* The caller can parse multiple schema documents, * JAXB external binding files (or potentially WSDL * and JSR-109.next mapping files in the future). * *
* All the errors found during this process will be sent * to the registered {@link ErrorListener}. * *
* Once all the documents are parsed, call the {@link #bind()} * method to get the compiled {@link JAXBModel} object. * * *
* The caller can feed the following synthesized schema * to achive the namespace URI {@code -> } Java package customization: *
{@code
*
*
*
*
*
* }
* Feed this synthesized schema document for each namespace URI
* you need to map.
*
* @author
* Kohsuke Kawaguchi (kohsuke.kawaguchi@sun.com)
*/
public interface SchemaCompiler {
/**
* Parses schemas or external bindings
* through SAX events by feeding events into
* SAX {@link ContentHandler}.
*
* @param systemId
* The system ID of the document to be read in.
*
* @see #parseSchema(String, XMLStreamReader)
*/
ContentHandler getParserHandler( String systemId );
/**
* Parses a schema or an external binding file
* from an external source.
*
* @param source
* Its system Id must be set to an absolute URI.
*/
void parseSchema( InputSource source );
/**
* Specifies the target spec version for this compilaion.
*
* @param version
* If null, XJC will generate the source code that
* takes advantage of the latest JAXB spec that it understands.
* @since 2.1 EA2
*/
void setTargetVersion( SpecVersion version );
/**
* Parses a schema or an external binding file
* from the specified DOM element.
*
* * The given DOM element is treated as if it's the root of a * virtual document. * *
* XJC will not be able to print location information for * errors found in this document, since DOM doesn't have them. * For this reason, use of this method is strongly discouraged. * * @param systemId * We need an absolute system ID that uniquely designates the virtual * document. This should be different from the system ID of * the document which contains this element. *
* One way to do that is by adding a fragment identifier * to the system ID of the document. For example, if the document * is "foo.wsdl" and you are passing in its types section, you * can use an unique identifier like "foo.wsdl#types" */ void parseSchema( String systemId, Element element ); /** * Parses a schema or an external binding file * from the given source. * *
* A stream reader must be pointing at the element or * at the start of the document. * XML is parsed until the corresponding end tag, then the * sub tree is processed as a schema document. * *
* When this method returns successfully, the parser is at * the next token of the end element. * * @param systemId * The absolute system ID of the document that is being parsed. * This information is necessary to avoid double-inclusion * and etc. * * Note that {@link XMLStreamReader#getLocation()} only * returns the system ID of the entity it is parsing, not * necessarily the system ID of the document itself. * * @throws XMLStreamException * If an error happens while parsing a document. * Note that not only the parser but also the XJC itself * may throw this error (as a result of the additional validation * for example.) */ void parseSchema( String systemId, XMLStreamReader reader ) throws XMLStreamException; void setErrorListener( ErrorListener errorListener ); void setEntityResolver( EntityResolver entityResolver ); /** * Sets the default Java package name into which the generated code will be placed. * *
* Customizations in the binding files/schemas will have precedence over this setting. * Set to null to use the default package name computation algorithm as specified by * the JAXB spec (which is the default behavior.) * *
* Initially this parameter is set to null. * * @param packageName * Java pckage name such as "org.foo.bar". Use "" to represent the root package, * and null to defer to the default computation algorithm. * * @see #forcePackageName(String) */ void setDefaultPackageName( String packageName ); /** * Forces all the JAXB-generated classes to go into the specific package. * *
* This setting takes precedence over the {@link #setDefaultPackageName(String)} * or any of the customization found in the JAXB binding files. This method * is designed to implement the semantics of the command-line '-p' option. * *
* This somewhat ugly semantics actually have a long history now and too late * to change. * * @see #setDefaultPackageName(String) */ void forcePackageName( String packageName ); /** * Sets the {@link ClassNameAllocator} to be used for the binding operation. * *
* This mechanism would allow the caller to participate in the binding operation. * * @see ClassNameAllocator */ void setClassNameAllocator( ClassNameAllocator allocator ); /** * Clears all the schema files parsed so far. * * @since 2.1.1 */ void resetSchema(); /** * Obtains the compiled schema object model. * * Once this method is called, no other method should be * invoked on the {@link SchemaCompiler}. * * @return * null if the compilation fails. The errors should have been * delivered to the registered error handler in such a case. */ S2JJAXBModel bind(); /** * Allows the calling code to tweak more schema compilation details. * *
* The caller can use this method to obtain an {@link Options} instance, * then tweak settings on it. The updated settings will be used when the * {@link #bind()} method is invoked. * *
* The returned {@link Options} object is useful for example to specify * command-line arguments. * * @since 2.0.2 * @deprecated * This method is not really "deprecated" (in the sense of being removed * from future versions), but the JAXB team is not committed to evolve * {@link Options} class in the compatible fashion. So please don't * use this method unless you know what you're doing. */ @NotNull Options getOptions(); }