{
  "title": "Sample Document",
  "author": "John Doe",
  "version": "1.0",
  "published": "2025-01-15",
  "tags": ["crawler", "extractor", "json"],
  "content": {
    "summary": "This is a sample JSON document for testing",
    "body": "The extractor should handle nested objects and arrays properly"
  },
  "metadata": {
    "created_at": "2025-01-01T00:00:00Z",
    "updated_at": "2025-01-15T12:00:00Z"
  }

    /**
     * The duration of the video in seconds.
     * Value must be from 1 to 28800 (8 hours).
     */
    private String duration;

    /**
     * The date the video was first published, in W3C format.
     */
    private String publicationDate;

    /**
     * An optional arbitrary string tag describing the video.
     * Maximum 32 characters.
     */
    private String category;

    }

    /**
     * Extracts text from the Publisher input stream.
     * @param in The input stream.
     * @param params The parameters.
     * @return The extracted data.
     */
    @Override
    public ExtractData getText(final InputStream in, final Map<String, String> params) {
        if (in == null) {
            throw new CrawlerSystemException("Microsoft Publisher input stream is null. Cannot extract text from null input.");

      <match value="@string{" type="stringignorecase" offset="0"/>
      <match value="@techreport{" type="stringignorecase" offset="0"/>
      <match value="@unpublished{" type="stringignorecase" offset="0"/>
    </magic>
    <magic priority="30">
      <match value="%" type="string" offset="0">
         <match value="\n@article{" type="stringignorecase" offset="2:128"/>