Official Documentation

LangGraph

Type Annotations

1. Typed Dictionary

   Used to define dictionaries with a fixed set of keys and their value types

   from typing import TypedDict
   
   class User(TypedDict):
       id: int
       name: str
       is_active: bool
   
   u: User = {"id": 1, "name": "Vignesh", "is_active": True}
   

2. Union

   Used when something can be more than one type

   from typing import Union
   
   def square(x: Union[int, float]) -> Union[int, float]:
       return x * x
   

3. Optional

   The value can be either a specific type or None

   from typing import Optional
   
   def get_name(user_id: int) -> Optional[str]:
       if user_id == 1:
           return "Vignesh"
       return None
   

4. Any

   Type can be anything, same as Typescript any

   from typing import Any
   
   def print_value(x: Any) -> None:
       print(x)
   

5. Lambda Function

   small function written in one line

   add = lambda a, b: a + b
   print(add(5, 3))  # 8
   

Elements

1. State
   • Definition: A dictionary-like object that represents the current memory/context of the graph during execution.
   • Purpose: Holds variables, messages, and data that flow between nodes.
2. Node
   • Definition: A single computation step in the graph
   • Purpose: Processes input from the state and outputs a result that updates the state.
3. Graph
   • Definition: The collection of nodes and edges forming a workflow.
   • Purpose: Defines the sequence and logic of execution between nodes.
4. Edges
   • Definition: Directed connections between nodes.
   • Purpose: Define how the graph moves from one node to another.
5. Conditional Edge
   • Definition: A branching edge that depends on conditions from the state.
   • Purpose: Enables if-else style decision-making inside the graph.
6. Start
   • Definition: The entry point of a graph.
   • Purpose: Defines where execution begins.
7. End
   • Definition: The exit point of the graph.
   • Purpose: Marks the completion of execution.
8. Tools
9. ToolNode