script.py

# -*- coding: utf-8 -*-
"""Pdf-Data 1.ipynb

Automatically generated by Colab.

Original file is located at
    https://colab.research.google.com/drive/1IB0DbFJbA27C0womZkoMQZ7oIkgJkHYU

# Install & import libs
"""

!pip install pypdf pandas tqdm

!apt-get install -y tesseract-ocr
!pip install pytesseract pdf2image pypdf pandas tqdm pillow
!apt-get install -y tesseract-ocr-ara

!apt-get install -y poppler-utils

import os
import pandas as pd
from pypdf import PdfReader
from tqdm import tqdm

"""# Mount Google Drive"""

from google.colab import drive
drive.mount('/content/drive')

"""# Config paths"""

BASE_FOLDER = "/content/drive/MyDrive/OitLab/Text"
OUTPUT_CSV = "/content/drive/MyDrive/OitLab/Text/pdf_dataset1.csv"

"""# Core extraction logic"""

# import os
# from pypdf import PdfReader
# from pdf2image import convert_from_path
# import pytesseract
# from tqdm.notebook import tqdm
# import re

# rows = []

# def clean_text(text):
#     text = re.sub(r'\s+', ' ', text)
#     return text.strip()

# for category in os.listdir(BASE_FOLDER):
#     category_path = os.path.join(BASE_FOLDER, category)

#     if not os.path.isdir(category_path):
#         continue

#     print(f"\nProcessing category: {category}")

#     files = [f for f in os.listdir(category_path) if f.lower().endswith(".pdf")]

#     for file in tqdm(files, desc="PDF files"):
#         pdf_path = os.path.join(category_path, file)

#         try:
#             reader = PdfReader(pdf_path)
#             total_pages = len(reader.pages)

#             for page_num, page in enumerate(
#                 tqdm(reader.pages, desc=f"{file}", total=total_pages, leave=False)
#             ):
#                 text = page.extract_text()
#                 text = "" if text is None else clean_text(text)

#                 # --------- OCR FALLBACK ----------
#                 if len(text) < 30:
#                     images = convert_from_path(
#                         pdf_path,
#                         first_page=page_num + 1,
#                         last_page=page_num + 1
#                     )
#                     ocr_text = pytesseract.image_to_string(
#                         images[0],
#                         lang="ara+eng"
#                     )
#                     text = clean_text(ocr_text)

#                 rows.append({
#                     "name": file,
#                     "page": page_num + 1,
#                     "content": text,
#                     "category": category,
#                     "char": len(text)
#                 })

#         except Exception as e:
#             print(f"Error with {pdf_path}: {e}")

import os
import pandas as pd
from pypdf import PdfReader
from pdf2image import convert_from_path
import pytesseract
from tqdm import tqdm
import re
from multiprocessing import Pool, cpu_count
from functools import partial

# ---------------- HELPERS ----------------
def clean_text(text):
    text = re.sub(r'\s+', ' ', text)
    return text.strip()

# ---------------- CONFIG ----------------
BASE_FOLDER = "/content/drive/MyDrive/OitLab/Text"
OUTPUT_CSV = "/content/drive/MyDrive/OitLab/Text/pdf_dataset1.csv"
PREFERRED_CATEGORIES = ["Historique","Religion","Muslim"]
N_WORKERS = max(1, cpu_count() - 1)

print(f"N_WORKERS: {N_WORKERS}")

# ---------------- LOAD EXISTING CSV ----------------
if os.path.exists(OUTPUT_CSV):
    df_existing = pd.read_csv(OUTPUT_CSV)
else:
    df_existing = pd.DataFrame(columns=["name","page","content","category","char"])

processed_set = set(zip(df_existing['category'], df_existing['name']))

# ---------------- PDF PROCESSOR ----------------
def process_pdf(task):
    category, pdf_path, file_name = task
    pdf_rows = []

    try:
        reader = PdfReader(pdf_path)
        total_pages = len(reader.pages)

        for page_num, page in enumerate(reader.pages):
            text = page.extract_text()
            text = "" if text is None else clean_text(text)

            # OCR fallback
            if len(text) < 30:
                images = convert_from_path(
                    pdf_path,
                    first_page=page_num + 1,
                    last_page=page_num + 1
                )
                ocr_text = pytesseract.image_to_string(
                    images[0],
                    lang="ara+eng"
                )
                text = clean_text(ocr_text)

            pdf_rows.append({
                "name": file_name,
                "page": page_num + 1,
                "content": text,
                "category": category,
                "char": len(text)
            })

    except Exception as e:
        print(f"Error processing {pdf_path}: {e}")

    return pdf_rows

# ---------------- BUILD TASK LIST ----------------
all_categories = [f for f in os.listdir(BASE_FOLDER)
                  if os.path.isdir(os.path.join(BASE_FOLDER, f))]

sorted_categories = []
for p_cat in PREFERRED_CATEGORIES:
    if p_cat in all_categories:
        sorted_categories.append(p_cat)
        all_categories.remove(p_cat)
sorted_categories.extend(all_categories)

tasks = []
for category in sorted_categories:
    category_path = os.path.join(BASE_FOLDER, category)
    files_in_category = [f for f in os.listdir(category_path)
                         if f.lower().endswith(".pdf")]

    for file_name in files_in_category:
        if (category, file_name) in processed_set:
            continue
        pdf_path = os.path.join(category_path, file_name)
        tasks.append((category, pdf_path, file_name))

print(f"Total PDFs to process: {len(tasks)}")
print(f"Using {N_WORKERS} workers")

# ---------------- MULTIPROCESSING ----------------
all_rows = []

with Pool(N_WORKERS) as pool:
    for pdf_rows in tqdm(pool.imap_unordered(process_pdf, tasks),
                         total=len(tasks)):
        if pdf_rows:
            all_rows.extend(pdf_rows)

            # incremental save (safe: only main process writes)
            df_temp = pd.DataFrame(pdf_rows)
            write_header = not os.path.exists(OUTPUT_CSV) or df_existing.empty
            df_temp.to_csv(
                OUTPUT_CSV,
                mode='a',
                header=write_header,
                index=False
            )

print("Processing complete!")