The org.xml.sax.parser system property is used to control the * processor used. It should be SAX2 parser which supports both of the * standardized extension handlers (for declaration and lexical events). * *
The first parameter to the command gives the name of the document that * will be given to that processor. If it is a file name, it is converted * to a URL first. * *
The second parameter describes a simple processing pipeline, and will * be used as input to {@link xml.pipeline.PipelineFactory} * methods which identify the processing to be done. Examples of such a * pipeline include
* * nsfix | validate to validate the input document * nsfix | write ( stdout ) to echo the file as XML text * dom | nsfix | write ( stdout ) parse into DOM, print the result ** *
Relatively complex pipelines can be described on the command line, but * not all interesting ones will require as little configuration as can be done * in that way. Put filters like "nsfix", perhaps followed by "validate", * at the front of the pipeline so they can be optimized out if a parser * supports those modes natively. * *
If the parsing is aborted for any reason, the JVM will exit with a * failure code. If a validating parse was done then both validation and * well formedness errors will cause a failure. A non-validating parse * will report failure on well formedness errors. * * @see xml.pipeline.PipelineFactory * * @author David Brownell * @version $Date: 2002/09/30 15:08:51 $ */ final public class DoParse { private DoParse () { /* no instances allowed */ } // first reported nonrecoverable error private static SAXParseException fatal; // error categories private static int errorCount; private static int fatalCount; /** * Command line invoker for this class; pass a filename or URL * as the first argument, and a pipeline description as the second. * Make sure to use filters to condition the input to stages that * require it; an nsfix filter will be a common requirement, * to restore syntax that SAX2 parsers delete by default. Some * conditioning filters may be eliminated by setting parser options. * (For example, "nsfix" can set the "namespace-prefixes" feature to * a non-default value of "true".) */ public static void main (String argv []) throws IOException { int exitStatus = 1; if (argv.length != 2) { // // XXX pipeline factory now has a pre-tokenized input // method, use it ... that way at least some params // can be written using quotes (have spaces, ...) // System.err.println ("Usage: DoParse [filename|URL] pipeline-spec"); System.err.println ("Example pipeline specs:"); System.err.println (" 'nsfix | validate'"); System.err.println ( " ... restore namespace syntax, validate"); System.err.println (" 'nsfix | xhtml ( stdout )'"); System.err.println ( " ... restore namespace syntax, write to stdout as XHTML" ); System.exit (1); } try { // // Get input source for specified document (or try ;-) // argv [0] = Resolver.getURL (argv [0]); InputSource input = new InputSource (argv [0]); // // Get the producer, using the system default parser (which // can be overridden for this particular invocation). // // And the pipeline, using commandline options. // EventProducer producer = new EventProducer (input); EventConsumer consumer; producer = new EventProducer (input); consumer = PipelineFactory.createPipeline (argv [1]); // // XXX want commandline option for tweaking error handler. // Want to be able to present warnings. // producer.setErrorHandler (new MyErrorHandler ()); // XXX need facility enabling resolving to local DTDs // // Parse. The pipeline may get optimized a bit, so we // can't always fail cleanly for validation without taking // a look at the filter stages. // producer.produce (consumer); try { if (producer.getFeature ( "http://org.xml/sax/features/validation")) exitStatus = ((errorCount + fatalCount) > 0) ? 1 : 0; else if (fatalCount == 0) exitStatus = 0; } catch (SAXException e) { if (hasValidator (consumer)) exitStatus = ((errorCount + fatalCount) > 0) ? 1 : 0; else if (fatalCount == 0) exitStatus = 0; } } catch (java.net.MalformedURLException e) { System.err.println ("** Malformed URL: " + e.getMessage ()); System.err.println ("Is '" + argv [0] + "' a non-existent file?"); e.printStackTrace (); // e.g. FNF } catch (SAXParseException e) { if (e != fatal) { System.err.print (printParseException ("Parsing Aborted", e)); e.printStackTrace (); if (e.getException () != null) { System.err.println ("++ Wrapped exception:"); e.getException ().printStackTrace (); } } } catch (SAXException e) { Exception x = e; if (e.getException () != null) x = e.getException (); x.printStackTrace (); } catch (Throwable t) { t.printStackTrace (); } System.exit (exitStatus); } // returns true if saw a validator (before end or unrecognized node) // false otherwise private static boolean hasValidator (EventConsumer e) { if (e == null) return false; if (e instanceof ValidationConsumer) return true; if (e instanceof TeeConsumer) { TeeConsumer t = (TeeConsumer) e; return hasValidator (t.getFirst ()) || hasValidator (t.getRest ()); } if (e instanceof WellFormednessFilter || e instanceof NSFilter ) return hasValidator (((EventFilter)e).getNext ()); // else ... gee, we don't know. Assume not. return false; } static class MyErrorHandler implements ErrorHandler { // dump validation errors, but continue public void error (SAXParseException e) throws SAXParseException { errorCount++; System.err.print (printParseException ("Error", e)); } public void warning (SAXParseException e) throws SAXParseException { // System.err.print (printParseException ("Warning", e)); } // try to continue fatal errors, in case a parser reports more public void fatalError (SAXParseException e) throws SAXParseException { fatalCount++; if (fatal == null) fatal = e; System.err.print (printParseException ("Nonrecoverable Error", e)); } } static private String printParseException ( String label, SAXParseException e ) { StringBuffer buf = new StringBuffer (); int temp; buf.append ("** "); buf.append (label); buf.append (": "); buf.append (e.getMessage ()); buf.append ('\n'); if (e.getSystemId () != null) { buf.append (" URI: "); buf.append (e.getSystemId ()); buf.append ('\n'); } if ((temp = e.getLineNumber ()) != -1) { buf.append (" line: "); buf.append (temp); buf.append ('\n'); } if ((temp = e.getColumnNumber ()) != -1) { buf.append (" char: "); buf.append (temp); buf.append ('\n'); } return buf.toString (); } }