Update app.py

app.py

@@ -1,126 +1,115 @@
-# pip install transformers torch accelerate pdfplumber fpdf
+# Hugging Face Spaces-ready version of your JobPrep.AI classic resume generator
+# Upload resume PDF + paste job description → get AI-rewritten resume PDF
 
+import gradio as gr
 from transformers import pipeline, AutoTokenizer, AutoModelForSeq2SeqLM
 import pdfplumber
 from fpdf import FPDF
 import re
+import os
 
-# === CONFIG ===
 MODEL_NAME = "MBZUAI/LaMini-Flan-T5-783M"
-PDF_INPUT = "Resume.Aryan-Salge.pdf"
-PDF_OUTPUT = "Final_Classic_Resume.pdf"
-
-JOB_DESCRIPTION = """
-Looking for a Customer Success Analyst skilled in CRM tools, stakeholder communication,
-data analysis, and reporting using Excel or Tableau.
-"""
-
-# === 1. Extract Resume Text ===
-def extract_text(pdf_path):
-    with pdfplumber.open(pdf_path) as pdf:
-        return "\n".join(page.extract_text() or "" for page in pdf.pages)
-
-raw_resume = extract_text(PDF_INPUT)
-
-# === 2. Load Model ===
 tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
 model = AutoModelForSeq2SeqLM.from_pretrained(MODEL_NAME)
 generator = pipeline("text2text-generation", model=model, tokenizer=tokenizer)
 
-# === 3. Define Sections to Rewrite ===
-sections = {
-    "Professional Summary": "",
-    "Skills": "",
-    "Work Experience": "",
-    "Projects": "",
-    "Education": "",
-    "Certifications": "",
-    "Leadership": ""
-}
-
-# === 4. Extract Section Texts from Resume ===
-for section in sections.keys():
-    match = re.search(rf"{section.upper()}\n(.+?)(?=\n[A-Z ]{{4,}}|\Z)", raw_resume, re.DOTALL)
-    if match:
-        sections[section] = match.group(1).strip()
-
-# === 5. Rewrite Each Section Using AI ===
-def rewrite_section(section_title, content):
-    if not content.strip():
-        return ""
+# Extract raw resume text from uploaded PDF
+def extract_text(pdf_file):
+    with pdfplumber.open(pdf_file.name) as pdf:
+        return "\n".join(page.extract_text() or "" for page in pdf.pages)
+
+# Extract name, email, phone, links from resume
+def extract_personal_info(text):
+    name = text.strip().split("\n")[0].strip()
+    email = re.search(r"[\w\.-]+@[\w\.-]+", text)
+    phone = re.search(r"(\+?\d[\d\s\-\(\)]{7,})", text)
+    links = re.findall(r"(https?://[^\s]+|www\.[^\s]+)", text)
+    location = re.search(r"\b(Dublin|Ireland|[A-Z][a-z]+,?\s?[A-Z][a-z]+?)\b", text)
+    return {
+        "name": name.upper(),
+        "contact": " | ".join(filter(None, [
+            email.group(0) if email else "",
+            phone.group(0).strip() if phone else "",
+            " | ".join(links),
+            location.group(0) if location else ""
+        ]))
+    }
+
+# Rewrite section using LLM
+def rewrite_section(title, content, job):
+    if not content.strip(): return ""
     prompt = f"""
-Rewrite the following section of a resume titled '{section_title}' to align with the job description below.
+Rewrite the following section of a resume titled '{title}' to align with the job description below.
 Make it concise, professional, and rich in job-relevant keywords.
 
 Section:
 {content}
 
 Job Description:
-{JOB_DESCRIPTION}
+{job}
 """
-    result = generator(prompt, max_length=512, do_sample=True, top_p=0.9, temperature=0.7)[0]["generated_text"]
+    result = generator(prompt, max_length=512, do_sample=True, top_p=0.9, temperature=0.7)[0]['generated_text']
     return result.strip()
 
-rewritten_sections = {title: rewrite_section(title, content) for title, content in sections.items()}
+# Generate and save the final resume PDF
+def generate_resume(pdf_file, job_description):
+    raw_text = extract_text(pdf_file)
+    personal = extract_personal_info(raw_text)
+
+    sections = {
+        "Professional Summary": "",
+        "Skills": "",
+        "Work Experience": "",
+        "Projects": "",
+        "Education": "",
+        "Certifications": "",
+        "Leadership": ""
+    }
+
+    for sec in sections.keys():
+        match = re.search(rf"{sec.upper()}\n(.+?)(?=\n[A-Z ]{{4,}}|\Z)", raw_text, re.DOTALL)
+        if match:
+            sections[sec] = match.group(1).strip()
 
-# === 6. PDF Generator (Classic Resume Format) ===
-def save_resume_to_pdf(filename, personal_info, rewritten):
+    rewritten = {title: rewrite_section(title, content, job_description) for title, content in sections.items()}
+
+    out_path = "final_resume.pdf"
     pdf = FPDF()
     pdf.add_page()
     pdf.set_auto_page_break(auto=True, margin=15)
-    pdf.set_font("Arial", size=12)
+    pdf.set_font("Arial", size=11)
 
-    # HEADER
     pdf.set_font("Arial", style="B", size=14)
-    pdf.cell(200, 10, txt=personal_info["name"], ln=True, align="C")
+    pdf.cell(200, 10, txt=personal['name'], ln=True, align="C")
     pdf.set_font("Arial", size=11)
-    pdf.cell(200, 10, txt=personal_info["contact"], ln=True, align="C")
+    pdf.multi_cell(0, 10, personal['contact'], align="C")
     pdf.ln(5)
 
-    # SECTIONS
     for title, body in rewritten.items():
-        if not body.strip():
-            continue
+        if not body.strip(): continue
         pdf.set_font("Arial", style="B", size=12)
         pdf.cell(200, 10, txt=title.upper(), ln=True)
         pdf.set_font("Arial", size=11)
-        for line in body.split("\n"):
-            pdf.multi_cell(0, 10, line.strip())
+        bullets = re.split(r"\s*-\s+", body.strip())
+        for bullet in bullets:
+            bullet = bullet.strip()
+            if bullet:
+                pdf.multi_cell(0, 10, f"- {bullet}")
         pdf.ln(4)
 
-    pdf.output(filename)
-    print(f"✅ Resume saved as: {filename}")
-
-# === 7. Personal Header (manually extracted or hardcoded for now)
-def extract_personal_info(text):
-    # Extract name: first non-empty line
-    name = text.strip().split("\n")[0].strip()
-
-    # Extract email
-    email_match = re.search(r"[\w\.-]+@[\w\.-]+", text)
-    email = email_match.group(0) if email_match else ""
-
-    # Extract phone number (very flexible pattern)
-    phone_match = re.search(r"(\+?\d[\d\s\-\(\)]{7,})", text)
-    phone = phone_match.group(0).strip() if phone_match else ""
-
-    # Extract LinkedIn / GitHub / site
-    links = re.findall(r"(https?://[^\s]+|www\.[^\s]+)", text)
-    linked_text = " | ".join(links)
-
-    # Extract location if found
-    location_match = re.search(r"\b(Dublin|Ireland|Mumbai|Pune|Hyderabad|New York|London|[A-Z][a-z]+,?\s?[A-Z][a-z]+?)\b", text)
-    location = location_match.group(0) if location_match else ""
-
-    contact_string = " | ".join(filter(None, [email, phone, linked_text, location]))
-
-    return {
-        "name": name.upper(),
-        "contact": contact_string
-    }
-
-# Use it:
-personal_info = extract_personal_info(raw_resume)
-
-# === 8. Generate the Final Resume PDF
-save_resume_to_pdf(PDF_OUTPUT, personal_info, rewritten_sections)
+    pdf.output(out_path)
+    return out_path
+
+# Gradio Interface
+demo = gr.Interface(
+    fn=generate_resume,
+    inputs=[
+        gr.File(label="Upload Your Resume PDF", type="file"),
+        gr.Textbox(label="Paste Job Description", lines=7, placeholder="Enter the job description here")
+    ],
+    outputs=gr.File(label="Download Final Resume PDF"),
+    title="JobPrep.AI | Resume Optimizer",
+    description="Upload your resume and tailor it instantly to any job description."
+)
+
+demo.launch()