Type Description
id string Stable book ID (e.g. book-00001)
file_name string Source PDF file name
title string Book title
authors list[string] Authors
source_type string Source type (e.g. کتاب)
total_pages int Number of pages in the source PDF
pages list Pages of the book
pages[].page_number int Page number
pages[].content string Text of the page
pages[].metadata dict Header lines of the page and Page number

Usage

from datasets import load_dataset

ds = load_dataset("QomSSLab/Legal_Books", split="train")

book = ds[0]

print(book["id"])
text = "\n\n".join(p["content"] for p in book["pages"])

License

Not specified. The books belong to their authors and publishers.