Update app.py

app.py

@@ -1,44 +1,68 @@
 import gradio as gr
+from transformers import pipeline
+import openai
 import os
-from huggingface_hub import InferenceClient, HfApi
 
-# Load token from Hugging Face Secrets (Hugging Face Spaces handles this)
-HF_TOKEN = os.getenv("HUGGINGFACEHUB_API_TOKEN")
+# Get OpenAI API key securely
+openai_api_key = os.getenv("OPENAI_API_KEY") or "sk-your-correct-key-here"
+client = openai.OpenAI(api_key=openai_api_key)
 
-# Check if token is valid
-if HF_TOKEN is None:
-    raise ValueError("❌ HUGGINGFACEHUB_API_TOKEN is not set. Go to your Space → Settings → Secrets.")
+# Load local model for code classification
+code_analyzer = pipeline("text-classification", model="microsoft/codebert-base")
 
-# Use FLAN-T5 for general QA
-MODEL_NAME = "google/flan-t5-base"
-client = InferenceClient(model=MODEL_NAME, token=HF_TOKEN)
+# Code Review Function
+def analyze_code(code):
+    if not code.strip():
+        return "No code provided.", "", ""
+    result = code_analyzer(code)
+    return result[0]["label"], "Consider refactoring for better performance", "Medium"
 
-# Define Q&A function
-def ask_ai(question):
-    if not question.strip():
-        return "⚠️ Please enter a valid question."
+# Metadata Validator (Mock)
+def validate_metadata(metadata):
+    if not metadata.strip():
+        return "No metadata provided.", "", ""
+    return "Field", "Unused field detected", "Remove it to improve performance"
 
+# Natural Language Processor using OpenAI GPT-3.5
+def process_nlp_query(query):
+    if not query.strip():
+        return "No query provided."
     try:
-        prompt = f"Answer the following question clearly:\n{question}"
-        response = client.text_generation(
-            prompt=prompt,
-            max_new_tokens=100,
-            temperature=0.7,
+        response = client.chat.completions.create(
+            model="gpt-3.5-turbo",
+            messages=[
+                {"role": "system", "content": "You are a helpful assistant specialized in Salesforce Apex and metadata."},
+                {"role": "user", "content": query}
+            ]
         )
-        return response.strip()
+        return response.choices[0].message.content.strip()
     except Exception as e:
-        # Return the full error message for debugging
-        return f"❌ Error: {str(e)}"
+        return f"❌ OpenAI API error: {str(e)}"
 
 # Gradio UI
-with gr.Blocks(title="AI Code Review & Metadata Validator") as demo:
-    gr.Markdown("## 🤖 AI Code Review & Metadata Validator")
-    gr.Markdown("Ask any technical question (e.g., Apex, SOQL, Metadata concepts)")
+with gr.Blocks() as demo:
+    gr.Markdown("# 🤖 AI Code Review & Metadata Validator (GPT-Powered)")
 
-    question = gr.Textbox(label="Your question", placeholder="What is a governor limit in Apex?")
-    answer = gr.Textbox(label="AI Response")
+    with gr.Tab("Code Review"):
+        code_input = gr.Textbox(label="Apex / LWC Code", lines=8)
+        issue_type = gr.Textbox(label="Issue Type")
+        suggestion = gr.Textbox(label="AI Suggestion")
+        severity = gr.Textbox(label="Severity")
+        code_button = gr.Button("Analyze Code")
+        code_button.click(analyze_code, inputs=code_input, outputs=[issue_type, suggestion, severity])
 
-    ask_btn = gr.Button("Ask")
-    ask_btn.click(fn=ask_ai, inputs=question, outputs=answer)
+    with gr.Tab("Metadata Validation"):
+        metadata_input = gr.Textbox(label="Metadata XML", lines=8)
+        mtype = gr.Textbox(label="Type")
+        issue = gr.Textbox(label="Issue")
+        recommendation = gr.Textbox(label="Recommendation")
+        metadata_button = gr.Button("Validate Metadata")
+        metadata_button.click(validate_metadata, inputs=metadata_input, outputs=[mtype, issue, recommendation])
+
+    with gr.Tab("Ask AI (Natural Language)"):
+        query_input = gr.Textbox(label="Your question", lines=2, placeholder="e.g. What is a governor limit in Apex?")
+        response_output = gr.Textbox(label="AI Response")
+        nlp_button = gr.Button("Ask")
+        nlp_button.click(process_nlp_query, inputs=query_input, outputs=response_output)
 
 demo.launch()