app.py

import csv
import json
import urllib.parse
import requests
import tempfile
import zipfile
import os
import re
import threading

from pathlib import Path

import gradio as gr

from fuzzywuzzy import fuzz
from pyarabic.araby import strip_tashkeel


def main() -> None:
  with gr.Blocks(
    theme=gr.themes.Default(font=[gr.themes.GoogleFont('Noto Sans Arabic'), 'Arial', 'sans-serif']),
    css='\n'.join([
      'html, body, .gradio-container { direction: rtl !important; }',
      'h1 { text-align: center; display: block; }',
      'th, td { text-align: right !important; }',
      'th span { white-space: nowrap !important; }',
      '.icon-wrap { right: unset !important; left: var(--size-3) !important; }',
    ])
  ) as demo:
    index_state = gr.State()
    results_data = gr.State()
    selected_book_id = gr.State()

    gr.Markdown('# ابحث في كتب مكتبة المسجد النبوي 🕌', rtl=True)

    title = gr.Textbox(label='عنوان الكتاب', placeholder='اكتب عنوان الكتاب', lines=1, rtl=True)

    with gr.Row():
      category = gr.Dropdown(choices=['جارٍ التحميل...'], label='التصنيف (اختياري)', interactive=False)
      author = gr.Dropdown(choices=['جارٍ التحميل...'], label='المؤلف (اختياري)', interactive=False)

    search_button = gr.Button('ابحث')

    gr.Markdown('## النتائج 🎯', rtl=True)

    results = gr.Dataframe(
      headers=[
        '#',
        'العنوان',
        'الصفحات',
        'المؤلف',
        'التصنيف',
        'درجة التطابق',
      ],
      interactive=False,
    )

    download_label = gr.Markdown('### تحميل ملفات الكتاب 📥', visible=False, rtl=True)

    with gr.Row(visible=False) as download_buttons_box:
      download_pdf_button = gr.Button('تحميل بصيغة PDF')
      download_txt_button = gr.Button('تحميل بصيغة TXT')
      download_docx_button = gr.Button('تحميل بصيغة DOCX')

    download_output = gr.File(label='تنزيل الملف 📥', visible=False)

    downloading_text = gr.Markdown('جارٍ تجهيز الملف للتحميل...', visible=False, rtl=True)

    demo.load(load_data, outputs=[index_state, category, author])

    search_button.click(
      fn=lambda t, c, a, idx: handle_search(idx, t, c, a),
      inputs=[title, category, author, index_state],
      outputs=[results, results_data, download_buttons_box, download_label, download_output, downloading_text],
    )

    results.select(
      fn=show_details,
      inputs=[results_data],
      outputs=[download_label, selected_book_id, download_buttons_box],
    )

    download_pdf_button.click(
      lambda: gr.update(visible=True), inputs=None, outputs=downloading_text
    ).then(
      fn=download_book,
      inputs=[index_state, selected_book_id, gr.State('pdf')],
      outputs=download_output,
    ).then(
      lambda: (gr.update(visible=False), gr.update(visible=True)),
      inputs=None,
      outputs=[downloading_text, download_output],
    )

    download_txt_button.click(
      lambda: gr.update(visible=True), inputs=None, outputs=downloading_text
    ).then(
      fn=download_book,
      inputs=[index_state, selected_book_id, gr.State('txt')],
      outputs=download_output,
    ).then(
      lambda: (gr.update(visible=False), gr.update(visible=True)),
      inputs=None,
      outputs=[downloading_text, download_output],
    )

    download_docx_button.click(
      lambda: gr.update(visible=True), inputs=None, outputs=downloading_text
    ).then(
      fn=download_book,
      inputs=[index_state, selected_book_id, gr.State('docx')],
      outputs=download_output,
    ).then(
      lambda: (gr.update(visible=False), gr.update(visible=True)),
      inputs=None,
      outputs=[downloading_text, download_output],
    )

  demo.launch()


def load_data() -> tuple[list[list[str | int]], gr.Dropdown, gr.Dropdown]:
  _index = load_index()
  _categories = get_categories(_index)
  _authors = get_authors(_index)

  return (
    _index,
    gr.update(choices=_categories, value=_categories[0], interactive=True),
    gr.update(choices=_authors, value=_authors[0], interactive=True),
  )


def load_index() -> list[list[str | int]]:
  with open('index.tsv', 'r', encoding='utf-8') as file:
    data = list(csv.reader(file, delimiter='\t'))[1:]

  for i in range(len(data)):
    data[i] = [i + 1] + data[i] + [normalize_text(data[i][2])]

  return data


def get_categories(index: list[list[str | int]]) -> list[str]:
  return [''] + sorted(set([row[1] for row in index if row[1]]))


def get_authors(index: list[list[str | int]]) -> list[str]:
  return [''] + sorted(set([row[2] for row in index if row[2]]))


def handle_search(
  index: list[list[str | int]],
  title: str,
  category: str,
  author: str,
) -> tuple:
  title = normalize_text(title)

  reset_buttons = gr.update(visible=False)
  reset_label = gr.update(visible=False)
  reset_download_file = gr.update(visible=False)
  reset_downloading_text = gr.update(visible=False)

  if not title.strip():
    return [
      [['', 'يرجى إدخال عنوان للبحث.', '', '', '']],
      [['', 'يرجى إدخال عنوان للبحث.', '', '', '']],
      reset_buttons,
      reset_label,
      reset_download_file,
      reset_downloading_text,
    ]

  filtered = index

  if category and category != '':
    filtered = [row for row in filtered if row[1] == category]

  if author and author != '':
    filtered = [row for row in filtered if row[2] == author]

  scored_results = []
  for row in filtered:
    score = fuzz.partial_ratio(title, row[-1])

    if score > 50:
      scored_results.append((score, row))

  if not scored_results:
    return [
      [['', 'لم يتم العثور على نتائج مطابقة.', '', '', '']],
      [['', 'لم يتم العثور على نتائج مطابقة.', '', '', '']],
      reset_buttons,
      reset_label,
      reset_download_file,
      reset_downloading_text,
    ]

  scored_results.sort(reverse=True)

  result_table = [[row[0], row[3], row[4], row[2], row[1], score] for score, row in scored_results[:100]]

  return [result_table, result_table, reset_buttons, reset_label, reset_download_file, reset_downloading_text]


def show_details(evt: gr.SelectData, results_data: list[list[str | int]]) -> tuple:
  return [gr.update(visible=True), results_data[evt.index[0]][0], gr.update(visible=True)]


def download_book(index_state: list[list[str | int]], selected_book_id: int, file_type: str) -> str:
  selected_book = index_state[selected_book_id - 1]
  title = f'{selected_book[3]} - {selected_book[2]}'
  file_paths = json.loads(selected_book[5 + {'pdf': 0, 'txt': 1, 'docx': 2}[file_type]].replace("'", '"'))

  if not file_paths:
    raise Exception('الملف غير متوفر')

  safe_title = sanitize_filename(title)

  if len(file_paths) == 1:
    url = f'https://huggingface.co/datasets/ieasybooks-org/prophet-mosque-library/resolve/main/{urllib.parse.quote(file_paths[0][2:])}'
    response = requests.get(url)
    ext = Path(file_paths[0]).suffix
    temp_file_name = os.path.join(tempfile.gettempdir(), f'{safe_title}{ext}')
    temp_file = open(temp_file_name, 'wb')
    temp_file.write(response.content)
    temp_file.close()

    schedule_file_deletion(temp_file_name)

    return temp_file_name
  else:
    temp_dir = tempfile.mkdtemp()
    zip_path = os.path.join(temp_dir, f'{safe_title}.zip')

    with zipfile.ZipFile(zip_path, 'w') as zipf:
      for path in file_paths:
        url = f'https://huggingface.co/datasets/ieasybooks-org/prophet-mosque-library/resolve/main/{urllib.parse.quote(path[2:])}'
        response = requests.get(url)
        filename = os.path.basename(path)
        full_path = os.path.join(temp_dir, filename)

        with open(full_path, 'wb') as f:
          f.write(response.content)

        zipf.write(full_path, arcname=filename)

    schedule_file_deletion(zip_path)

    return zip_path


def schedule_file_deletion(file_path: str, delay: int = 3):
  def delete_file():
    if os.path.exists(file_path):
      os.remove(file_path)

  threading.Timer(delay, delete_file).start()


def sanitize_filename(name: str) -> str:
  return re.sub(r'[\\/:"*?<>|]+', '', name).strip()


def normalize_text(text: str) -> str:
  text = strip_tashkeel(text)

  text = text.replace('أ', 'ا')
  text = text.replace('إ', 'ا')
  text = text.replace('آ', 'ا')
  text = text.replace('ي', 'ى')
  text = text.replace('ة', 'ه')

  return text


if __name__ == '__main__':
  main()