Document parseDocument(String filename) throws Exception {
FileReader reader = new FileReader(filename);
String firstLine = new BufferedReader(reader).readLine();
reader.close();
Document document = null;
if (firstLine.startsWith("<?xml")) {
System.err.println("XML detected; using default XML parser.");
} else {
try {
Class nekoParserClass = Class.forName("org.cyberneko.html.parsers.DOMParser");
Object parser = nekoParserClass.newInstance();
Method parse = nekoParserClass.getMethod("parse", new Class[] {String.class});
Method getDocument = nekoParserClass.getMethod("getDocument", new Class[0]);
parse.invoke(parser, filename);
document = (Document) getDocument.invoke(parser);
} catch (Exception e) {
System.err.println("NekoHTML HTML parser not found; HTML4 support disabled.");
}
}
if (document == null) {
DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
try { // http://www.w3.org/blog/systeam/2008/02/08/w3c_s_excessive_dtd_traffic
factory.setFeature("http://apache.org/xml/features/nonvalidating/load-external-dtd", false);
} catch (ParserConfigurationException e) {
System.err.println("Warning: Could not disable external DTD loading");
}
DocumentBuilder builder = factory.newDocumentBuilder();
document = builder.parse(filename);
}
return document;
}
