File size: 6,886 Bytes
f3997d4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
from sqlalchemy.orm import Session
from typing import List, Dict, Optional
import json

from app.database.models import Message, Session as ChatSession
from app.services.session_service import SessionService
from app.utils.helpers import generate_id


class ChatService:
    """Service for chat operations and agent orchestration."""
    
    @staticmethod
    def process_message(
        db: Session,
        message: str,
        session_id: Optional[str] = None,
        user_id: Optional[str] = None,
        policy_ids: Optional[List[str]] = None
    ) -> Dict:
        """
        Process a user message and generate response using LangGraph workflow.
        
        Args:
            db: Database session
            message: User message content
            session_id: Optional session ID
            user_id: Optional user ID
            policy_ids: Optional list of policy IDs to search within
            
        Returns:
            Dictionary with response message and metadata
        """
        from app.llm.graph import multi_agent_graph
        
        # Create or get session
        if not session_id:
            chat_session = SessionService.create_session(db, user_id)
            session_id = chat_session.id
        else:
            chat_session = SessionService.get_session_by_id(db, session_id)
            if not chat_session:
                raise ValueError("Session not found")
        
        # Save user message
        user_message = Message(
            id=generate_id(),
            session_id=session_id,
            role="user",
            content=message
        )
        db.add(user_message)
        db.commit()
        
        # Get chat history for context
        chat_history = ChatService.get_chat_history(db, session_id, limit=10)
        
        # Format chat history for the graph
        history_messages = [
            {"role": msg["role"], "content": msg["content"]}
            for msg in chat_history
        ]
        
        # Process query through LangGraph workflow
        response_data = multi_agent_graph.process_query(
            query=message,
            user_id=user_id,
            chat_history=history_messages,
            policy_ids=policy_ids
        )
        
        # Prepare metadata
        meta = {
            "agent": response_data.get("agent", "unknown"),
            "routing_reasoning": response_data.get("routing_reasoning", ""),
            "sources": response_data.get("sources", []),
            "metadata": response_data.get("metadata", {}),
            "policy_names": response_data.get("policy_names", [])  # Policy document names
        }
        
        print(f"[Chat Service] Saving meta with policy_names: {meta.get('policy_names', [])}")
        
        # Save assistant message
        assistant_message = Message(
            id=generate_id(),
            session_id=session_id,
            role="assistant",
            content=response_data.get("answer", "I apologize, but I couldn't generate a response."),
            meta=json.dumps(meta)
        )
        db.add(assistant_message)
        
        # Update session timestamp
        SessionService.update_session_timestamp(db, session_id)
        
        db.commit()
        db.refresh(assistant_message)
        
        # Auto-generate session title if this is the first exchange
        messages_count = db.query(Message).filter(Message.session_id == session_id).count()
        if messages_count == 2 and chat_session.title == "New Conversation":
            # Generate title from first user message
            title = ChatService._generate_session_title(message)
            SessionService.update_session_title(db, session_id, title)
        
        return {
            "message": assistant_message,
            "session_id": session_id,
            "agent": meta["agent"],
            "sources": meta.get("sources", [])
        }
    
    @staticmethod
    def get_chat_history(
        db: Session,
        session_id: str,
        limit: Optional[int] = None
    ) -> List[Dict]:
        """
        Get chat history for a session.
        
        Args:
            db: Database session
            session_id: Session ID
            limit: Optional limit on number of messages
            
        Returns:
            List of message dictionaries
        """
        query = db.query(Message).filter(
            Message.session_id == session_id
        ).order_by(Message.created_at.asc())
        
        if limit:
            # Get last N messages
            total = query.count()
            if total > limit:
                query = query.offset(total - limit)
        
        messages = query.all()
        
        return [
            {
                "id": msg.id,
                "role": msg.role,
                "content": msg.content,
                "meta": json.loads(msg.meta) if msg.meta else {},
                "created_at": msg.created_at.isoformat()
            }
            for msg in messages
        ]
    
    @staticmethod
    def _generate_session_title(first_message: str) -> str:
        """
        Generate a ChatGPT-style session title using LLM.
        
        Args:
            first_message: The first user message in the conversation
            
        Returns:
            A concise, descriptive title (max 50 characters)
        """
        from app.llm.client import llm_client
        
        try:
            prompt = f"""Generate a very short, concise title for a chat conversation that starts with this message:

"{first_message}"

Requirements:
- Maximum 50 characters
- Be specific and descriptive
- Capture the main topic/question
- Professional tone
- No quotes around the title
- Examples: "Building Code Requirements", "Fire Safety Regulations", "Basement Definition"

Return ONLY the title, nothing else:"""

            title = llm_client.get_completion(
                messages=[{"role": "user", "content": prompt}],
                temperature=0.7,
                max_tokens=20
            )
            
            # Clean up the title
            title = title.strip().strip('"').strip("'")
            
            # Ensure it's not too long
            if len(title) > 50:
                title = title[:47] + "..."
            
            # Fallback if empty or too short
            if len(title) < 3:
                title = first_message[:50].strip()
                if len(first_message) > 50:
                    title += "..."
            
            return title
            
        except Exception as e:
            print(f"[Chat Service] Error generating title: {e}")
            # Fallback to simple truncation
            title = first_message[:50].strip()
            if len(first_message) > 50:
                title += "..."
            return title


# Global chat service instance
chat_service = ChatService()