Methods for creating and using a relational description of a formatted transaction while maintaining data element context are provided. The formatted transaction can be in electronic data interchange (EDI) format or another structured transaction format. A computer receiving the formatted transaction generates or is provided with a set of assignments of unique persistent tags to all possible data element contexts that are of interest. As the transaction is being parsed, each of its data elements is tagged with the appropriate unique persistent tag that identifies its context. Each data element is then written to a field in a relation in which the column name of the field matches the unique persistent tag. Data elements that can occur only once in a transaction are written to a root relation, while data elements that can occur multiple times are written to a subrelation. Data elements that can occur multiple times within a group of data elements that itself can iterate as a group are written to a nested subrelation, up to any depth of nesting allowed by the transaction format definition. The invention also provides a method for constructing a transaction from a relational description.
The select list for the query is thus constructed such that it produces a result set having a predefined structure. Columns representing identifier properties of an entity are added to the select list. If the entity is an inheritance entity or a collection, then entities from a least derived entity to all descendants are identified and columns representing their identifiers are added to the select list and columns representing non-inherited properties for those entities are added. If the entity is not an inheritance entity or a collection, then columns representing properties of the entity are added to the select list.
A message broker data processing apparatus for receiving messages from a sender application, processing the received messages and deciding which receiver application to forward the processed messages. An incoming stream of messages is received from a sender application, with each message being arranged as a tuple having at least one field. The incoming stream of messages is collated with data stored in a database, the data being also arranged as tuples having at least one field. The stream of message is processed so that a resultant stream of messages becomes directed to at least one receiving application.
A relevance information setting unit imparts the same relevance code to related elements in a standard message to indicate a relevance. A mapping processing unit displays a standard message and a local message side by side on a display and while referring to the relevance codes imparted to the standard elements, defines link relations with local elements on the display to create a conversion table for use in a message format conversion in a conversion unit. A link line appears when a standard element and a local element are selected to be linked.
A method of operating and communicating with a host computer system is provided using mark-up language inputs and outputs directed by an intermediary which has been previously constructed by formulation of the host data and program structures and host machine characteristics into a system of metadata which allows the reconstruction of requests and results into and out of mark-up languages binary data streams.
A system for processing a data file which includes versioned records of a fixed length, each record having one or more data fields and a reserve area for accommodating a change in the record format by creating a further field using a portion of the reserve area such that the length of the record is not changed. Additional records may be added to the record set to accommodate new or expanded data fields that exceed the fixed length of the records. Each record of the data file has a version number. The system includes an input vertical stack processor for transitioning previous versions of the records into a current version of the record for processing by an application processor. An output vertical stack processor creates outbound versioned records including the current version and all previous versions for each record of the data file.
