app.py

import requests
from bs4 import BeautifulSoup
import openai
import gradio as gr
import os
from dotenv import load_dotenv
import time

# Load environment variables
load_dotenv()
openai.api_key = os.getenv("OPENAI_API_KEY")

# Supported languages and their codes
LANGUAGES = {
    "English": "en",
    "Hindi": "hi",
    "Telugu": "te",
    "Kannada": "kn",
    "Malayalam": "ml",
    "Tamil": "ta"
}

# Function to scrape and summarize content from a URL
def scrape_and_summarize(url):
    try:
        response = requests.get(url, timeout=10)  # Set timeout to 10 seconds
        response.raise_for_status()
        soup = BeautifulSoup(response.content, 'html.parser')
        title = soup.find('title').get_text() if soup.find('title') else "No Title"
        paragraphs = soup.find_all('p')
        content = '\n'.join([para.get_text() for para in paragraphs])
        
        # Truncate content if too large for API
        if len(content) > 2000:
            content = content[:2000] + "..."
        
        # Summarize the content using OpenAI
        prompt = f"Summarize the following content:\n\n{content}"
        response = openai.chat.completions.create(
            model="gpt-4o-mini",
            messages=[
                {"role": "system", "content": "You are an expert summarizer."},
                {"role": "user", "content": prompt}
            ]
        )
        summary = response.choices[0].message.content.strip()
        return title, summary
    except requests.exceptions.RequestException as e:
        return "Error", f"Failed to scrape content from {url}: {str(e)}"
    except Exception as e:
        return "Error", f"Failed to summarize content from {url}: {str(e)}"

# Function to translate content using OpenAI
def translate_content(content, target_language):
    if target_language == "en":
        return content  # No translation needed
    
    try:
        prompt = f"Translate the following content to {target_language}:\n\n{content}"
        response = openai.chat.completions.create(
            model="gpt-4o-mini",
            messages=[
                {"role": "system", "content": "You are a multilingual translator."},
                {"role": "user", "content": prompt}
            ]
        )
        return response.choices[0].message.content.strip()
    except Exception as e:
        return f"Translation failed: {str(e)}"

# Function to create newsletter
def create_newsletter(contents, language):
    try:
        prompt = "Create a newsletter with the following content:\n\n"
        for title, summary, url in contents:
            translated_summary = translate_content(summary, LANGUAGES[language])
            prompt += f"Title: {title}\nURL: {url}\n\n{translated_summary}\n\n"
        
        response = openai.chat.completions.create(
            model="gpt-4o-mini",
            messages=[
                {"role": "system", "content": "You are a helpful assistant expert in making newsletters."},
                {"role": "user", "content": prompt}
            ]
        )
        return response.choices[0].message.content.strip()
    except Exception as e:
        return f"Failed to create newsletter: {str(e)}"

# Function to process URLs and generate newsletter
def process_urls(url1, url2, url3, url4, url5, language):
    urls = [url for url in [url1, url2, url3, url4, url5] if url]
    if not urls:
        return "No URLs provided.", None
    
    with gr.update() as progress:
        progress(0.2)
        time.sleep(1)
    
    contents = []
    for url in urls:
        title, summary = scrape_and_summarize(url)
        contents.append((title, summary, url))
    
    with gr.update() as progress:
        progress(0.6)
        time.sleep(1)
    
    newsletter = create_newsletter(contents, language)
    with gr.update() as progress:
        progress(1.0)
        time.sleep(1)
    
    file_path = "newsletter.txt"
    try:
        with open(file_path, "w", encoding="utf-8") as file:
            file.write(newsletter)
    except Exception as e:
        return f"Failed to save newsletter: {str(e)}", None
    
    return newsletter, file_path

# Gradio interface
iface = gr.Interface(
    fn=process_urls,
    inputs=[
        gr.Textbox(label="URL 1"),
        gr.Textbox(label="URL 2"),
        gr.Textbox(label="URL 3"),
        gr.Textbox(label="URL 4"),
        gr.Textbox(label="URL 5"),
        gr.Dropdown(choices=list(LANGUAGES.keys()), label="Select Language", value="English")
    ],
    outputs=["html", gr.File(label="Download Newsletter")],
    title="Multilingual AI Newsletter Generator",
    description="Enter up to 5 URLs to generate a summarized newsletter in your preferred language. Copy and paste the output into your CMS tool for further editing. A progress indicator will show the process, and you can download the newsletter as a text file."
)

iface.launch()