ReAct 与交互轨迹：建立可学习的 Agent 环境

import java.util.*;
public class Main {
    static void near(double expected,double actual) {
        if (!Double.isFinite(actual) || Math.abs(expected-actual)>1e-9*Math.max(1,Math.abs(expected)))
            throw new AssertionError(expected+" != "+actual);
    }
    static void check(boolean ok) { if(!ok) throw new AssertionError(); }
    static void bad(Runnable r) {
        try { r.run(); } catch(IllegalArgumentException e) { return; }
        throw new AssertionError("invalid input accepted");
    }
enum Action { LEFT, RIGHT, FINISH }
    record Step(Action action,int observation,int reward,boolean terminal) {}
    static String audit(List<Step> steps,int budget) {
        if(steps==null||budget<1||steps.size()>budget) throw new IllegalArgumentException("budget");
        int state=0;boolean ended=false;boolean success=false;
        for(Step s:steps) {
            if(s==null||s.action()==null||ended) throw new IllegalArgumentException("after end");
            if(s.action()==Action.LEFT) state=Math.max(0,state-1);
            if(s.action()==Action.RIGHT) state=Math.min(2,state+1); // 环境重算状态，不信任日志自述。
            ended=s.action()==Action.FINISH;
            success=ended&&state==2;
            int reward=success?1:0; // 只有在目标处提交结束才算成功。
            if(s.observation()!=state||s.reward()!=reward||s.terminal()!=ended)
                throw new IllegalArgumentException("forged transition");
        }
        if(ended) return success?"SUCCESS":"FAILED";
        return steps.size()==budget?"TRUNCATED":"PARTIAL"; // 预算耗尽与正常终止分开。
    }
    public static void main(String[] args) {
Step r1=new Step(Action.RIGHT,1,0,false),r2=new Step(Action.RIGHT,2,0,false);
        check(audit(List.of(r1,r2,new Step(Action.FINISH,2,1,true)),3).equals("SUCCESS"));
        check(audit(List.of(new Step(Action.FINISH,0,0,true)),3).equals("FAILED"));
        check(audit(List.of(r1,r2),2).equals("TRUNCATED"));
        check(audit(List.of(),3).equals("PARTIAL"));
        bad(()->audit(List.of(new Step(Action.RIGHT,2,0,false)),3));
        bad(()->audit(List.of(new Step(Action.FINISH,0,0,true),r1),3));
        bad(()->audit(List.of(r1,r2),1));
        check(audit(List.of(new Step(Action.LEFT,0,0,false)),2).equals("PARTIAL"));
        System.out.println("PASS 8 cases");
    }
}