utils/pdf_processor.py

import PyPDF2
import unicodedata
import re
import logging

from .gpt_processor import Translator


class PDFProcessor:
    def __init__(self, file_path: str) -> None:
        self.file_path = file_path
        self.file_info = {
            "file_name": self.file_path.split("/")[-1],
            "file_format": "PDF",
            "total_pages": 0,
            "file_content": {},
            "file_full_content": "",
            "is_chinese": "",
        }
        self.__build_info()

    def __build_info(self) -> None:
        try:
            with open(self.file_path, "rb") as pdf_file:
                pdf_reader = PyPDF2.PdfReader(pdf_file)
                pages = len(pdf_reader.pages)
                self.file_info["total_pages"] = pages
                for i, page in enumerate(pdf_reader.pages):
                    text = page.extract_text()
                    text = unicodedata.normalize("NFKD", text)
                    text = text.replace("\n", " ").replace("\r", "")
                    text = re.sub(" +", " ", text)
                    self.file_info["is_chinese"] = self.__is_chinese(text)

                    page_info = {}
                    logging.info(f"Processing page {i + 1}...")
                    page_info["page_num"] = i + 1
                    page_info["page_content"] = text
                    self.file_info["file_content"][i + 1] = page_info
                    self.file_info["file_full_content"] = (
                        self.file_info["file_full_content"] + page_info["page_content"]
                    )
        except FileNotFoundError:
            print(f"File not found: {self.file_path}")
        except Exception as e:
            print(f"An error occurred: {str(e)}")

    def __is_chinese(self, text: str) -> bool:
        for char in text:
            if char >= "\u4e00" and char <= "\u9fff":
                return True
        return False