Create app.py

app.py

@@ -0,0 +1,184 @@
+import streamlit as st
+import os
+import tempfile
+import json
+import random
+from PyPDF2 import PdfReader
+import openai
+
+# ---------------------------
+# Set up OpenAI API Key
+# ---------------------------
+openai.api_key = os.getenv("OPENAI_API_KEY")  # Ensure this is set in your environment/Space secrets
+
+# ---------------------------
+# Helper Function: Extract text from PDF
+# ---------------------------
+def extract_text(uploaded_file):
+    pdf_reader = PdfReader(uploaded_file)
+    text = ""
+    for page in pdf_reader.pages:
+        text += page.extract_text()
+    return text
+
+# ---------------------------
+# OpenAI Helper Functions
+# ---------------------------
+def generate_summary_from_text(text):
+    prompt = f"Summarize the following document in a concise manner, highlighting the key points that a student should know:\n\n{text}"
+    messages = [
+        {"role": "system", "content": "You are an educational assistant."},
+        {"role": "user", "content": prompt}
+    ]
+    response = openai.ChatCompletion.create(model="gpt-4o-mini", messages=messages)
+    return response.choices[0].message.content.strip()
+
+def chat_with_document(text, conversation_history, user_query):
+    # Append the new query to the conversation history.
+    messages = conversation_history + [
+        {"role": "user", "content": f"Document content:\n\n{text}\n\nQuestion: {user_query}"}
+    ]
+    response = openai.ChatCompletion.create(model="gpt-3.5-turbo", messages=messages)
+    return response.choices[0].message.content.strip()
+
+def generate_flashcards_from_text(text, num_cards):
+    prompt = (
+        f"Generate {num_cards} flashcards based on the following document. "
+        "Return a Python dictionary (in valid JSON format) where each key is a flashcard question and its value is the corresponding answer. "
+        f"Document:\n\n{text}"
+    )
+    messages = [
+        {"role": "system", "content": "You are an educational assistant that creates study flashcards."},
+        {"role": "user", "content": prompt}
+    ]
+    response = openai.ChatCompletion.create(model="gpt-3.5-turbo", messages=messages)
+    output = response.choices[0].message.content.strip()
+    try:
+        # Try to parse the output as JSON and convert it into a Python dict.
+        flashcards = json.loads(output)
+        if isinstance(flashcards, dict):
+            return flashcards
+        else:
+            return {}
+    except Exception as e:
+        st.error(f"Error parsing flashcards: {e}")
+        return {}
+
+# ---------------------------
+# Sidebar: Upload and Mode Selection
+# ---------------------------
+st.sidebar.title("Study Companion Setup")
+
+uploaded_pdf = st.sidebar.file_uploader("Upload your study PDF", type="pdf")
+mode = st.sidebar.radio("Select Mode", ("Summary", "Chat", "Flashcards"))
+
+# For Flashcards mode, ask how many flashcards to generate.
+num_flashcards = None
+if mode == "Flashcards":
+    num_flashcards = st.sidebar.number_input("Number of flashcards to generate:", min_value=1, max_value=20, value=5, step=1)
+
+# ---------------------------
+# Session State Initialization
+# ---------------------------
+if "pdf_text" not in st.session_state:
+    st.session_state.pdf_text = None
+if "summary" not in st.session_state:
+    st.session_state.summary = None
+if "chat_history" not in st.session_state:
+    st.session_state.chat_history = [{"role": "assistant", "content": "Hi, how can I help you with your study material?"}]
+if "flashcards" not in st.session_state:
+    st.session_state.flashcards = {}
+if "current_card" not in st.session_state:
+    st.session_state.current_card = 0
+if "score" not in st.session_state:
+    st.session_state.score = 0
+if "show_answer" not in st.session_state:
+    st.session_state.show_answer = False
+
+# ---------------------------
+# Process PDF Upload
+# ---------------------------
+if uploaded_pdf is not None:
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp:
+        tmp.write(uploaded_pdf.read())
+        pdf_file_path = tmp.name
+    # Extract the full text from the PDF and store it in session state.
+    st.session_state.pdf_text = extract_text(pdf_file_path)
+    st.sidebar.success("PDF uploaded and processed successfully!")
+
+# ---------------------------
+# Main Area: Mode-Based Display
+# ---------------------------
+st.title("Study Companion: PDF-based Learning")
+
+if st.session_state.pdf_text is None:
+    st.info("Please upload a PDF from the sidebar to begin.")
+else:
+    if mode == "Summary":
+        st.header("Summary & Key Points")
+        # Generate summary only once
+        if st.session_state.summary is None:
+            with st.spinner("Generating summary..."):
+                st.session_state.summary = generate_summary_from_text(st.session_state.pdf_text)
+        st.write(st.session_state.summary)
+
+    elif mode == "Chat":
+        st.header("Chat with Your Study Companion")
+        # Display chat history
+        for msg in st.session_state.chat_history:
+            st.chat_message(msg["role"]).write(msg["content"])
+        user_question = st.chat_input("Ask a question about the document:")
+        if user_question:
+            # Preserve current history and then update
+            st.session_state.chat_history.append({"role": "user", "content": user_question})
+            st.chat_message("user").write(user_question)
+            with st.spinner("Processing your question..."):
+                response = chat_with_document(st.session_state.pdf_text, st.session_state.chat_history, user_question)
+            st.session_state.chat_history.append({"role": "assistant", "content": response})
+            st.chat_message("assistant").write(response)
+
+    elif mode == "Flashcards":
+        st.header("Practice Flashcards")
+        # Provide a button to generate flashcards on demand.
+        if st.button("Generate Flashcards"):
+            with st.spinner("Generating flashcards..."):
+                flashcards = generate_flashcards_from_text(st.session_state.pdf_text, num_flashcards)
+            st.session_state.flashcards = flashcards
+            st.session_state.current_card = 0
+            st.session_state.score = 0
+            st.session_state.show_answer = False
+            st.success("Flashcards generated successfully!")
+        
+        if not st.session_state.flashcards:
+            st.info("No flashcards available. Click the button above to generate flashcards.")
+        else:
+            # Loop through flashcards stored as a dictionary.
+            flashcards = st.session_state.flashcards
+            total_cards = len(flashcards)
+            # Check if all flashcards have been practiced.
+            if st.session_state.current_card >= total_cards:
+                st.success(f"You've completed all flashcards! Final Score: {st.session_state.score} / {total_cards}")
+                st.info("Restart the session or generate new flashcards from the sidebar.")
+            else:
+                # Get the current flashcard key-value pair.
+                current_keys = list(flashcards.keys())
+                current_key = current_keys[st.session_state.current_card]
+                current_answer = flashcards[current_key]
+                st.write(f"**Question:** {current_key}")
+                if st.button("Show Answer"):
+                    st.session_state.show_answer = True
+                if st.session_state.show_answer:
+                    st.write(f"**Answer:** {current_answer}")
+                    col1, col2 = st.columns(2)
+                    with col1:
+                        if st.button("Correct"):
+                            st.session_state.score += 1
+                            st.success("Correct!")
+                    with col2:
+                        if st.button("Wrong"):
+                            st.error("Incorrect!")
+                    if st.button("Next Card"):
+                        st.session_state.current_card += 1
+                        st.session_state.show_answer = False
+                        st.rerun()
+                st.write(f"**Current Score:** {st.session_state.score} / {total_cards}")