Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Allow specifying streaming option for public_stream route #141

Merged
merged 1 commit into from
Sep 11, 2024
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion backend/app/api/routes/teams.py
Original file line number Diff line number Diff line change
Expand Up @@ -224,6 +224,7 @@ async def public_stream(
team_chat: TeamChatPublic,
thread_id: str,
team: CurrentTeam,
streaming: bool = True,
) -> StreamingResponse:
"""
Stream a response from a team using a given message or an interrupt decision. Requires an API key for authentication.
Expand All @@ -233,6 +234,7 @@ async def public_stream(
Parameters:
- `team_id` (int): The ID of the team to which the message is being sent. Must be a valid team ID.
- `thread_id` (str): The ID of the thread where the message will be posted. If the thread ID does not exist, a new thread will be created.
- `streaming` (bool, optional): A flag to enable or disable streaming mode. If `True` (default), the messages will be streamed in chunks.

Request Body (JSON):
- The request body should be a JSON object containing either the `message` or `interrupt` field:
Expand Down Expand Up @@ -277,6 +279,6 @@ async def public_stream(

messages = [team_chat.message] if team_chat.message else []
return StreamingResponse(
generator(team, members, messages, thread_id, team_chat.interrupt),
generator(team, members, messages, thread_id, team_chat.interrupt, streaming),
media_type="text/event-stream",
)
3 changes: 2 additions & 1 deletion backend/app/core/graph/build.py
Original file line number Diff line number Diff line change
Expand Up @@ -502,6 +502,7 @@ async def generator(
messages: list[ChatMessage],
thread_id: str,
interrupt: Interrupt | None = None,
streaming: bool = True,
) -> AsyncGenerator[Any, Any]:
"""Create the graph and stream responses as JSON."""
formatted_messages = [
Expand Down Expand Up @@ -602,7 +603,7 @@ async def generator(
]
}
async for event in root.astream_events(state, version="v2", config=config):
response = event_to_response(event)
response = event_to_response(event, streaming)
if response:
formatted_output = f"data: {response.model_dump_json()}\n\n"
yield formatted_output
Expand Down
22 changes: 14 additions & 8 deletions backend/app/core/graph/messages.py
Original file line number Diff line number Diff line change
Expand Up @@ -38,24 +38,30 @@ def get_message_type(message: Any) -> str | None:
return None


def event_to_response(event: StreamEvent) -> ChatResponse | None:
def event_to_response(event: StreamEvent, streaming: bool) -> ChatResponse | None:
"""Convert event to ChatResponse"""
kind = event["event"]
id = event["run_id"]
if kind == "on_chat_model_stream":
# Either listen to stream or end based on streaming arg
chat_model_event_kind = "on_chat_model_stream" if streaming else "on_chat_model_end"
if kind == chat_model_event_kind:
name = event["metadata"]["langgraph_node"]
message_chunk: AIMessageChunk = event["data"]["chunk"]
type = get_message_type(message_chunk)
chat_message: AIMessage | AIMessageChunk = (
event["data"]["chunk"]
if kind == "on_chat_model_stream"
else event["data"]["output"]
)
type = get_message_type(chat_message)
content: str = ""
if isinstance(message_chunk.content, list):
for c in message_chunk.content:
if isinstance(chat_message.content, list):
for c in chat_message.content:
if isinstance(c, str):
content += c
elif isinstance(c, dict):
content += c.get("text", "")
else:
content = message_chunk.content
tool_calls = message_chunk.tool_calls
content = chat_message.content
tool_calls = chat_message.tool_calls
if content and type:
return ChatResponse(
type=type, id=id, name=name, content=content, tool_calls=tool_calls
Expand Down