Anforderungen  |   Konzepte  |   Entwurf  |   Entwicklung  |   Qualitätssicherung  |   Lebenszyklus  |   Steuerung
 
 
 
 


Quelle  feedback-reflection.test.ts

  Sprache: JAVA
 

import { mkdtemp, rm, writeFile } from "node:fs/promises";
import os from "node:os";
import path from "node:path";
import { afterEach, describe, expect, it, vi } from "vitest";
import { storeSessionLearning } from "./feedback-reflection-store.js";
import {
  buildFeedbackEvent,
  buildReflectionPrompt,
  clearReflectionCooldowns,
  isReflectionAllowed,
  loadSessionLearnings,
  parseReflectionResponse,
  recordReflectionTime,
} from "./feedback-reflection.js";

describe("buildFeedbackEvent", () => {
  it("builds a well-formed custom event", () => {
    const event = buildFeedbackEvent({
      messageId: "msg-123",
      value: "negative",
      comment: "too verbose",
      sessionKey: "msteams:user1",
      agentId: "default",
      conversationId: "19:abc",
    });

    expect(event.type).toBe("custom");
    expect(event.event).toBe("feedback");
    expect(event.value).toBe("negative");
    expect(event.comment).toBe("too verbose");
    expect(event.messageId).toBe("msg-123");
    expect(event.ts).toBeGreaterThan(0);
  });

  it("omits comment when not provided", () => {
    const event = buildFeedbackEvent({
      messageId: "msg-123",
      value: "positive",
      sessionKey: "msteams:user1",
      agentId: "default",
      conversationId: "19:abc",
    });

    expect(event.comment).toBeUndefined();
    expect(event.value).toBe("positive");
  });
});

describe("buildReflectionPrompt", () => {
  it("includes the thumbed-down response", () => {
    const prompt = buildReflectionPrompt({
      thumbedDownResponse: "Here is a long explanation...",
    });

    expect(prompt).toContain("previous response wasn't helpful");
    expect(prompt).toContain("Here is a long explanation...");
    expect(prompt).toContain("reflect");
  });

  it("truncates long responses", () => {
    const longResponse = "x".repeat(600);
    const prompt = buildReflectionPrompt({
      thumbedDownResponse: longResponse,
    });

    expect(prompt).toContain("...");
    expect(prompt.length).toBeLessThan(longResponse.length + 500);
  });

  it("includes user comment when provided", () => {
    const prompt = buildReflectionPrompt({
      thumbedDownResponse: "Some response",
      userComment: "Too wordy",
    });

    expect(prompt).toContain('User\'s comment: "Too wordy"');
  });

  it("works without optional params", () => {
    const prompt = buildReflectionPrompt({});
    expect(prompt).toContain("previous response wasn't helpful");
    expect(prompt).toContain('"followUp":false');
  });
});

describe("parseReflectionResponse", () => {
  it("parses strict JSON output", () => {
    expect(
      parseReflectionResponse(
        '{"learning":"Be more direct next time.","followUp":true,"userMessage":"Sorry about that. I will keep it tighter."}',
      ),
    ).toEqual({
      learning: "Be more direct next time.",
      followUp: true,
      userMessage: "Sorry about that. I will keep it tighter.",
    });
  });

  it("parses JSON inside markdown fences", () => {
    expect(
      parseReflectionResponse(
        '```json\n{"learning":"Ask a clarifying question first.","followUp":false,"userMessage":""}\n```',
      ),
    ).toEqual({
      learning: "Ask a clarifying question first.",
      followUp: false,
      userMessage: undefined,
    });
  });

  it("falls back to internal-only learning when parsing fails", () => {
    expect(parseReflectionResponse("Be more concise.\nFollow up: yes.")).toEqual({
      learning: "Be more concise.\nFollow up: yes.",
      followUp: false,
    });
  });
});

describe("reflection cooldown", () => {
  afterEach(() => {
    clearReflectionCooldowns();
    vi.restoreAllMocks();
  });

  it("allows first reflection", () => {
    expect(isReflectionAllowed("session-1")).toBe(true);
  });

  it("blocks reflection within cooldown", () => {
    recordReflectionTime("session-1");
    expect(isReflectionAllowed("session-1"60_000)).toBe(false);
  });

  it("allows reflection after cooldown expires", () => {
    // Manually set a past timestamp
    recordReflectionTime("session-1");
    // Override the map entry to simulate time passing
    clearReflectionCooldowns();
    expect(isReflectionAllowed("session-1"1)).toBe(true);
  });

  it("tracks sessions independently", () => {
    recordReflectionTime("session-1");
    expect(isReflectionAllowed("session-1"60_000)).toBe(false);
    expect(isReflectionAllowed("session-2"60_000)).toBe(true);
  });

  it("keeps longer custom cooldown entries during pruning", () => {
    vi.spyOn(Date, "now").mockReturnValue(0);
    recordReflectionTime("target"600_000);

    vi.spyOn(Date, "now").mockReturnValue(301_000);
    for (let index = 0; index <= 500; index += 1) {
      recordReflectionTime(`session-${index}`, 600_000);
    }

    expect(isReflectionAllowed("target"600_000)).toBe(false);
  });
});

describe("loadSessionLearnings", () => {
  let tmpDir: string;

  afterEach(async () => {
    if (tmpDir) {
      await rm(tmpDir, { recursive: true, force: true });
    }
  });

  it("returns empty array when file doesn't exist", async () => {
    tmpDir = await mkdtemp(path.join(os.tmpdir(), "learnings-test-"));
    const learnings = await loadSessionLearnings(tmpDir, "nonexistent");
    expect(learnings).toEqual([]);
  });

  it("reads existing learnings", async () => {
    tmpDir = await mkdtemp(path.join(os.tmpdir(), "learnings-test-"));
    const safeKey = Buffer.from("msteams:user1""utf8").toString("base64url");
    const filePath = path.join(tmpDir, `${safeKey}.learnings.json`);
    await writeFile(filePath, JSON.stringify(["Be concise""Use examples"]), "utf-8");

    const learnings = await loadSessionLearnings(tmpDir, "msteams:user1");
    expect(learnings).toEqual(["Be concise""Use examples"]);
  });

  it("keeps distinct session keys isolated across the filename persistence boundary", async () => {
    tmpDir = await mkdtemp(path.join(os.tmpdir(), "learnings-test-"));

    await storeSessionLearning({
      storePath: tmpDir,
      sessionKey: "msteams:user1",
      learning: "Use bullets",
    });
    await storeSessionLearning({
      storePath: tmpDir,
      sessionKey: "msteams/user1",
      learning: "Avoid bullets",
    });

    await expect(loadSessionLearnings(tmpDir, "msteams:user1")).resolves.toEqual(["Use bullets"]);
    await expect(loadSessionLearnings(tmpDir, "msteams/user1")).resolves.toEqual(["Avoid bullets"]);
  });

  it("reads and migrates legacy sanitized session learning files", async () => {
    tmpDir = await mkdtemp(path.join(os.tmpdir(), "learnings-test-"));
    const legacyFile = path.join(tmpDir, "msteams_user1.learnings.json");
    await writeFile(legacyFile, JSON.stringify(["Legacy learning"]), "utf-8");

    await expect(loadSessionLearnings(tmpDir, "msteams:user1")).resolves.toEqual([
      "Legacy learning",
    ]);

    await storeSessionLearning({
      storePath: tmpDir,
      sessionKey: "msteams:user1",
      learning: "New learning",
    });

    const migratedFile = path.join(
      tmpDir,
      `${Buffer.from("msteams:user1""utf8").toString("base64url")}.learnings.json`,
    );
    await expect(loadSessionLearnings(tmpDir, "msteams:user1")).resolves.toEqual([
      "Legacy learning",
      "New learning",
    ]);
    await expect(rm(legacyFile, { force: false })).rejects.toMatchObject({ code: "ENOENT" });
    await expect(loadSessionLearnings(tmpDir, "msteams:user1")).resolves.toEqual([
      "Legacy learning",
      "New learning",
    ]);
    await expect(loadSessionLearnings(tmpDir, "msteams/user1")).resolves.toEqual([]);
    await expect(
      import("node:fs/promises").then((fs) => fs.readFile(migratedFile, "utf-8")),
    ).resolves.toContain("Legacy learning");
  });
});

Messung V0.5 in Prozent
C=100 H=98 G=98

¤ Dauer der Verarbeitung: 0.11 Sekunden  (vorverarbeitet am  2026-05-26) ¤

*© Formatika GbR, Deutschland






Wurzel

Suchen

Beweissystem der NASA

Beweissystem Isabelle

NIST Cobol Testsuite

Cephes Mathematical Library

Wiener Entwicklungsmethode

Haftungshinweis

Die Informationen auf dieser Webseite wurden nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit, noch Qualität der bereit gestellten Informationen zugesichert.

Bemerkung:

Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.






                                                                                                                                                                                                                                                                                                                                                                                                     


Neuigkeiten

     Aktuelles
     Motto des Tages

Software

     Produkte
     Quellcodebibliothek

Aktivitäten

     Artikel über Sicherheit
     Anleitung zur Aktivierung von SSL

Muße

     Gedichte
     Musik
     Bilder

Jenseits des Üblichen ....

Besucherstatistik

Besucherstatistik

Monitoring

Montastic status badge