Update app.py

app.py

@@ -1,54 +1,52 @@
-import requests
-from bs4 import BeautifulSoup
+import fitz  # PyMuPDF for PDF handling
+import docx
 from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
 import gradio as gr
 
-# Function to scrape LinkedIn profile
-def scrape_linkedin_profile(url):
-    response = requests.get(url)
-    if response.status_code == 200:
-        soup = BeautifulSoup(response.content, 'html.parser')
-        profile = {}
+# Function to extract text from PDF
+def extract_text_from_pdf(pdf_file):
+    doc = fitz.open(pdf_file)
+    text = ""
+    for page in doc:
+        text += page.get_text()
+    return text
 
-        # Example scraping logic (adjust based on actual LinkedIn page structure)
-        profile['name'] = soup.find('title').text.strip()
-        profile['headline'] = soup.find('div', {'class': 'ph5'}).text.strip()
-        profile['about'] = soup.find('section', {'id': 'about'}).text.strip() if soup.find('section', {'id': 'about'}) else "No About Section"
+# Function to extract text from DOCX
+def extract_text_from_docx(docx_file):
+    doc = docx.Document(docx_file)
+    full_text = []
+    for paragraph in doc.paragraphs:
+        full_text.append(paragraph.text)
+    return '\n'.join(full_text)
 
-        return profile
-    else:
-        return f"Failed to fetch LinkedIn page, status code: {response.status_code}"
-
-# Function to generate roast using a Gen AI model
-def generate_roast(profile_data):
-    # Initialize the Gen AI model from Hugging Face
+# Function to generate roast based on resume text
+def generate_roast(resume_text):
     tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neo-1.3B")
     model = AutoModelForCausalLM.from_pretrained("EleutherAI/gpt-neo-1.3B")
     
-    # Format profile into a prompt
-    prompt = f"Roast this LinkedIn profile:\nName: {profile_data['name']}\nHeadline: {profile_data['headline']}\nAbout: {profile_data['about']}\n\nRoast:"
+    prompt = f"Roast this resume:\n\n{resume_text}\n\nRoast:"
     
     generator = pipeline('text-generation', model=model, tokenizer=tokenizer)
-    
-    # Generate roast
     roast = generator(prompt, max_length=150, num_return_sequences=1)
     
     return roast[0]['generated_text']
 
 # Gradio interface function
-def roast_linkedin(url):
-    profile_data = scrape_linkedin_profile(url)
-    
-    if isinstance(profile_data, dict):
-        roast = generate_roast(profile_data)
-        return roast
+def roast_resume(file):
+    if file.name.endswith('.pdf'):
+        resume_text = extract_text_from_pdf(file.name)
+    elif file.name.endswith('.docx'):
+        resume_text = extract_text_from_docx(file.name)
     else:
-        return profile_data
+        return "Unsupported file format. Please upload a PDF or DOCX file."
+    
+    roast = generate_roast(resume_text)
+    return roast
 
 # Create Gradio interface
-interface = gr.Interface(fn=roast_linkedin, inputs="text", outputs="text", 
-                         title="LinkedIn Profile Roaster",
-                         description="Enter the LinkedIn profile URL and get a humorous roast generated by AI!")
+interface = gr.Interface(fn=roast_resume, inputs="file", outputs="text", 
+                         title="Resume Roaster",
+                         description="Upload your resume in PDF or DOCX format, and let the AI roast it!")
 
-# Launch Gradio app
+# Launch the app
 interface.launch()