by Thomas Erl
Although originally conceived as a technology to bridge the gap between disparate RPC-based communication platforms, SOAP has evolved into the most widely supported messaging format and protocol for use with XML Web services. Hence the SOAP acronym is frequently referred to as the Service-Oriented Architecture (or Application) Protocol, instead of the Simple Object Access Protocol.
The SOAP specification establishes a standard message format that consists of an XML document capable of hosting RPC and document-centric data. This facilitates synchronous (request and response) as well as asynchronous (process-driven) data exchange models. With WSDL establishing a standard endpoint description format for applications, the document-centric message format is much more common.
SOAP institutes a method of communication based on a processing model that is in relative alignment with the overall Web services framework. It differs only in that it introduces terms and concepts that relate specifically to the manner in which SOAP messages need to be handled (within a technical implementation of this framework).
A SOAP node represents the processing logic responsible for transmitting, receiving, and performing a variety of processing tasks on SOAP messages. An implementation of a SOAP node is typically platform specific, and is commonly labeled as a SOAP server or a SOAP listener. Specialized variations also exist, such as SOAP routers. Conceptually, all are considered SOAP nodes. The SOAP node establishes as the underlying transport mechanism for a Web service.
SOAP nodes are best viewed as the technical infrastructure that powers all communication between Web services. When discussing the involvement of SOAP nodes, however, a slight departure from the terms and concepts established by the Web services framework needs to be incorporated.
Like Web services, SOAP nodes can exist as initial senders, intermediaries, and ultimate receivers. Whereas Web services are also classified as requestors and providers, SOAP nodes performing the equivalent tasks (sending, receiving) are referred to as SOAP senders and SOAP receivers. The SOAP specification, however, does not classify these as roles. For the purpose of this book, we'll refer to them as "types" of SOAP nodes.
A node type of SOAP initial sender is also a SOAP sender, and the SOAP ultimate receiver is also a SOAP receiver.
A SOAP node can act as an intermediary, transitioning through both SOAP sender and receiver types during the processing of a SOAP message.
(See diagrams below.)
As with Web service roles, the same SOAP node can act as different types, depending on its position within the message path and the state of the current business activity. For instance, a SOAP node transmitting a message as the initial sender can later receive a response as the ultimate receiver.
The container of SOAP message information is referred to as a SOAP envelope. Let's open it, and take a brief look at the underlying structure of a typical SOAP message.
The root Envelope element that frames the message document consists of a mandatory body section and an optional header area.
The SOAP header is expressed using the Header construct, which can contain one or more sections or blocks.
Common uses of header blocks include:
- implementation of (predefined or application-specific) SOAP extensions, such as those introduced by second-generation specifications
- identification of target SOAP intermediaries
- providing supplementary meta information about the SOAP message
While a SOAP message progresses along a message path, intermediaries may add, remove, or process information in SOAP header blocks. Although an optional part of a SOAP message, the use of the header section to carry header blocks is commonplace when working with second-generation Web services specifications.
The one part of a SOAP message that is not optional is the body. As represented by the Body construct, this section acts as a container for the data being delivered by the SOAP message. Data within the SOAP body is often referred to as the payload or payload data.
A Field Guide to Integrating XML
and Web services
The Body construct can also be used to host exception information within nested Fault elements. Although fault sections can reside alongside standard data payloads, this type of information is often sent separately in response messages that communicate error conditions.
The Fault construct consists of a series of system elements used to identify characteristics of the exception.
versions do not match
Now that you've had a look at the internal structure and syntax of a SOAP message, let's finish by briefly introducing SOAP node roles. When discussing SOAP nodes, roles relate to an optional env:role attribute that a SOAP message can use to identify header blocks intended for specific types of SOAP receivers. Therefore, SOAP roles are associated only to types of SOAP nodes that perform a receiving function. In other words, intermediaries and ultimate receivers.
The two most common env:role attribute values are next and ultimateReceiver. An intermediary node will process only header blocks identified with the next role, whereas a node acting as the ultimate receiver will process both.
Note: The env:role attribute was introduced in version 1.2 of the SOAP specification. It was previously named env:actor.
- The WS-Coordination Context Management Framework
- Cross-Service Transactions with WS-AtomicTransaction
- Long-Running Transactions with WS-BusinessActivity
- An Overview of the WS-Security Framework
- Service-Oriented Business Processes with BPEL
- Building a Communications Framework with WS-Policy and WS-ReliableMessaging
- Defining the Web Service with WSDL
- An Inside Look into SOAP Messaging
- UDDI In and Out of the Enterprise
- A W3C Web Services Glossary
- Understanding WS-Policy Part I: Operator Composition Rules and Attachments
- Understanding WS-Policy Part II: Operator Composition Rules and Attachments
- Web Service Contract Versioning Fundamentals Part I: Version Identifiers and Versioning Strategies
- Web Service Contract Versioning Fundamentals Part II: Version Identifiers andVersioning Strategies
Foreword by Grady Booch
With contributions from David Chappell, Jason Hogg, Anish Karmarkar, Mark Little, David Orchard, Satadru Roy, Thomas Rischbeck, Arnaud Simon, Clemens Utschig, Dennis Wisnosky, and others.
Governing Shared Services On-Premise & in the Cloud by Stephen Bennett, Thomas Erl, Clive Gee, Anne Thomas Manes, Robert Schneider, Leo Shuster, Andre Tost, Chris Venable
For more information about these books, visit: www.servicetechbooks.com