@Override
protected String getDocumentText(String absolutePath, String encoding) throws IOException {
String text;
try {
SAXParserFactory spf = SAXParserFactory.newInstance();
spf.setValidating(false);
spf.setNamespaceAware(true);
spf.setFeature("http://xml.org/sax/features/namespaces", true);
spf.setFeature("http://xml.org/sax/features/validation", false);
spf.setFeature("http://apache.org/xml/features/nonvalidating/load-dtd-grammar", false);
spf.setFeature("http://apache.org/xml/features/nonvalidating/load-external-dtd", false);
SAXParser saxParser = spf.newSAXParser();
XMLReader xmlReader = saxParser.getXMLReader();
TeiToTxtSaxHandler handler = new TeiToTxtSaxHandler();
xmlReader.setContentHandler(handler);
xmlReader.parse(new InputSource(absolutePath));
text = handler.getText();
} catch (ParserConfigurationException | SAXException e) {
logger.error("Could not read TEI document: {}", absolutePath);
logger.warn("Ignoring TEI document " + absolutePath);
text = "";
this.failedFileCounter++;
}
return text;
}
/**
* {@link SAXParserFactoryUtil#setXIncludeAware}のテストです。
*
* @throws Exception
*/
public void testSetXIncludeAware() throws Exception {
SAXParserFactory spf = SAXParserFactoryUtil.newInstance();
SAXParserFactoryUtil.setXIncludeAware(spf, true);
spf.setNamespaceAware(true);
SAXParser parser = SAXParserFactoryUtil.newSAXParser(spf);
InputSource is =
new InputSource(ResourceUtil.getResourceAsStream("org/seasar/framework/util/include.xml"));
is.setSystemId("include.xml");
parser.parse(
is,
new DefaultHandler() {
@Override
public void startElement(
String uri, String localName, String qName, Attributes attributes)
throws SAXException {
if ("bar".equals(qName)) {
included = true;
}
}
@Override
public InputSource resolveEntity(String publicId, String systemId)
throws IOException, SAXException {
InputSource is =
new InputSource(
ResourceUtil.getResourceAsStream("org/seasar/framework/util/included.xml"));
is.setSystemId("included.xml");
return is;
}
});
assertTrue(included);
}
/**
* Create a SAX parser from the JAXP factory. The result can be used to parse XML files.
*
* <p>See class Javadoc for hints on setting an entity resolver. This parser has its entity
* resolver set to the system entity resolver chain.
*
* @param validate if true, a validating parser is returned
* @param namespaceAware if true, a namespace aware parser is returned
* @throws FactoryConfigurationError Application developers should never need to directly catch
* errors of this type.
* @throws SAXException if a parser fulfilling given parameters can not be created
* @return XMLReader configured according to passed parameters
*/
public static XMLReader createXMLReader(boolean validate, boolean namespaceAware)
throws SAXException {
SAXParserFactory factory;
if (!validate && useFastSAXParserFactory) {
try {
factory = createFastSAXParserFactory();
} catch (ParserConfigurationException ex) {
factory = SAXParserFactory.newInstance();
} catch (SAXException ex) {
factory = SAXParserFactory.newInstance();
}
} else {
useFastSAXParserFactory = false;
factory = SAXParserFactory.newInstance();
}
factory.setValidating(validate);
factory.setNamespaceAware(namespaceAware);
try {
return factory.newSAXParser().getXMLReader();
} catch (ParserConfigurationException ex) {
throw new SAXException(
"Cannot create parser satisfying configuration parameters", ex); // NOI18N
}
}
/** Private constructor */
private XMLReaders(int validate) {
SAXParserFactory fac = SAXParserFactory.newInstance();
boolean val = false;
// All JDOM parsers are namespace aware.
fac.setNamespaceAware(true);
switch (validate) {
case 0:
fac.setValidating(false);
break;
case 1:
fac.setValidating(true);
val = true;
break;
case 2:
fac.setValidating(false);
try {
SchemaFactory sfac = SchemaFactory.newInstance(XMLConstants.W3C_XML_SCHEMA_NS_URI);
Schema schema = sfac.newSchema();
fac.setSchema(schema);
val = true;
} catch (SAXException se) {
// we could not get a validating system, set the fac to null
fac = null;
}
break;
}
jaxpfactory = fac;
validates = val;
}
public GpxReader()
throws javax.xml.parsers.ParserConfigurationException, org.xml.sax.SAXException {
javax.xml.parsers.SAXParserFactory factory = javax.xml.parsers.SAXParserFactory.newInstance();
factory.setNamespaceAware(true);
this.parser = factory.newSAXParser();
}
public static SAXParser getSaxParser() throws ParserConfigurationException, SAXException {
SAXParserFactory parserFactory = SAXParserFactory.newInstance();
parserFactory.setNamespaceAware(false);
parserFactory.setValidating(true);
parserFactory.setXIncludeAware(false);
return parserFactory.newSAXParser();
}
protected void validateXmlString(final String xml) throws Exception {
if (getSchemaFile() == null) {
LOG.warn("skipping validation, schema file not set");
return;
}
final SchemaFactory schemaFactory =
SchemaFactory.newInstance("http://www.w3.org/2001/XMLSchema");
final File schemaFile = new File(getSchemaFile());
LOG.debug("Validating using schema file: {}", schemaFile);
final Schema schema = schemaFactory.newSchema(schemaFile);
final SAXParserFactory saxParserFactory = SAXParserFactory.newInstance();
saxParserFactory.setValidating(true);
saxParserFactory.setNamespaceAware(true);
saxParserFactory.setSchema(schema);
assertTrue("make sure our SAX implementation can validate", saxParserFactory.isValidating());
final Validator validator = schema.newValidator();
final ByteArrayInputStream inputStream = new ByteArrayInputStream(xml.getBytes());
final Source source = new StreamSource(inputStream);
validator.validate(source);
}
public static BehavioralPatternsCatalog load(final Reader reader) {
BehavioralPatternsCatalog catalog = null;
try {
final BehavioralPatternsCatalogSaxHandler handler = new BehavioralPatternsCatalogSaxHandler();
final SAXParserFactory factory = SAXParserFactory.newInstance();
factory.setValidating(true);
factory.setNamespaceAware(true);
final XMLReader xmlReader = factory.newSAXParser().getXMLReader();
xmlReader.setContentHandler(handler);
xmlReader.setErrorHandler(handler);
xmlReader.setEntityResolver(handler);
xmlReader.parse(new InputSource(reader));
catalog = handler.getCatalog();
reader.close();
} catch (final SAXException e) {
BehavioralAnalysisPlugin.logError("Error parsing Behavioral Patterns Catalog.", e);
} catch (final ParserConfigurationException e) {
BehavioralAnalysisPlugin.logError("Error parsing Behavioral Patterns Catalog.", e);
} catch (final IOException e) {
BehavioralAnalysisPlugin.logError("Error opening Behavioral Patterns Catalog.", e);
}
return catalog;
}
public static ParentalControlConcern parse(File file) {
try {
log.info("Parsing '" + file.getCanonicalPath() + "'...");
SAXParserFactory factory = SAXParserFactory.newInstance();
factory.setNamespaceAware(true);
factory.setValidating(true);
SAXParser parser = factory.newSAXParser();
parser.setProperty(JAXP_SCHEMA_LANGUAGE, W3C_XML_SCHEMA);
ParentalControlConcernHandler handler = new ParentalControlConcernHandler();
parser.parse(file, handler);
log.info("Parsed '" + file.getCanonicalPath() + "'");
return handler.result;
} catch (SAXParseException ex) {
log.error(
"Could not validate the parental control concern XML file! Validation error follows.");
log.error(ex.getMessage());
ex.printStackTrace();
return null;
} catch (Exception ex) {
log.error(ex);
return null;
}
}
/**
* Return a <code>SAXParserFactory</code> instance that is non-validating and is namespace
* aware.
*
* @return configured <code>SAXParserFactory</code>
*/
private SAXParserFactory getConfiguredFactory() {
SAXParserFactory factory = SAXParserFactory.newInstance();
factory.setValidating(false);
factory.setNamespaceAware(true);
return factory;
} // END getConfiguredFactory
/**
* Test XMLSequence reading using XML Parsers.
*
* @throws Exception if something goes wrong
*/
public void testXMLSequence() throws Exception {
File testFile = new File(testDir, "XMLSequence.txt");
XMLSequence sequence = new XMLSequence(new BufferedInputStream(new FileInputStream(testFile)));
if (!sequence.markSupported())
throw new AssertionFailedError("Mark is not supported for XMLSequence");
sequence.mark((int) testFile.length() + 1);
SAXParserFactory factory = SAXParserFactory.newInstance();
factory.setNamespaceAware(true);
XMLReader xmlReader = factory.newSAXParser().getXMLReader();
int i = 0;
while (sequence.hasNext()) {
InputStream input = sequence.next();
InputSource source = new InputSource(input);
xmlReader.parse(source);
input.close();
i++;
}
sequence.reset();
i = 0;
while (sequence.hasNext()) {
InputStream input = sequence.next();
InputSource source = new InputSource(input);
xmlReader.parse(source);
input.close();
i++;
}
sequence.close();
}
private XMLReader getXMLReader(ContentHandler contentHandler, ErrorHandler errorHandler)
throws ParserConfigurationException, SAXException {
// setup sax factory ; be sure just one instance!
SAXParserFactory saxFactory = SAXParserFactory.newInstance();
// Enable validation stuff
saxFactory.setValidating(true);
saxFactory.setNamespaceAware(true);
// Create xml reader
SAXParser saxParser = saxFactory.newSAXParser();
XMLReader xmlReader = saxParser.getXMLReader();
// Setup xmlreader
xmlReader.setProperty(
XMLReaderObjectFactory.APACHE_PROPERTIES_INTERNAL_GRAMMARPOOL, grammarPool);
xmlReader.setFeature(Namespaces.SAX_VALIDATION, true);
xmlReader.setFeature(Namespaces.SAX_VALIDATION_DYNAMIC, false);
xmlReader.setFeature(XMLReaderObjectFactory.APACHE_FEATURES_VALIDATION_SCHEMA, true);
xmlReader.setFeature(XMLReaderObjectFactory.APACHE_PROPERTIES_LOAD_EXT_DTD, true);
xmlReader.setFeature(Namespaces.SAX_NAMESPACES_PREFIXES, true);
xmlReader.setContentHandler(contentHandler);
xmlReader.setErrorHandler(errorHandler);
return xmlReader;
}
@Override
public XMLLoader init(SolrParams args) {
// Init StAX parser:
inputFactory = XMLInputFactory.newInstance();
EmptyEntityResolver.configureXMLInputFactory(inputFactory);
inputFactory.setXMLReporter(xmllog);
try {
// The java 1.6 bundled stax parser (sjsxp) does not currently have a thread-safe
// XMLInputFactory, as that implementation tries to cache and reuse the
// XMLStreamReader. Setting the parser-specific "reuse-instance" property to false
// prevents this.
// All other known open-source stax parsers (and the bea ref impl)
// have thread-safe factories.
inputFactory.setProperty("reuse-instance", Boolean.FALSE);
} catch (IllegalArgumentException ex) {
// Other implementations will likely throw this exception since "reuse-instance"
// isimplementation specific.
log.debug("Unable to set the 'reuse-instance' property for the input chain: " + inputFactory);
}
// Init SAX parser (for XSL):
saxFactory = SAXParserFactory.newInstance();
saxFactory.setNamespaceAware(true); // XSL needs this!
EmptyEntityResolver.configureSAXParserFactory(saxFactory);
xsltCacheLifetimeSeconds = XSLT_CACHE_DEFAULT;
if (args != null) {
xsltCacheLifetimeSeconds = args.getInt(XSLT_CACHE_PARAM, XSLT_CACHE_DEFAULT);
log.info("xsltCacheLifetimeSeconds=" + xsltCacheLifetimeSeconds);
}
return this;
}
/**
* Contact the remote WebDAV server and do a PROPFIND lookup, returning a {@link List} of all
* scavenged resources.
*/
private List propfind(Location location) throws IOException {
/* Create the new HttpClient instance associated with the location */
final HttpClient client = new HttpClient(location);
client.addRequestHeader("Depth", "1");
client.setAcceptableStatus(207).connect("PROPFIND", true);
/* Get the XML SAX Parser and parse the output of the PROPFIND */
try {
final SAXParserFactory factory = SAXParserFactory.newInstance();
factory.setValidating(false);
factory.setNamespaceAware(true);
final SAXParser parser = factory.newSAXParser();
final String systemId = location.toString();
final InputSource source = new InputSource(systemId);
final Handler handler = new Handler(location);
source.setByteStream(client.getResponseStream());
parser.parse(source, handler);
return handler.list;
} catch (ParserConfigurationException exception) {
Exception throwable = new IOException("Error creating XML parser");
throw (IOException) throwable.initCause(exception);
} catch (SAXException exception) {
Exception throwable = new IOException("Error creating XML parser");
throw (IOException) throwable.initCause(exception);
} finally {
client.disconnect();
}
}
/**
* @see MessageBodyReader#readFrom(Class, Type, MediaType, Annotation[], MultivaluedMap,
* InputStream)
*/
@Override
public Object readFrom(
Class<Object> type,
Type genericType,
Annotation[] annotations,
MediaType mediaType,
MultivaluedMap<String, String> httpHeaders,
InputStream entityStream)
throws IOException {
try {
SAXParserFactory spf = SAXParserFactory.newInstance();
spf.setXIncludeAware(isXIncludeAware());
spf.setNamespaceAware(true);
spf.setValidating(isValidatingDtd());
spf.setFeature(XMLConstants.FEATURE_SECURE_PROCESSING, isSecureProcessing());
spf.setFeature(
"http://xml.org/sax/features/external-general-entities", isExpandingEntityRefs());
spf.setFeature(
"http://xml.org/sax/features/external-parameter-entities", isExpandingEntityRefs());
XMLReader reader = spf.newSAXParser().getXMLReader();
JAXBContext jaxbContext = getJaxbContext(type);
Unmarshaller um = jaxbContext.createUnmarshaller();
return um.unmarshal(new SAXSource(reader, new InputSource(entityStream)));
} catch (Exception e) {
throw new IOException("Could not unmarshal to " + type.getName());
}
}
/**
* 通过XSD（XML Schema）校验XML
*
* @throws FileNotFoundException
*/
public boolean validateXMLByXSD(String testXSD, String xmlFileName) throws FileNotFoundException {
// String xmlFileName = "Q:\\_dev_stu\\xsdtest\\src\\note.xml";
String xsdFileName = "E:/datashare/" + testXSD;
FileInputStream fls = new FileInputStream(new File(xmlFileName));
try {
// 创建默认的XML错误处理器
XMLErrorHandler errorHandler = new XMLErrorHandler();
// 获取基于 SAX 的解析器的实例
SAXParserFactory factory = SAXParserFactory.newInstance();
// 解析器在解析时验证 XML 内容。
factory.setValidating(true);
// 指定由此代码生成的解析器将提供对 XML 名称空间的支持。
factory.setNamespaceAware(true);
// 使用当前配置的工厂参数创建 SAXParser 的一个新实例。
SAXParser parser = factory.newSAXParser();
// 创建一个读取工具
SAXReader xmlReader = new SAXReader();
// 获取要校验xml文档实例
Document xmlDocument = (Document) xmlReader.read(fls);
// 设置 XMLReader 的基础实现中的特定属性。核心功能和属性列表可以在
// [url]http://sax.sourceforge.net/?selected=get-set[/url] 中找到。
parser.setProperty(
"http://java.sun.com/xml/jaxp/properties/schemaLanguage",
"http://www.w3.org/2001/XMLSchema");
parser.setProperty(
"http://java.sun.com/xml/jaxp/properties/schemaSource", "file:" + xsdFileName);
// 创建一个SAXValidator校验工具，并设置校验工具的属性
SAXValidator validator = new SAXValidator(parser.getXMLReader());
// 设置校验工具的错误处理器，当发生错误时，可以从处理器对象中得到错误信息。
validator.setErrorHandler(errorHandler);
// 校验
validator.validate(xmlDocument);
XMLWriter writer = new XMLWriter(OutputFormat.createPrettyPrint());
// 如果错误信息不为空，说明校验失败，打印错误信息
if (errorHandler.getErrors().hasContent()) {
System.out.println("XML文件校验失败！");
writer.write(errorHandler.getErrors());
return false;
} else {
System.out.println("XML文件校验成功！");
return true;
}
} catch (Exception ex) {
System.out.println(
"XML文件: " + xmlFileName + " 通过XSD文件:" + xsdFileName + "检验失败。\n原因： " + ex.getMessage());
// ex.printStackTrace();
return false;
} finally {
try {
fls.close();
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
}
private SAXParser getParser() {
SAXParserFactory saxParserFactory = SAXParserFactory.newInstance();
saxParserFactory.setNamespaceAware(true);
try {
return saxParserFactory.newSAXParser();
} catch (Exception e) {
return null;
}
}
/**
* Creates a new instance of the validator.
*
* @param grammarCache system grammar cache for retrieving grammars used for validation
* @throws ParserConfigurationException if the JAXP parser factory is misconfigured.
* @throws SAXException if the JAXP parser factory is misconfigured.
*/
public XMLValidator(XMLGrammarCache grammarCache)
throws ParserConfigurationException, SAXException {
this.grammarCache = grammarCache;
SAXParserFactory parserFactory = SAXParserFactory.newInstance();
parserFactory.setNamespaceAware(true);
parserFactory.setValidating(false);
saxParser = parserFactory.newSAXParser();
}
/**
* Main import method. Pass in an input stream (perhaps from an HTTP POST) and get a list of
* PVConfigs.
*
* @param configFileName
* @return
* @throws Exception
*/
public static LinkedList<PVConfig> importEngineConfig(InputStream is) throws Exception {
SAXParserFactory sfac = SAXParserFactory.newInstance();
sfac.setNamespaceAware(false);
sfac.setValidating(false);
SAXParser parser = sfac.newSAXParser();
EngineConfigParser econfig = new EngineConfigParser();
parser.parse(is, econfig);
return econfig.pvConfigs;
}
/**
* Creates a non-validating, non-namespace-aware {@link XMLReader} using the specified {@link
* ContentHandler}.
*
* <p>If the given {@link ContentHandler} is <code>null</code>, the {@link XMLReader} is not
* initialised.
*
* @param handler The content handler to use.
* @return The requested {@link XMLReader}
* @throws SAXException Should a SAX exception occur
* @throws ParserConfigurationException Should a parser config exception occur
*/
public static XMLReader makeXMLReader(ContentHandler handler)
throws SAXException, ParserConfigurationException {
SAXParserFactory factory = SAXParserFactory.newInstance();
factory.setNamespaceAware(false);
factory.setValidating(false);
XMLReader reader = factory.newSAXParser().getXMLReader();
if (handler != null) reader.setContentHandler(handler);
return reader;
}
private static void parse(final InputSource input, final JDOParser recognizer)
throws SAXException, ParserConfigurationException, IOException {
SAXParserFactory factory = SAXParserFactory.newInstance();
factory.setValidating(true);
factory.setNamespaceAware(true);
XMLReader parser = factory.newSAXParser().getXMLReader();
parser.setEntityResolver(new JDOEntityResolver());
parser.setContentHandler(recognizer);
parser.setErrorHandler(recognizer.getDefaultErrorHandler());
parser.parse(input);
}
public AssemblingParser(EntityResolver er, InputSource in) throws ConfigException {
this.er = er;
this.main = in;
factory = SAXParserFactory.newInstance();
factory.setNamespaceAware(true);
TransformerFactory tf = TransformerFactory.newInstance();
if (!tf.getFeature(SAXSource.FEATURE) || !tf.getFeature(SAXResult.FEATURE))
throw new ConfigException("XSLT not supported");
transfactory = (SAXTransformerFactory) tf;
}
protected static SAXParser createParser() throws Exception {
SAXParserFactory spf = SAXParserFactory.newInstance();
spf.setNamespaceAware(true);
spf.setValidating(true);
SAXParser parser = spf.newSAXParser();
parser.setProperty(
"http://java.sun.com/xml/jaxp/properties/schemaLanguage",
"http://www.w3.org/2001/XMLSchema");
return parser;
}
static {
// SAXParserFactory factory = SAXParserFactory.newInstance();
factory = SAXParserFactory.newInstance();
factory.setNamespaceAware(true); // We add it to pick up the element of the msd namespace.
factory.setValidating(false);
invalidCharset.add("EUC_JP");
invalidCharset.add("EUCJP");
invalidCharset.add("EUC_KR");
invalidCharset.add("EUCKR");
}
/** Reads a reply from the socket and makes sure its shape is in order. */
private void receiveAndVerify(DatagramSocket s)
throws IOException, SAXException, ParserConfigurationException {
DatagramPacket p = new DatagramPacket(new byte[1024], 1024);
s.receive(p);
String xml = new String(p.getData(), 0, p.getLength(), "UTF-8");
System.out.println(xml);
// make sure at least this XML parses
SAXParserFactory spf = SAXParserFactory.newInstance();
spf.setNamespaceAware(true);
spf.newSAXParser().parse(new InputSource(new StringReader(xml)), new DefaultHandler());
}
/**
* Return the registered SAX parser factory or null if one does not exist. * @return
*
* @return returns a SAXParserFactory
*/
public SAXParserFactory getFactory() {
if (saxParserFactory == null) {
saxParserFactory = new SAXParserFactoryImpl();
}
SAXParserFactory theFactory = saxParserFactory;
if (theFactory != null) {
theFactory.setNamespaceAware(true);
theFactory.setXIncludeAware(false);
theFactory.setValidating(false);
}
return theFactory;
}
private SAXParser buildSaxParser() throws JoranException {
try {
SAXParserFactory spf = SAXParserFactory.newInstance();
spf.setValidating(false);
spf.setNamespaceAware(true);
return spf.newSAXParser();
} catch (Exception pce) {
String errMsg = "Parser configuration error occurred";
addError(errMsg, pce);
throw new JoranException(errMsg, pce);
}
}
private void init() {
boolean success = false;
// If JAXP is prefered (comes with Sun JVM 1.4.0 and higher)
if (!success) {
try {
javax.xml.parsers.SAXParserFactory spf = javax.xml.parsers.SAXParserFactory.newInstance();
spf.setNamespaceAware(true);
javax.xml.parsers.SAXParser saxParser = spf.newSAXParser();
parser = saxParser.getXMLReader();
logger.info("Using JAXP/SAX XML parser.");
success = true;
} catch (ParserConfigurationException | SAXException e) {
logger.warn("Could not instantiate JAXP/SAX XML reader: ", e.getMessage());
logger.debug(e);
}
}
// Aelfred is first alternative.
if (!success) {
try {
parser =
(XMLReader)
this.getClass()
.getClassLoader()
.loadClass("gnu.xml.aelfred2.XmlReader")
.newInstance();
logger.info("Using Aelfred2 XML parser.");
success = true;
} catch (ClassNotFoundException | InstantiationException | IllegalAccessException e) {
logger.warn("Could not instantiate Aelfred2 XML reader!");
logger.debug(e);
}
}
// Xerces is second alternative
if (!success) {
try {
parser =
(XMLReader)
this.getClass()
.getClassLoader()
.loadClass("org.apache.xerces.parsers.SAXParser")
.newInstance();
logger.info("Using Xerces XML parser.");
success = true;
} catch (ClassNotFoundException | InstantiationException | IllegalAccessException e) {
logger.warn("Could not instantiate Xerces XML reader!");
logger.debug(e);
}
}
if (!success) {
logger.error("Could not instantiate any XML parser!");
}
}
/** Loads a Configuration from an XMLFile... */
public void loadConfig(String XMLFile) {
try {
sectionHandling = new HashMap<String, EnumMap<SIT, EnumMap<CIT, Boolean>>>();
SAXParserFactory factory = SAXParserFactory.newInstance();
factory.setNamespaceAware(true);
SAXParser sp = factory.newSAXParser();
DefaultHandler handler = new ConfigLoader(this);
sp.parse(XMLFile, handler);
} catch (Exception e) {
System.err.println(e);
loadConfig();
}
}
/** Attaches the reader to the catalog. */
private void attachReaderToCatalog(Catalog catalog) {
SAXParserFactory spf = new SAXParserFactoryImpl();
spf.setNamespaceAware(true);
spf.setValidating(false);
SAXCatalogReader saxReader = new SAXCatalogReader(spf);
saxReader.setCatalogParser(
OASISXMLCatalogReader.namespaceName,
"catalog",
"com.sun.org.apache.xml.internal.resolver.readers.OASISXMLCatalogReader");
catalog.addReader("application/xml", saxReader);
}
