app.py

import streamlit as st
import os
from groq import Groq
import numpy as np
import re
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
from docx import Document
from PyPDF2 import PdfReader
from transformers import pipeline

# Initialize Groq client
client = Groq(
    api_key=os.environ.get("GROQ_API_KEY"),
)

# Initialize HuggingFace summarization pipeline
summarizer = pipeline("summarization")

# Function to get Groq analysis of the job description
def groq_chat_completion(prompt):
    chat_completion = client.chat.completions.create(
        messages=[
            {
                "role": "user",
                "content": prompt,
            }
        ],
        model="llama3-8b-8192",
    )
    return chat_completion.choices[0].message.content

# Function to extract text from uploaded files
def extract_text(file):
    if file.type == "text/plain":
        return file.read().decode("utf-8")
    elif file.type == "application/pdf":
        pdf_reader = PdfReader(file)
        text = ""
        for page in pdf_reader.pages:
            text += page.extract_text() or ""
        return text
    elif file.type == "application/vnd.openxmlformats-officedocument.wordprocessingml.document":
        doc = Document(file)
        text = ""
        for para in doc.paragraphs:
            text += para.text + "\n"
        return text
    else:
        return ""

# Function to extract keywords and calculate similarity
def extract_keywords(documents):
    vectorizer = TfidfVectorizer(stop_words="english")
    tfidf_matrix = vectorizer.fit_transform(documents)
    return vectorizer, tfidf_matrix

def calculate_similarity(tfidf_matrix):
    similarity_matrix = cosine_similarity(tfidf_matrix)
    return similarity_matrix

# Function to generate summary for each resume
def generate_summary(text):
    if len(text.split()) > 200:  # Summarize only if the text is long enough
        summary = summarizer(text, max_length=150, min_length=50, do_sample=False)
        return summary[0]['summary_text']
    return text  # Return original text if it's short

# Streamlit UI
st.title("Detail Job Creator and Resume Scanner")
st.write("Analyze resumes and match them with job descriptions.")

# Upload job description and display Groq analysis first
st.subheader("Job Description")
job_description = st.text_area(
    "Paste the job description here:",
    height=150,
)

if job_description:
    st.subheader("Groq Analysis")
    groq_response = groq_chat_completion(job_description)
    st.write("Groq's analysis of the job description:")
    st.write(groq_response)

    # Proceed with resume upload only if job description is provided
    st.subheader("Upload Resumes")
    uploaded_files = st.file_uploader(
        "Upload resume files (Text, Word, or PDF):", 
        accept_multiple_files=True, 
        type=["txt", "docx", "pdf"]
    )

    if st.button("Analyze Resumes"):
        if not uploaded_files:
            st.error("Please upload at least one resume.")
        else:
            # Extract text from resumes
            resumes = [extract_text(file) for file in uploaded_files]
            resumes = [resume for resume in resumes if resume.strip()]  # Filter out empty files

            if not resumes:
                st.error("No valid text extracted from resumes. Please check your files.")
            else:
                # Combine job description and resumes for analysis
                documents = [job_description] + resumes

                # Extract keywords and calculate similarity
                vectorizer, tfidf_matrix = extract_keywords(documents)
                similarities = calculate_similarity(tfidf_matrix)

                # Display match scores and summaries
                st.subheader("Resume Match Scores and Summaries")
                for i, file in enumerate(uploaded_files):
                    st.write(f"**Resume {i+1}: {file.name}**")
                    st.write(f"Match Score: {similarities[0][i + 1] * 100:.2f}%")
                    
                    # Generate and display summary
                    summary = generate_summary(resumes[i])
                    st.write("**Summary:**")
                    st.write(summary)
                    st.write("---")