Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| from PyPDF2 import PdfReader | |
| import google.generativeai as genai | |
| import re | |
| # Configure the Generative AI API | |
| genai.configure(api_key="AIzaSyCtN-Ontp3bOU0g209_dXbDCMCfCULMFCQ") | |
| # Updated Prompt Template | |
| input_prompt = """ | |
| You are an AI question generator. Generate a technical mock test with 30 questions based on the following instructions: | |
| 1. Analyze the resume to identify the candidate's skills, experience, projects, and certifications. | |
| 2. Generate questions that thoroughly test the candidate's knowledge and expertise in these areas. | |
| 3. Analyze the job description to extract key requirements and generate questions that evaluate whether the candidate meets these requirements. | |
| 4. Ensure the questions assess the candidate's capability and alignment with the job's expectations, aiding the hiring team in evaluating the validity of the candidate's claims. | |
| Resume: | |
| {resume_text} | |
| Job Description: | |
| {jd_text} | |
| """ | |
| answer_prompt_template = """ | |
| You are an AI evaluator. Below are questions and a student's answers. Score each answer on a scale of 0 to 1, where 1 is completely correct and 0 is incorrect. Provide the correct answer for each question as well. | |
| Questions and Answers: | |
| {questions_and_answers} | |
| Provide your response in the following format: | |
| 1. [Score] - [Probable Correct Answer] | |
| 2. [Score] - [Probable Correct Answer] | |
| ... | |
| """ | |
| # Function to extract text from PDF | |
| def input_pdf_text(file): | |
| try: | |
| pdf = PdfReader(file) | |
| text = "".join(page.extract_text() for page in pdf.pages) | |
| return text | |
| except Exception as e: | |
| return f"Error extracting text: {e}" | |
| # Function to generate questions using Generative AI | |
| def get_gemini_response(prompt): | |
| model = genai.GenerativeModel('gemini-pro') | |
| response = model.generate_content(prompt) | |
| if hasattr(response.candidates[0], "content") and hasattr(response.candidates[0].content, "parts"): | |
| return response.candidates[0].content.parts[0].text | |
| else: | |
| return None | |
| # Initialize session state for persistence | |
| if "questions" not in st.session_state: | |
| st.session_state.questions = None | |
| if "user_answers" not in st.session_state: | |
| st.session_state.user_answers = {} | |
| if "test_submitted" not in st.session_state: | |
| st.session_state.test_submitted = False | |
| # Streamlit UI | |
| st.title("Technical Mock Test Generator") | |
| st.write("Upload a resume and a job description to generate a technical mock test based on the skills and requirements.") | |
| resume_file = st.file_uploader("Upload Resume (PDF)", type=["pdf"]) | |
| jd_file = st.file_uploader("Upload Job Description (PDF)", type=["pdf"]) | |
| if resume_file and jd_file: | |
| # Reset session state when new files are uploaded | |
| if st.session_state.questions is None or st.session_state.test_submitted: | |
| resume_text = input_pdf_text(resume_file) | |
| jd_text = input_pdf_text(jd_file) | |
| if "Error" in resume_text or "Error" in jd_text: | |
| st.error("Error reading one or both files. Please check the files and try again.") | |
| else: | |
| prompt = input_prompt.format(resume_text=resume_text, jd_text=jd_text) | |
| st.write("Generating questions...") | |
| try: | |
| # Generate questions | |
| questions_content = get_gemini_response(prompt) | |
| if not questions_content: | |
| st.error("Unexpected response format. Please check the Generative AI configuration.") | |
| else: | |
| # Extract questions using regular expression | |
| questions = re.findall(r'\d+\.\s.*', questions_content) | |
| questions = [re.sub(r'^\d+\.\s', '', question).strip() for question in questions] | |
| # Save questions to session state | |
| st.session_state.questions = questions | |
| st.session_state.user_answers = {idx + 1: "" for idx in range(len(questions))} | |
| st.session_state.test_submitted = False | |
| except Exception as e: | |
| st.error(f"Error generating questions: {e}") | |
| # Display questions | |
| if st.session_state.questions: | |
| st.write("### Mock Test Questions") | |
| for idx, question in enumerate(st.session_state.questions, 1): | |
| st.write(f"**Q{idx}:** {question}") | |
| st.session_state.user_answers[idx] = st.text_input( | |
| f"Your Answer for Q{idx}:", | |
| value=st.session_state.user_answers[idx], | |
| key=f"q_{idx}" | |
| ) | |
| # Submit button | |
| if st.button("Submit Test"): | |
| # Prepare input for LLM evaluation | |
| qa_pairs = "\n".join( | |
| f"{idx}. Q: {question} A: {st.session_state.user_answers[idx]}" | |
| for idx, question in enumerate(st.session_state.questions, 1) | |
| ) | |
| eval_prompt = answer_prompt_template.format(questions_and_answers=qa_pairs) | |
| try: | |
| # Get LLM evaluation | |
| eval_response = get_gemini_response(eval_prompt) | |
| # Debug: Display raw response | |
| st.write("### Raw Evaluation Response") | |
| st.write(eval_response) | |
| # Parse response | |
| results = re.findall(r'(\d+)\.\s\[([\d\.]+)\]\s-\s(.*)', eval_response) | |
| # Display score and feedback | |
| if results: | |
| total_score = sum(float(score) for _, score, _ in results) | |
| attempted = sum(1 for idx in st.session_state.user_answers.values() if idx.strip()) | |
| st.subheader(f"Your Score: {total_score:.2f} / {len(results)}") | |
| st.write(f"Total Questions Attempted: {attempted} / {len(results)}") | |
| # Display probable correct answers | |
| st.subheader("Probable Correct Answers") | |
| for idx, (_, score, correct_answer) in enumerate(results, 1): | |
| st.write(f"Q{idx}: {correct_answer} (Score: {score})") | |
| else: | |
| st.error("Unexpected evaluation response format.") | |
| except Exception as e: | |
| st.error(f"Error evaluating answers: {e}") |