Wait for User Input¶

In [1]:

var userHomeDir = System.getProperty("user.home");
var localRespoUrl = "file://" + userHomeDir + "/.m2/repository/";
var langchain4jVersion = "1.0.1";
var langchain4jbeta = "1.0.1-beta6";
var langgraph4jVersion = "1.6-SNAPSHOT";

var userHomeDir = System.getProperty("user.home"); var localRespoUrl = "file://" + userHomeDir + "/.m2/repository/"; var langchain4jVersion = "1.0.1"; var langchain4jbeta = "1.0.1-beta6"; var langgraph4jVersion = "1.6-SNAPSHOT";

In [2]:

%%bash 
rm -rf \{userHomeDir}/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/org/bsc/langgraph4j

%%bash rm -rf \{userHomeDir}/Library/Jupyter/kernels/rapaio-jupyter-kernel/mima_cache/org/bsc/langgraph4j

In [ ]:

%dependency /add-repo local \{localRespoUrl} release|never snapshot|always
// %dependency /list-repos
%dependency /add org.bsc.langgraph4j:langgraph4j-core:\{langgraph4jVersion}
%dependency /add org.bsc.langgraph4j:langgraph4j-postgres-saver:\{langgraph4jVersion}
%dependency /add org.bsc.langgraph4j:langgraph4j-langchain4j:\{langgraph4jVersion}
%dependency /add dev.langchain4j:langchain4j:\{langchain4jVersion}
// %dependency /add dev.langchain4j:langchain4j-open-ai:\{langchain4jVersion}
%dependency /add dev.langchain4j:langchain4j-ollama:\{langchain4jbeta}
%dependency /add net.sourceforge.plantuml:plantuml-mit:1.2024.6
%dependency /list-dependencies
%dependency /resolve

%dependency /add-repo local \{localRespoUrl} release|never snapshot|always // %dependency /list-repos %dependency /add org.bsc.langgraph4j:langgraph4j-core:\{langgraph4jVersion} %dependency /add org.bsc.langgraph4j:langgraph4j-postgres-saver:\{langgraph4jVersion} %dependency /add org.bsc.langgraph4j:langgraph4j-langchain4j:\{langgraph4jVersion} %dependency /add dev.langchain4j:langchain4j:\{langchain4jVersion} // %dependency /add dev.langchain4j:langchain4j-open-ai:\{langchain4jVersion} %dependency /add dev.langchain4j:langchain4j-ollama:\{langchain4jbeta} %dependency /add net.sourceforge.plantuml:plantuml-mit:1.2024.6 %dependency /list-dependencies %dependency /resolve

utility to render graph respresentation in PlantUML

In [4]:

import net.sourceforge.plantuml.SourceStringReader;
import net.sourceforge.plantuml.FileFormatOption;
import net.sourceforge.plantuml.FileFormat;
import org.bsc.langgraph4j.GraphRepresentation;

void displayDiagram( GraphRepresentation representation ) throws IOException { 
    
    var reader = new SourceStringReader(representation.getContent());

    try(var imageOutStream = new java.io.ByteArrayOutputStream()) {

        var description = reader.outputImage( imageOutStream, 0, new FileFormatOption(FileFormat.PNG));

        var imageInStream = new java.io.ByteArrayInputStream(  imageOutStream.toByteArray() );

        var image = javax.imageio.ImageIO.read( imageInStream );

        display(  image );

    }
}

import net.sourceforge.plantuml.SourceStringReader; import net.sourceforge.plantuml.FileFormatOption; import net.sourceforge.plantuml.FileFormat; import org.bsc.langgraph4j.GraphRepresentation; void displayDiagram( GraphRepresentation representation ) throws IOException { var reader = new SourceStringReader(representation.getContent()); try(var imageOutStream = new java.io.ByteArrayOutputStream()) { var description = reader.outputImage( imageOutStream, 0, new FileFormatOption(FileFormat.PNG)); var imageInStream = new java.io.ByteArrayInputStream( imageOutStream.toByteArray() ); var image = javax.imageio.ImageIO.read( imageInStream ); display( image ); } }

Define graph with interruption¶

In [5]:

import org.bsc.langgraph4j.*;
import org.bsc.langgraph4j.prebuilt.MessagesState;
import org.bsc.langgraph4j.state.Channel;
import dev.langchain4j.data.message.AiMessage;
import dev.langchain4j.data.message.ChatMessage;
import org.bsc.langgraph4j.action.AsyncNodeAction;
import org.bsc.langgraph4j.action.AsyncEdgeAction;
import static org.bsc.langgraph4j.action.AsyncNodeAction.node_async;
import static org.bsc.langgraph4j.action.AsyncEdgeAction.edge_async;
import org.bsc.langgraph4j.checkpoint.MemorySaver;
import org.bsc.langgraph4j.checkpoint.PostgresSaver;
import org.bsc.langgraph4j.CompileConfig;
import static org.bsc.langgraph4j.StateGraph.END;
import static org.bsc.langgraph4j.StateGraph.START;

public class State extends MessagesState<String> {

    public State(Map<String, Object> initData) {
        super( initData  );
    }

    public Optional<String> humanFeedback() {
        return value("human_feedback");
    }

}

AsyncNodeAction<State> step1 = node_async( state -> {
    return Map.of( "messages", "Step 1" );
});

AsyncNodeAction<State> humanFeedback = node_async( state -> {
    return Map.of();
});

AsyncNodeAction<State> step3 = node_async( state -> {
    return Map.of( "messages", "Step 3" );
});

AsyncEdgeAction<State> evalHumanFeedback = edge_async( state -> {
    var feedback = state.humanFeedback().orElseThrow();
    return ( feedback.equals("next") || feedback.equals("back") ) ? feedback : "unknown";
});

var builder = new StateGraph<>(State.SCHEMA, State::new)
    .addNode("step_1", step1)
    .addNode("human_feedback", humanFeedback)
    .addNode("step_3", step3)
    .addEdge(START, "step_1")
    .addEdge("step_1", "human_feedback")
    .addConditionalEdges("human_feedback", evalHumanFeedback, 
        Map.of( "back", "step_1", "next", "step_3", "unknown", "human_feedback" ))
    .addEdge("step_3", END)
    ;

// Set up memory
// var saver = new MemorySaver();

// Set up Postgres saver
var saver = PostgresSaver.builder()
                .host("localhost")
                .port(5432)
                .user("admin")
                .password("bsorrentino")
                .database("lg4j-store")
                .stateSerializer( builder.getStateSerializer() ) 
                .dropTablesFirst(true)
                .build();

// Add
var compileConfig = CompileConfig.builder()
                        .checkpointSaver(saver)
                        .interruptBefore("human_feedback")
                        .releaseThread(true)
                        .build();

var graph = builder.compile(compileConfig);


displayDiagram( graph.getGraph(GraphRepresentation.Type.PLANTUML, "Human in the Loop", false) );

import org.bsc.langgraph4j.*; import org.bsc.langgraph4j.prebuilt.MessagesState; import org.bsc.langgraph4j.state.Channel; import dev.langchain4j.data.message.AiMessage; import dev.langchain4j.data.message.ChatMessage; import org.bsc.langgraph4j.action.AsyncNodeAction; import org.bsc.langgraph4j.action.AsyncEdgeAction; import static org.bsc.langgraph4j.action.AsyncNodeAction.node_async; import static org.bsc.langgraph4j.action.AsyncEdgeAction.edge_async; import org.bsc.langgraph4j.checkpoint.MemorySaver; import org.bsc.langgraph4j.checkpoint.PostgresSaver; import org.bsc.langgraph4j.CompileConfig; import static org.bsc.langgraph4j.StateGraph.END; import static org.bsc.langgraph4j.StateGraph.START; public class State extends MessagesState { public State(Map initData) { super( initData ); } public Optional humanFeedback() { return value("human_feedback"); } } AsyncNodeAction step1 = node_async( state -> { return Map.of( "messages", "Step 1" ); }); AsyncNodeAction humanFeedback = node_async( state -> { return Map.of(); }); AsyncNodeAction step3 = node_async( state -> { return Map.of( "messages", "Step 3" ); }); AsyncEdgeAction evalHumanFeedback = edge_async( state -> { var feedback = state.humanFeedback().orElseThrow(); return ( feedback.equals("next") || feedback.equals("back") ) ? feedback : "unknown"; }); var builder = new StateGraph<>(State.SCHEMA, State::new) .addNode("step_1", step1) .addNode("human_feedback", humanFeedback) .addNode("step_3", step3) .addEdge(START, "step_1") .addEdge("step_1", "human_feedback") .addConditionalEdges("human_feedback", evalHumanFeedback, Map.of( "back", "step_1", "next", "step_3", "unknown", "human_feedback" )) .addEdge("step_3", END) ; // Set up memory // var saver = new MemorySaver(); // Set up Postgres saver var saver = PostgresSaver.builder() .host("localhost") .port(5432) .user("admin") .password("bsorrentino") .database("lg4j-store") .stateSerializer( builder.getStateSerializer() ) .dropTablesFirst(true) .build(); // Add var compileConfig = CompileConfig.builder() .checkpointSaver(saver) .interruptBefore("human_feedback") .releaseThread(true) .build(); var graph = builder.compile(compileConfig); displayDiagram( graph.getGraph(GraphRepresentation.Type.PLANTUML, "Human in the Loop", false) );

SLF4J: No SLF4J providers were found.
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See https://www.slf4j.org/codes.html#noProviders for further details.

Start graph until interruption¶

In [6]:

// Input
Map<String,Object> initialInput = Map.of("messages", "Step 0");

// Thread
var invokeConfig = RunnableConfig.builder()
                        .threadId("Thread1")
                        .build();

// Run the graph until the first interruption
for (var event : graph.stream(initialInput, invokeConfig)) {
    System.out.println(event);
}

// Input Map initialInput = Map.of("messages", "Step 0"); // Thread var invokeConfig = RunnableConfig.builder() .threadId("Thread1") .build(); // Run the graph until the first interruption for (var event : graph.stream(initialInput, invokeConfig)) { System.out.println(event); }

NodeOutput{node=__START__, state={
	messages=[
	Step 0
	]
	}}
NodeOutput{node=step_1, state={
	messages=[
	Step 0
	Step 1
	]
	}}

Wait for user input and update state¶

⚠️ The Java notebook, until now, doesn't support user input (take a look issue #39) so we could simulate input ⚠️

In [7]:

// We can check the state
System.out.printf("--State before update--\n%s\n", graph.getState(invokeConfig));

// Simulate user input
var userInput = "back"; // back means we want to go back to the previous node
System.out.printf("\n--User Input--\nTell me how you want to update the state: '%s'\n\n", userInput);

// We now update the state as if we are the human_feedback node
var updateConfig = graph.updateState(invokeConfig, Map.of("human_feedback", userInput), null);

// We can check the state
System.out.printf("--State after update--\n%s\n", graph.getState(invokeConfig) );

// We can check the next node, showing that it is node 3 (which follows human_feedback)
System.out.printf("\ngetNext()\n\twith invokeConfig:[%s]\n\twith updateConfig:[%s]\n", 
            graph.getState(invokeConfig).getNext(),  
            graph.getState(updateConfig).getNext());
;

// We can check the state System.out.printf("--State before update--\n%s\n", graph.getState(invokeConfig)); // Simulate user input var userInput = "back"; // back means we want to go back to the previous node System.out.printf("\n--User Input--\nTell me how you want to update the state: '%s'\n\n", userInput); // We now update the state as if we are the human_feedback node var updateConfig = graph.updateState(invokeConfig, Map.of("human_feedback", userInput), null); // We can check the state System.out.printf("--State after update--\n%s\n", graph.getState(invokeConfig) ); // We can check the next node, showing that it is node 3 (which follows human_feedback) System.out.printf("\ngetNext()\n\twith invokeConfig:[%s]\n\twith updateConfig:[%s]\n", graph.getState(invokeConfig).getNext(), graph.getState(updateConfig).getNext()); ;

--State before update--
StateSnapshot{node=step_1, state={
	messages=[
	Step 0
	Step 1
	]
	}, config=RunnableConfig{ threadId=Thread1, checkPointId=6b4937e2-68c3-43f0-ae9b-96ef4c05048e, nextNode=human_feedback, streamMode=VALUES }}

--User Input--
Tell me how you want to update the state: 'back'

--State after update--
StateSnapshot{node=step_1, state={
	messages=[
	Step 0
	Step 1
	]
	human_feedback=back
	}, config=RunnableConfig{ threadId=Thread1, checkPointId=a684697c-57db-40fc-9cd1-8231a39bbab9, nextNode=human_feedback, streamMode=VALUES }}

getNext()
	with invokeConfig:[human_feedback]
	with updateConfig:[human_feedback]

Continue graph execution after interruption¶

In [8]:

import org.bsc.langgraph4j.GraphInput;
// Continue the graph execution
for (var event : graph.stream(GraphInput.resume(), updateConfig)) {
    System.out.println(event);
}

import org.bsc.langgraph4j.GraphInput; // Continue the graph execution for (var event : graph.stream(GraphInput.resume(), updateConfig)) { System.out.println(event); }

NodeOutput{node=human_feedback, state={
	messages=[
	Step 0
	Step 1
	]
	human_feedback=back
	}}
NodeOutput{node=step_1, state={
	messages=[
	Step 0
	Step 1
	]
	human_feedback=back
	}}

Waif for user input (again) and update state¶

⚠️ The Java notebook, until now, doesn't support user input (take a look issue #39) so we could simulate input ⚠️

In [9]:

var userInput = "next"; // 'next' means we want to go to the next node
System.out.printf("\n--User Input--\nTell me how you want to update the state: '%s'\n", userInput);

// We now update the state as if we are the human_feedback node
var updateConfig = graph.updateState(invokeConfig, Map.of("human_feedback", userInput), null);

System.out.printf("\ngetNext()\n\twith invokeConfig:[%s]\n\twith updateConfig:[%s]\n", 
            graph.getState(invokeConfig).getNext(),  
            graph.getState(updateConfig).getNext());
            ;   

var userInput = "next"; // 'next' means we want to go to the next node System.out.printf("\n--User Input--\nTell me how you want to update the state: '%s'\n", userInput); // We now update the state as if we are the human_feedback node var updateConfig = graph.updateState(invokeConfig, Map.of("human_feedback", userInput), null); System.out.printf("\ngetNext()\n\twith invokeConfig:[%s]\n\twith updateConfig:[%s]\n", graph.getState(invokeConfig).getNext(), graph.getState(updateConfig).getNext()); ;

--User Input--
Tell me how you want to update the state: 'next'

getNext()
	with invokeConfig:[human_feedback]
	with updateConfig:[human_feedback]

Continue graph execution after the 2nd interruption¶

In [10]:

// Continue the graph execution
for (var event : graph.stream(GraphInput.resume(), updateConfig)) {
    System.out.println(event);
}

// Continue the graph execution for (var event : graph.stream(GraphInput.resume(), updateConfig)) { System.out.println(event); }

NodeOutput{node=human_feedback, state={
	messages=[
	Step 0
	Step 1
	]
	human_feedback=next
	}}
NodeOutput{node=step_3, state={
	messages=[
	Step 0
	Step 1
	Step 3
	]
	human_feedback=next
	}}
NodeOutput{node=__END__, state={
	messages=[
	Step 0
	Step 1
	Step 3
	]
	human_feedback=next
	}}