A portion of the disclosure of this patent document contains material that is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or patent disclosure as it appears in the U.S. Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever.
Generally, software systems provide software architectures that organize and interconnect software components within a computer system. The software components perform operations that give a software application its function. Generally, an application operates with several features, each having one or more components. The components and hence, the features may be formed from one or more underlying software components that parse the operation into smaller and simpler tasks.
When constructing a software application, a software developer must create the software components using a development language, such as C#. In creating the imperative code, the software code that gives the software component its function, the developer must create links between all the components through calls or other constructs. For any functionality provided by an application, the software developer generally creates the code for all the different software components that perform the application specific functions and manually codes the interconnections between the software components that rely on each other to complete tasks. The software developer creates user interfaces (UI), data constructs, and all the required operations for the interactions between the user and the application.
In many applications, the UI, operations, and data constructs are application-specific. Thus, to create an application, a software developer typically creates enormous amounts of code. In addition, the software developer generally organizes and creates all the interrelationships between the different pieces of software. To create any application, a software developer must be very skilled because the underlying languages used to create applications are complicated.
Due to the complexity of the current methods for generating applications and due to the specificity of the code for a given application and its inevitable interconnectedness, software developers cannot easily modify or expand on current applications without great effort and without risking damage to existing applications.
It is with respect to these considerations and others that the present invention has been made.
This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This summary is not intended to identify key features or essential features of the claimed subject matter.
Embodiments of the present invention solve the above and other problems by providing a software application framework for generating a runtime structure for a given software application and for managing execution of the software application. An application generated by the application framework of the present invention is comprised of a collection or graph of connected application components. Functionalities of an application configured according to the application framework are enabled by dynamically configuring groups of application components into component domains where each domain is configured to enable a given functionality of the application, for example, displaying a picture in a word processing document.
In order to generate a runtime structure for the application, the application passes an application description for each required application functionality to an application description engine. The application description provides declarative rules for structuring and composing the component domains, and the application description engine is operative to interpret the declarative rules for creating and reconfiguring the component domains as required based on data events received by the application. Data events received by the application may be generated by user action, for example, in response to a user selection of a functionality button or control or data object in a user interface. Data events also may be generated by external changes, for example, data received by the application as a result of the running of an external process or from another application or third party source. According to one embodiment, the application description and the application description engine are structured and/or operate according to the Extensible Markup Language (XML). While XML may be used within the application framework to provide a standard and simple means of stating actions or other occurrences within the framework, the application framework is not limited to the use of only XML.
The application description engine interprets the application description for each required functionality and then obtains application components needed to construct a component domain for each required functionality. According to one embodiment, the application description engine obtains application components from a component library maintained in association with the application. For example, the application description engine may construct one domain for displaying text in a document, a second domain for displaying a picture object in a document, a third domain for formatting functionality of the application, and so on.
According to embodiments, the application framework further includes an execution management model for managing execution of the application and the domains comprising the application. When a data event, for example, the deletion of a data item in a table of data, occurs during execution of the application that requires invocation of a given application functionality, and thus requires invocation of a particular component domain that enables the functionality, a single processing thread is dispatched by the application to the subject domain for executing the components of the domain according to the data-driven event.
Once the single processing thread enters the subject domain, components of the domain are executed by the thread according to an execution phase model. In a first phase, a read/request phase, a reading of required data, for example, a value to be deleted from a table object, is performed, followed by a request for required changes, for example, deletion of a requested value from the example table object. During a second phase, a revalidation/reconfiguration phase, the subject domain is revalidated or reconfigured by the application description engine according to the requested changes. That is, the application description engine reconfigures the domain, if necessary, according to a new configuration applicable to the change in data.
During the reconfiguration phase, the application description engine may discard some components of the domain, obtain new components for the domain, or destroy the domain altogether. Thus, a given domain is a collection of components that have a lifetime cycle equal to the time between initial configuration and a subsequent reconfiguration, which in turn, starts a lifetime cycle for the next configuration, if any, of the given domain. Accordingly, each component domain generated by the application description engine serves as a collection of components needed for performing a given data-driven functionality of the application, and the domain has a lifetime determined by the need of the application for the services of the domain.
A more complete appreciation of the present invention and its improvements can be obtained by reference to the accompanying drawings, which are briefly summarized below, to the following detailed description of exemplary embodiments of the invention, and to the appended claims.
The present invention will now be described more fully with reference to the accompanying drawings in which embodiments of the invention are shown. The invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided in an effort to make the disclosure thorough and complete and such that it will fully convey the scope of the invention to those skilled in the art.
As briefly described above, embodiments of the present invention provide a software application framework for generating a runtime structure for a given software application and for managing execution of a software application. The runtime structure of a given software application is composed of one or more domains of application components required for performing one or more respective functionalities of the software application. The component domains are generated by an application description engine in response to an application description received from the application for each domain. During execution of the application, each component domain serves as a unit of application execution management where each domain is utilized to execute a respective functionality of the application. As data changes are received by the application associated with a given component domain, the component domain is reconfigured by the application description engine as required to respond to the associated data. Data changes or events received by the application may be generated by user action, for example, in response to a user selection of a functionality button or control or data object in a user interface. Data events also may be generated by external changes, for example, data received by the application as a result of the running of an external process or from another application or third party source. Thus, the application is a collection or graph of connected components that are organized and dynamically reconfigured based on data received by and/or processed by the application.
An example of a suitable computing system environment 100 on which the invention may be implemented is illustrated in
The invention is operational with numerous other general purpose or special purpose computing system environments or configurations. Examples of well known computing systems, environments, and/or configurations that may be suitable for use with the invention include, but are not limited to, personal computers, server computers, hand-held or laptop devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like.
The invention may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
With reference to
Computer 110 typically includes a variety of computer readable media. Computer readable media can be any available media that can be accessed by the computer 110 and includes both volatile and nonvolatile media and removable and non-removable media. By way of example, and not limitation, computer readable media may comprise computer storage media and communication media. Computer storage media includes volatile, nonvolatile, removable, and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by computer 110. Communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media. The term “modulated data signal” means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media includes wired media, such as a wired network or direct-wired connection, and wireless media, such as acoustic, RF, infrared, and other wireless media. Combinations of the any of the above should also be included within the scope of computer readable media.
The system memory 130 includes computer storage media in the form of volatile and/or nonvolatile memory, such as read only memory (ROM) 131 and random access memory (RAM) 132. A basic input/output system 133 (BIOS), containing the basic routines that help to transfer information between elements within computer 110, such as during start-up, is typically stored in ROM 131. RAM 132 typically contains data and/or program modules, such as those modules constructed or executed under the XML application framework, that are immediately accessible to and/or presently being operated on by processing unit 120. By way of example, and not limitation,
The computer 110 may also include other removable/non-removable, volatile/nonvolatile computer storage media. By way of example only,
The drives and their associated computer storage media discussed above and illustrated in
A monitor 191 or other type of display device is also connected to the system bus 121 via an interface, such as a video interface 190. In addition to the monitor 191, the computer 110 may also include other peripheral output devices such as speakers 194 and printer 193, which may be connected through an output peripheral interface 192.
The computer 110 may operate in a networked environment using logical connections to one or more remote computers, such as a remote computer 180. The remote computer 180 may be a personal computer, a server, a router, a network PC, a peer device, or other common network node, and typically includes many or all of the elements described above relative to the computer 110, although only a memory storage device 181 has been illustrated in
When used in a LAN networking environment, the computer 110 is connected to the LAN 171 through a network interface or adapter 170. When used in a WAN networking environment, the computer 110 typically includes a modem 172 or other means for establishing communications over the WAN 173, such as the Internet. The modem 172, which may be internal or external, may be connected to the system bus 121 via the user input interface 160, or other appropriate mechanism. In a networked environment, program modules depicted relative to the computer 110, or portions thereof, may be stored in the remote memory storage device 181. By way of example, and not limitation, the remote application programs 185 reside on memory device 181. It will be appreciated that the network connections shown are exemplary and other means of establishing a communications link between the computers may be used.
For purposes of understanding and illustration, it is advantageous to describe embodiments of the present invention in terms of an example application user interface in which is displayed example data objects.
The user interface 195 includes a scrollbar 199 for scrolling data contained in the user interface work area up or down as desired by a user. An example word processing document is illustrated in the work area of the user interface 195. The example document, for example, a letter, memorandum, report, and the like, includes a text object 196, an embedded picture object 197, and an embedded table object 198. Referring still to
As will be described in detail below, according to embodiments of the present invention, the software application responsible for displaying the example user interface 195 and the example data objects 196, 197, 198 is a collection or graph of related components that are grouped together in logical groupings or domains for providing the functionalities required by the application. At any given instance, the application includes a runtime structure comprised of those logical groupings of related components, referred to herein as domains, which are required for providing the functionality of the application at that instance. As the application is executed, the component domains are dynamically reconfigured based on changes in data received by the application, and associated functionality required by the application for dealing with the changes in data. For example, referring still to
As described in detail below, during execution of the application where data events are received by the application, the runtime structure of component domains are dynamically revalidated and/or reconfigured as required in response to data-driven events. For example, if a user selects the picture object 197 for deletion or other modification, the application will pass the change in data to an application description engine, which will dynamically reconfigure the component domain responsible for displaying the picture object 197 so that the component domain will include the appropriate application components for displaying the picture object 197 based on the data event received, for example, deletion or other modification of the picture object 197. In the case of a new function required by the application, for example, the entry of a new data object, such as the table object 198, into view in the user interface 195, the application will call on the application description engine to generate a new component domain comprised of one or more application components required for displaying the incoming data object.
According to embodiments of the present invention, the application 204, 304 is comprised of a number of application components, each of which have been created by a software developer for providing a given functionality independently of other components or in association with other components. For example, a given application component may be included in the application 204, 304 for providing a print function of the associated application. Another application component may be included in the application for providing a particular formatting function of the associated application. Typically, a given function of an application 204, 304, for example, formatting, printing, editing, data display, and the like, is enabled by a collection of application components which when operating together enable and provide the desired functionality. For example, the functionality of a given software application for displaying a text object 196 or picture object 197, illustrated above in
As briefly described above, at any given time, the application 204, 304 comprises a runtime structure, which is an instantiation of connected application components required by the application at that time. As described below, during execution of the application, the instantiation of application components is dynamically reconfigured as the needs of the application change based on data received by the application. In order to generate a given runtime structure of application components, the application 204, 304 passes an application description for each required functionality from an application description 318 to an application description engine 320. A detailed description of the application description 318 an the operation of the application description engine are provided below. For example, referring back to
The application 204, 304 passes an application description for the functions currently required by the application to the application configuration or description engine 320 for generating and configuring groupings of application components required for providing each of those functions. According to one embodiment of the present invention, the application description passed from the application to the application description engine includes declarative rules structured and formatted according to the Extensible Markup Language (XML), as described below with reference to
Once the application description engine 320 receives application descriptions from the application 204, 304 associated with functions required by the application, the application description engine interprets the application descriptions in terms of application components that are required to fulfill each described function, for example, displaying a picture object 197, as illustrated in
Once the application description engine obtains the required application components (for example, individual modules of software programming), the application description engine creates component domains for each of the functionalities presently required by the application 204, 304. As illustrated in
According to embodiments, an execution management model is provided where each component domain serves as a unit of execution management for the application 204, 304. As data events are received by the application 204, 304, as described in detail below with reference to
Similarly, if the document illustrated in
During execution of the application 204, 304, execution management is accomplished by managing the operation and component structure of each component domain. Referring still to
In response to the data event, the application 204, 304 dispatches a single processing thread 275 to the appropriate domain 260 responsible for processing data events associated with the table object 198. At the subject domain, also referred to herein as a concurrency domain, the single processing thread 275 enters the domain of components and processes the data event in a series of phases. For a detailed description of the dispatch of the single processing thread to the subject domain for processing a data-driven event via the components of the domain, see the description set out below under the heading “Multithreading with Concurrency Domains.”
Following from the present example, once the delete control is selected for deleting the selected data item from the table object 198, the single processing thread is dispatched to the domain 260 in response to the action data of deleting the table item. Once the processing thread enters the domain, a first processing phase, a read data/request data phase is initiated. For example, during the first phase, the data event is read by the component domain 260 to determine the data item or data value in the table object 198 that is to be affected by the data-driven event. Next, as part of the first operating phase, changes to the data are requested, for example, a request that the selected data item be deleted from the table object 198. In response, a data connector component in the table object domain sets up a request that the selected data item be deleted from the table object 198. A detailed description of components of the application 204, 205 including data connector components is provided below with respect to
The next processing phase includes a revalidate/reconfigure phase in which changes to the component domain responsive to the data event are made by the application description engine. During the revalidate/reconfigure phase, all data connector components associated with data that will be affected, for example, a data connector component associated with the data item to be deleted from the table object 198, are notified to mark those data items for revalidation/reconfiguration. According to one embodiment, the marking of affected data occurs during a sub-phase of the read data/request data phase.
Once all affected data items are marked, the revalidate/reconfigure phase proceeds, and each marked data item is processed, as required. For example, the data item to be deleted from the table object 198 is deleted, and any associated domain components are deleted or added, as required. For example, if as a result of the deletion of the selected data item one or more application components comprising the affected domain 260 need to be removed, the application description engine will reconfigure the domain to remove those components in response to an application description 318 passed to the application description engine during the revalidate phase in response to the data event. On the other hand, if new application components are required for the affected domain, the application description engine will likewise obtain new application components for integration into and reconfiguration of the affected domain. Thus, the affected component domain is dynamically reconfigured during the execution of the application based on a data event associated with the domain. A detailed description of the execution phase model described herein for processing data events in the component domains 250, 255, 260 is provided below under the heading “Application Framework Phasing Model.”
As described above, each component domain is configured by the application description engine 320 based on the present need for the component domain by the application 204, 304. And, as described above, as the needs of the application 204, 304 change in response to a data event, for example, the deletion of a data object, the addition of a data object, or the modification of a data object, new component domains are created or existing component domains are reconfigured as required. Each new component in a given component domain has a lifetime equal to other components in the component domain because a given component domain lifetime begins at its instantiation by the application description engine or its reconfiguration by the application description engine and ends when the component domain is discarded or is subsequently reconfigured by the application description engine. Thus, a given component domain is a unit of application functionality and management that lives during the time of a particular configuration of the component domain and terminates or dies upon a subsequent reconfiguration of the component domain where a subsequently reconfigured version of the component domain is considered a new component domain for purposes of application execution management.
Having described above an application framework for configuring a software application as a collection or graph of connected components and for managing the structure and execution of an application via one or more component domains, a detailed description of the components comprising a given application and of communications between the components is provided below with reference to
In an exemplary embodiment, XAF employs XML (Extensible Markup Language) to represent data, as XML data, and applications, as XML applications. Hereinafter, XML will be used in the description to represent the standard language. However, one skilled in the art will recognize that the present invention is not limited to the use of XML.
The XAF software system is data focused, data driven, and offers powerful data transformation capabilities including, but not limited to, interacting with heterogeneous data from mutually unaware sources. All parts of XAF can be described as events directed at or involving data, including the behavior of the application and the application's user interface. XAF abandons the paradigm that the application is the data creator and data is the creation. Rather, XAF builds the application based upon the data. In embodiments of the present invention, the XAF application components are actually instantiated and interconnected based on how the data is displayed, on what the data is, and how the data is stored. Thus, XAF can create and stitch together the “right” components, at runtime, to process the specific data in the correct fashion.
In building the applications, XAF allows an application “author” to provide rules one how components are created and connected together based on the data type and the user interface (UI). These rules and the components may be written in XML. XML allows for rich and composable UIs that are sensitive to the current context and allows software developers to build applications in a compositional approach. The software author can build features or applications end-to-end in a declarative model that reduces the need for hard-coded application logic. For example, the software author simply declares that a certain UI connector goes with a certain UI and a certain action. XAF then connects the UI connector together with the UI and the action.
Additionally, the XML application framework provides a rich library of pre-built components, application features, and example applications. While some imperative code is still required, XAF places the components in a library. The library allows other software authors to use or employ those components in the declarative model. Thus, a software developer may create their own components by generating new imperative code for the component, but the software author can also employ pre-built components to compose new or modified applications without generating any new imperative code.
An exemplary XML application framework 202 is shown in
UI connectors 206 provide the connection between the application 204 and the one or more UIs 218 and/or 220. In one embodiment, the UIs 218 and/or 220 are graphical user interfaces that display information to the user. The UI connectors 206 present and support the ability for the application user to edit data within the XAF 202. A UI connector 206 maps data between the XAF data layer (represented by the data connectors 210 and the data stores 222 and/or 224) and the specific UI 218 and/or 220. In addition, the UI connectors 206 map UI events from the UIs 218 and/or 220 into data editing operations within the data stores 222 and/or 224.
The UI connector 206 is specific to the type of UI element in the UI 218 and/or 220. Thus, for every UI representation of an item of data within a UI 218 and/or 220, there is a corresponding UI connector 206. In one example, a data element, such as a cell value within a spreadsheet, displayed in the UI 218 and/or 220 will have a specific UI connector 206, coupled to the data element, which translates a user modification to the data element into a UI event to connect to an action 208 on the XAF data. Thus, the UI connectors 206 translate specific UI events into XAF actions 208 that are generically represented within the XAF application 204. In one embodiment, all data changes within the application 204 are represented as XML data changes. In another example, a user may operate a user interface control, such as a scroll bar. The change in the scroll bar creates a change in the state of the UI 218 and/or 220. The UI state change may also be represented as a change in data. Thus, a UI connector 206 can receive the UI event, e.g. the scroll bar operation, translate the UI event into an XML data change event for the XAF data layer.
The UI connector 206 connects the UI event to an action 208. An action 208 is a declarative statement for a data change in the XAF data layer. For example, a user operation of the scroll bar can create a “click on the scroll bar” event. The UI connector 206 connects this UI event to an action 208, such as “increment scroll bar down one position.” In one embodiment, the data change action is represented in XML. For example, a data change may appear as the XML statement:
Data connectors 210 marshal data to and from external data stores 222 and/or 224. Similar to the UI connectors 206, data connectors 210 translate between the internal XAF data representations and external data types in the different data stores 222 and/or 224. Thus, there are data-specific data connectors 210 for each type of external data. The data connectors 210 receive the actions 208 and translate the standard format XAF data action 208 into a data-specific action to affect the data within the data stores 222 and/or 224. In one embodiment, the data connector 210 translates the XML data action into a data-specific action. For example, the scroll bar operation represented by a standard format XML statement is translated into an interface state-specific data change and sent to the data store 222 or 224 storing the specific interface state data.
In addition, the data connectors 210 translate changes in the data stores 222 and/or 224 into standard format XAF data representations that are sent to the UI connectors 206. The UI connectors 206 translate the standard format XAF data representations into UI events that are sent to the UIs 218 and/or 220. Thus, for any change within a UI 218 and/or 220, the UI connector 206 connects the user interface event into an action 208. The action 208 is sent to a data connector 210 that translates the action 208 into a data-specific change in a data store 222 and/or 224. Once the data is changed, the data connector 210 translates the changed data in the data store 222 and/or 224 into a standard format XAF data representation. The data connector 210 sends the XAF data representation back to the UI connector 206, which translates the XAF data representation into a UI-specific display event. The UI connector 206 sends the UI-specific display event to the UI 218 and/or 220 for display, where the UI 218 and/or 220 displays the changed data created from the UI event.
The application 204 can continue to process these cyclical changes between data and UI events for all actions occurring within the application 204. In some embodiments, changes in data at the data store 222 and/or 224 also force changes in the UIs 218 and/or 220 without a user event occurring, such as a data changed forced from another operation outside of the user's control.
Some optional components within the application 204 include accessors 212 and transformers 214 and/or 216. Accessors 212 mediate between actions 208 and data connectors 210. To access data of a certain type in a certain data store 222 and/or 224, the action 208 can specify an accessor 212 or “access model” to use for the data. The accessor 212 allows the application 204 to access specific types of data that are not homogenous, such as JPEG, MPEG, etc. Thus, regardless of the type of data within the data store 222 and/or 224, the application 204 interfaces and changes that data. The accessors 212 ensure that the data connectors 210 manage any type of data including data types that are not yet devised or developed. The accessor 212 transforms the standard format action into a standard format, data-customized action.
In other embodiments, one or more transformers 214 and/or 216 mediate between data connectors 210 and UI connectors 206. A transformer 214 and/or 216 changes the data output from the data connector 210 into a UI-customized format required by the UI connector 206 for display. For example, if the UI connector 206 requires data in a list, the transformer 214 and/or 216 may modify tabular data through one or more simple changes, each simple change performed by a single transformer 214 and/or 216, into a list of data. Thus, any form of data output by the data connector 210 can be transformed through one or more canonical transformers 214 and/or 216 into a form acceptable and usable by the UI connector 206. For a detailed discussion of the operation of transaction transforms, such as the transformers 214, 216, see the description below under the heading “Transaction Transforms.”
Other types of transforms may be utilized in accordance with embodiments of the present invention. For example functional transforms may be described using a functional language such as XQuery. A proxy transform takes a list of locations and proxies the data found at those locations in the form of a sequence of aliased data. A solver transform embodies dedicated solver technology and supports caching for efficiency. For example, one solver transform solves algebraic equation systems numerically, and another solves those systems symbolically. A caching and indexing transform is a pass-through transform at data level (i.e., the output data is equal to the input data), and the caching and index transform adds caching and indexing in specified dimensions to accelerate downstream indexed access.
According to embodiments of the present invention, the application framework employs common interfaces between components of an application, described herein. Because the framework relies on data communication between the various components described herein, a uniform data structure, for example, XML, and common interfaces between the components allow for efficient data exchange for application configuration and management.
Referring now to
In embodiments of the present invention, the XAF 300 includes an XML application description (XAD) 318 and a XML application description (XAD) engine 320. The XAD 318 includes the declarative rules and descriptions of all applications included within XAF 300, such as application 304. A software author authors the XAD 318 for the application desired by the software author. The XAD 318 comprises schemas and tags for pulling or instantiating certain software components and binding those components together. In embodiments of the invention, the schemas and tags pertain to how the components interact. A component may be a UI connector, an action, an accessor, a data connector, a transformer, or other software component. In some embodiments, the components are written in imperative code and provide the point-wise functionality to the application. Thus, the component may be the underlying software that provides the operability to the XAF components, such as the UI connector, action, data connector, etc.
While the application description 318 can be represented using other formats, in one embodiment, the XAD 318 is represented in XML. The tags are XML tags. The schemas and tags provide a syntactic and semantic framework for declaring rules. XAD allows the application author to declare application wide resources and commands. Commands are the bits of code that are executed based on events, such as mouse clicks, keyboard selections, voice commands, etc. In embodiments, the user creates the events in a user interface, but, in other embodiments, another action by some process may create the event. The commands may be in the form of named actions, for example, a change selection action, an edit view action, a retrieve data action and an edit entry action, as described below with respect to
The XAD 318 is parsed by the XAD engine 320. Initially, the XAD engine 320 parses the XAD 318 when the application is launched. Thereafter, the XAD engine 320 parses the XAD 318 when needed. For example, according to one embodiment, the XAD 318 is parsed in runtime by the XAD engine 320 to reconfigure an application, such as when instantiating and connecting a plug-in component. According to another embodiment, a compilation model is provided where the XAD is parsed at compile time and is converted to a sequence of tokens, which are interpreted by the XAD engine at runtime. When not compiling, the textual parsing is done when the XAD is loaded into memory, and it is stored as a data structure.
The XAD engine 320 processes the rules, identifies the resources, instantiates the components, and connects the components, composite, and commands. The components and resources may be imperative code, such as executable code written in C#. The imperative code characterizes the actual software components and software classes that underlie the UI connectors, actions, data connectors, and other XAF components. The components are pulled from a component library, such as component library 322. The component library 322 holds all components used in any application with XAF 300.
The entire collection of components within the library 322 is available to any application. Thus, XAF 300 allows for prolific sharing of application components and the code that creates those components. In addition, XAF is very extensible because new components can be created by software developers and stored into the component library 322. The XAD engine 320 can then be call, instantiate, and connect the new components by reading new rules a software author authors in the XAD 318. The new rules and new components may then form new XAF components, new data types, new data stores, new UIs, or even new applications. The XAD and the XAD engine are described in further detail below.
An embodiment of a method 400 for executing a user event is shown in
Provide operation 406 provides a standard format action for a data change represented in the action. In one embodiment, an XML statement represents the action for the data change. In one embodiment, optional aid operation 408 connects the action to an accessor, such as accessor 212, to help configure the action for the specific data being changed. The accessor sends the data-specific action to a data connector, such as data connector 210. Translate operation 410 receives the data-specific action and translates this standard format action, which may be an XML statement, into a code and data-specific operation for a specific data store, such as data store 222.
A data change is made in the data store. The data connector then reads the changed data from the data store. The changed data is translated into a standard format data representation, such as a XML data representation. Optionally, the standard format data representation is then sent to a transformer, such as transformer 214. Optional change operation 412 changes the data record received by the transformer into a UI-specific data statement. For example, the data change may require the data to be displayed in a sorted list. One or more transformers may sort the data retrieved and create a final data statement that has the data presented in a sorted list. The transformer then provides the changed and transformed data to a UI connector, such as UI connector 206.
Connect operation 414 then receives the changed data and connects with a UI component to display and provide the changed data. Thus, the UI now provides a view of the changed data. In one embodiment, the data change may be a user interface state and the display of the changed data is a display of the changed user interface. As represented by the flow continuing back to the user event operation 402, the process 400 of receiving events, affecting a change in data, and displaying the changed data is iterative and can be accomplished repeatedly for each user event or data change. If data changes or events cause or require the reconfiguration of the application, new components may be created and connected and/or reconfigured as described below with respect to
An embodiment of a method 500 for creating and configuring an application is shown in
Identify operation 506 identifies the resources and/or components necessary for the application. In one embodiment, the components and resources are found within a component library, such as component library 322. Instantiate operation 508 instantiates the components. In one embodiment, helper code in a XAD engine, such as XAD engine 320, formally instantiates the components identified in the component library. In a further embodiment, a temporary set of creator components is instantiated. The creator components then instantiate the processing components with regard to the input data and rules associated with the tags in the XAD.
Connect operation 510 connects the instantiated components. In one embodiment, the XAD engine connects the components according to data used or the user event processed. Once the application components are instantiated and connected, wait operation 512 waits for a user event. Thus, the XAD engine maintains a passive presence with the application. The XAD engine continually monitors the data changes and is responsive to changes in data provided by a data connector, such as data connector 210. At a data change, determine operation 514 determines if the data change requires a reconfiguration of the application, such as inserting a plug-in component. If a reconfiguration is required, flow continues Yes to parse operation 504. If a reconfiguration is not required, flow continues No to wait operation 512.
As described herein, an application framework is provided for generating a runtime structure for a given software application and for managing execution of a software application. The following are detailed descriptions of various aspects of embodiments of the invention described above, including the operation and functionality of transaction transforms, component domain processing via phase modeling and threading, and operation of the application description engine for generating and reconfiguring application components.
Transaction Transforms
As described above, one or more data stores are provided for receiving and storing data via one or more connectors. According to embodiments of the present invention, data may be selectively isolated from data stores through the use of one or more “transaction transforms.”
Data stores 602 and 604 contain data on a database server or on a computer-readable medium. Data may be read from and written to a data store as described above. A check or some other request may be sent by an agent connected to a given data store to see if writing data to the data store is possible. If writing the data is possible, a positive indication is passed back to the agent. Likewise, if writing the data is not possible or not known to be possible, a negative indication is passed back to the agent.
Data connectors 606 and 608 connect data transforms 610, 612 and 614 to data stores 602 and 604. In an embodiment, protocols are implemented within data connectors 606 and 608 for dealing with one or more types of data stores. Each protocol handles one or more types of data store such as a database file encoded using a particular format. Data connector protocols atomically commit data modifications or changes received from data clients 622 and 624. The UI connectors 206, described above with reference to
In one embodiment, data connectors 606 and 608 achieve atomicity by way of pessimistic concurrency. Pessimistic concurrency involves locking subsets of data (e.g., one or more records) at a data store to prevent a client from modifying data in a way that affects other clients. In a pessimistic concurrency model, when a client performs an action that causes a lock to be applied, other clients cannot perform actions that would conflict with the lock until the client that owns the lock releases the lock. This model is useful in environments where there may be heavy contention for data, and/or where the cost of protecting data with locks is less than the cost of rolling back transactions if concurrency conflicts occur. Pessimistic concurrency is best used when lock times will be short, such as in programmatic processing of records.
In another embodiment, data connectors 606 and 608 achieve atomicity using optimistic concurrency with compensation actions. Optimistic concurrency does not make use of locks. When a first client needs to update a record, the protocol determines whether another client has changed the record since it was last read by the first client. Optimistic concurrency is useful in environments with little data contention. A compensating action is an action that compensates for the effects of a transaction. For example, the compensating effect for a bank transfer of funds from account A to account B is the transfer of the same amount of funds from account B back to account A. Similarly, the compensating action for reserving a hotel room would be the cancellation of the reservation. Compensating actions allow a transaction to be “rolled back,” or backed out with no lingering side effects or other negative consequences. A compensating action may be used to roll back a transaction when a conflict between two clients arises (e.g., a first client reads a data value, and then a second client modifies the data value before the first client has attempted to modify the value). In an embodiment, data connectors 606 and 608 cache data from data stores 602 and 604, respectively. If data requested by a data client is cached by a data connector, the corresponding data store need not be queried for the requested data. Data connectors are discussed in more detail below.
Data transforms 610, 612, 614, and 620 encode and/or decode data according to predefined rules. Data transforms 610, 612, 614, and 620 implement functions that can perform arbitrarily complex processing. Data transforms 612 and 620 are in series with one another. Because of this, the results of the function implemented by data transform 612 are used as input to the function implemented by data transform 620 or vice versa. In an embodiment, data transforms may be serially connected to buffer the results of previous transforms, and allow complex functions to be implemented more easily and modularly. Likewise, subsets of data within a view can be committed or refreshed (see below) without affecting the remainder of the data in the view.
Data transform 614 is a special kind of data transform known as a transaction transform, and as such, contains a buffering module 616 and a control module 618. In an embodiment, a transaction transform allows isolated data to be committed atomically and in such a way that data is consistent between data stores. Buffering module 616 holds isolated data from data clients 622 and 624. In an example, data clients 622 and 624 are dialog boxes for editing data in data store 604. Edits made to the data may be stored in buffering module 616 substantially as the edits are made. In another embodiment, transaction transforms map edit requests against their output back to edit requests against their input(s). As a result, when edits are made, the input/requesting entity recognizes such an edit as complete. Further, these transaction transforms use their buffer 616 to enable delayed and controlled mapping of such edits, as discussed below.
Several types of control operations may be performed by a transaction transform. When a user wishes to commit the data in data client 622 or data client 624 (e.g., when an “Apply” button associated with a data client is activated), control module 618 receives a commit control operation from that data client and attempts to commit data using the two-phase concurrency protocol discussed below, in conjunction with
In some situations, it is beneficial to update the data in a buffer without destroying uncommitted changes one or more users have made using one or more data clients. In such a case, a synchronize control operation may be issued by the data client. When control module 618 receives a synchronize control operation from a data client, the isolated data in buffering module 616 is updated based on the latest version in the data store(s), and the uncommitted changes to the isolated data are merged with the updated data. The process of synchronization of two or more sets of changes to data is well understood in the art, and synchronization may be used in a variety of circumstances. The details of how synchronization may be implemented in a given context are quite specific to that context. For example, in the context of a product database, it may be acceptable to merge a line of products in the database, but not acceptable to merge two versions of changes to a product's name. Other contexts may have other rules for synchronization, any and all of which may be used in conjunction with the present invention. Resolving conflicts is also known. Predetermined rules may be provided and followed as to which updates should govern in case of a conflict. Other embodiments may alert the user(s) to resolve such conflicts.
Buffering module 616 maintains status information that may vary based on the type of operation currently in progress. For example, if a commit is in progress, status information may include whether the commit is pending, was successful, or failed. Alternatively, if a refresh is in progress, status information may include whether the refresh is pending, or complete. If a synchronization operation is in progress, status information may include whether the synchronization is pending, was successful, or failed. The specific control operations are discussed in more detail below.
Some data accessible to a data client may not require isolation. For example, alterations to a financial database (such as that used to track bank accounts) are typically committed immediately. In an embodiment, a given data view may include both isolated and non-isolated data. In another embodiment, a given data client may include read-only data that may not be edited within the data client.
In response to the receipt of an application specification from an application developer, an embodiment of the present invention identifies which parts of the application require isolated data and implements transaction transforms as required.
Implement operation 706 implements one or more transaction transforms corresponding to each data entity that is marked as a data entity to be isolated. Transaction transforms are then connected and activated in substantially the same way as a data transform.
Determine operation 806 determines whether the control operation request is requesting a synchronize operation. If the control operation request is for a synchronize operation, flow branches YES to save operation 808. If the control operation request is not for a synchronize operation, flow branches NO to refresh operation 810.
If a synchronize operation was requested, save operation 808 saves any uncommitted changes to the buffered data. The uncommitted changes may be saved to a file on a computer-readable medium, to volatile or non-volatile memory, or other form of computer storage, or committed to an external database or similar service. Flow then proceeds to refresh operation 810.
Refresh operation 810 retrieves the latest copy of the buffered data that exists in the data stores, and places it in the buffer associated with the transaction transform. Refresh operation 810 may need to query several data stores to determine which data store contains the most up-to-date copy of the data. In another embodiment, refresh operation 810 may alternatively or additionally check the data connector associated with each data store to see if the data connector contains a cached copy of the buffered data.
Determine operation 812 determines whether the control operation request is requesting a synchronize operation. In an embodiment, determine operation 812 simply checks the results of determine operation 806. If the control operation request is for a synchronize operation, flow branches YES to merge operation 814. If the control operation request is not for a synchronize operation, flow branches NO to the end of the operational flow.
If a synchronize operation was requested, merge operation 814 merges the changes to the buffered data that were saved by save operation 808 with the buffered data refreshed by refresh operation 810. The rules for merging two versions of a body of data vary based on the context of the application. Some exemplary rules are presented in conjunction with
When receive operation 906 receives a request for a commit operation, a modified two-phase commit protocol is invoked. First, issue operation 908 issues commit requests to a plurality of data connectors associated with data stores. In an embodiment, a commit request includes the data to be committed. The commit requests are received by one or more data connectors that have cached the data, and/or one or more data stores that hold the data, and are treated as a request for assurance that the data can be atomically committed. Receive operation 910 receives responses from the data connectors and/or the data stores that hold the data. Once all the responses have been received, determine operation 912 determines whether all the recipients of the commit request agreed to the commit. If all the recipients agreed, flow branches YES to send operation 914. If not all the recipients agreed, flow branches NO to send operation 916.
If all the recipients agreed to commit the data, send operation 914 sends commit commands to the all the recipients of commit requests issued by issue operation 908. The recipients commit the data (that is, update their own data with the changes).
If not all of the recipients agreed to commit the data, send operation 916 sends abort commands to the recipients. The recipients are thus made aware that the proposed commit will not take place, and that any data associated with the commit may be discarded. In an embodiment, send operation 916 only sends abort commands to the recipients that agreed to commit the data, while the recipients that did not agree to commit the data automatically discard the data without being explicitly ordered to. Next, refresh operation 918 executes a refresh control operation to refresh the buffered data. In an embodiment, refresh operation 918 instead executes a synchronize control operation to synchronize the buffered data.
Other embodiments of the present invention are also envisioned. In one embodiment, each transaction transform may include a policy module that enforces synchronization rules. In another embodiment, the policy module sets the concurrency policy for refreshing and synchronization. Exemplary concurrency policies include conservative concurrency (wherein a complete copy of all buffered data is always made), and optimistic concurrency (wherein data is copied into the buffer only as needed). A policy module may also allow for modular addition of policy to handle other types of control operations not discussed herein.
In another embodiment, transaction transforms can be used to access and manipulate data in conjunction with a query language. SQL (Structured Query Language) is one such query language that is in wide use. XQuery is another such query language. The use of other query languages in conjunction with the present invention is also envisioned.
In yet another embodiment, the data connectors 610 may contribute and, to some extent, manage the parts of the buffer, e.g., buffer and logic 616 in
Application Framework Phasing Model
As briefly described above, applications and/or application components may be constrained to certain phases. Generally, phasing constrains the execution of software methods within a computer system by subjecting the software components to a multi-tiered phasing model. A software component can be a class, an object, a method, or other software code construct that is within a computer system. A phase is an operating state that is simultaneously and collectively shared by a set of software components. The computer system executes a top-level phasing model, also referred to as a master phasing model, with one or more sub-phases occurring during one or more of the phases of the master phasing model. The operations within the computer system are constrained to a set of phases or sub-phases.
An exemplary embodiment of the multi-tiered phasing model 1000 is shown in
Each software component is constrained to operate within a certain phase. A constraint is placed upon each of the software methods to execute or to be called only during the phases to which the software method is constrained. Software methods that can create conflicts or contradictory results are constrained to different phases, where the software methods cannot be legally called from the current phase. As such, each software method is executed in a known fashion without conflict between the methods that accomplish contradictory tasks. All methods execute under a particular phase constraint such that the software system is known to be in a state compatible with the present phase constraint.
Referring again to
The phase model 1000 illustrates a phase space. A phase space is a finite directed graph determining valid phases (graph nodes) and valid phase transitions (graph edges). A phase space therefore determines valid sequences of phases. The phase space 1000 is defined over the phase set Phase 11002, Phase 21004, and Phase 31006. The phase space 1000 also has three phase transitions 1018a, 1018b, and 1018c. A phase transition represents when the simultaneous change of phase occurs by all software components that share the pre-transition phase.
When software components share a phase space, those software components are part of a phase domain. A phase domain is a set of software components agreeing on a common phasing model as defined by a particular phase space. For example, all software components that agree to be constrained by the master phase space having master phases 1002, 1004, and 1006 are part of the master phase domain. Thus, all software components associated with the software components in the master phase domain include a phase constraint associated with at least one of the master phases 1002, 1004, and 1006.
A phase constraint is a static constraint that limits the phases valid in a certain program context. In particular, constraints can be applied to a program section, asserting that the program section will only execute during a phase honoring the constraint. In one embodiment, phase constraints are written as an attribute in brackets, such as [Phase 1]. This data structure is explained in more detail below.
A computer environment 1100 having one or more components occupying one or more phase domains is shown in
The phase domain 1100 can be characterized by a multi-tiered phase space. In embodiments of the present invention, one or more components, such as component 11104, register with the master director 1102. Component 11104 represents any type of software construct, including software components or methods. The software component 1104 is constrained to one of the phases in the master phase space.
In other embodiments, one or more sub-directors, such as sub-director 11106 and sub-director 21108, register with the master director 1102. The sub-directors control one or more other phase domains with one or more different phase spaces. Thus, the phase domain 1100 has one or more nested phase domains. All components, such as component 21110 registered with sub-director 11106, are constrained to one or more of the sub-phases within a sub-phase space and within one or more of the master phases of the master phase space. In one embodiment, the sub-phase domain controlled by the sub-director operates within a single master phase. The operations of the sub-phases can occur repeatedly during the single master phase.
In embodiments of the present invention, the sub-directors, such as sub-director 2, register other sub-directors, such as sub-director 3, to create further nested sub-phase domains. In some embodiments, the sub-director 21108 controls the operation of component 31112 and the sub-director 31114. In further embodiments, a director, such as sub-director 31114, controls more than one component, such as component 41116 and component 51118. Each sub-director may control a phase space having unique phases. Thus, sub-director 11106 operates a first sub-phase space while sub-director 31114 operates a second sub-phase space. If two-phase spaces do not interact, then the phase spaces are called orthogonal spaces. A combination of orthogonal phase spaces can form a Cartesian phase space, which can be used to form the phase domain for a single product. An underlying phase set is the Cartesian product of the orthogonal phase sets, wherein the valid phase transitions are also the Cartesian products of the valid transitions for the orthogonal phase sets.
A director, in embodiments of the present invention, is a logical clock. The director cycles through the phases similar to a clock in a hardware system. At each phase transition, the director simultaneously changes the phase for all software components within the phase domain. In one embodiment, any sub-directors may change the sub-phase in a sub-phase domain at the same time. The logical clock awaits the completion of an operation constrained to the phase or to an operation executing within a sub-phase constrained to the phase.
An exemplary embodiment of a phase space 1200 that may be used for a master phase domain is shown in
An Update phase 1204 directs the commands and requests to the appropriate software component. In embodiments of the present invention, during the Update phase 1204, the software components fulfill commands or requests. In one embodiment, the Update phase 1204 has a sub-phase space 500 occurring during the Update phase 1204. An exemplary sub-phase 500 is shown in
A third phase, the Revalidate phase 1206, directs and executes other methods not processed during the Update phase 1204. In one embodiment, all requests to retrieve data are completed during the Revalidate phase 1206. For example, after data is updated in the Update phase 1204, all software components are informed that data changes have occurred, and the informed software components retrieve the updated data. In one embodiment, the Revalidate phase 1206 operates a sub-phase space 600. An exemplary embodiment of the sub-phase space 1300 is shown in
To change phases, the phase space 1200 proceeds through a phase transition. In the exemplary embodiment, there are three phase transition 1208a, 1208b, and 1208c representing the transitions between the three phases 1202, 1204, and 1206. As explained above, a phase transition is the point in time at which the director, such as director 1102, changes the phase clock, and the phase for all software components in the phase domain changes simultaneously.
Alerting or notifying software components, within the phase domain, of the current phase or the transition to a new phase may occur. In one embodiment, the director notifies all software components of the phase. In other embodiments, a requesting method asks the director for the phase. In embodiments of the present invention, a transition notification is sent to one or more software components within the phase domain. In one embodiment, transition notifications occur either during the current phase or at the beginning of the next phase. In other embodiments, separate phases are employed for the notification process. For example, phase space 1200 would have three notifying phases positioned at the transitions 1208a, 1208b, and 1208c that are for notifying software components within the phase domain.
An exemplary sub-phase space 1300 of the Update phase 1204 is shown in
In another embodiment, both the Agreement phase and the Commit or Abort phase are sub-phases of the “Commit or Abort” sub-phase 1304 and sub-phase space 1300 has a Mark phase 1302 instead of an Agreement sub-phase 1302. Here, the data changes are made in the Commit or Abort phase 1304, and all software components using the data are marked for update in the Mark phase 1302. Marking a software component is setting a flag in the software component that signals the software component to retrieve the updated data in an appropriate later phase. In one embodiment, the marked software components retrieve the data in a Revalidate phase 1206. In another embodiment, the Mark phase 1302 has two sub-phases: a mark sub-phase and a final mark sub-phase. Here, the software components using the data are marked in the mark sub-phase and retrieve the data in the final mark sub-phase.
Another exemplary sub-phase space 1400, occurring during the Revalidate phase 1206, is shown in
An exemplary embodiment of a module reconfiguration is shown in
In one embodiment, the Plug sub-phase 1402 has further sub-phases. A Construct sub-phase 1410 creates new software instances by instantiating a known class, calling a software component, or using an interface on an existing instance to acquire a clone or specialized derived instance. A Configure sub-phase 1412 adds or removes connections between instances. Finally, an Initialize sub-phase 1414 sets properties and requires negotiation between properly connected instances. The sub-phases in the Plug sub-phase 1402 may deviate from those presented here. In addition, the Play sub-phase 1404 may also contain sub-phases.
Other phasing spaces are contemplated. For example, a sub-phase space for user interface changes is contemplated. In the user interface sub-phase space, an invalidate sub-phase can allow the execution of methods for building structures. A draw sub-phase then draws the built structures. Other phase spaces can be used for other types of operations, as one skilled in the art will recognize. In addition, one skilled in the art will recognize that the exemplary phase spaces presented above may be changed as to the number of phases or sub-phases, to the number of tiers or levels, and to the type of phases or sub-phases. As such, the present invention is extensible. In one embodiment, new superphases are overlaid on existing phase spaces. In another embodiment, new phases are added to existing phase spaces. In still other embodiments, more sub-phases or new tiers of sub-phase spaces are added to existing phase spaces.
An exemplary embodiment of a data structure 1500 having a phase constraint constraining the execution of an item of the code is shown in
In embodiments of the present invention, a data structure contains a form of constraint that depends on the software component and the type of operation performed. In one embodiment, the constraint is a call constraint. The call constraint constrains the call of a method to a specified phase. Thus, the execution of methods in other software components or the same software components is constrained by limiting the initiation of those methods only during specified phases. In another embodiment, the constraint is a constructor constraint. A constructor is a special form of method that instantiates software components. Thus, the instantiation of software components is constrained to a specified phase, such as explained with the Construct sub-phase 1410 in
The constraint is expressed by a phase constraint attribute in the software code that can be assigned to any target software component. In embodiments of the present invention, a phase constraint attribute is assigned to an entire class and is inheritable. Thus, child components inherit constraints from their parent components. In some embodiments, a phasing scheme places multiple phase constraint attributes on the same target. Thus, the software target is constrained by the conjunction of the multiple phase constraints.
Each constraint is a constraint on a “Type” associated with the level of the phase specified. As such, a constraint specifying a superphase is a constraint on “Superphase.” A constraint specifying a sub-phase is a constraint on “Sub-phase.” Constraints on types that are sub-phases are constraints on the union of the all the constraints on the “Super-Types.” The relationships between constraints on types is used by compilers or used at runtime to check for the validity of the constraint relationships amongst different software components.
Enforcing the constraints may occur at runtime or at compile time. At compile time, the constraints on types can be checked. A compiler can check the constraints on Types and constraints on sub-Types against a set of soundness rules for methods with a constraint on a Type calling to methods with a constraint on a sub-Type. A constraint scheme is valid if the constraint on a sub-Type is the same or weaker than the constraint on a Type, for example, if the constraint on a Type specifies the Plug phase 1402 and the constraint on a sub-Type specifies the Initialize sub-phase 1414. In this embodiment, the Initialize sub-phase constraint 1414 executes within the Plug sub-phase 1402, and therefore, is a weaker constraint. A constraint scheme is invalid if the constraint on a sub-Type is mutually disjoint with the constraint on a Type, for example, if the constraint on a Type specifies a Play sub-phase 1404 and the constraint on a sub-Type specifies the opposed Plug sub-phase 1402. A constraint scheme is valid, but must undergo some dynamic checking, if the constraint on a sub-Type is stronger than or overlapping with the constraint on a Type, for example, if the constraint on a Type specifies the Plug sub-phase 1402 and the constraint on a sub-Type specifies the Initialize sub-phase 1414. In this embodiment, if the phase domain currently operates in both the Plug sub-phase 1402 and the Initialize sub-phase 1414, the call scheme is valid. However, if the domain is not within one of the two phases, the scheme is invalid. Other soundness rules are contemplated and incorporated into the present invention.
An exemplary embodiment of a method 1600 for operating a computer environment within a multi-tiered phasing domain is shown in
Determine operation 1604 determines if any software components, such as component 1104, are constrained to the first of the master phases. If a software component is constrained to the first phase, an execute operation 1606 executes the software component during the first phase. If there are no software components to execute or during the execution of the software components, determine operation 1608 determines if there is a sub-phase space, such as sub-phase space 1300, that occurs during the first phase. If there are no sub-phase spaces occurring during the first phase, the process proceeds through connector 1 to a transition operation 1622 shown in
If there is a sub-phase space that occurs during the first phase, an identify operation 1610 identifies the sub-phase space and the applicable sub-phases. In one embodiment, a sub-director, such as sub-director 1106, is initiated and registers with the master director controlling the master phase space. The sub-director begins a sub-phase logical clock to cycle through the sub-phases in the sub-phase space. Determine operation 1612 determines if there are any software components, such as component 1112, constrained to the current sub-phase. In one embodiment, the software components constrained to the sub-phase space register with the sub-director. Thus, a nested sub-phase domain is created under the master phase domain. If there are software components in the sub-phase domain constrained to the current sub-phase, execute operation 1614 executes those software components during the current sub-phase. Determine operation 1616 determines if there are further sub-phase spaces, such as sub-phases 1410, 1412, and 1414, that occur within the current sub-phase. If there are further sub-phases, the process returns to identify operation 1610 to identify the further sub-phases.
If there are no further sub-phase spaces to identify, determine operation 1618 determines if there are any other sub-phases left to occur in the current sub-phase space. If there is another sub-phase to occur in the current sub-phase space, transition operation 1620 transitions to the next sub-phase in the sub-phase space. In one embodiment, the sub-director waits until all threads in the current sub-phase are executed and then transitions to the next sub-phase. Then, the process proceeds to the determine operation 1612 again. If there are no sub-phases left in the current sub-phase space, then determine operation 1618 determines if there is another superphase to transition to in any superphase space. If there is another superphase, transition operation 1620 transitions to the next superphase. The process (determining the sub-phases within a superphase; executing software components within the sub-phases; transitioning to the next sub-phase until all sub-phases are complete; and then transitioning to the next superphase) repeats until all sub-phase spaces are cycled through and a transition to a next master phase is required. Once the sub-phase loop ends for the first master phase, the process proceeds through connector 1 to transition operation 1622 shown in
Transition operation 1622 transitions to a next master phase, such as the Update phase 1204. In one embodiment, the master director waits for all threads executing in the first phase to end. Then, the master director changes the logical phase clock to the next phase. In some embodiments, the master director follows the transition rules outlined above with reference to
Determine operation 1624 determines if any software components are constrained to the current master phase. If there are software components constrained to the next master phase, execute operation 1626 executes the software components. In embodiments, the software components have already registered with the master director. The software components continue to check with the master director for the current phase. When the phase transitions and the master director reports that the domain is now in the next master phase, the software components, constrained to the next master phase, begin to execute.
If there are no software components constrained to the next master phase or during the execution of the constrained software components, determine operation 1628 determines if there are any sub-phase spaces within the current master phase. If there are sub-phase spaces, identify operation 1630 identifies the sub-phase space and transitions to a first sub-phase. Determine operation 1632 determines if any software components are constrained to the current sub-phase. If there are software components constrained to the current sub-phase, execute operation 1634 executes the software components
If there are no software components constrained to the current sub-phase or during the execution of those software components, a determine operation 1636 determines if there are further sub-phase spaces within the current sub-phase. If there are further sub-phase spaces, the process returns to identify operation 1630. If there are no further sub-phase spaces within the current sub-phase, determine operation 1638 determines if there is a next sub-phase in the current sub-phase space or a next superphase in the superphase space. If there is a next sub-phase or superphase, transition operation 1640 transitions to the next sub-phase or superphase. If there is not a next sub-phase or superphase under the current master phase, then determine operation 1642 determines if there is a next master phase, such as the Revalidate phase 1206. If there is a next master phase, the process returns to the transition operation 1622. If there is not another master phase in the master phase space, the process returns through connector 2 to the transition operation 1602 and starts the phase cycle over by transitioning to the first phase.
To further explain the present invention, an exemplary computer system operating within a multi-tiered phasing domain is described below with reference to
The computer environment 1700 operates under a phase space. For purposes of explanation and not limitation, the entire system 1700 operates under the phase space 1200 shown in
A master director 1102 changes the phase in the computer system domain 1700 to an Update phase 1204. Here, the commands 1716 and 1718 are sent to the Selection State module 1706 and the Address book 1708, respectively. A user interface sub-phase space for changes to the user interface, as explained above with reference to
A master director 1102 transitions to the Revalidate phase 1206 after all the invalidating and drawing constrained operations have been completed in the user interface sub-phase space. A data retrieve operation is executed during the Revalidation phase 1206. The data retrieve operation, in one embodiment, is constrained to a data retrieve sub-phase space. The data retrieve sub-phase space has two sub-phases, mark and final mark, as explained with reference to
Upon completing all operations constrained to the Revalidate phase 1206, the master director 1102 transitions back to the Request phase 1202. Here, the user interface again accepts commands and requests from the user input devices. A user inputs a change to the contact information within the detailed view 1704. For example, the user changes the address of a contact. A command 1720 is sent from the detailed view 1704 to the address book 1704 to change the data. In addition, a command 1722 is sent to the selection state 1706 to update the views of the master view 1702 and the detailed view 1704. In this embodiment, the commands would be queued, and the master director 1102 would transition to an Update phase 1204. The Update phase 1204 initiates a data write sub-phase space, such as the sub-phase space 1300.
In an Agreement sub-phase 1302, the address book sends a data change request to the several data stores 1710, 1712, and 1714. One or more of the data stores may contain a copy of the data changed in the detailed view 1704. Thus, each data store that has the data must agree to change the data. Thus, a voting procedure occurs during the Agreement sub-phase 1302. If all data stores agree to commit the change, the agreement is sent back to the address book 1708. The sub-director changes the phase to the Commit or Abort phase 1304. Here, the data change is sent to the data stores and is used to update the data.
Meanwhile, during the Update phase 1204, a user interface sub-phase space occurs. The selection state 1706 invalidates sections of the master view 1702 and the detailed view 1704 containing old data during an invalidate sub-phase. In a draw sub-phase, the master view 1702 and the detailed view 1704 are redrawn holding spaces for the changed data. Upon all sub-phases completing in the Update phase 1204, the phase domain 1700 transitions to a Revalidate phase 1206.
In the Revalidate phase 1206, further sub-phase spaces include sub-phases for Mark and Final Mark. Transitioning into the Mark phase, the address book 1708 marks the master view 1702 and the detailed view 1704 as requiring the changed data. In the Final Mark sub-phase, the changed data is written into the master view 1702 and the detailed view 1704. These changes can occur in a very short amount of time and in a fine granularity. For instance, the phases are cycled after every character entered by the user and in minute fractions of a second. Thus, the changes would appear to happen instantaneously.
The present example demonstrates how multi-tiered phasing constrains on the execution of methods within a system. If the commands and the changes happened without phasing, it is possible that the master view 1702 and the detailed view 1704 may be updated before all the data store have changed the data. Therefore, the user may view mixed results in the detailed view 1704 or the master view 1702 depending on the order of the methods attempting to change the data and to update the user interface views.
Multithreading with Concurrency Domains
As briefly described above, applications and/or application components may be constrained to certain phases or certain phase domains. According to embodiments of the present invention components of a software application may be partitioned into separate domains for providing synchronization and thread isolation within partitions of components and for providing for improved concurrent operations between partitions of components.
According to one embodiment, the same thread need not be used as the internal thread 1802 throughout the life of the concurrency domain 1801. Rather, when no objects need to execute on the internal thread 1802, the thread serving as the internal thread 1802 may return to a thread pool (not shown). When a thread is once again needed, a new thread may be pulled from the thread pool to act as the internal thread 1802. According to another embodiment, one of the single-threaded objects 1803 has thread affinity, meaning that the single-threaded object 1803 needs to execute on the same thread. In this embodiment, the same thread serves as the internal thread 1802 throughout the life of the concurrency domain 1801. According to one embodiment, secondary threads 1804, discussed in more detail herein, are also allocated from the thread pool.
Referring still to
The objects 1803 within a concurrency domain 1801 are isolated from the rest of the secondary threads 1804 and external objects 1810 in the program. Secondary threads 1804 do not execute any single-threaded object 1803 included within the concurrency domain 1801. Each external object 1810 is configured for execution on one or more of the secondary threads 1804. External objects 1810 asynchronously communicate with the single-threaded objects 1803 within a concurrency domain 1801. Communication includes the passing of data between objects or the invocation of one object's methods (e.g., or tasks) by another object.
Asynchronous communication across concurrency domain 1801 boundaries is achieved through the use of boundary objects 1807. Each concurrency domain 1801 is associated with one or more boundary objects 1807. These boundary objects 1807 can be viewed as a membrane or gated wall enclosing the concurrency domain 1801. Examples of boundary objects 1807 include data connectors and objects that implement custom protocols between concurrency domains 1801 or between a concurrency domain 1801 and an external object 1810.
Single-threaded objects 1803 within the concurrency domain 1801 use one or more boundary objects 1807 to asynchronously communicate with the external objects 1810. The single-threaded objects 1803 communicate with the boundary object 1807 using the internal thread 1802. The boundary object 1807 then communicates with the external object 1810 using one or more secondary threads 1804. The boundary object 1807 thereby passes information and invocations across the boundaries of the concurrency domain 1801. According to another embodiment, one boundary object 1807 communicates with another boundary object 1807 using the secondary thread 1804 before passing information to the external object 1810.
The boundary object 1807 acts as an interface between the internal thread 1802 of the concurrency domain 1801 and each of the secondary threads 1804. According to one embodiment, the boundary object 1807 receives an inbound communication from an external object 1810 using a secondary thread 1804 and filters the communication to the appropriate internal objects 1803. The filtering method will be described in more detail herein. According to another embodiment, the boundary object 1807 receives an outbound communication from an internal object 1803 using the internal thread 1802 and transmits the communication to the appropriate external objects 1810 using a secondary thread 1804. According to one embodiment, boundary objects may call out to external objects on an internal thread, but the boundary objects doing so are under constraints. That is, allowing boundary objects to call out to external objects must not cause unbounded delays or deadlocks by doing so. Another constraint prevents external objects from holding references to internal objects that prevents direct reentrancy of the concurrency domain under the control of external objects.
Synchronous communication occurs when the thread on which a first object is executing enters a second object to execute a method of the second object. External objects 1810 do not synchronously communicate with the single-threaded objects 1803 within the concurrency domain 1801. Accordingly, a secondary thread executing an external object 1810 does not directly call into or enter a single-threaded object 1803 within the concurrency domain 1801.
According to one embodiment, the external object 1910 includes a multithreaded object 1905 configured for execution on two or more secondary threads 1904. One portion 1905A of the multithreaded object 1905 is shown executing on one secondary thread 1904A and another portion 1905B of the multithreaded object 1905 is shown executing on another secondary thread 1904B. According to another embodiment, the external object 1910 includes a plurality of multithreaded objects 1905 or a single-threaded object (not shown) configured for execution on one secondary thread 1904.
The concurrency domain 1901 in the system 1900 maintains a work queue 1908. The work queue 1908 is a multi-element data structure on which tasks (e.g., invocations of methods of internal, single-threaded objects 1903, data updates, and other executable methods) are posted (e.g., inserted) and from which tasks are removed. According to one embodiment, tasks are removed from the work queue 1908 only in the same order in which they were posted; that is, according to a first in, first out constraint. According to another embodiment, tasks posted to the work queue 1908 are assigned a priority and each task is removed according to its priority.
Incoming communications are posted to the work queue 1908 by the boundary object 1907. These posted communications form work items 1911, which are requests (e.g., invocations or calls) for the execution of tasks of an internal, single-threaded object 1903 or a boundary object 1907 associated with the concurrency domain 1901. The request that forms the work item 1911 can be communicated to the boundary object 1907 by an external object 1910 or by another boundary object 1907. For example, in
According to one embodiment, to conserve resources when preparing to post a new task to the work queue 1908, the boundary object 1907 checks the work queue 1908 and determines whether any of the queued work items 1911 include related tasks. If there are related tasks, the boundary object 1907 can selectively bundle the new task with a previously queued related task as a subtask rather than posting the new task as an entirely new work item 1911.
Referring still to
According to one embodiment, posting a work item 1911 to the work queue 1908 does not force the dispatcher 1909 to act. Rather, execution of work items 1911 is deferred to a point in time dictated by a top-level cycle logic of the concurrency domain 1901. Once the work item 1911 is posted to the work queue 1908, the internal thread 1902 executes the requested task in the next appropriate cycle of the concurrency domain 1901 as determined by the dispatcher 1909. Accordingly, external objects 1910 do not determine when a work item 1911 is removed and hence when a task of an internal, single-threaded object 1903 is invoked and executed. External objects 1910 also do not determine when boundary objects 1907 execute tasks on the internal thread 1902 of the concurrency domain 1901.
Once a task is dispatched and completed, the out-bound result is passed to the boundary object 1907 as a callback. The boundary object 1907 then communicates the callback to the external object 1910 that originally posted the work item 1911 that invoked the task that achieved the result. Examples of callbacks include data flags indicating the task is complete, method calls, and the like.
The boundary object 2003 then communicates 2004 with a work queue 2005. This communication 2004 generally includes posting a work item (not shown) to the work queue 2005. The work queue 2005 then communicates 2006 with a dispatcher 2007. This communication 2006 generally includes the dispatcher 2007 sequentially dispatching each work item posted on the work queue 2005. Finally, the dispatcher 2007 communicates 2008 with the internal, single-threaded object 2009 whose task is being invoked. This communication 2008 generally includes the invocation of the task of the internal, single-threaded object 2009. In another embodiment, the external object 2001 is communicating with another boundary object (not shown) of the concurrency domain.
The asynchronous communication across concurrency domain boundaries, described above with reference to
Referring back to
Furthermore, if the internal thread 1902 is allowed to execute the task of the external object 1910 and if the execution of that task included an invocation of another task of the internal, single-threaded object 1903, the internal thread 1902 would not be allowed to reenter the concurrency domain 1901. Rather, the internal thread 1902 would enter a boundary object 1907 of the concurrency domain 1901 to invoke the task responsible for posting work items 1911. Alternatively, as described above, under certain constraints, boundary objects may call out to external objects on an internal thread for invocation of task. After invocation of the task, the internal thread 1902 would return to executing the task of the external object 1910 and subsequently return to executing the first, original task of the internal, single-threaded object 1903. In other words, the internal thread 1902 would not execute the invocation of the second task by the external object 1910 until execution of the first task is complete and until directed to do so by the dispatcher 1909 of the concurrency domain 1901.
Referring now to
The paths of execution of the internal thread 2102 and the secondary thread 2104 are shown in both figures. In
Referring still to
Method 2202 also leads to method 2211, which is executed simultaneously with operation 2202. In method 2211, the boundary object 2107 invokes a task associated with the data source 2112. The invocation is performed on one of the secondary threads 2104. Next, the method proceeds to operation 2212 in which the task of the data source 2112 is executed on one or more of the secondary threads 2104. Next, operation 2213 includes the database 2112 transmitting the result of the execution back to the boundary object 2107 as a callback. The transmission of the result occurs on one or more of the secondary threads 2104. Then, in operation 2214, the boundary object 2107 posts the callback to the work queue 2108 as a work item 2111. The post is executed on one or more of the secondary threads 2104.
From operation 2214, the method proceeds to operation 2204. Operation 2203 also leads into operation 2204. Operation 2204 occurs when the dispatcher 2109, which was sequentially executing the work items 2111 in the work queue 2108 in operation 2203, reaches the callback work item 2111 added by the boundary object 2107 in operation 2214. The dispatcher 2109 dispatches the callback using the internal thread 2102. Once the callback has been dispatched, the dispatcher 2109 continues to sequentially dispatch each work item 2111 in the work queue 2108 in operation 2205. The method ends at 2206.
Referring now to
Referring still to
From operation 2402, the method proceeds to both operations 2403 and 2412. In operation 2403, the dispatcher 2309 of the first concurrency domain 2301 sequentially dispatches each work item 2311 on the work queue 2308. In operation 2412, the boundary object 2307 of the first concurrency domain 2301 uses one or more of the secondary threads 2304 to communicate with the boundary object 2327 of the second concurrency domain 2321. The communication includes the request to invoke the task. Then, in operation 2413 the second boundary object 2327 posts the requested task to the work queue 2328 as a work item 2331. The post is executed using one or more of the secondary threads 2304.
Both operations 2413 and operations 2422 lead to operation 2423. In operation 2423, the dispatcher 2329 reaches and dispatches the work item 2331 including the requested task. The dispatch is executed on the internal thread 2322 of the second concurrency domain 2321. Then, the task is executed as a callback to the single-threaded object 2303 in the first concurrency domain 2301 in operation 2424. At this point, the method again splits, proceeding to both operations 2425 and 2414. In operation 2425, the dispatcher 2329 continues to sequentially dispatch each work item 2331 on the work queue 2328.
Operation 2414 occurs concurrently with operation 2425. In operation 2414, the boundary object 2327 of the second concurrency domain 2321 uses one or more secondary threads 2304 to request the boundary object 2307 of the first concurrency domain 2301 to post the callback to the work queue 2308 as a work item 2311. Next, in operation 2415, the boundary object 2307 posts the call back to the work queue 2308. The post is executed on one or more of the secondary threads 2304.
Operation 2404 occurs when the dispatcher 2309 of the first concurrency domain 2301 reaches the callback posted on the work queue 2308. The dispatcher 2309 uses the internal thread 2302 of the first concurrency domain 2301 to dispatch the callback. The callback executes in operation 2405. Next, the method proceeds to operation 2406 in which the dispatcher 2309 continues to sequence through the work queue 2308, dispatching each work item 2311 in order. The method ends at 2406.
Another example of a system (not shown) includes three or more concurrency domains interfacing with each other and with other external objects. Such a system would function substantially according to the same operations as described herein. Each concurrency domain in the system would include an internal thread, one or more single-threaded objects, and a dispatcher. Each concurrency domain would be associated with at least one boundary object and would maintain a work queue. All communications across the boundaries of the concurrency domains would be asynchronous (e.g., filtered through the respective boundary objects, work queues, and dispatchers).
Application Description Language
As briefly described above, embodiments of the present invention include an application description (XAD) and an application description (XAD) engine for providing declarative rules and descriptions of all applications and components included within an application framework, such as the application 304, described above with reference to
The use of such a declarative language is different from the use of an imperative language by declaratively modeling an application's data, processing and conversion of the data, user interface (UI) and the way the UI interacts with the data. Embodiments of the present invention relate to a particular declarative language used to characterize XML application framework (XAF) applications. The declarative language is sometimes referred to herein as XAF Application Definition (XAD). Although much of the implementation details described herein refer to the use of XML, those skilled in the art will recognize that other syntaxes could be used to implement aspects of the present invention.
As may be appreciated, in order to provide such separation between the declarative application file and the configuration of instantiated objects that perform the data processing, in an embodiment, a platform may be used, such as a platform that builds and executes applications. In accordance with aspects of the present invention, the platform used is referred to herein as the XAF platform, which is a platform for building and executing such software applications. In particular, XAF also uses XML to uniformly represent data, and employs a highly componentized approach to building each application. In an embodiment, XAF leverages XML to provide powerful data transformation capabilities and rich data. Both external data and application state can be accessed and manipulated using XML's capabilities. User interfaces and complex data flows may be defined in XAD.
As stated, XAD is a declarative language used to author XAF applications.
In the exemplary embodiments herein, the application description 2502 is expressed as a set of declarative constructs, such as well-formed XML. The application description 2502, therefore describes how one or more objects 2510, 2512 and 2514 should be configured within the executing application. In essence, the application description 2502 is used to compute the configuration of the application but not the results of the application, which is done by objects 2510, 2512 and 2514. Since the objects 2510, 2512 and 2514 may be used by other applications written in non-declarative languages, XAD improves application development by reuse of such objects. Further, XAD enables the dynamic creation and modification of applications in ways not possible using traditional, static declarative application definitions as discussed in more detail below.
With respect to the data 2504, 2506 and 2508 shown in
The objects 2510, 2512 and 2514 represent well-factored components, such as viewers, editors, transformers, etc. The application description can call “pre-programmed” components and/or add new components to the set.
Objects have well-defined types, so the corresponding XAD tags and parameters have corresponding types in the XAD language. XAD is strongly typed and the XAD compiler detects errors statically. Tags and parameters have return types and cardinality constraints. Additionally, data-types support several additional constraints: schema, conversion requirements, and accessor availability (accessors are strongly typed object facades for XML data).
XAD not only allows existing applications to be modified or augmented simply by altering one or more text files, but also allows developers to constrain modifications and add-ins (modules that may be added to existing applications, roughly analogous to browser plug-ins) to preserve security, and user model integrity as discussed in more detail below.
Sample XAD Application
In order to understand many of the nuances of XAD and in particular an XAD application description, such as application description 2502, a sample description will be discussed. Table 1 illustrates the source code of a sample XAD application in accordance with one embodiment of the invention. Sample XAD application in Table 1 is relatively simple and many other, complex XAD are contemplated. Indeed, the sample shown in Table 1 has a sole display a “Hello World” message. <Application>
As may be appreciated, the sample shown in Table 1 comprises several XML tags in order to describe the resulting objects that process data. Line 1 identifies the version of XML and Line 2 identifies the system tag for XAD application description, which is described in more detail below. Lines 3-4 identify some namespaces of tags available in XAF. The primary portion of the sample in Table 1 for the purpose of this discussion is the “Application” tag shown in Line 8. The application tag is an example of an entity tag that causes the creation of an application object at runtime that orchestrates the application, and controls the highest levels of application functionality. Similarly, the “Window” tag in line 9 (another example of an entity tag) causes the creation of a window object, i.e., a user interface entity that is responsible for the display of an application's frame. In this case “Window” is a parameter of “Application.” Finally, a “TextBlock” entity, the resulting text object created as a result of the text tag in line 10, is a user interface entity created at runtime that is responsible for the display of text (in this case, “Hello World”). The use of tags to create objects (via builders) is described in much more detail below. The foregoing example is provided simply to show a more simple sample XAD application.
In operation, the XAD application shown in Table 1 is compiled and executed to create the application, window and text objects, as generally shown in
Initially, a developer authors the application description 2602. In an embodiment, the first tag of the XAD file 2602 is “<sys:XAD>”. Also, in order to operate correctly, the computer system can have access to XAF or any suitable equivalent, such as by installing the XAF SDK for example. Upon issuing an app launch command, the XAD 2602 is automatically verified and, in one embodiment, compiled to a more efficient form. A compiler or XAF engine 2604 parses the XAD 2602 and executes the compiled form which instantiates the necessary objects. Also, the engine 2604 creates and connects the graph of object components (also referred to as entities) 2612, 2614, 2616, 2618 together to create the application 2606 that ultimately processes the data. Meanwhile the engine 2604 connects, i.e., binds, the various objects with the necessary data 2608. Engine 2604 may also be considered a configuration service in that it instantiates the objects and connects the objects together to create the application 2606.
The application 2606 is a graph of connected components 2612, 2614, 2616, 2618. For instance, the application 2606 may have one or more data connector entities 2612. The data connector entities 2612 represent the objects that connect data to the system and communicate with the various data stores available. Another type of entity relates to the UI connectors 2614. The UI connectors 2614 represent the objects that actually process user interface features, and in fact, communicate with the user interface 2620. Also included in the application 2606 are transforms 2616 that typically provide some transformation of data prior to relaying information to the UI connectors, and actions 2618 that process data to perform some type of action such as modifying data. More details of the process that creates the necessary components and the process of binding objects to the data are described below.
As may be appreciated from this discussion, the tags used in the XAD 2602 are not used to refer to specific objects, but instead, XAD tags refer to object factories. An object factory is a description that XAD uses in conjunction with its tag builder class to build one or more entities using parameters specified in the tag. The engine 2604 provides the actual creation of such objects. In an embodiment, the engine has access to a component library 2622 that provides some details of existing and/or reusable tags and objects.
Consequently, applications 2602 are defined using tags, allowing modular construction of applications.
With these general principles in mind, some of the specifics of XAD will be discussed. For more details of the language, a specification is attached as Appendix A which is expressly incorporated into this specification.
Objects and Factories
Objects are instantiated through factories, which may be builder components, or the XAD engine customized with fragments of XAD. Factories permit deferred instantiation for purposes such as lazy evaluation or virtualization (saving the cost of objects that need not be instantiated immediately). This is achieved through parameters that have the NewScope=“True” attribute. Together with conditional instantiation through the sys:Switch construct, factories enable dynamic data-dependent application configuration. Use of dependency management and revalidation in the XAF platform enables the XAD engine to automatically reconfigure the application as needed when data changes. This eliminates the need for imperative reconfiguration code which is common in conventionally written applications, and which is costly and error-prone. The XAD engine enables factories to be defined in the XAD language, and constructs the corresponding factory implementation automatically.
A variety of object types may be instantiated using framework tags in XAD, such as objects 2510, 2512 and 2514 shown in
Another type of object that may be instantiated using a framework tag in XAD is a user interface object. User interface objects provide functionality relating to control display, and to input handling. A user interface object can be used to trigger an action (see actions, discussed below).
Yet another type of object that may be instantiated using a framework tag in XAD is an action. Actions provide “hooks” (functions attached to a user interface element) to handle for user events. For example, clicking on a button on the user interface will activate an action object associated with the button.
When user event occurs, the corresponding action can be used to store and pass event details. Event details may be stored in main memory, on a hard drive, or other memory media. Event details are passed using “scope” variables, which are discussed below. For example, if the user event was a mouse click, it may make sense to store the x and y positions of the mouse at the time of the event. Other details that may be stored when a user event occurs include, but are not limited to, which key was pressed, which index of a list of data elements was selected at the time of the event, which row and/or column of a table were selected at the time of the event, which rows and/or columns of a table were visible to the user at the time of the event, etc. Through the use of these events, application functionality is realized.
Entity Tags
As may be appreciated from the example shown in Table 1 above, XAD applications comprise the use of tags. In an embodiment, the one kind of tag is the “entity tag.” Other tags, such as grouping or selector tags are discussed in more detail below. Entity tags map to common language runtime (CLR) objects at runtime. The only restriction on the CLR object is that it must (in an embodiment) be a XAF entity. To be a XAF entity, a CLR object derives from one of two base classes, e.g., “Microsoft.Xaf.Core.BaseTypes.EntityElement” or “Microsoft.Xaf.Core.BaseTypes.EntityComposite”, and adheres to certain protocols and contracts defined by XAF.
The builder 2704 (
Exemplary entity tags are shown in Table 1 above, in lines 8, 9 and 10. For instance, in line 8, an application entity tag is shown. As stated above, this tag causes the instantiation of an entity that orchestrates the application, and controls the highest levels of application functionality. A window entity tag (line 9 of Table 1) causes the creation of a user interface entity that is responsible for the display of an application's frame. Finally, a text entity tag (line 10 of Table 1) causes the creation of a user interface entity responsible for the display of text (in this case, “Hello World”). As may be appreciated by those skilled in the art, many other entity types may be defined and used.
Before using an entity tag, such as the application entity tag, window entity tag or text entity tag shown in Table 1, each entity tag is defined, e.g., within the XAF platform. Indeed, every entity tag has a corresponding tag definition. Every tag definition contains enough information to eventually map every entity tag to an entity. Entity tags are defined by declarations of the form, for instance, the “Text” tag definition begins as follows “<sys:TagDefinition Name=“Text” . . . > . . . ” and ends with the following: “</sys:TagDefinition>”.
Included in the definition is the “name” for the new tag, e.g., “Text” and the “type” for the particular entity tag. Entity tags may be classified into various types and all entity tags have a type. These types are an extensible set and really correspond to the CLR type of the entity associated with a tag. Some types used by the XAF framework include “sys:Data”, “sys:EventHandler”, “fwk:UIElement” and “sys:Selector”. Whether the tag is added to the framework or the system typically depends on whether the tag draws on internal engine interfaces and/or versions at the same rate as the core language. If so, then the tag should be added to the system otherwise it is added to the framework. The type is declared in its definition. Continuing with the Text definition provided above, the tag definition can be expanded shown below in Table 2.
In Table 2, the tag definition indicates that a new entity tag named Text should be created of type fwk:UIElement (a framework tag indicating a UI Element type of entity). The Text tag may subsequently be used anywhere a fwk:UIElement entity may be invoked. An example of the Text entity tag can be seen in the HelloWorld.xad example appearing in Table 1.
Parameterization
Typically, each entity tag in a XAD application description has some parameters to describe how to create, connect, or configure the resulting object or entity. A parameter may be a reference to a piece of data, or a reference to a factory or object. For instance, a scope attribute may be associated with a parameter to indicate whether the parameter refers to a factory or an object. If a parameter requires a certain type of object to be applicable, the parameter may be marked with that type for type consistency checking. Such type checking ensures that execution of an application will not result in a type mismatch, which could cause an incorrect computation.
XAD allows many different ways of declaring the parameters for an entity. A common form is shown below in Table 3.
Initially, “Param” is a reserved name within embodiments of XAD and is used to define parameters. Also, as indicated by the ellipses in Table 3, one or more other parameters may also be added to a Text tag definition. In this example, a new entity of type UI element named Text should be created. Further, a single parameter named “FontFamily” of type “Data” (and more specifically “String Data”) should be created. The resulting Text tag can be used with a static content string, e.g., “<sys:Text FontFamily=“Arial” . . . />” which will cause the text to be in Arial font. Importantly, tag definition parameters correspond to an attribute within the actual instance of the tag that has the same name as the parameter.
Alternatively, a tag may be defined to take parameters that are other entities, e.g., “<Text FontSize=“25”> <foo:RandomWordGenerator Param=“Text”/> </Text>”. In an embodiment, the above definition will cause the Text tag's Text attribute to use the value returned by the foo:RandomWordGenerator function object, and render the resulting value on the screen in a twenty-five-point font. In another embodiment, this definition could have been written as: “<Text FontSize=“25”> <Text.Text> <foo:RandomWordGenerator /> </Text.Text> </Text>”. The latter definition will instantiate a Text entity and define its Text attribute as the value of foo:RandomWordGenerator.
In an embodiment, parameters themselves may contain a plurality of its own entities. Table 4 illustrates example code portions setting parameters.
The code shown in Table 4 (Row 1) defines the UI Element named FlowPanel and has several parameters, thus causing the display of a flow of UI elements. The set of UI elements to be displayed can be specified via a single parameter named “Children” which are defined in Table 4 (Row 2). That is, Row 2 illustrates the application description calling/describing the flow panel defined in Row 1.
One default parameter may be specified for each entity in XAD. Table 5 illustrates an example line of code defining a default parameter, whose name is “Children”. If no parameter name is specified on a child tag, the child will be associated with the parameter whose DefaultParam attribute is set to “true”.
Data Binding
In an embodiment, XAD leverages data binding (the binding of XML data to an object designed especially for the data). The name of an object may therefore be used to reference data more specifically when specifying parameters. Unlike static values (which, by definition, remain constant for the lifetime of an application), a data bound value can be updated through modification of the data bound value's associated data. In one embodiment, this modification may be performed by an application. In another embodiment, this modification may be performed by a user or an external process. In an exemplary embodiment, the line “<Text Text=“$TextValue” FontSize=“25” />” causes the Text parameter to be assigned a value of “$TextValue”. In this embodiment, “$” is a reserved character and use of this character in this manner invokes data binding during the creation and connecting of the components. Such data binding causes the indirect pointing or associating of data from another source or object to be bound to this entity tag.
Data binding may be used to refer to literal data. That is, literal data may be specified using the sys:InlineData tag. For example, “<sys:InlineData Name=“TextValue”>Hello World</sys:InlineData>” defines a piece of inline data, namely the string “Hello World.” This exemplary literal data tag is given the name TextValue. Consequently, by using the name $TextValue in a XAD description, the XAD code can bind to the data within the literal data tag, and use it for the value of any parameter of type sys:Data.
Data binding may also be used to refer to a relative data location. For example, <Text Text=“$SomeLocation/myd:Employees/myd:Employee/myd:Name” /> uses a dynamic prefix, i.e., “$SomeLocation” followed by relative data references, i.e., “myd:Employees”, “myd:Employee”, and “myd:Name” without using the “$” character. The result is a data binding to the absolute location formed from the value of $SomeLocation appended with the given relative location.
Additionally, data binding may be used in conjunction with grouping tags.
For example, the sample code in Table 6 (Row 1) binds the MergeTreesTransform entity to the set of entities under the Resources grouping tag.
Table 6 (Row 2) illustrates that XAD allows binding actions to data values. For example, the following definition binds actions to the value of $ScratchData. Under the definition, the initial value of $ScratchData is “Bye.” “MouseLeftButtonDown” is a parameter of type “sys:EventHandler.” The action corresponding to the “MouseLeftButtonDown” parameter is invoked whenever the Text entity encounters a Left Mouse Button Down event. This event will cause the SetTextValueAction to be invoked, which will change the text value of the <Greeting> element to “Caio.” Likewise, a Right Mouse Button Down event will result in the value of $ScratchData changing to “Au Revoir.”
In yet another example, data binding allows an application to receive command line parameters. For example, the sample code in Table 7 allows specification of command line parameters. The sys:Main tag is the entry point for applications.
The application defined in Table 7 accepts command line parameters such as “xad CommandLine.xad /TextData=Testing” where CommandLine.xad is the name of the application, TextData is the parameter name, and Testing is the value. In an embodiment, a plurality of parameters can be passed into XAD in the form of “xad ApplicationName /ParamName1=Value1 /ParamName2=Value2 . . . /ParamNamen=Valuen”. One skilled in the art will appreciate that while the exemplary embodiment above uses a string as an input value, command line parameters may alternatively or additionally include alphanumeric values, names of files, image or other resource locations, etc. Also, a command line parameter may be given a default value for use when no command line parameter value was specified for the parameter.
Data Source Tag
In an embodiment, external data is exposed to an XAD application by way of a DataSource tag. Further, the XmlFileDataSource tag may be used to establish a connection between an application and an XML file so that the latter may be used as a data source by the application. External data sources may be used in a fashion similar to inline data provided by sys:InlineData tags. External data sources, however, may be saved out to disk using XAD actions, and thus changes to data by other users can be consumed by the application if desired.
Derived Tag Definition
XAD allows the use of derived tag definitions to leverage abstraction and reuse of definitions. For example, in an application that requires a series of twelve point Verdana font headings, a user might place a definition such as “<Text FontFamily=“Verdana” FontSize=“12” Text=“heading text here” />” at each position where a heading is desired. However, if many hundreds or thousands of headings are needed, each one being potentially different, derived tag definitions allow for greater efficiency in specification. More specifically, a derived tag definition can be defined and an example is shown in Table 8.
The definition shown in Table 8 will cause XAD to create a new tag named TitleText of type fwk:UIElement. The Text parameter within the tag is bound to the value passed in for $Text, which can be passed in differently for each instantiation of the heading.
As another example, consider the following example shown in Table 9, wherein the first row is a sample snippet of XAD, the second row illustrates a sample derived tag definition to replace the snippet of XAD and the third row illustrates an instantiation call using the derived tag definition.
As is shown in Table 9, derived tags can increase the functionality and simplicity in calling or instantiating the same items multiple times.
Scope and New Scope Parameters
Scope parameters provide defined boundaries for certain sub-graphs of objects of varying lifetimes, in particular for delayed and/or repeated instantiation of those entities. Factories are the means of deferring and later invoking instantiation. Scope variables provide values that are parameters provided by the entity that invokes the instantiation. One example of use of scope variables is to allow for the receipt of different event properties, such as mouse button down events.
XAD uses new scope parameters to pass data between application components, between objects, and/or between combinations thereof. In an embodiment, new scope parameters may be used by a parent component to pass parameters to newly created child components. XAD thus allows users to leverage code reuse. In one embodiment, a list entity is constructed which includes references to the data in each list entry. The same parameter that defines what to do with a single list entry is applied to each new list entry, allowing arbitrarily large list of entries to be created using a single parameter or set of parameters. Further, parameter passing between parent and child objects may likewise be optimized. For example, a parent object can pass list indices to a child object, said list indices corresponding to list elements that were passed to the child, via new scope parameters.
Parameters that introduce a new scope have the “NewScope” attribute set to true. For example, the line “<sys:Param Name=” . . . ” NewScope=“true” . . . />” is typically used to set the new attribute to true. New scope parameters may be used to introduce one or more variables within the scope. The variables are parameters that are declared implicitly. Variables are declared along with the declaration of the new scope parameter, as can be seen in the example provided in Table 10 (Row 1). As discussed above, new scope is introduced for any entity that is to be instantiated multiple times. Table 10 (Row 2) illustrates a definition of such a factory:
In the above example (Table 10, Row 2), the scope introduced by the RepeatedChild parameter has a scope variable named DataContext. This variable is set to the item corresponding to the sub-view to be instantiated. The above definition would be applied as shown in Table 11 (Row 1), assuming List.xml contains the XML shown in Table 11 (Row 2).
In the example shown in Table 11, for each list item, the FlowPanel entity will create a corresponding Text entity. The resulting Flowpanel entity would display a list of the Employees A-H on the monitor. One skilled in the art will appreciate that each line corresponds to an instantiation of a Text entity. What varies from one instantiation to the next is the value of the “DataContext” scope variable. The resulting pattern is referred to as a repeater pattern. It is the basis for iteration over datasets in XAD. In an embodiment, repeater patterns may be nested for multidimensional iteration.
Attached Parameters
Attached parameters are parameters associated by a parent object with a child, that the child may not know how to process. For example, a “triangle” object that is a child of a parent “drawing” object may not specifically know how to deal with “position” coordinates, since position coordinates relate to general location within the drawing and not specifically to shaped such as triangles. Attached parameters allow parent objects to access data and/or functionality associated with a child object. XAD checks attached parameters for correct usage based on the definition of said parameters in the parent object. Table 12 (row 1) illustrates an exemplary tag definition.
The XAD code definition in Table 12 (row 1) defines an entity called DockPanel, which includes an attached parameter called “Dock” which may be referred to as “DockPanel.Dock”. Table 12 (row 2) illustrates exemplary XAD code using the DockPanel entity.
As shown in Table 12, DockPanel.Dock is specified as an attribute to each of the Text entities, even though DockPanel.Dock is not necessarily one of the parameters specified in the tag definitions for the Text entity. Since DockPanel.Dock is an attached parameter, it is used in conjunction with a Text entity, but is actually defined in the declaration of the DockPanel entity shown in Table 12 (row 1).
Resource Parameters
Another type of parameter is the resource parameter. Resource parameters provide a place to store objects that do not have a logical position in the object hierarchy. Resource parameters can be used to anchor sets of arbitrary entities under a given entity. Actions and state associated with resource parameters may be used throughout the application. An exemplary definition of resource parameters is shown in Table 13.
UI elements may be bound to state and/or actions via XAD's data binding capability, discussed above.
Structures
In XAD, a “structure” is a group of entities that may be passed as one name. Structures in XAD may also be nested. An example of structures and nested structures is shown in Table 14.
In this example, the $ActionsAndData.Actions.Print and $ActionsAndData.Actions.Save bindings would be valid for all parameters of type sys:EventHandler. Similarly, $ActionsAndData.RecentDocuments, $ActionsAndData.Preferences.User, and $ActionsAndData.Preferences.Enterprise would be valid for all parameters of type sys:Data. $ActionsAndData would be valid for all parameters of Signature app:ActionsAndDataSignature. $ActionsAndData.Actions would be valid for all parameters of Signature app:ActionsSignature. Finally, $ActionsAndData.Preferences would be valid for all parameters of Signature app:PreferencesSignatures.
Manifests
In an embodiment, XAD uses “Manifests” to bring together the parts of an application, such as an application description. A manifest is a specification capable of being processed by a computer that describes aspects of the application, e.g., which files are needed by the application, what dependencies exist between objects, and what user-environment-specific data is part of the application (such as user interface element text in the user's native language, icons, etc., in contrast to data which the application operates upon).
Built-In Tags
Built-in tags are tags that are pre-defined and provide significant capabilities to XAD developers. Conventionally built-in tags are prefixed with “sys:”. For instance, these built-in tags may allow conditional operations, repeating operations, and the creation of objects. One such type of built-in tag is referred to as a “dynamic entity” tag, which may be used to change, within a data description, what types of objects are used. This allows real-time modification of the configuration of an object on the screen. For example, in response to a user selecting a given user interface element, the function and/or purpose of the element may be altered in real time. For example, one output window may be used to display output from arbitrarily many different application views based on receipt of a “cycle to the next application view” user command. The dynamic entity tag, in an embodiment, provides for dynamic entity instantiation by tag name as shown in the following Table 15:
In the example shown in Table 15, the built-in entity “Dynamic Entity” provides a developer the ability to dynamically define an entity. In this case, $DateTag resolves to a local name.
Another type of built-in tag called a “switch entity” tag allows applications to be dynamically reconfigured by way of conditional construction of an object or of multiple, related objects. While XAD is a declarative language, it is still expressive enough to allow conditional construction of entities. A single construct is provided to enable conditional construction: Switch. The switch entity tag can be parameterized by “cases” or “conditions” tags. Cases can be thought of as similar to a C# switch, while Conditions are similar to a set of C# if-else clauses. In an embodiment, the use of Cases and Conditions is mutually exclusive, meaning that a Switch cannot mix Cases and Conditions within its body. In addition, in this embodiment, Cases may only be used when a Switch specifies a value for its Data param, otherwise Conditions must be used.
Using Switch allows an author to conditionally return 1 or many entities as the result of evaluating the switch. Table 16 illustrates some sample XAD data (row 1), sample uses of conditional switching (rows 2 and 3).
As may be appreciated, row 2 of Table 16 shows the syntax, in an embodiment, to create entities based on values greater than 50.00 dollars. Row 3 of Table 16 furthers this example of showing the text block entities for other conditional tests. The resultant view would display all the elements items in row 1, i.e., “Toaster,” “VCR,” “DVD Player,” “Receiver,” and “Sub-woofer.” However, “Toaster” would be in Georgia bold font and in green, where “Receiver” would be in Arial bold font and in red.
Switch together with dependency tracking and revalidation enables automatic implementation of dynamic reconfiguration. Dynamic reconfiguration is discussed in more detail below.
Framework Tags
XAD and XAF allow for another type of tag referred to as the framework tag (abbreviated “fwk”). A framework tag may be used to create an object that is part of an application. Framework tags may be used to reference actions, data sources, transformations, and association managers (each discussed below). In an embodiment, application debugging may also be performed via framework tags.
At the most simplistic level, UI connectors provide the ability to visualize XAF data and to respond to events that modify that data. Any one particular UI connector actually owns an “element tree,” or set of visual elements that may be displayed, that the UI connector is presenting. Associated with each UI connector is code to translate data into visual elements and code to translate events into editing operations.
In an embodiment, the UI connectors simply expose properties and events of a client platform, such as the Windows® Presentation Foundation (WPF) elements. Moreover, a UI connector provides a specific form of data-binding, binding data from the XAF Data Layer to the WPF element. Different elements are designed to visualize different types of data. For example, an Image element visualizes bitmapped display data, a Text element visualizes string data, and a ListBox visualizes an ordered list of possibly heterogeneous data. Some elements, like ListBox and TextBox, also can visualize a ‘selection’ on that data. The concept of a selection is basically another data input.
In general, specific UI connectors are designed to work with a specific type of data to be visualized. If two elements use exactly the same mechanism to accept the input (or parameters) for what they visualize then a developer can write a single UI connector to handle both of them. The parameters are used to give the UI connector a factory object, which is used to create children as required. The UI connector typically provides a data context to the factory. The factory is normally a fragment of XAD, which is wrapped in an object by the XAD engine. Table 17 illustrates a set of UI connectors that may be provided.
One type of framework tag, the association manager, may be used to associate an object with a data structure or element of a data structure. For example, a list association manager may be used to associate an object with an item in a list. The resulting list item will be dynamic, in that it will contain a reference the object and its associated data versus a static value.
In Operation
Create operation 2802 creates an object. In one embodiment, create operation 2802 creates data structures necessary for the object. In another embodiment, create operation 2802 also allocates memory as needed for the data structures.
Once create operation 2802 has created an object, connect operation 2804 connects the object to other objects according to the relationship of that object to others. In one embodiment, a graph of connections between the created object and one or more preexisting objects is built. The create-connect-initialize sequence enables the XAD engine to support cyclic connections among a group of objects. In the case of data and transforms, such cycles are prohibited, and the XAD compiler performs some checking to prevent such cycles.
A wiring or connecting graph produced by connect operation 2804 defines interconnections and interrelationships between objects. In an embodiment, these connections will remain until altered by dynamic reconfiguration (discussed below), or until the object is erased. Which connections are necessary is determined by the engine as it executes the application markup code. In an embodiment, connect operation 2804 connects data provider objects to the application.
Initialize operation 2806 initializes the created objects. In an embodiment, initialization involves setting one or more nondeterministic values to zero, or other predefined value. In another embodiment, initialize operation 2806 calls constructors associated with objects. In yet another embodiment, initialize operation 2806 instructs created objects to initialize themselves. One skilled in the art will appreciate that initialize operation 2806 may also perform other operations typically associated with object initialization without departing from the scope of the invention. In an embodiment, initialize operation 2806 occurs conditionally based on whether a given object contains data to be initialized. In an embodiment, objects may be initialized with different values based on conditionals. For example, the following definition shown in Table 16 would cause XAD to display items in a list in red if they are priced above $500, and display items in green if priced below $100.
In one embodiment, the engine performs the operations 2802, 2804, and 2806 using just-in-time compilation. In another embodiment, the application is precompiled. In yet another embodiment, a combination of pre-compilation and just-in-time compilation is used to achieve a balance of performance and flexibility.
XAD allows applications to be dynamically reconfigured. Dynamic reconfiguration may be triggered by a change in data flowing into an application, or modification of data existing with an application.
A listener is a computer agent, which monitors an application or group of applications for certain events. The XAD listener monitors certain registered objects' data within a XAD application for a change in value. Registration operation 2902 registers objects and their data with such a listener.
When a change in a data value occurs within a registered object, a change notification is received from the listener by receive operation 2904. Mark operation 2906 then identifies one or more data values that depend on the changed value, and marks them as affected by the change. Analyze operation 2908 analyzes the effects of the changed value on each of the marked data values. Based on the results of this analysis, determine operation 2910 determines whether any objects were rendered invalid by the changed value.
If one or more objects were rendered invalid by the change, flow branches YES to remove operation 2912, which removes the one or more now-invalid objects. Flow then branches to determine operation 2914. If determine operation 2910 determined that no objects were rendered invalid by the change, flow branches, flow branches NO to determine operation 2914.
Determine operation 2914 determines whether any new objects are required as a result of the changed value. If one or more new objects are required, flow branches YES to creation operation 2802 (
To simplify creation of new objects and deletion of invalid objects, XAD is capable of lifetime management of objects. A lifetime domain is a unit of reconfiguration in a XAD application. Objects whose life spans are tied to each other (for example, a first object may have no purpose without the existence of a second object, and the second object may have no purpose without the existence of the first object) are placed into the same lifetime domain. There can be many objects in a given lifetime domain. When the objects are no longer needed, all objects in a lifetime domain can be easily deleted by referencing the lifetime domain for deletion (versus potentially referencing thousands upon thousands of individual objects in a lifetime domain for deletion). One skilled in the art will appreciate that objects within a given lifetime domain need not be functionally related; rather, they need only share a creation time and a codependence such that they can be destroyed at the same time.
In an embodiment, XAD can leverage conditionally created objects in conjunction with dynamic reconfiguration of applications. Objects of varying types can be therefore be created, depending on the nature of the data value change.
Determine operation 3004 determines if a tag contains any parameters that must be instantiated. If no parameter tags are present, flow branches NO to operation 3008. If parameter tags are present, flow branches YES to instantiate operation 3006.
Instantiate operation 3006 instantiates the parameters found by determine operation 3004. As discussed above, parameters are used to describe how to create, connect, or configure a factory or object. As described earlier, parameters may include one or more properties of a factory or object, a reference to a piece of data, or a reference to another factory or object. Properties may be either static or dynamic.
Process operation 3010 processes a tag according to its data type. Several different kinds of tags exist, and thus several possible approaches are implemented by process operation 3010. A primitive tag (a tag that actually corresponds to an object) is processed by creating a new object. In an embodiment, all objects created correspond to entities.
Some tags define scope boundaries. Such tags cause process operation 3010 to create a “factory,” which may be invoked at a later time for on-demand instantiation of entities. When a factory is processed, a new scope domain is created. Factories can be checked by entities for dependency information, which can be useful during dynamic reconfiguration.
Process operation 3010 processes a switch tag by creating a switch factory. This is a factory that supports branching of application control based on various user-specified conditions. The factory may be used to monitor various conditions which, when met, will trigger a branch in application flow. In an embodiment, just-in-time compilation is used to modify objects when necessitated by a tag being reprocessed as part of dynamic reconfiguration.
Process operation 3010 may also process other types of tags. For example, a binding tag triggers process operation 3010 to create a factory to indicate that a certain parameter value is actually a reference to some other entity. This is roughly analogous to a pointer in a procedural language. A dynamic entity tag allows late binding of an entity to be created. In one embodiment, the late binding is deferred until application runtime.
After process operation 3010 has processed a tag, flow returns to determine operation 3002.
In an embodiment, add-ins may be added to an XAD application. Add-ins are self-contained software packages which can extend the basic feature set of a given application without requiring the application to be upgraded. Add-ins allow new features or third-party extensions to be added incrementally as they become available, without requiring users to go through a potentially expensive upgrade cycle.
XAD add-ins are packaged in assemblies. In an embodiment, assemblies are loaded through static dependencies specified in the manifest (discussed above). In one embodiment, XAD applications communicate with add-ins via an extensibility protocol. An extensibility protocol may be defined in a separate assembly so that the protocol can be published and references by both the application and third parties developing the extension.
An extensibility protocol has two parts. The first part is the metadata schema. Metadata is data that describes data. In the context of XAD, metadata can be used to describe the features implemented by an add-in. The metadata may include tag names, parameters to be passed, human-readable text for the UI, etc. The extensible application defines the metadata schema (what to expect from the add-in), while the add-in provides the content. The second part of an extensibility protocol is the tag signature (discussed above). An add-in implements its features by defining tags. An application uses these tags to access the add-in features. The application defines the tag signatures (what type of entities the tags should result in, what parameters the tags should take, etc.), while the add-in provides the tag definitions (the tag name and the actual implementation). Tag signatures are therefore roughly analogous to interfaces, while tag definitions are roughly analogous to interface implementations.
In an embodiment, XAD supports a global metadata tree accessible via a Metadata tag. A metadata tree is a union of all metadata constituent files from all dynamically loaded assemblies. At runtime, the metadata tree is automatically updated whenever an assembly with metadata constituent files is loaded or unloaded. Proper change notifications are sent so the application can update itself dynamically (as discussed previously in conjunction with
In an embodiment, an add-in defines tags for accessing the features the add-in implements. Since the add-ins may not be available at the time an application is written, it may not be possible for the application author to know and use the tags defined by an add-in. The author may only know, and thus may only statically define, the tag signatures. The actual tag names and eventual parameter values can be provided by the add-in using metadata. Since the metadata scheme is part of the application-defined extensibility protocol, metadata scheme information is accessible to the application. The application can therefore read the metadata, and dynamically instantiate an entity by tag name. XAD supports a DynamicEntity tag for this purpose.
Although the present invention has been described in language specific to structural features, methodological acts, and computer readable media containing such acts, it is to be understood that the present invention defined in the appended claims is not necessarily limited to the specific structure, acts, or media described. One skilled in the art will recognize other embodiments or improvements that are within the scope and spirit of the present invention. Therefore, the specific structure, acts, or media are disclosed as exemplary embodiments of implementing the claimed invention. The invention is defined by the appended claims.
This patent application claims priority to U.S. Patent Application Ser. No. 60/676,137 filed Apr. 29, 2005 and entitled “Application Description Language,” and U.S. Patent Application Ser. No. 60/703,220 filed Jul. 28, 2005 and entitled “Markup Language Based Application Framework with Application Description Language, Concurrency Domains, Application Framework Phasing Models, and Application Framework Transaction Transforms”, the disclosures of which are expressly incorporated herein, in their entirety, by reference. This patent application is also related to and filed concurrently with U.S. patent application Ser. No. ______, entitled “Multithreading with Concurrency Domains,” bearing attorney docket number 14917.0258USU1/MS310186.01; U.S. patent application Ser. No. ______, entitled “Application Framework Phasing Model,” bearing attorney docket number 14917.0258USU2/MS310224.01; U.S. patent application Ser. No. ______, entitled “Application Description Language,” bearing attorney docket number 14917.0258USU4/MS309853.01; and U.S. patent application Ser. No. ______, entitled “Transaction Transforms,” bearing attorney docket number 14917.0258USU5/MS311262.01; U.S. patent application Ser. No. ______, entitled “XML Application Framework”, bearing attorney docket number 14917.0258USU6/MS316537.01; U.S. patent application Ser. No. ______, entitled “XML Application Framework”, bearing attorney docket number 14917.0258USU7/MS316538.01; U.S. patent application Ser. No. ______, entitled “XML Application Framework”, bearing attorney docket number 14917.0258USU8/MS316539.01; which are assigned to the same assignee as the present application and expressly incorporated herein, in their entirety, by reference.
Number | Date | Country | |
---|---|---|---|
60676137 | Apr 2005 | US | |
60703220 | Jul 2005 | US |