Quelle  web-tools.readability.test.ts

import { beforeEach, describe, expect, it, vi } from "vitest";

const { resolvePluginWebContentExtractorsMock } = vi.hoisted(() => ({
  resolvePluginWebContentExtractorsMock: vi.fn(),
}));

vi.mock("../../plugins/web-content-extractors.runtime.js", () => ({
  resolvePluginWebContentExtractors: resolvePluginWebContentExtractorsMock,
}));

import { extractReadableContent } from "../../web-fetch/content-extractors.runtime.js";

describe("web fetch readability", () => {
  beforeEach(() => {
    resolvePluginWebContentExtractorsMock.mockReset();
  });

  it("dispatches to enabled web content extractors", async () => {
    resolvePluginWebContentExtractorsMock.mockReturnValue([
      {
        id: "readability",
        pluginId: "web-readability",
        label: "Readability",
        extract: vi.fn().mockResolvedValue({
          text: "extracted text",
          title: "Extracted",
        }),
      },
    ]);

    const result = await extractReadableContent({
      html: "<article><p>raw html</p></article>",
      url: "https://example.com/article",
      extractMode: "text",
      config: {},
    });
    expect(result).toMatchObject({
      extractor: "readability",
      text: "extracted text",
      title: "Extracted",
    });
  });

  it("reuses extractor resolution for repeated calls with the same config object", async () => {
    const config = {};
    resolvePluginWebContentExtractorsMock.mockReturnValue([
      {
        id: "readability",
        pluginId: "web-readability",
        label: "Readability",
        extract: vi.fn().mockResolvedValue({
          text: "cached resolver text",
        }),
      },
    ]);

    await extractReadableContent({
      html: "<article><p>first</p></article>",
      url: "https://example.com/first",
      extractMode: "text",
      config,
    });
    await extractReadableContent({
      html: "<article><p>second</p></article>",
      url: "https://example.com/second",
      extractMode: "text",
      config,
    });

    expect(resolvePluginWebContentExtractorsMock).toHaveBeenCalledTimes(1);
    expect(resolvePluginWebContentExtractorsMock).toHaveBeenCalledWith({ config });
  });

  it("returns null when no extractor produces content", async () => {
    resolvePluginWebContentExtractorsMock.mockReturnValue([
      {
        id: "readability",
        pluginId: "web-readability",
        label: "Readability",
        extract: vi.fn().mockResolvedValue(null),
      },
    ]);

    const result = await extractReadableContent({
      html: "<article><p>Main content starts here with enough words to satisfy readability.</p><p>Second paragraph for signal.</p></article>",
      url: "https://example.com/article",
      extractMode: "text",
      config: {},
    });
    expect(result).toBeNull();
  });

  it("continues when a plugin extractor throws", async () => {
    resolvePluginWebContentExtractorsMock.mockReturnValue([
      {
        id: "broken",
        pluginId: "broken-plugin",
        label: "Broken",
        extract: vi.fn().mockRejectedValue(new Error("boom")),
      },
      {
        id: "readability",
        pluginId: "web-readability",
        label: "Readability",
        extract: vi.fn().mockResolvedValue({
          text: "fallback text",
        }),
      },
    ]);

    const result = await extractReadableContent({
      html: "<article><p>raw html</p></article>",
      url: "https://example.com/article",
      extractMode: "text",
      config: {},
    });
    expect(result).toMatchObject({
      extractor: "readability",
      text: "fallback text",
    });
  });

  it("returns null when extractor loading throws", async () => {
    resolvePluginWebContentExtractorsMock.mockImplementation(() => {
      throw new Error("loader boom");
    });

    await expect(
      extractReadableContent({
        html: "<article><p>raw html</p></article>",
        url: "https://example.com/article",
        extractMode: "text",
        config: {},
      }),
    ).resolves.toBeNull();
  });
});