Update app.py

app.py

@@ -9,14 +9,16 @@ from dotenv import load_dotenv
 # Set Streamlit page configuration
 st.set_page_config(page_title="Chat with Notes and AI", page_icon=":books:", layout="wide")
 
-# Load environment variables
+# Load environment variables from .env file
 load_dotenv()
 
-# OpenAI API Key (set in .env or directly in your environment)
-OPENAI_API_KEY = os.getenv("OPENAI_API_KEY", "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")
+# Retrieve OpenAI API key from environment
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+if not OPENAI_API_KEY:
+    raise ValueError("OpenAI API key not found. Set it in the .env file or environment variables.")
 openai.api_key = OPENAI_API_KEY
 
-# Template for instruction-only prompts
+# Function to generate response from OpenAI API
 def generate_openai_response(instruction, context=None):
     try:
         messages = [
@@ -35,7 +37,7 @@ def generate_openai_response(instruction, context=None):
     except Exception as e:
         return f"Error: {str(e)}"
 
-# Extracting text from .txt files
+# Extract text from .txt files in a folder
 def get_text_files_content(folder):
     text = ""
     for filename in os.listdir(folder):
@@ -44,7 +46,7 @@ def get_text_files_content(folder):
                 text += file.read() + "\n"
     return text
 
-# Converting text to chunks
+# Convert raw text into manageable chunks
 def get_chunks(raw_text):
     text_splitter = CharacterTextSplitter(
         separator="\n",
@@ -54,13 +56,13 @@ def get_chunks(raw_text):
     )
     return text_splitter.split_text(raw_text)
 
-# Using OpenAI embeddings model and FAISS to create vectorstore
+# Create a FAISS vectorstore using OpenAI embeddings
 def get_vectorstore(chunks):
     embeddings = OpenAIEmbeddings()  # Uses OpenAI Embeddings
     vectorstore = FAISS.from_texts(texts=chunks, embedding=embeddings)
     return vectorstore
 
-# Generating response from user queries
+# Handle user queries by fetching relevant context and generating responses
 def handle_question(question, vectorstore=None):
     if vectorstore:
         # Retrieve relevant chunks using similarity search
@@ -72,37 +74,31 @@ def handle_question(question, vectorstore=None):
         # Fallback to instruction-only prompt if no context is found
         return generate_openai_response(question)
 
+# Main function for the Streamlit app
 def main():
     st.title("Chat with Notes :books:")
 
-    # Initialize session state
+    # Initialize session state for vectorstore
     if "vectorstore" not in st.session_state:
         st.session_state.vectorstore = None
 
     # Define folders for Current Affairs and Essays
-    data_folder = "data"  # Current Affairs folders
-    essay_folder = "essays"  # Essays folder
+    data_folder = "data"  # Folder for Current Affairs notes
+    essay_folder = "essays"  # Folder for Essays
 
     # Content type selection
     content_type = st.sidebar.radio("Select Content Type:", ["Current Affairs", "Essays"])
 
-    # Handle Current Affairs (each subject has its own folder)
+    # Populate subject list based on selected content type
     if content_type == "Current Affairs":
-        if os.path.exists(data_folder):
-            subjects = [f for f in os.listdir(data_folder) if os.path.isdir(os.path.join(data_folder, f))]
-        else:
-            subjects = []
-    # Handle Essays (all essays are in a single folder)
+        subjects = [f for f in os.listdir(data_folder) if os.path.isdir(os.path.join(data_folder, f))] if os.path.exists(data_folder) else []
     elif content_type == "Essays":
-        if os.path.exists(essay_folder):
-            subjects = [f.replace(".txt", "") for f in os.listdir(essay_folder) if f.endswith('.txt')]
-        else:
-            subjects = []
+        subjects = [f.replace(".txt", "") for f in os.listdir(essay_folder) if f.endswith('.txt')] if os.path.exists(essay_folder) else []
 
     # Subject selection
     selected_subject = st.sidebar.selectbox("Select a Subject:", subjects)
 
-    # Process selected subject
+    # Load and process the selected subject
     raw_text = ""
     if content_type == "Current Affairs" and selected_subject:
         subject_folder = os.path.join(data_folder, selected_subject)
@@ -113,12 +109,12 @@ def main():
             with open(subject_file, "r", encoding="utf-8") as file:
                 raw_text = file.read()
 
-    # Display preview of notes
+    # Display notes preview
     if raw_text:
         st.subheader("Preview of Notes")
-        st.text_area("Preview Content:", value=raw_text[:2000], height=300, disabled=True)  # Show a snippet of the notes
+        st.text_area("Preview Content:", value=raw_text[:2000], height=300, disabled=True)
 
-        # Create vectorstore for Current Affairs or Essays
+        # Generate vectorstore for the selected notes
         text_chunks = get_chunks(raw_text)
         vectorstore = get_vectorstore(text_chunks)
         st.session_state.vectorstore = vectorstore
@@ -136,5 +132,6 @@ def main():
         else:
             st.warning("Please load the content for the selected subject before asking a question.")
 
+# Run the app
 if __name__ == '__main__':
     main()