Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| from youtube_transcript_api import YouTubeTranscriptApi | |
| import re | |
| import tempfile | |
| import os | |
| import openai | |
| # Set up OpenAI API key | |
| openai.api_key = os.getenv("OPENAI_API_KEY") | |
| # Supported file types for OpenAI API | |
| SUPPORTED_FILE_TYPES = ["mp3", "mp4", "mpeg", "mpga", "m4a", "wav", "webm"] | |
| # Function to transcribe audio using OpenAI Whisper API | |
| def transcribe_audio(file_path): | |
| file_extension = os.path.splitext(file_path)[1][1:] | |
| if file_extension not in SUPPORTED_FILE_TYPES: | |
| return f"Error: Unsupported file type '{file_extension}'. Please upload a valid file." | |
| try: | |
| with open(file_path, "rb") as file: | |
| response = openai.Audio.transcribe( | |
| model="whisper-1", | |
| file=file, | |
| response_format="json" | |
| ) | |
| return response['text'] | |
| except Exception as e: | |
| return f"Error during transcription: {e}" | |
| # Function to get transcript from YouTube | |
| def get_transcript(url): | |
| try: | |
| video_id_match = re.search(r"(?:v=|\/)([0-9A-Za-z_-]{11}).*", url) | |
| if not video_id_match: | |
| return "Error: Invalid YouTube URL" | |
| video_id = video_id_match.group(1) | |
| transcript = YouTubeTranscriptApi.get_transcript(video_id) | |
| transcript_text = ' '.join([entry['text'] for entry in transcript]) | |
| return transcript_text | |
| except Exception as e: | |
| return str(e) | |
| # Function to summarize text using OpenAI API | |
| def summarize_text(text): | |
| try: | |
| response = openai.ChatCompletion.create( | |
| model="gpt-3.5-turbo", | |
| messages=[ | |
| {"role": "system", "content": "You are a helpful assistant."}, | |
| {"role": "user", "content": f"Summarize the following text:\n\n{text}"} | |
| ], | |
| max_tokens=100, | |
| ) | |
| summary = response.choices[0].message['content'].strip() | |
| return summary | |
| except Exception as e: | |
| return f"Error summarizing text: {e}" | |
| # Function to generate quiz questions using OpenAI API | |
| def generate_quiz_questions(text): | |
| try: | |
| response = openai.ChatCompletion.create( | |
| model="gpt-3.5-turbo", | |
| messages=[ | |
| {"role": "system", "content": "You are a helpful assistant."}, | |
| {"role": "user", "content": f"Generate 10 quiz questions from the following text. Make sure the questions cover all key points and are varied in type (e.g., multiple choice, short answer). Number each question and provide the correct answer after the question number:\n\n{text}"} | |
| ], | |
| max_tokens=1000, | |
| ) | |
| quiz_questions = response.choices[0].message['content'].strip() | |
| return quiz_questions | |
| except Exception as e: | |
| return f"Error generating quiz questions: {e}" | |
| # Function to parse quiz questions from generated text | |
| def parse_quiz_questions(quiz_text): | |
| questions = [] | |
| current_question = None | |
| current_choices = [] | |
| correct_answer = None | |
| lines = quiz_text.split("\n") | |
| for line in lines: | |
| line = line.strip() | |
| if re.match(r'^\d+\.', line): # This line is a question number | |
| if current_question: | |
| questions.append({ | |
| "question": current_question, | |
| "choices": current_choices, | |
| "correct_answer": correct_answer | |
| }) | |
| current_question = line | |
| current_choices = [] | |
| correct_answer = None | |
| elif line.startswith("Answer:"): | |
| correct_answer = line.split(":", 1)[1].strip() | |
| elif line: | |
| current_choices.append(line) | |
| # Add the last question if it exists | |
| if current_question: | |
| questions.append({ | |
| "question": current_question, | |
| "choices": current_choices, | |
| "correct_answer": correct_answer | |
| }) | |
| return questions | |
| # Function to generate explanation for quiz answers using OpenAI API | |
| def generate_explanation(question, correct_answer, user_answer): | |
| try: | |
| response = openai.ChatCompletion.create( | |
| model="gpt-3.5-turbo", | |
| messages=[ | |
| {"role": "system", "content": "You are a helpful assistant."}, | |
| {"role": "user", "content": f"Explain why the correct answer to the following question is '{correct_answer}' and not '{user_answer}':\n\n{question}"} | |
| ], | |
| max_tokens=100, | |
| ) | |
| explanation = response.choices[0].message['content'].strip() | |
| return explanation | |
| except Exception as e: | |
| return f"Error generating explanation: {e}" | |
| # Function to check answers and provide feedback | |
| def check_answers(questions, user_answers): | |
| feedback = [] | |
| correct_count = 0 | |
| for i, question in enumerate(questions): | |
| correct_answer = question.get('correct_answer') | |
| user_answer = user_answers.get(f"question_{i+1}", "") | |
| if user_answer == correct_answer: | |
| feedback.append({ | |
| "question": question['question'], | |
| "user_answer": user_answer, | |
| "correct_answer": correct_answer, | |
| "status": "Correct" | |
| }) | |
| correct_count += 1 | |
| else: | |
| explanation = generate_explanation(question['question'], correct_answer, user_answer) | |
| feedback.append({ | |
| "question": question['question'], | |
| "user_answer": user_answer, | |
| "correct_answer": correct_answer, | |
| "status": "Incorrect", | |
| "explanation": explanation | |
| }) | |
| return feedback | |
| # Function to handle uploaded files | |
| def handle_uploaded_file(uploaded_file): | |
| file_path = tempfile.mktemp(suffix=os.path.splitext(uploaded_file.name)[1]) | |
| with open(file_path, "wb") as f: | |
| f.write(uploaded_file.read()) | |
| return file_path | |
| # Streamlit app layout and functionality | |
| st.title("YouTube Transcript Quiz Generator") | |
| st.markdown("**Instructions:** Paste a YouTube link or upload a media file to generate a quiz.") | |
| option = st.selectbox("Choose input type", ("YouTube URL", "Upload audio/video file")) | |
| if "generated_quiz" not in st.session_state: | |
| st.session_state.generated_quiz = False | |
| if option == "YouTube URL": | |
| url = st.text_input("YouTube URL", value="") | |
| if url: | |
| if st.button("Generate Quiz"): | |
| transcript_text = get_transcript(url) | |
| if "Error" not in transcript_text: | |
| summary = summarize_text(transcript_text) | |
| quiz_text = generate_quiz_questions(transcript_text) | |
| questions = parse_quiz_questions(quiz_text) | |
| if len(questions) < 10: | |
| st.error("No valid quiz questions could be generated.") | |
| else: | |
| st.session_state.summary = summary | |
| st.session_state.questions = questions[:10] # Ensure only 10 questions are taken | |
| st.session_state.user_answers = {} | |
| st.session_state.generated_quiz = True | |
| else: | |
| st.error(transcript_text) | |
| if option == "Upload audio/video file": | |
| uploaded_file = st.file_uploader("Choose an audio or video file", type=SUPPORTED_FILE_TYPES) | |
| if uploaded_file: | |
| if st.button("Generate Quiz"): | |
| tmp_file_path = handle_uploaded_file(uploaded_file) | |
| transcript_text = transcribe_audio(tmp_file_path) | |
| os.remove(tmp_file_path) | |
| if "Error" not in transcript_text: | |
| summary = summarize_text(transcript_text) | |
| quiz_text = generate_quiz_questions(transcript_text) | |
| questions = parse_quiz_questions(quiz_text) | |
| if len(questions) < 10: | |
| st.error("No valid quiz questions could be generated.") | |
| else: | |
| st.session_state.summary = summary | |
| st.session_state.questions = questions[:10] # Ensure only 10 questions are taken | |
| st.session_state.user_answers = {} | |
| st.session_state.generated_quiz = True | |
| else: | |
| st.error(transcript_text) | |
| if st.session_state.generated_quiz: | |
| st.write("## Summary") | |
| st.write(st.session_state.summary) | |
| st.write("## Quiz Questions") | |
| for i, question in enumerate(st.session_state.questions): | |
| st.write(f"### Question {i+1}") | |
| st.write(question['question']) | |
| st.session_state.user_answers[f"question_{i+1}"] = st.radio( | |
| label="", | |
| options=question['choices'], | |
| key=f"question_{i+1}" | |
| ) | |
| if st.button("Submit Answers"): | |
| if "questions" in st.session_state and st.session_state.questions: | |
| with st.spinner('Processing your answers...'): | |
| feedback = check_answers(st.session_state.questions, st.session_state.user_answers) | |
| st.write("## Feedback") | |
| for i, item in enumerate(feedback): | |
| with st.expander(f"Question {i+1} Feedback"): | |
| st.write(f"### {item['question']}") | |
| st.write(f"**Your answer:** {item['user_answer']}") | |
| st.write(f"**Correct answer:** {item['correct_answer']}") | |
| if item['status'] == "Incorrect": | |
| st.write(f"**Explanation:** {item['explanation']}") | |
| else: | |
| st.write("Please generate the quiz first.") | |