Spaces:
Sleeping
Sleeping
File size: 6,468 Bytes
16d5a75 | 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 | from math import log
import os
from typing import TypedDict, Optional, List, Literal
from langchain_core.documents import Document
from langchain_core.tools import tool
from src.utils.helper import (
fake_token_counter,
convert_list_context_source_to_str,
convert_message,
)
from src.utils.logger import logger
from langchain_core.messages import trim_messages, AnyMessage
from .prompt import entry_chain, build_lesson_plan_chain
from .data import primary_level_format, high_school_level_format
from typing import Annotated, Sequence
from langgraph.graph.message import add_messages
from langchain_core.messages import HumanMessage
class State(TypedDict):
user_query: str | AnyMessage
messages_history: list
document_id_selected: Optional[List]
lesson_name: str
subject_name: str
class_number: int
entry_response: str
build_lesson_plan_response: list[AnyMessage]
final_response: str
messages: Annotated[Sequence[AnyMessage], add_messages]
def trim_history(state: State):
history = (
convert_message(state["messages_history"])
if state.get("messages_history")
else None
)
if not history:
return {"messages_history": []}
chat_message_history = trim_messages(
history,
strategy="last",
token_counter=fake_token_counter,
max_tokens=int(os.getenv("HISTORY_TOKEN_LIMIT", 4000)),
start_on="human",
end_on="ai",
include_system=False,
allow_partial=False,
)
return {"messages_history": chat_message_history}
async def entry(state: State):
logger.info(f"Entry {state['messages']}")
entry_response: AnyMessage = await entry_chain.ainvoke(state)
logger.info(f"Entry response: {entry_response}")
logger.info(f"Entry response tool_calls: {entry_response.content}")
# Check if entry_response has tool_calls attribute and it's not empty
if (
hasattr(entry_response, "tool_calls")
and entry_response.tool_calls
and len(entry_response.tool_calls) > 0
and any(
tool_call["name"] == "EntryExtractor"
for tool_call in entry_response.tool_calls
)
and "args" in entry_response.tool_calls[0]
and "class_number" in entry_response.tool_calls[0]["args"]
and "subject_name" in entry_response.tool_calls[0]["args"]
and "lesson_name" in entry_response.tool_calls[0]["args"]
):
logger.info("V么 膽芒y")
class_number = str(int(entry_response.tool_calls[0]["args"]["class_number"]))
subject_name = str(entry_response.tool_calls[0]["args"]["subject_name"])
lesson_name = str(entry_response.tool_calls[0]["args"]["lesson_name"])
return {
"entry_response": entry_response.content,
"class_number": class_number,
"subject_name": subject_name,
"lesson_name": lesson_name,
}
logger.info("kh么ng v么")
return {
"entry_response": entry_response.content,
"class_number": None,
"subject_name": None,
"lesson_name": None,
"final_response": entry_response.content,
}
async def build_lesson_plan(state: State):
logger.info(f"build_lesson_plan {state['messages']}")
has_change_lesson = any(
hasattr(message, "tool_calls")
and any(tool_call["name"] == "ChangeLesson" for tool_call in message.tool_calls)
for message in state["messages"]
)
has_extract_lesson = any(
hasattr(message, "tool_calls")
and any(
tool_call["name"] == "extract_lesson_content"
for tool_call in message.tool_calls
)
for message in state["messages"]
)
if has_change_lesson and not has_extract_lesson:
state["messages"] = []
state["messages_history"] = []
if has_extract_lesson and has_change_lesson:
state["messages"] = [
message
for message in state["messages"]
if not hasattr(message, "tool_calls")
or not any(
tool_call["name"] == "ChangeLesson" for tool_call in message.tool_calls
)
]
class_number = state["class_number"]
if int(class_number) > 5:
lesson_plan_format = high_school_level_format
else:
lesson_plan_format = primary_level_format
state["lesson_plan_format"] = lesson_plan_format
build_lesson_plan_response = await build_lesson_plan_chain.ainvoke(state)
return {
"build_lesson_plan_response": [build_lesson_plan_response],
"messages": build_lesson_plan_response,
"final_response": build_lesson_plan_response.content,
}
def change_lesson(state: State):
build_lesson_plan_response = state["build_lesson_plan_response"][-1]
print("Build lesson plan response: ", build_lesson_plan_response)
logger.info(f"Build lesson plan response: {build_lesson_plan_response}")
# Check if there are tool calls in the response
if (
hasattr(build_lesson_plan_response, "tool_calls")
and build_lesson_plan_response.tool_calls
):
# Extract values from tool_calls
try:
# Get the first tool call (should be ChangeLesson)
tool_call = build_lesson_plan_response.tool_calls[0]
# Extract class_number (handle float conversion if needed)
class_number_value = tool_call["args"]["class_number"]
if isinstance(class_number_value, float):
class_number = str(int(class_number_value))
else:
class_number = str(class_number_value)
# Extract subject and lesson name
subject_name = str(tool_call["args"]["subject_name"])
lesson_name = str(tool_call["args"]["lesson_name"])
logger.info(
f"Extracted lesson data: class={class_number}, subject={subject_name}, lesson={lesson_name}"
)
return {
"class_number": class_number,
"subject_name": subject_name,
"lesson_name": lesson_name,
}
except (KeyError, IndexError, TypeError) as e:
logger.error(f"Error extracting lesson data: {e}")
# Default values if extraction fails
logger.warning("Could not extract lesson data from response")
return {
"class_number": None,
"subject_name": None,
"lesson_name": None,
}
|