Method extractParserTips

src/main/java/gate/corpora/TikaFormat.java:166–190

Tries to extract tips for the parser as specified here - http://tika.apache.org/0.7/parser.html . The tips are not critical for successful parsing. @param doc @return metadata, not null but may be empty

(Document doc)

   * @return metadata, not null but may be empty
   */
  private Metadata extractParserTips(Document doc) {
    Metadata metadata = new Metadata();
    Object inputMime = doc.getFeatures().get("MimeType");
    if (inputMime instanceof String) {	
      if (!"application/tika".equals(inputMime)) {
        metadata.add(Metadata.CONTENT_TYPE, (String) doc.getFeatures().get("MimeType"));
      }
    }
    if (doc instanceof DocumentImpl) {
      if (((DocumentImpl)doc).getMimeType() != null) {
        metadata.add(Metadata.CONTENT_TYPE, ((DocumentImpl)doc).getMimeType());
      }
    }
    if (doc.getSourceUrl() != null && doc.getSourceUrl().getProtocol().startsWith("file")) {
      try {
        File fn =new File(doc.getSourceUrl().toURI());
        metadata.add(Metadata.RESOURCE_NAME_KEY, fn.getName());
      } catch (URISyntaxException e) {
        log.debug("Could not extract filename from uri: " + doc.getSourceUrl(), e);
      } catch (IllegalArgumentException e) {
        log.debug("Could not extract filename from uri: " + doc.getSourceUrl(), e);
      }
    }
    return metadata;
  }
}