This application claims the benefit of European Patent Application No. 07150350.2 filed 21 Dec. 2007, entitled “A METHOD, APPARATUS AND COMPUTER PROGRAM FOR DERIVING OVERLAY INFORMATION FROM A USER INPUT FOR A BASE MESSAGE SCHEMA ASSOCIATED WITH A NODE IN A MESSAGE FLOW”, which is assigned to the assignee of the present application, and the teachings of which are hereby incorporated by reference in their entirety.
The invention relates to message processing and more particularly to the use of schema information in the processing of a message.
It is known for business applications to use web services, messaging etc. to communicate with one another over an Enterprise Service Bus (ESB). For example, a certain class of applications may send messages to each other using messaging systems, which are connected together via the bus. In this way, one application connected to one messaging system may talk to a different application connected to another messaging system.
Applications interacting across an ESB commonly use messages such as XML messages to communicate with one another. Messages are normally described by type definitions (schema) that define the message structure and allowable content. Message type definitions may include weakly-typed fields, where the actual message contents may vary and more than one definition may apply for that part of the message. For XML messages and in some other cases, XML Schema provides the type definition, and weakly-typed fields are most commonly encountered where the “any”, “anyType”, or “anySimpleType” constructs are used. These are constructs that are well known in the art.
Tasks such as viewing message structure, for example, in order to define transformations between one message type and another usually depend upon a visualization of the message structure, which is derived directly from the message type definition. Visualizing the content of weakly-typed fields is difficult, because their content structure is variable, and the XML Schema or other type information does not describe the content fully. In these situations, the result is an incomplete visualization of limited assistance for defining how these messages should be handled, and the end user must often write custom logic to handle the weakly-typed content.
According to first aspect, there is provided a method for deriving overlay information from a user input for a base message schema associated with a node in a message flow. The method can present the base message schema to the user. The base message schema can have at least one variable portion. A user selection of a variable portion of the base schema can be received. The user can be provided with possible overlay schema information for the selected variable portion. A user selection of the schema to be overlaid on the selected variable portion; can be received. The selected overlay schema can be associated with the base schema.
In one embodiment, an overlay instruction can be generated that contains information for locating the selected variable portion. In one embodiment, the overlay schema can be associated with the overlay instruction. In one embodiment, a message structure can be visualized by providing an amalgamated view of the base message schema and any defined overlay schemas. In one embodiment, a second message structure is visualized and the two message structures can be used to define mappings between fields in both structures. In one embodiment, a connection between a first node in the message flow and a second node can be permitted based on one or more schema propagation rules. In one embodiment, the two nodes can have the same base schema associated therewith and the rules can assert that connection is acceptable if only the first node (and not the second node) has an overlay schema associated therewith.
According to a second aspect, an apparatus can be provided for deriving overlay information from a user input for a base message schema associated with a node in a message flow. The apparatus can include a means for presenting the base message schema to the user. The base message schema can have at least one variable portion. The apparatus can include a means for receiving a user selection of a variable portion of the base schema, a means for providing the user with possible overlay schema information for the selected variable portion, a means for receiving a user selection of the schema to be overlaid on the selected variable portion, and a means for associating the selected overlay schema with the base schema.
According to a third aspect, a computer program can be provided that includes a program code means adapted to perform the method of the first aspect when said program is run on a computer. The computer program can be stored in a storage medium, such as a volatile or non-volatile memory.
Each node has input and output terminals (not illustrated). Each terminal has message schema information associated with it. An input terminal's schema information defines the format of messages received by the terminal's node, while an output terminal's schema information defines the format of messages sent by the terminal's node.
As discussed previously, such schema information may be weakly typed. In other words, a schema may not fully define the format of a message. Such a format may depend upon the processing route taken by a message. For example, a local order message may need to look different to a European order message.
In the example of
The nodes in a mediation constitute a message flow which a user designs visually on a blank canvas. When a flow is first created, its input and output nodes (e.g. receive and send nodes of
As shown in
In
As shown in
Propagation also works in the reverse direction. When the output of the transform node is wired to the output node, the transform nodes output terminal receives type information from the output nodes output terminal. This is shown in
In this way type or schema information propagates through a message workflow. To clarify, if two transform nodes (or other nodes which change the type information) are placed in succession, then the type is known based on the configuration of those nodes. For example, once a transformation has been configured, then the input and output message types are known.
As previously mentioned, schema information may contain portions which vary dependent upon the specific type of message. A new type of node, the SET node, is defined which allows a user to manipulate variable schema information. This is discussed with reference to
At step 120, the input terminal of the set node instance receives schema information from the previous node in the message flow. This is as discussed above. It should be clarified however that the schema information is preferably not actually sent between terminals but rather a pointer enabling access to such information. This provides for a more robust solution as will be explained later on. Thus at step 120 a pointer to the base message schema is received.
At step 130, the user is presented with a graphical view of the base schema. Mechanisms for achieving this are already known in the art and so will not be discussed in any detail below.
Such a base schema may include variable portions as discussed above. The user is permitted to select a variable portion in order to overlay that portion with specific schema definition information. The mediation component stores possible schema (type) definitions from which the user may select and presents these at step 150. At step 160, the mediation component receives the user's type definition selection. At step 165, an overlay instruction is generated which indicates which element within the base schema is to be overlaid. At step 170, a pointer is provided to the selected type definition which is to be used as the overlay. Processing loops round for all variable portions within the base schema.
The set node 180, also has an overlay id (Oid) associated with it. This points to an instruction 185 in storage which provides an instruction for locating the element (variable portion) within the base schema that is to be overlaid and the instruction for overlaying this portion. Note that the element may be located using an XPath expression.
The overlay instruction in turn points to the information 190 that is to be overlaid. In this instance the order is a European order and so the address information includes country information.
Thus, the output terminal of the set node now has all the information necessary to more specifically define messages received. Such type information is propagated along the message flow as previously discussed. The propagation is a ‘tooling-time’ propagation. The tooling needs to be told precisely what to expect at a certain point in the flow so that when the message representation is shown it has the correct form.
Whether base and overlay schema information is permitted to be propagated is defined by some validation rules. When wiring two terminals together (connecting an arc between them) that do not have the same type map (i.e. they have a type mapping of their own, either because they are part of a dataflow node that asserts a type mapping or because they are connected to a dataflow node that asserts a type mapping), connection is either allowed or disallowed according to the following rules;
Having more strongly defined the schema (type) information, it is now possible to use such information to visualize the contents of a message or to define the mappings required when transforming between two message types.
When visualizing a message, the current node's input terminal's message schema is typically used to render the message appropriately. (For a mapping (transformation) visualization is performed of both input and output message structures, based upon the input and output terminal information respectively.) This is discussed in more detail with reference to
The base schema, overlay instruction and overlay information are however stored separately as discussed and do not need to be amalgamated even when the internal structure of the base or overlay type is modified. That is, although what looks like an amalgamated view of the overall message structure is produced, that view will automatically reflect any changes made to its constituent parts.
It should be appreciated that a new schema is not actually created. Rather, the message in accordance with a virtual view of an underlying message schema is presented. This provides for a much more robust solution since the parts of the schema can be individually edited. If a new schema was actually created, then it would not be possible to edit the constituent parts without the overall schema becoming out of date.
It should further be appreciated that it is optionally possible to validate messages at runtime to determine whether they conform to the amalgamated ‘virtual view’ of the underlying schema. For validation it is the input schema that is used.
Number | Date | Country | Kind |
---|---|---|---|
07150350 | Dec 2007 | EP | regional |
Number | Name | Date | Kind |
---|---|---|---|
6077313 | Ruf | Jun 2000 | A |
6636855 | Holloway et al. | Oct 2003 | B2 |
7007033 | Rothschiller | Feb 2006 | B1 |
7281018 | Begun et al. | Oct 2007 | B1 |
8135755 | Selca | Mar 2012 | B2 |
20020152244 | Dean et al. | Oct 2002 | A1 |
20020194227 | Day et al. | Dec 2002 | A1 |
20030046282 | Carlson | Mar 2003 | A1 |
20030050931 | Harman et al. | Mar 2003 | A1 |
20030154154 | Sayal et al. | Aug 2003 | A1 |
20030182452 | Upton | Sep 2003 | A1 |
20040030993 | Hong Huey | Feb 2004 | A1 |
20040107401 | Sung et al. | Jun 2004 | A1 |
20040172584 | Jones et al. | Sep 2004 | A1 |
20050066317 | Alda et al. | Mar 2005 | A1 |
20050091290 | Cameron et al. | Apr 2005 | A1 |
20050132276 | Panditharadhya et al. | Jun 2005 | A1 |
20050209876 | Kennis | Sep 2005 | A1 |
20050246629 | Hu | Nov 2005 | A1 |
20050262049 | Somppi | Nov 2005 | A1 |
20050262440 | Stanciu et al. | Nov 2005 | A1 |
20060074736 | Shukla et al. | Apr 2006 | A1 |
20060090129 | Collie et al. | Apr 2006 | A1 |
20060101068 | Stuhec et al. | May 2006 | A1 |
20060107252 | Smuga et al. | May 2006 | A1 |
20060190814 | Collie et al. | Aug 2006 | A1 |
20060288270 | Gaurav et al. | Dec 2006 | A1 |
20070014305 | Assad | Jan 2007 | A1 |
20070106933 | Nene et al. | May 2007 | A1 |
20070150806 | Hartmann | Jun 2007 | A1 |
20070234210 | Bukovec et al. | Oct 2007 | A1 |
20080040360 | Meijer et al. | Feb 2008 | A1 |
20080046453 | Kostoulas | Feb 2008 | A1 |
20080052287 | Stanciu et al. | Feb 2008 | A1 |
20080104105 | Heifets et al. | May 2008 | A1 |
20080104579 | Hartmann | May 2008 | A1 |
20080162529 | Stuhec | Jul 2008 | A1 |
20090132912 | Sulistio et al. | May 2009 | A1 |
20100115394 | Oshima et al. | May 2010 | A1 |
Number | Date | Country | |
---|---|---|---|
20090164593 A1 | Jun 2009 | US |