The Universal Business Language Technical Committee (UBL TC) has released the 1.0 Committee-Draft version in 2004 Sep for implementation trials and reviews (also known as UBL 1.0-cd2). Incorporating wide range of review comments, inputs from e.Gov, RosettaNet and OAG/UNCEFACT, the TC revised the models and datatypes thoroughly, resulting in a highly polished form of schema set that is currently undergoing the due process to become an OASIS Standard. The corresponding UBLish that generates UBL 1.0-cd2 schemas based on UBL 1.0-cd2 spreadsheet models will soon be made available. The new version, UBLish v1.0.10, has been completed and is undergoing final packaging for release soon. There are a number of new features incorporated that allows UBLish users to work on generating customized schemas based on UBL-like schema generation processes. In using UBLish to customize schemas, the relevant naming and design rules would then be automatically incorporated in various applicable parts of the generated schemas. This helps in reducing the amount of work involved in producing the schemas, and allows UBLish users to focus on the data modeling processes. The schemas and instances generated by UBLish v1.0.10 can be found in this page, or by clicking here . When these schemas are compared to those found in UBL 1.0-cd2 package, there are some discrepancies and problems detected and reported in http://lists.oasis-open.org/archives/ubl-dev/200409/msg00056.html . Background on UBLish The UBL TC uses the UBL inter-schema helper (UBLish) as an integration and schema generation tool during their development of UBL 1.0-beta schemas. The UBL TC was then given a boost in terms of cutting down tremendously the turn-around time between a change in the model spreadsheets and the resulting change in the schemas that ought to mirror the spreadsheets' inputs. Developed by SoftML, UBLish relieves UBL standards workers from the myriads of details associated with the verification and assembly of multiple sources of inputs into producing XML schemas. The task of incorporating Naming and Design Rule requirements, implementation of code list references, and inclusion of Core Component Types that are implemented using UBL-edited schemas has now gone into the UBLish engine, which expeditiously integrates these desparate sources of inputs, verifies them on-the-fly, and churns out readily usable XML schemas that conform to UBL standard requirements. As a result of using UBLish, the UBL TC had:
Background on UBL UBL uses XML and related technologies to elevate the level of interoperability and common digital understanding among electronic business entities. This major effort is spearheaded by well-known XML Architect, Jon Bosak from Sun Microsystem, and joined by a number of multi-national corporates and individual members of the Organisation for the Advancement of Structured Information Standards (OASIS) in the UBL Technical Committee (TC). The effort has gained international momentum lately, thanks to its harmonisation and synchronization efforts with other major international standards bodies and major industrial verticals that are using XML. The UBL TC establishes liaison work with other organisations through its Liaison Sub-Committee (LSC) , which is also chaired by Jon Bosak. Already, it has built links with organisations such as
Software Utilities The software scripts and utilities below are authored by SoftML or staff of SoftML in support of UBL's efforts. All information entities here are to be used on an AS-IS basis without warranty whatsoever. |
8 Document Types
Other components being included in the UBLish schema generation process are:
The Code List schema files were managed via a Code List Catalogue in UBLish v1.0b.10. In the new UBLish v1.0.10 (without the 'b'), code list schema files will be generated from Specialized Datatype spreadsheet. UBL TC has organized both Code List and Non-Code-List specialized datatypes together within the Specialized Datatype spreadsheet. So in UBLish v1.0.10, we're looking at a similar processing model to synchronize schema generation. The Common Core Parameters schema file was manually edited during UBLish v1.0b.10. In UBLish v1.0.10 (without the 'b') which generates UBL 1.0 schemas, the Common Core Parameters schema file is being automated and generated. This produces much more consistent layout and ensures that errors, if any, become systematic and manageable instead of time-consuming sporadic occurrences resulting from human errors. Sample Schemas and Instances for UBL 1.0-cd2 (2004 Sep release) Generated by UBLish (2004 Sep 08) Using UBLish v1.0.10, a set of UBL 1.0-cd2 schemas and sample instances are generated for each UBL 1.0-cd2 model spreadsheet. The schemas are transformed directly from UBL 1.0-cd2 model spreadsheets without intermediate steps, generating schemas that reflect completely the model values stored in spreadsheets. When compared to the corresponding schemas found in UBL 1.0-cd2 package, the discrepancies between UBL 1.0-cd2 spreadsheets and schemas were revealed. For the set of schemas found here generated by UBLish v1.0.10, the values and constituent data are taken from the model spreadsheets with the assumption that the model spreadsheets are correct. The instance documents basically "span" all the fields, filling each leaf element with a random value casted to the proper XML Schema scalar type. Thus, each document "touches" each and every defined element in the schema at least once, and is therefore useful in pitching against other UBL schema validators to see if all the fields are validated. Note, however, that just one set of instance document cannot be a conclusive compatibility or qualifying test for correctness nor accuracy. In addition, there is no semantic correctness built into these random samples. In other words, certain fields that ought to have values dependent or calculated based on other fields would not manifest such relationships. For example, the total amount of all the line items should be the sum of each amount of each line item. But because random values are used, such summation relationship will not exist in these random samples. All this set of random documents provides is a sample in the UBL instance space for a candidate UBL schema validator to work on, and potentially revealing some shortcomings of the validator. Another use for these random documents on the qualitative side will be to provide a visual hint on how each UBL document would look like when filled with data. A brief note on how the samples are generated: Each schema is used as a template to fill in the relevant types and random scalar data values. For elements with minOccurs and/or maxOccurs, the generated sample contains 1 occurrence regardless. For schemas that invoked some types that resulted in chain infinite recursion, the recursion truncation method used is to truncate on first detection of type recursion. Any attribute values are skipped.
Below this, we list the schemas and randomly filled instances
for your browsing.
Sample Instance for UBL 1.0cd Generated by UBLish (developed on 2004 May 26, published on 2004 Jul 28) Using UBLish v1.0.10, a set of sample instance documents has been generated for each UBL 1.0cd schema. Below this, we list the entire archive, and each randomly filled sample instance documents for your browsing.
Important Dates
Past UBL Release
UBL - 1.0beta Description Page
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||