app.py

# 安裝需要的庫
# 在 Hugging Face Spaces 中運行時，這些依賴應放在 requirements.txt 文件中

import fitz  # PyMuPDF
import gradio as gr
from openai import OpenAI

# Function to extract text from the PDF
def extract_text_from_pdf(pdf_path):
    document = fitz.open(pdf_path)
    all_text = []

    for page_num in range(len(document)):
        page = document.load_page(page_num)
        page_text = page.get_text("text")
        all_text.append(page_text)

    return "\n".join(all_text)

# Function to calculate movie protagonist potential score
def calculate_movie_potential(extracted_text, api_key):
    client = OpenAI(api_key=api_key)
    prompt = (
        "請根據以下內容分析這個人是否適合成為電影主角，並給出百分比評分 (滿分100%)。請依以下分析維度評估：\n"
        "1. 戲劇性（是否使用強烈的情緒用詞，如「驚喜」、「崩潰」）。\n"
        "2. 成長潛力（是否提到挑戰或困難）。\n"
        "3. 幽默感（是否有幽默的描述或笑話）。\n"
        "4. 英雄氣質（是否有「幫助他人」的內容）。\n"
        "5. 劇情代入感（是否有豐富的細節描述）。\n\n"
        f"日記或聊天內容:\n{extracted_text[:4000]}\n\n"
        "請給出總評分（百分比）並分項說明每個維度的得分與理由："
    )
    response = client.chat.completions.create(
        model="gpt-3.5-turbo",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=1000,
        temperature=0.7
    )
    response_text = response.choices[0].message.content.strip()

    return response_text

# Function to process PDF, generate content and calculate movie potential
def process_pdf_and_generate(api_key, pdf_file):
    try:
        # Extract text from the uploaded PDF file
        extracted_text = extract_text_from_pdf(pdf_file)

        # Check if extracted text is empty
        if not extracted_text.strip():
            return "Error: The uploaded PDF appears to be empty or contains unreadable text. Please upload a valid document."

        # Calculate movie protagonist potential
        movie_potential = calculate_movie_potential(extracted_text, api_key)

        # Return the movie potential analysis
        return movie_potential
    except Exception as e:
        return f"Error: {e}"

# Define the Gradio interface
interface = gr.Interface(
    fn=process_pdf_and_generate,
    inputs=[
        gr.Textbox(label="API Key", type="password"),
        gr.File(label="Upload PDF (Diary, Notes, Chat Logs)", type="filepath"),
    ],
    outputs=gr.Textbox(label="Movie Protagonist Potential Analysis"),
    title="Movie Protagonist Analyzer with RAG",
    description=(
        "Upload a PDF (diary, notes, or chat logs) and provide your OpenAI API key. The system will analyze the content "
        "and evaluate how suitable you are to be a movie protagonist based on key dimensions using LangChain (RAG)."
    )
)

# Launch the Gradio app
interface.launch()