reverting frontend

app.py

@@ -2,51 +2,11 @@ import os
 import json
 import gradio as gr
 import requests
-from dotenv import load_dotenv
 import gradio.components as gc
 import uuid
 
-# Load environment variables
-load_dotenv()
-
-# Get sensitive config from environment variables (set these in your .env file)
-ELASTICSEARCH_URL = os.getenv("ELASTICSEARCH_URL")
-ELASTICSEARCH_USER = os.getenv("ELASTICSEARCH_USER")
-ELASTICSEARCH_PASSWORD = os.getenv("ELASTICSEARCH_PASSWORD")
-OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
-AWS_LAMBDA_URL = os.getenv("AWS_LAMBDA_URL")
-GRADIO_AUTH_USERNAME = os.getenv("GRADIO_AUTH_USERNAME")
-GRADIO_AUTH_PASSWORD = os.getenv("GRADIO_AUTH_PASSWORD")
-
-# Check required env vars (skip Elasticsearch-related)
-missing_vars = []
-for var in ["OPENAI_API_KEY", "AWS_LAMBDA_URL", "GRADIO_AUTH_USERNAME", "GRADIO_AUTH_PASSWORD"]:
-    if not os.getenv(var):
-        missing_vars.append(var)
-if missing_vars:
-    raise RuntimeError(f"Missing required environment variables: {', '.join(missing_vars)}. Please set them in your .env file.")
-
-# Initialize clients
-if ELASTICSEARCH_URL:
-    es = None #Elasticsearch(
-        #ELASTICSEARCH_URL,
-        #basic_auth=(str(ELASTICSEARCH_USER), str(ELASTICSEARCH_PASSWORD))
-    #)
-else:
-    es = None
-
-# Initialize OpenAI
-#openai_client = OpenAI(api_key=OPENAI_API_KEY)
-def chat_completion(messages, model="gpt-3.5-turbo", temperature=0.1):
-    #return openai_client.chat.completions.create(
-    #    model=model,
-    #    messages=messages,
-    #    temperature=temperature
-    #)
-    return None
-
 def process_faq(question, user_id="anonymous", model="claude-sonnet"):
-    """Process FAQ by calling AWS Lambda function with streaming response"""
+    """Process FAQ by calling AWS Lambda function"""
     try:
         # Determine the correct Lambda URL and model parameter based on selection
         if model.startswith("nova-"):
@@ -68,129 +28,91 @@ def process_faq(question, user_id="anonymous", model="claude-sonnet"):
         print(f"DEBUG: Sending to {lambda_url}")
         print(f"DEBUG: Payload: {json.dumps(payload, indent=2)}")
         
-        # Make the API call
-        response = requests.post(
+        # Make the API call with streaming
+        with requests.post(
             lambda_url,
             headers={"Content-Type": "application/json"},
-            json=payload
-        )
-        
-        if response.status_code != 200:
-            return f"Error: Lambda function returned status code {response.status_code}"
-        
-        return response.text
+            json=payload,
+            stream=True
+        ) as response:
+            if response.status_code != 200:
+                return f"Error: Lambda function returned status code {response.status_code}"
+            
+            # Process the response (Lambda returns complete response as plain text)
+            full_response = ""
+            for chunk in response.iter_content(chunk_size=1024, decode_unicode=True):
+                if chunk:
+                    full_response += chunk
+            
+            # Simulate streaming by yielding progressively
+            if full_response:
+                words = full_response.split()
+                current_text = ""
+                for i, word in enumerate(words):
+                    current_text += word + " "
+                    if i % 2 == 0 or i == len(words) - 1:  # Yield every 2 words or at the end
+                        yield current_text.strip()
+            
+            return full_response
             
     except Exception as e:
         return f"Error processing FAQ: {str(e)}"
 
-def natural_to_query(natural_query):
-    """Convert natural language to Elasticsearch query body"""
-    try:
-        prompt = f"""Convert the following natural language query into an Elasticsearch query body.\nThe query should be in JSON format and follow Elasticsearch query DSL syntax.\n\nNatural language query: {natural_query}\n\nReturn only the JSON query body, nothing else."""
-        response = chat_completion([
-            {"role": "system", "content": "You are an expert in Elasticsearch query DSL. Convert natural language to Elasticsearch queries."},
-            {"role": "user", "content": prompt}
-        ], model="gpt-3.5-turbo", temperature=0.1)
-        # Extract and format the query
-        if hasattr(response, 'choices'):
-            # For OpenAI v1.x
-            content = response.choices[0].message.content.strip()
-        else:
-            # For OpenAI v0.x
-            content = response["choices"][0]["message"]["content"].strip()
-        try:
-            query_json = json.loads(content)
-            return json.dumps(query_json, indent=2)
-        except json.JSONDecodeError:
-            return content
-    except Exception as e:
-        return f"Error generating query: {str(e)}"
-
-def execute_elasticsearch_query(query_body):
-    """Execute the Elasticsearch query"""
-    try:
-        # Parse the query body
-        query_json = json.loads(query_body)
-        
-        # Execute the query
-        response = es.search(
-            index="your_index_name",  # Replace with your actual index name
-            body=query_json
-        )
-        
-        # Format the response
-        return json.dumps(response, indent=2)
-    except json.JSONDecodeError:
-        return "Error: Invalid JSON query body"
-    except Exception as e:
-        return f"Error executing query: {str(e)}"
-
 # --- Gradio v4.x UI ---
 def faq_wrapper(question, user_id, model):
-    return process_faq(question, user_id, model)
-
-def elasticsearch_generate(natural_input):
-    return natural_to_query(natural_input)
-
-def elasticsearch_execute(query_body):
-    return execute_elasticsearch_query(query_body)
+    # Gradio expects a non-generator for Interface
+    result = ""
+    for chunk in process_faq(question, user_id, model):
+        result = chunk
+    return result
 
 with gr.Blocks() as demo:
-    gc.Markdown("# MCP Tools - Local Version")
-    with gr.Tab(label="FAQ"):  # type: ignore
-        faq_input = gc.Textbox(label="Enter your question", lines=3)
-        model_selector = gc.Dropdown(
-            label="Select Model",
-            choices=["nova-micro", "nova-lite", "nova-pro", "claude-haiku", "claude-sonnet"],
-            value="claude-sonnet",
-            interactive=True
-        )
-        # Generate random user ID for this session
-        session_user_id = str(uuid.uuid4())[:8]
-        faq_button = gc.Button("Process")
-        faq_output = gc.Textbox(label="Response", lines=10)
-        with gr.Row():  # type: ignore
-            thumbs_down = gc.Button("Report bad response", elem_id="thumbs-down", interactive=True)
-        feedback_msg = gc.Markdown(visible=False)
-
-        def report_bad_response():
-            return gr.update(value="Bad response reported. Thank you for your feedback.", visible=True), gr.update(interactive=False)
-
-        thumbs_down.click(report_bad_response, outputs=[feedback_msg, thumbs_down])
-
-        # Re-enable report button and clear feedback when a new FAQ is processed
-        def reset_feedback(*args):
-            return gr.update(interactive=True), gr.update(value="", visible=False)
-        faq_button.click(reset_feedback, outputs=[thumbs_down, feedback_msg], preprocess=False)
-        faq_button.click(lambda q, m: faq_wrapper(q, session_user_id, m), inputs=[faq_input, model_selector], outputs=faq_output)
-    with gr.Tab(label="Elasticsearch"):  # type: ignore
-        gc.Markdown("### Step 1: Natural Language to Query")
-        natural_input = gc.Textbox(label="Describe what you want to search for", lines=3, placeholder="Example: Find all documents containing 'machine learning' in the title")
-        generate_button = gc.Button("Generate Query")
-        query_output = gc.Textbox(label="Generated Query Body", lines=10, placeholder="The generated Elasticsearch query will appear here")
-        generate_button.click(elasticsearch_generate, inputs=natural_input, outputs=query_output)
-        gc.Markdown("### Step 2: Execute Query")
-        gc.Markdown("You can modify the query above if needed, then click Execute")
-        execute_button = gc.Button("Execute Query")
-        result_output = gc.Textbox(label="Query Results", lines=10, placeholder="The query results will appear here")
-        execute_button.click(elasticsearch_execute, inputs=query_output, outputs=result_output)
+    gc.Markdown("# MCP Chatbot")
+    faq_input = gc.Textbox(label="Enter your question", lines=3)
+    model_selector = gc.Dropdown(
+        label="Select Model",
+        choices=["nova-micro", "nova-pro", "claude-haiku", "claude-sonnet"],
+        value="claude-sonnet",
+        interactive=True
+    )
+    # Generate random user ID for this session
+    session_user_id = str(uuid.uuid4())[:8]
+    faq_button = gc.Button("Process")
+    faq_output = gc.Textbox(label="Response", lines=10)
+    with gr.Row():  # type: ignore
+        thumbs_down = gc.Button("Report bad response", elem_id="thumbs-down", interactive=True)
+    feedback_msg = gc.Markdown(visible=False)
+
+    def report_bad_response():
+        return gr.update(value="Bad response reported. Thank you for your feedback.", visible=True), gr.update(interactive=False)
+
+    thumbs_down.click(report_bad_response, outputs=[feedback_msg, thumbs_down])
+
+    # Re-enable report button and clear feedback when a new FAQ is processed
+    def reset_feedback(*args):
+        return gr.update(interactive=True), gr.update(value="", visible=False)
+    faq_button.click(reset_feedback, outputs=[thumbs_down, feedback_msg], preprocess=False)
+    faq_button.click(lambda q, m: faq_wrapper(q, session_user_id, m), inputs=[faq_input, model_selector], outputs=faq_output)
 
 if __name__ == "__main__":
     # Check if running in Hugging Face Spaces
     is_spaces = os.getenv("SPACE_ID") is not None
     
-    # Get auth credentials from environment variables
-    auth_username = GRADIO_AUTH_USERNAME
-    auth_password = GRADIO_AUTH_PASSWORD
-    
     # Configure launch parameters
-    launch_params = {
-        "server_name": "0.0.0.0",
-        "server_port": int(os.getenv("PORT", 7860)),
-        "share": not is_spaces,  # Only enable sharing when not in Spaces
-        "auth": (auth_username, auth_password),  # Add basic auth
-        "auth_message": "Please enter your credentials to access the application."
-    }
+    if is_spaces:
+        # No auth in Hugging Face Spaces
+        launch_params = {
+            "server_name": "0.0.0.0",
+            "server_port": int(os.getenv("PORT", 7860)),
+            "share": False
+        }
+    else:
+        # Local development - no auth required
+        launch_params = {
+            "server_name": "0.0.0.0",
+            "server_port": int(os.getenv("PORT", 7860)),
+            "share": True
+        }
     
     # Launch the app
-    demo.launch(**launch_params) 
+    demo.launch(**launch_params)