<inceptionYear>2010</inceptionYear>

  <licenses>
    <license>
      <name>Apache License, Version 2.0</name>
      <url>http://www.apache.org/licenses/LICENSE-2.0.txt</url>
      <distribution>repo</distribution>
    </license>
  </licenses>

  <developers>
    <!-- Sonatype requires that we list someone here. Guava has many contributors: https://github.com/google/guava/graphs/contributors -->
    <developer>

    String ALT_USER_SETTINGS_XML_LOCATION = "org.apache.maven.user-settings";
    String ALT_GLOBAL_SETTINGS_XML_LOCATION = "org.apache.maven.global-settings";
    String ALT_LOCAL_REPOSITORY_LOCATION = "maven.repo.local";

    Settings buildSettings(MavenExecutionRequest request) throws IOException, XmlPullParserException;

    /**
     * @return a <code>Settings</code> object from the user settings file.

    scm:
      - git:
          name: origin
          # master node jenkins user ~/.ssh
          credentials-id: f6c7695a-671e-4f4f-a331-acdce44ff9ba
          reference-repo: "/var/lib/jenkins/.git-references/elasticsearch.git"
          branches:
            - "${branch_specifier}"
          url: "https://github.com/elastic/elasticsearch.git"
          basedir: ""
          wipe-workspace: true

    }

    private Path getArtifactPath(
            String groupId, String artifactId, String version, String classifier, String extension) {
        Path repo = getProjectLocalRepo();
        return repo.resolve(groupId)
                .resolve(artifactId)
                .resolve(version)
                .resolve(artifactId
                        + '-'
                        + version

        return new MavenProject(model);
    }

    protected List<ArtifactRepository> getRemoteRepositories() throws InvalidRepositoryException {
        File repoDir = new File(getBasedir(), "src/test/remote-repo").getAbsoluteFile();

        RepositoryPolicy policy = new RepositoryPolicy();
        policy.setEnabled(true);
        policy.setChecksumPolicy("ignore");
        policy.setUpdatePolicy("always");

    }

    @Override
    public ArtifactRepository createDefaultRemoteRepository() throws InvalidRepositoryException {
        return new MavenArtifactRepository(
                DEFAULT_REMOTE_REPO_ID,
                "file://"
                        + new File(System.getProperty("basedir", "."), "src/test/remote-repo")

def main() -> None:
    logging.basicConfig(level=logging.INFO)
    settings = Settings()
    logging.info(f"Using config: {settings.model_dump_json()}")
    g = Github(settings.github_token.get_secret_value())
    repo = g.get_repo(settings.github_repository)

    discussion_nodes = get_discussion_nodes(settings=settings)
    experts_results = get_discussions_experts(discussion_nodes=discussion_nodes)

    // ----------------------------------------------------------------------

    protected File getLocalRepositoryPath() throws FileNotFoundException, URISyntaxException {
        File markerFile = getFileForClasspathResource("local-repo/marker.txt");

        return markerFile.getAbsoluteFile().getParentFile();
    }

    protected static File getFileForClasspathResource(String resource)
            throws FileNotFoundException, URISyntaxException {

**Fess Crawler** is a Java-based web crawling framework for enterprise content extraction.

### Essential Info

- **Language**: Java 21+
- **Build**: Maven 3.x
- **License**: Apache 2.0
- **DI**: LastaFlute DI
- **Repo**: https://github.com/codelibs/fess-crawler

### Tech Stack

- **HTTP**: Apache HttpComponents 4.5+
- **Extraction**: Apache Tika 3.0+, POI 5.3+, PDFBox 3.0+
- **Testing**: JUnit 4, UTFlute, Mockito 5.7.0

    steps:
      - name: Dump GitHub context
        env:
          GITHUB_CONTEXT: ${{ toJson(github) }}
        run: echo "$GITHUB_CONTEXT"
      - uses: actions/checkout@v5
        name: Checkout PR for own repo
        if: env.HAS_SECRETS == 'true'
        with:
          # To be able to commit it needs to fetch the head of the branch, not the
          # merge commit
          ref: ${{ github.head_ref }}