XML/Value RFP (Request For Proposal). Object Management Group. OMG Document: orbos/99-08-20. 24 pages. Framingham Corporate Center, 492 Old Connecticut Path, Framingham, MA 01701-4568. [Telephone: +1-508-820 4300, Facsimile: +1-508-820 4303.] "Objective of this RFP: This RFP asks for a standard way to represent XML values (documents) using OMG IDL non-object (value) types. The RFP solicits proposals for the following: A standard way to represent XML documents using OMG IDL data types -- primitive data types, constructed data types (structs, sequence, unions), and value types." Problem Statement: XML is emerging as an important technology for the web and the internet. XML can be viewed as essentially a meta-syntax that is used to define the syntax of arbitrary markup languages. An XML Document Type Definition (DTD) is a description of a particular language. (Basically the 'sentences' in the language defined by a DTD constitute the set of XML documents that satisfy the DTD.) A particular DTD combined with the XML rules for interpreting it, define an abstract syntax tree for a string (document) that is a member of the language (type) defined by the DTD. Hence XML can be viewed as a way to describe and encode complex data structures. It can encode more complex data structures than can 'traditional' IDL (structs, unions, sequences, etc.), as well as arbitrary repetitions and optional elements and can have conditional dependencies. Although it was originally designed to represent document structures, many people are using its ability to encode complex data as a way to encode complex 'business' data such as EDI data, employee payroll information, molecular structures, etc. The 'data definition' part of IDL can also be used to encode complex data structures. IDL provides definition facilities for object types (interfaces) which have identity and reference semantics, and value types which do not. 'Traditional' IDL, pre CORBA 2.3, was limited in expressive power to primitive scalar data types, structs, sequences, unions, etc. It could not easily express trees, graphs, recursive structures, etc. However with the adoption of the new IDL valuetype, CORBA 2.3 IDL is much more powerful. It can encode arbitrary graphs and has essentially the same expressive power of XML. Please note that the use of the word value and value types in this RFP denotes the non-interface IDL types. It does not just refer to IDL valuetypes, but also includes other kinds of values as described above. This situation provides us with the opportunity to define a standard way to encode and carry XML data structures as IDL values--primitive types, constructed types (structs, unions, sequences) and valuetypes. It would then be possible to efficiently pass and process XML documents as parameters to IDL operations. Currently one can only pass an XML document as an unstructured string. A set of IDL value definitions (types) would be defined that correspond to an XML DTD. An instance of the set would represent an XML document and could be used as a parameter type for an IDL operation. There are currently many ongoing initiatives which are attempting to define standard XML DTDs for a variety of specialized uses, e.g., EDI, healthcare records, payroll information, etc. Having a standard way to leverage and use their work within the CORBA world would be a great benefit. It would also greatly shorten the 'adoption cycle' within an OMG context, since currently each such initiative would also require a corresponding RFP to define its representation in IDL. . ." [excerpted]
[August 31, 2000] As part of the standardization effort for OMG IDL representation of XML documents, several companies have released a joint submission specifying XMLDOM: A DOM/Value Mapping. This revised submission provides a specification for DOM-based XML/Value Mapping [XMLDOM], in response to the ORBOS RFP - XML/Value Mapping. The relevant OMG RFP requests a "standard way to represent XML documents using OMG IDL data types -- primitive data types, constructed data types (structs, sequence, unions), and value types." The joint submission has been provided by: BEA Systems, Cape Clear Software Ltd, Hewlett-Packard Company, International Business Machines Corporation, IONA Technologies PLC, Oracle Corporation, PeerLogic, Inc., Persistence Software, Rogue Wave Software, and Unisys Corporation. The RFP requests "a standard way to represent XML values (documents) using OMG IDL non-object (value) types." This response provides an XML to IDL mapping leveraging the Document Object Model (DOM) technical recommendation from the World Wide Web Consortium (W3C). The DOM is an extensively used standard mechanism for defining access to XML content. The DOM includes a set of interfaces defined in IDL with mappings to Java and C++. The purpose here is to enable IDL users to access XML content using IDL valuetypes while maintaining maximum DOM compatibility. To this end, DOM level 1 and level 2 interfaces are re-declared as IDL valuetypes instead of the IDL interfaces in the DOM standard. The RFP does not request a mapping from IDL to XML. Mapping from IDL to XML is already accomplished using the MOF and XMI OMG standards." The new submission "provides two essential scenarios for using XML to create IDL valuetypes. The first scenario, where dynamic information is present, leverages existing standards to provide access to the full contents of an XML document in terms of IDL valuetypes. The second scenario builds upon the first where additional static information is present from XML DTDs and (in the future) XML Schemas. The DTDs / Schemas are metadata used to generate Valuetypes that match the types of information expected to be present in XML documents. The metadata from the DTDs / Schemas and Valuetypes may be imported into CORBA Interface Repositories and the Meta Object Facility, providing wide metadata distribution through OMG standards. The dynamic information scenario is the processing of an XML document when the meaning of the XML elements found in the document is not defined. In this case, only minimal information is known - what is in the XML document and little else. The DOM is a standard representation for the complete contents of an XML document. The DOM satisfies the requirement of the W3C XML Information Set (Infoset) to provide an access mechanism to the document contents. By expressing the DOM in terms of IDL valuetypes, a CORBA implementation has practical, standardized, and direct access the full information in of the XML document." For further description and references, see (1) the extended abstract and (2) "Object Management Group XML/Value RFP". [cache]
"CORBA & XML Resource Page" - Links to OMG's XML activities and references to articles on CORBA/XML.
| Last modified: August 31, 2000 |
| Object Management Group XML/Value RFP |
