Update app.py

app.py

@@ -14,6 +14,9 @@ from langchain.schema.runnable import RunnablePassthrough
 from langchain.schema.output_parser import StrOutputParser
 from langchain.llms import HuggingFacePipeline
 import gradio as gr
+import re
+from bs4 import BeautifulSoup
+import inflection
 
 
 """
@@ -32,10 +35,13 @@ pdf_files = ["Apple-10K-2023.pdf", "Apple-10K-2024.pdf"]
 all_documents = []
 
 def preprocess_text(text):
-    text = text.replace("\n", " ").strip()
-    text = ' '.join(text.split())  # Remove extra spaces
+    # Remove HTML tags
+    text = BeautifulSoup(text, "html.parser").get_text()
+    
+    # Remove extra whitespace and newlines
+    text = re.sub(r'\s+', ' ', text).strip()
     return text
-
+    
 for pdf_path in pdf_files:
     loader = PyPDFLoader(pdf_path)
     documents = loader.load()
@@ -99,13 +105,17 @@ conversation_chain = (
     | StrOutputParser()
 )
 
+def to_camel_case(text):
+    """Convert normal text to camelCase using inflection package."""
+    camel_text = inflection.camelize(text, uppercase_first_letter=True)  # Ensure lowerCamelCase
+    return camel_text
+    
 """### 📌 Step 8: Create a function to get the confidence score"""
 
 # Function to Get Confidence Score
 def get_confidence_score(question):
     retrieved_docs_with_scores = vectordb.similarity_search_with_score(question, k=5)
     max_score = max([doc[1] for doc in retrieved_docs_with_scores]) if retrieved_docs_with_scores else 0
-    print(max_score)
     return min(1.0, round(max_score, 2))  # Normalize to 0-1 scale
 
 
@@ -116,11 +126,32 @@ def chat_with_rag(message, history):
     try:
         response = conversation_chain.invoke(message)
         confidence_score = get_confidence_score(message)
-        formatted_response = f"**Answer:** {response}\n\n**Confidence Score:** {confidence_score:.2f}"
+        formatted_response = f"**Answer:** {to_camel_case(response)}\n\n**Confidence Score:** {confidence_score:.2f}"
         return formatted_response
     except Exception as e:
         return f"Error: {str(e)}"
 
+#  A relevant financial question (high-confidence).
+
+user_input = "what are the biggest challenges for Apple?"
+confidence_score = get_confidence_score(user_input)
+output = conversation_chain.invoke(user_input)
+print(f"📌 **Answer:** {to_camel_case(output)}\n\n**Confidence Score:** {confidence_score:.2f}")
+
+# A relevant financial question (low-confidence).
+
+user_input = "what was apple's Total revenue in 2023?"
+confidence_score = get_confidence_score(user_input)
+output = conversation_chain.invoke(user_input)
+print(f"📌 **Answer:** {to_camel_case(output)}\n\n**Confidence Score:** {confidence_score:.2f}")
+
+# An irrelevant question (e.g., "What is the capital of France?") to check system robustness.
+
+user_input = "What is the capital of France?"
+output = conversation_chain.invoke(user_input)
+confidence_score = get_confidence_score(user_input)
+print(f"📌 **Answer:** {to_camel_case(output)}\n\n**Confidence Score:** {confidence_score:.2f}")
+
 # Create Gradio Chatbot UI with Auto-Clearing Input
 demo = gr.ChatInterface(
     fn=chat_with_rag,  # Function to generate responses
@@ -128,9 +159,9 @@ demo = gr.ChatInterface(
     description="Ask questions about financial reports and get AI-powered answers!",
     theme="soft",  # Aesthetic theme
     examples=[
+        ["What are the biggest challenges for Apple?"],
         ["What was Apple's total revenue in 2024?"],
         ["What are the biggest financial risks for Apple?"],
-        ["What are the biggest challenges for Apple?"],
         ["What is the capital of France?"]
     ],
     submit_btn="Ask",  # Customize the submit button text
@@ -138,25 +169,5 @@ demo = gr.ChatInterface(
 )
 
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-# demo = gr.ChatInterface(
-#     respond,
-#     additional_inputs=[
-#         gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-#         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-#         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-#         gr.Slider(
-#             minimum=0.1,
-#             maximum=1.0,
-#             value=0.95,
-#             step=0.05,
-#             label="Top-p (nucleus sampling)",
-#         ),
-#     ],
-# )
-
-
 if __name__ == "__main__":
     demo.launch()