|
|
|
Sponsored Link •
|
|
Advertisement
|
Next, we need to package these artifacts into a WAR file, along with one or more deployment descriptors. J2EE deployment descriptors may convey information not only to an application server, but possibly to other tools and even to people interacting with the service.
For a JAX-RPC 2.0 Web app, a deployment descriptor bundled in the
WAR should be called jaxrpc-ri.xml. This file
conveys both runtime information about a Web service, and also
communicates information to a deployment compiler tool discussed
in the final deployment step. The content of the
jaxrpc-ri.xml file may be as simple as the
following, fashioned after a sample in the JAX-RPC 2.0 reference
implementation:
<?xml version="1.0" encoding="UTF-8"?>
<webServices
xmlns="http://java.sun.com/xml/ns/jax-rpc/ri/dd"
version="1.0"
targetNamespaceBase="http://artima.com"
typeNamespaceBase="http://artima.com"
urlPatternBase="/ws">
<endpoint
name="sayhello"
displayName="A Java Web service"
interface="server.HelloImpl"
implementation="server.HelloImpl"
wsdl="/WEB-INF/HelloImplService.wsdl"/>
<endpointMapping
endpointName="sayhello"
urlPattern="/sayhello"/>
</webServices>
This descriptor defines a single endpoint, sayhello,
and maps that endpoint to the URL pattern /sayhello.
Note that the service interface and implementation refer to the
same class.
To make this WAR file work inside a servlet container, you would
also want to package a web.xml file in the WAR, too:
<!DOCTYPE web-app
PUBLIC "-//Sun Microsystems, Inc.//DTD Web Application 2.3//EN"
"http://java.sun.com/j2ee/dtds/web-app_2_3.dtd">
<web-app>
<display-name>Hello, JAX-RPC 2.0!</display-name>
<description>A hello, world, Web service.</description>
<session-config>
<session-timeout>60</session-timeout>
</session-config>
</web-app>
Creating a WAR file is easily accomplished with the Ant
war task. Note that in the resulting Web archive,
the schema file is located in the Web app's root directory so it
can be downloaded by clients that require access to the WSDL. The
WSDL itself will be served up by the JAX-RPC servlet, which will
attach the service endpoint address to the WSDL at runtime. The
following is the structure of the resulting WAR:
schema1.xsd WEB-INF/web.xml WEB-INF/jaxrpc-ri.xml WEB-INF/HelloImplService.wsdl WEB-INF/classes/server/HelloImpl.class WEB-INF/classes/server/jaxrpc/SayHello.class WEB-INF/classes/server/jaxrpc/SayHello.java WEB-INF/classes/server/jaxrpc/SayHelloResponse.class WEB-INF/classes/server/jaxrpc/SayHelloResponse.java
The final step is to deploy the Web service WAR archive. That requires "cooking" the initial WAR: modifying the deployment descriptor, and adding several runtime artifacts to facilitate the XML-Java data binding.
The JAX-RPC 2.0 reference implementation includes the
wsdeploy tool that operates an a single "raw" WAR
file, performs the deployment tasks on elements in that WAR, and
produces a new, deployment-ready, or "cooked," WAR file.
wsdeploy reads the jaxrpc-ri.xml inside
the WAR, and then invokes another JAX-RPC tool,
wscompile with an option to generate deployment-time
Web service implementation artifacts. Running
wsdeploy is as simple as invoking the
wsdeploy Ant task, which is also part of the JAX-RPC
2.0 reference implementation. I will not discuss the individual
files generated for the deployment-ready WAR, partly because they
are dependent on the current, early-access release of the
reference implementation.
The WAR "cooked" by wsdeploy can be copied to the
webapps directory of a servlet container. The Web
service is now ready to be used.
In order for a servlet container to serve JAX-RPC 2.0-based Web
services, all JAR files from the reference implementation must be
copied to a library directory the servlet engine looks in for
classes. With Jakarta Tomcat, for instance, you can copy the
JAX-RPC reference implementation JARs to the
/shared/lib directory.
The JAX-RPC servlet provides a summary page for the newly installed Web service. For this example app, that summary page can be accessed at the following URL:
http://localhost:8080/jaxrpc-hello/hello
(substitute the
host name, the port number, and the context root with appropriate
values for your server). The following Web page is displayed at
that URL:
The Web service's WSDL, in turn, is accessible via this URL, which is where clients would download that WSDL. Note that clients would also need the XML schema file imported in the WSDL, which is why that schema file had to reside in the Web archive's root:
http://localhost:8080/jaxrpc-hello/hello?WSDL
If you inspect the resulting WSDL with a browser, you will note that the WSDL correctly indicates the service's endpoint address:
<service name="HelloImplService">
<port name="HelloImplPort"
binding="tns:HelloImplBinding">
<soap:address
location="http://localhost:8080/jaxrpc-hello/hello"/>
</port>
</service>
This brief tutorial demonstrated that JAX-RPC 2.0's use of
annotations greatly simplifies writing a Web service. Typing in
the example code, and running the Ant tasks defined in JAX-RPC
2.0, should yield a fully working Web service in less than
three minutes. All the while, you can concern yourself with the
Java code that implements the business logic, and let JAX-RPC 2.0
tools handle the rest of the Web-service-related classes and
files. Download the JAX-RPC 2.0 reference implementation for an
example of how to write Web service client for a similar Web
service.
Have an opinion about JAX-RPC 2.0's programming model?
Discuss this article in the Articles Forum topic,
Three Minutes to a Web Service.
Resources
The JAX-RPC 2.0 reference implementation and community Web site
https://jaxrpc.dev.java.net/
JSR 175: A Metadata Facility for the Java Programming Language
http://www.jcp.org/en/jsr/detail?id=175
JSR 181: Web Services Metadata for the Java Platform
http://www.jcp.org/en/jsr/detail?id=181
JSR 222: Java Architecture for XML Binding (JAXB) 2.0
http://www.jcp.org/en/jsr/detail?id=222
JSR 224: Java API for XML-Based RPC (JAX-RPC) 2.0
http://www.jcp.org/en/jsr/detail?id=224
Annotation Processing Tool (apt)
http://java.sun.com/j2se/1.5.0/docs/guide/apt/index.html
Frank Sommers is a Senior Editor with Artima Developer. He also serves as chief editor of the Web zine ClusterComputing.org, the IEEE Technical Committee on Scalable Computing's newsletter, and is an elected member of the Jini Community's Technical Advisory Committee. Prior to joining Artima, Frank wrote the Jiniology and Web services columns for JavaWorld.
|
Sponsored Links
|