Create services/claude_service.py

services/claude_service.py

@@ -0,0 +1,115 @@
+# src/services/claude_service.py
+
+import base64
+from anthropic import Anthropic
+import streamlit as st
+from config.settings import ANTHROPIC_API_KEY, CLAUDE_MODEL
+
+class ClaudeService:
+    def __init__(self):
+        """Initialize Claude service with API key from HuggingFace secrets"""
+        if not ANTHROPIC_API_KEY:
+            raise ValueError("Anthropic API key not found in HuggingFace secrets. Please ensure ANTHROPIC_API_KEY is set in your space's secrets.")
+        self.client = Anthropic(api_key=ANTHROPIC_API_KEY)
+        self.model = CLAUDE_MODEL
+
+    def analyze_image(self, image_data, prompt):
+        """Analyze image using Claude Vision"""
+        try:
+            encoded_image = base64.b64encode(image_data).decode('utf-8')
+            
+            message = self.client.messages.create(
+                model=self.model,
+                max_tokens=1000,
+                messages=[{
+                    "role": "user",
+                    "content": [
+                        {
+                            "type": "text",
+                            "text": prompt
+                        },
+                        {
+                            "type": "image",
+                            "source": {
+                                "type": "base64",
+                                "media_type": "image/jpeg",
+                                "data": encoded_image
+                            }
+                        }
+                    ]
+                }]
+            )
+            
+            return message.content[0].text
+        except Exception as e:
+            st.error(f"Error in image analysis: {str(e)}")
+            return None
+
+    def detect_chart_type(self, image_data):
+        """Detect chart type from image"""
+        prompt = """What type of financial chart is this? 
+        Choose from: Candlestick, Line, OHLC, Area, or Other. 
+        Just respond with one word."""
+        
+        result = self.analyze_image(image_data, prompt)
+        return result.strip() if result else "Other"
+
+    def get_educational_content(self, concept):
+        """Get educational content about trading concepts"""
+        try:
+            message = self.client.messages.create(
+                model=self.model,
+                max_tokens=1000,
+                messages=[{
+                    "role": "user",
+                    "content": f"""Please explain the trading concept '{concept}' 
+                    in a clear, educational way. Include:
+                    1. Basic Definition
+                    2. How it Works
+                    3. Key Characteristics
+                    4. When to Look for It
+                    5. Trading Implications
+                    6. Common Mistakes to Avoid
+                    7. Real-World Example"""
+                }]
+            )
+            
+            return message.content[0].text
+        except Exception as e:
+            st.error(f"Error getting educational content: {str(e)}")
+            return None
+
+    def continue_analysis(self, question, previous_analysis, image_data=None):
+        """Continue analysis based on follow-up question"""
+        try:
+            content = [
+                {
+                    "type": "text",
+                    "text": f"""Previous analysis: {previous_analysis}
+                    User's follow-up question: {question}
+                    Please provide a detailed answer to the follow-up question, 
+                    maintaining the context of the previous analysis."""
+                }
+            ]
+            
+            if image_data:
+                encoded_image = base64.b64encode(image_data).decode('utf-8')
+                content.append({
+                    "type": "image",
+                    "source": {
+                        "type": "base64",
+                        "media_type": "image/jpeg",
+                        "data": encoded_image
+                    }
+                })
+            
+            message = self.client.messages.create(
+                model=self.model,
+                max_tokens=1000,
+                messages=[{"role": "user", "content": content}]
+            )
+            
+            return message.content[0].text
+        except Exception as e:
+            st.error(f"Error in follow-up analysis: {str(e)}")
+            return None