public static WebDocument getDocument(Page page) {
InputStream stream = new ByteArrayInputStream(page.getContentBytes());
try {
Metadata metadata = new Metadata();
metadata.set(Metadata.CONTENT_TYPE, page.getContentType());
metadata.set(Metadata.CONTENT_LOCATION, page.getURL().toString());
metadata.set(Metadata.LOCATION, page.getURL().toString());
metadata.set(Metadata.MIME_TYPE_MAGIC, page.getContentType());
metadata.set(Metadata.CONTENT_ENCODING, page.getContentEncoding());
metadata.set(Metadata.TIKA_MIME_FILE, page.getContentType());
String text = getTika().parseToString(stream, metadata);
WebDocument wd =
new WebDocument(0, metadata.get(Metadata.TITLE), text, page.getURL().toString());
stream.close();
return wd;
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} catch (TikaException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} finally {
try {
stream.close();
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} // close the stream
}
return null;
}
private void manageDetails(final GetItemResponse response, final DataHandler stream) {
InputStream is = null;
ContentHandler contenthandler = new BodyContentHandler();
Metadata metadata = new Metadata();
// metadata.set(Metadata.RESOURCE_NAME_KEY, f.getName());
Parser parser = new AutoDetectParser();
ParseContext context = new ParseContext();
try {
is = stream.getInputStream();
parser.parse(is, contenthandler, metadata, context);
is.close();
is.reset();
} catch (IOException e) {
e.printStackTrace();
} catch (SAXException e) {
e.printStackTrace();
} catch (TikaException e) {
e.printStackTrace();
}
String contentAuthorValue = metadata.get(Metadata.AUTHOR);
String contentAuthorKey = currentProperties.getProperty(KpeopleLabel.getCorePropertiesAuthor());
if (contentAuthorValue != null) {
eventResult.setDetail(contentAuthorKey, contentAuthorValue);
}
String contentCreationDateValue = metadata.get(Metadata.CREATION_DATE);
String contentCreationDateKey =
currentProperties.getProperty(KpeopleLabel.getCorePropertiesCreationDate());
if (contentCreationDateValue != null) {
eventResult.setDetail(contentCreationDateKey, contentCreationDateValue);
}
String contentKeywordsValue = metadata.get(Metadata.KEYWORDS);
String contentKeywordsKey =
currentProperties.getProperty(KpeopleLabel.getCorePropertiesKeywords());
if (contentKeywordsValue != null) {
eventResult.setDetail(contentKeywordsKey, contentKeywordsValue);
}
String[] names = metadata.names();
/*
* for (int i = 0; i < names.length; i++) {
* System.out.println(names[i]); }
*/
}
/** @param args */
public static void main(String[] args) {
// String fileLocation = "G:/asas/album/song.mp3";
String fileLocation = "C:\\Users\\Public\\Music\\Sample Music\\Kalimba.mp3";
try {
InputStream input = new FileInputStream(new File(fileLocation));
ContentHandler handler = new DefaultHandler();
Metadata metadata = new Metadata();
Parser parser = new Mp3Parser();
ParseContext parseCtx = new ParseContext();
parser.parse(input, handler, metadata, parseCtx);
input.close();
// List all metadata
String[] metadataNames = metadata.names();
for (String name : metadataNames) {
System.out.println(name + ": " + metadata.get(name));
}
// Retrieve the necessary info from metadata
// Names - title, xmpDM:artist etc. - mentioned below may differ
// based
System.out.println("----------------------------------------------");
System.out.println("Title: " + metadata.get("title"));
System.out.println("Artists: " + metadata.get("xmpDM:artist"));
System.out.println("Composer : " + metadata.get("xmpDM:composer"));
System.out.println("Genre : " + metadata.get("xmpDM:genre"));
System.out.println("Album : " + metadata.get("xmpDM:album"));
} catch (FileNotFoundException e) {
e.printStackTrace();
} catch (IOException e) {
e.printStackTrace();
} catch (SAXException e) {
e.printStackTrace();
} catch (TikaException e) {
e.printStackTrace();
}
}
