phase-iii / src /utils /context_builder.py
umarwaqar's picture
something deploying
cf699d4
Raw History Blame Contribute Delete
2.94 kB
from typing import List, Dict, Any
from ..models.message import Message
class ContextBuilder:
"""Utility for rebuilding conversation context from database messages"""
@staticmethod
def build_context(messages: List[Message], max_messages: int = 100) -> List[Dict[str, str]]:
"""
Build conversation context from message objects for agent processing
Args:
messages: List of Message objects from database
max_messages: Maximum number of messages to include (default 100)
Returns:
List of dicts with 'role' and 'content' keys formatted for OpenAI
"""
# Limit to max_messages (most recent)
limited_messages = messages[-max_messages:] if len(messages) > max_messages else messages
# Format for OpenAI API
context = []
for msg in limited_messages:
context.append({
"role": msg.role,
"content": msg.content
})
return context
@staticmethod
def get_context_summary(messages: List[Message]) -> Dict[str, Any]:
"""
Get summary statistics about conversation context
Args:
messages: List of Message objects
Returns:
Dict with context statistics
"""
total_messages = len(messages)
user_messages = sum(1 for msg in messages if msg.role == "user")
assistant_messages = sum(1 for msg in messages if msg.role == "assistant")
system_messages = sum(1 for msg in messages if msg.role == "system")
total_chars = sum(len(msg.content) for msg in messages)
return {
"total_messages": total_messages,
"user_messages": user_messages,
"assistant_messages": assistant_messages,
"system_messages": system_messages,
"total_characters": total_chars,
"average_message_length": total_chars // total_messages if total_messages > 0 else 0
}
@staticmethod
def truncate_context(messages: List[Message], max_tokens: int = 4000) -> List[Message]:
"""
Truncate conversation context to fit within token limit
Keeps most recent messages and removes oldest ones
Args:
messages: List of Message objects
max_tokens: Approximate maximum tokens (using 4 chars per token estimate)
Returns:
Truncated list of messages
"""
# Rough estimate: 1 token ≈ 4 characters
max_chars = max_tokens * 4
total_chars = 0
truncated = []
# Start from most recent and work backwards
for msg in reversed(messages):
msg_chars = len(msg.content)
if total_chars + msg_chars <= max_chars:
truncated.insert(0, msg)
total_chars += msg_chars
else:
break
return truncated