State schema definition in LangChain - Step-by-Step Execution

Concept Flow - State schema definition

Define schema fields

↓

Set field types and defaults

↓

Create schema object

↓

Use schema to validate or store state

↓

Access or update state via schema

This flow shows how to define a state schema by specifying fields and types, creating the schema object, and then using it to manage state.

Execution Sample

LangChain

from langchain.schema import State

state = State(
  user_name=str,
  count=int,
  active=bool
)

state.user_name = "Alice"

Defines a state schema with fields user_name, count, and active, then sets user_name to 'Alice'.

Execution Table

Step	Action	Field/Variable	Value/Type	Result/State
1	Define State schema	user_name	str	Schema created with user_name:str
2	Define State schema	count	int	Schema created with count:int
3	Define State schema	active	bool	Schema created with active:bool
4	Create State instance	state	State object	state with fields user_name, count, active
5	Assign value	state.user_name	"Alice"	state.user_name set to 'Alice'
6	Access value	state.user_name	"Alice"	Returns 'Alice'
7	Assign value	state.count	5	state.count set to 5
8	Assign value	state.active	True	state.active set to True
9	Access value	state.count	5	Returns 5
10	Access value	state.active	True	Returns True
11	Assign wrong type	state.count	"five"	Error or type mismatch
12	End	-	-	State schema enforces types; invalid assignment rejected

💡 Execution stops because the schema enforces types and rejects invalid assignments.

Variable Tracker

Variable	Start	After 5	After 7	After 8	Final
state.user_name	undefined	"Alice"	"Alice"	"Alice"	"Alice"
state.count	undefined	undefined	5	5	5
state.active	undefined	undefined	undefined	True	True

Key Moments - 2 Insights

Why does assigning a string to state.count cause an error?

Because the schema defines count as an int type, assigning a string violates the type rule, as shown in step 11 of the execution_table.

Can we access state fields before assigning values?

No, fields start undefined until assigned, as seen in variable_tracker where initial values are undefined.

Visual Quiz - 3 Questions

Look at the execution_table at step 5, what value is assigned to state.user_name?

A undefined

B "Bob"

C "Alice"

D 5

💡 Hint

Check the 'Value/Type' column at step 5 in execution_table.

At which step does the schema reject an invalid type assignment?

A Step 7

B Step 11

C Step 3

D Step 9

💡 Hint

Look for the step mentioning 'Error or type mismatch' in execution_table.

If we set state.active to False instead of True at step 8, what changes in variable_tracker?

A state.active would be False after step 8

B state.active would remain undefined

C state.active would cause an error

D state.active would be True anyway

💡 Hint

Refer to variable_tracker row for state.active and consider assignment effects.

Concept Snapshot

State schema definition in langchain:
- Define fields with types inside State()
- Create a State object with these fields
- Assign values matching the declared types
- Access values via state.field_name
- Schema enforces type safety, rejects wrong types

Full Transcript

This visual execution trace shows how to define a state schema in langchain by specifying fields and their types. We create a State object with fields user_name as string, count as integer, and active as boolean. Assigning values to these fields updates the state. Accessing fields returns their current values. The schema enforces type safety, so assigning a wrong type like a string to count causes an error. The variable tracker shows how each field changes over time. Key moments clarify why type enforcement matters and when fields start undefined. The quiz tests understanding of assignments, errors, and state changes.