94. Copy (COPY) a stream relationship used in interaction, the target object becomes a copy of the source object, and then the two are independent of each other. See the becomes. Semantic replication relationships are a stream relationship that represents an object to be born another object in interaction. It represents the activities of making a copy. After the copy stream is executed, the value of the two independent objects can be independent. The replication conversion in interaction can be serial numbers, indicating that it is related to other activities. The representation of the replication stream is represented by a dotted arrow from the original object to the newly generated object. The arrow is marked with a structure type keyword "copy", but also serial numbers. The replication conversion can appear in a cooperative diagram, sequence diagram, and active diagram. Examples Explanations 13-83 show that a file creates a backup on another node. First, replication ("Copy"), then move the copy to the next node ("become") position. Figure 13-83 Copying and Become stream 95. Creating an object or other instance (such as the use case instance) implementation and initialization. Antonyms: destroyed. See instantiation. Semantic creation objects are the results of instantiation object messages. Creating actions can have parameters that will be used for initialization of new instances. After creating actions, new objects comply with their constraints and can receive messages. Creating an operation or constructor can be declared as an operation of a class scope. The goal of such operations (at least in the overview) is the class itself. The programming language is as implemented as objects in SmallTalk, the class, and runtime, so the regular message of the creation operation and class is as implemented. In C , there is no actual runtime object, and the creation operation can be considered a conceptual message optimized at runtime. The calculation of the instantiation class is excluded in the method of C . Otherwise, each method can be modeled to pass the message to the class. The initial value expression of class properties (on concept) is calculated by the creation operation, and its result is used for initialization of attributes. The code for creating an operation can be implicitly replaced, so the initial value expression is an overloaded default. In the state machine, the parameters of the constructor of the creation of objects are used as the default events on the conversion of the initial state of the top layer. The representation is in the class diagram, the creation operation (constructor) is used as an operation containing the table's operational table. It can have a parameter table, and the return value is an instance of the class, which can be omitted. As the operational operation, its name is underlined (Figure 13-84). Represents the "constructor" structure. Figure 13-84 Creating the execution of the creation operation in the operation sequence diagram indicates that the arrow points to the object symbol (with a rectangle with the underscore object name). The object symbol is issued is the lifeline of the object. (If the object is active, you can be a dotted or double solid line). It continues until the object is destroyed or the end (Figure 13-85). Figure 13-85 Creating Sequence Figure 13-86 Creating actions in the Create Collaboration in the Collaboration Diagram In the Create Operation in Figure 13-86 Creating actions in the figure in the figure in the figure in the figure. The first message passed to this object instance is implicit to create a message of the object. Although the message actually points to the class itself, it is usually omitted. The message is used for the initialization of instances (new renegan, but not initialization). As shown in Figure 13-86. See a collaborative diagram, sequence diagram, where there is a representation created in a process implementation. 96. Current Event triggers events running to the completion step in the execution of the state machine. See Run to Completion, State Machine, Transition Semantics For convenience, the state machine can turn several connected conversion signals into an event response.
In addition to all transformations outside the final conversion into the pseudo-state - in order to construct the dummy state, it does not wait for external events. In principle, all messages can be collected on a conversion, however, with a pseudo-state division multiplex state allows multiple transformations to share ordinary successive events. The execution of the conversion chain is atomic - ie, is a portion that cannot be run to the completion step by an external event. In such a column of conversion, the activity and monitoring conditions of the message attachment have the first conversion of the first conversion of the trigger and the implicit entry of its parameters. This event is the current event in the conversion. The type of current event can be identified by polymorphism or conditional statement. Once the exact type knows, you can access its parameters. The current event is particularly helpful to help the initial conversion of the newly created object get the initialization parameters. After the new object is created, the creation event becomes the current event, and its parameters can be used by the initial conversion of the new object state machine. Examples Figure 13-87 is a conversion from an IDLE state to a PURCHASE state, which is triggered by the Request event. PURCHASE's entry activity calls the setup operation, which uses the current event. The program can access the current event to get the Request event and its parameter product. If there is a current event multiple bindings, the program requires a conditional statement to get the correct trigger event. The syntax is described by the programming language description. Represents the naming of the current event, can be completed in an expression with keyword currenteentevent. Specific expression languages provide a more detailed grammar. Figure 13-87 Using the current event 97. Data type (DATA TYPE) There is no identifier of a set of values (independent existence, there may be side effects). Data types include raw predefined types and user-defined types. Original type right: numbers, strings, and passengers. User-defined types are enumerated types. The anonymous data type used for implementation in the program language can be defined by the language type. See Classifier, Identity. Semantic data types are the predefined foundations required for the user to define the type. They are semantically defined outside of language structure. Numbers are predefined, including real and integers. Strings are also predefined. These types are not defined by the user. The enumeration type is a user-defined group of naming elements, and there is a predefined order between them and there are no other properties. Enumeration Types have names and a series of enumerated values. The enumeration type Boolean is predefined, and its enumeration value is false and true. Operation can be defined on the data type, and the parameters of the operation can have a data type. Because the data type has no identifier, only simple values, the operation on the data type does not modify the data type; they directly return the value. Because there is no identifier, it is meaningless to create a new data type. The value of all data types (conceptually) is predefined. The operation on the data type is a request for the system, which does not change the system status, only one value returns. Data types can also be defined by language type - data type expressions in programming languages. Such expressions specify a named data type with the target language. For example, expression of Person * (*) is a data type expression in C , which is not a simple data type. 98. Data Value is an instance of the data type and is a value without an identifier. See Data Type, object (Object). Semantic data is just a number of mathematics - a pure value. Because the data value has no identifier, the same data value as the two representations cannot distinguish. In programming languages, data values are delivered. If the data value is not meaningful, the transformation data value is meaningless, and its value is never changed. In fact, the data value is the value itself. The usual conversion data value refers to the content that transforms a variable with a data value, making it a new data value. The data value itself is constant. 99. Default Value Automatically provides values as part of some programming languages or tools.
The default value of the element properties is not an integral part of the UML semantic, and does not appear in the model. 100. Deferred Event When the object is in a particular state, an event that is delayed by the authenticated time. See the state machine, transition. A series of events can be specified as latency. When the object is in a state of delaying an event, if the event occurs, it will not trigger the corresponding conversion, ie the event does not have timely effect. Until the object is transferred to a state that does not require the event delay, the event is valid. Other events appear in this state activity. After the object enters a new state, no longer appears one by one, and triggering the conversion in a new state (the order of the previously delayed event is unprinted, the specific appearance order of the dependency is dangerous). If the event does not trigger transition in a new state, it will be ignored and lost. In a regular state machine, you should carefully use the delay event. They can be more direct to be constructed as a parallel status, during which, the main operating part can be operated. Used in the active diagram, you can guarantee the order in which the asynchronous information is not lost. If there is a conversion that is triggered in the state, the conversion will be overloaded. The representation delay event is expressed as internal conversion with a special retention action delay. Delayed use for status and its nested status. (Fig. 13-88) 101. Delegation an object response ability to the information sent by another object. Authorization can be used as an option for inheritance. In some languages (such as Self), it is supported by the inheritance structure of the language. In most other languages (such as C and SmallTalk), it can be implemented by aggregation or association with another object. The operation of the first object requires the operation of the second object to continue its work. Contrast: inherits. See Association. Figure 13-88 Delay event 102. Dependency relationship between two elements, one of the elements (servers) will affect another element (customer), or provide the information you need to it (customer) . It is a simplicity of different model relationships. Semantic dependencies are statements that represent two elements between one or several models. It can combine several different elements. It seems that there is a biological use of "invertebrate" in combination with all the creatures that are not "vertebrate". When an asymmetric knowledge system is expressed, the independent element becomes a service, and an independent element is a customer. To illustrate the role of dependencies in the model, it can have a name. However, dependencies themselves are usually enough to explain the problem without additional names. Dependencies can have a structure type to establish an accurate attribute of dependencies. Dependencies can also have a text description for a detailed description of it. There is a dependency between the two packs, indicating that there is a dependency between at least one pair of elements in the two packages. For example, the dependence between the two classes can be expressed as two dependencies between the packages containing them. There is a dependency between two packages that cannot be explained that all elements in the two packages have dependencies - in fact, this situation is very rare. See the package. Dependencies can include references to subordinate dependencies. For example, the dependency between the two packages can reference the dependencies between two classes. Dependency does not have to be passed. Note that the association and generalization relationship meet the general definition of dependencies, but they have their own model representations and meanings, usually not considered as a dependency. Realization relationships have independent representations, but is considered a dependency.
There are several different dependencies: abstract, binding, combination, permission, use (usage). Abstract dependencies represent different abstractions. Two elements represent the same concept with different signs. Usually an element is more abstract, another more actual; it may also be a different representation of the same abstraction level. From not specific to specific order, abstraction includes structural type, description, refinement (keyword refine), realization (by independent expression), derived (keyword Derive). Binding. Binding dependencies will be connected by a template constraint element domain template. The requirements for template parameters are used as a list of bound attached. Permission (Permission). Licensing (usually as a specific constructor) allows the package or class to allow it to use some of the packages or classes of some content. The system type of license-dependent relationship has access, friend, input. Use (USAGE). Use dependencies (keyword "USE") to connect customer elements with the server elements. The changes in the server will lead to changes in customers. Use usually represent an implementation dependency, one of which relies on the service of another element to achieve its own operation. The constructed type used includes calling, instance (keyword "instantiate"), parameter, transmission. This is an open list, there are other dependence on different programming languages. Indicates the dashed arrow between the two model elements. Model elements (customers) at the end of arrows depends on model elements (servers) at the arrow. On the arrow, you can have a keyword that depends on the type of relationship, and you can have a name (Figure 13-89). Some dependencies between classes in Figure 13-89 There are several other relationships that also use a virtual line arrow with keywords, but they are not dependent. These elements have stream (becoming and replicated), combined (extension, and included), constraints, and comments. If the comment or constraint is one of the elements, the arrow can be omitted because constraints or comments are always at the end of the arrow. Standard Elements, Binding (Bind), Call, Create, Derive, Extend, Include, Import, Friends (INSTANCEO), Instantiate, PowerType, Send, Track, Usage (USE). 103. Configuration (Deployment) Describes the development steps of the configuration of the real world environment operation system. In this step, you must decide configuring parameters, implement, resource allocation, distribution, and parallelism. The result of this step is reflected in the expansion map and structure file. See the development process, modeling step (deployment diagram) represents a view of the runtime process node structure, component instance, and its object structure. The component represents the performance of the code unit at runtime. The components that do not act as runtime do not appear in the deployment diagram, but in the component diagram. Deployment diagram performance example; component diagram expression component type definition. See Components, Interface, Node (NODE). The semantic deployment diagram contains a node instance that is connected with the communication chain. Node instances include instances of runtime, such as components instances and objects. Components instances and objects can also contain objects. The model can represent the dependencies between the instance and their interfaces, and can also express the movement between the node or other containers. The deployment diagram has a descriptor form and an instance form. Example forms (previously described) expressed the location of a specific component instance as part of a part of the system structure. This is a common form of deployment diagrams. Descriptor form An explanation of which component can be present on which node can be connected, similar to class diagram.
Representation Figure 13-90 The deployment diagram deployment diagram of the client - server system is a mesh map of node symbols and path representing communication. (Figure 13-90). Node symbols can have components instances, and the means exist or run on the node. The component symbol can be object, indicating that the object is part of the component. The components are connected to the dashed arrow (possibly passing through the interface), indicating that a component uses another component. The dependency can be explained using the structure type when necessary. The deployment diagram is similar to the object map, usually used to represent an example of each node in the system. I rarely use the deployment diagram to define the relationship between nodes and nodes. The movement between the components between the nodes and the movement between the components can be represented by a dotted arrow with keyword "become". In this case, the member or object is only in a respective node or component within a portion of time. The objects in the deployment diagram of Figures 13-133 move between nodes. See "Deployment View) represents the node in the distributed system, the view of the objects in the components on each node, and the objects in the node. See Configuration (Deployment Diagram). 106. A relationship between two elements of derived (DeriVation) can calculate another element from one element. Detecting can be built into an abstract dependent structure with keyword derive. See the export element (Derived Element). 107. Derived Element can be calculated from other elements, which does not increase semantic information, existed for the convenience of system design. See constraint, dependency. Export elements in the semantic model are redundant because it can be calculated from one or several other elements. The formula for calculating the export element can be given as a constraint. There are several reasons for the use of export elements in the model. From the analysis level, the export element is not necessary. But it can be used as a macro definition or name a useful concept. It must be remembered that the export element does not add semantic information for the model. In the model of the design level, the export element represents a selection-model to avoid overhead of repeated calculations. For example, the calculated intermediate result or reference to a column value. Export elements are refreshed after using export elements, and exports elements depending on the elements depending on the elements. Figure 13-91 Derived attributes and derived association representations are derived before the element names (/) represent the export elements. Such as attributes, role names, or associated names (Figures 13-91). The specific method of calculating the export element can be explained with a dependent relationship with the structure type "Derive". Typically, the arrow depending on the relationship is removed from the constraint arrow of the element, and a constraint character is added to the edge of the export element. Discussing derived associations may be the most common export elements. It provides a virtual association that can be calculated from two or more basic associations. The derived association is WorksForCompany, which can be calculated by the WorksFordEctor and Employer combination. Using WorksForCompany in an example can avoid duplication, but this association does not provide any additional information. Associated generalization (Figure 13-30) is slightly different. It provides two different levels of details associated with the same association. It is usually not implemented at the same time, but only sub-association is achieved. Sometimes the parent is implemented and the sub-association is intended as a constraint. 108. Descendant refers to a child or an element containing a sub-relational chain, a transmission closure of a specific relationship. Antonym: Ancestor. 109. Descriptor (Descriptor) is used to describe model elements of a common attribute of a series of instances. Attributes include structural, relationships, behaviors, constraints, destination, and more. Comparison: instance. Semantic descriptors are model elements that describe a column instance. Most elements in the model are descriptors. This word has almost all contents of the model - class, association, status, use case, cooperation, etc. Sometimes "Type" indicates the same meaning.
However, the application range of usually "type" is narrow, just refers to elements similar to the class. The descriptor includes all the elements that say the effect. It has connotation and extension. The connotation is a structural description and conventional rules; epitaxial is an example in which the descriptor is explained. Extension is not necessarily physically accessible when running. One pair of models in the model is the descriptor - instance. The representation represents the relationship between the descriptors and examples, indicating that both the same set symbol, and painting a line at the name of the instance. The descriptor has a name; the instance has both its own name and the name of the descriptor. The two use the colon, the name is underlined. 110. A stage of design (Design) system, how it will be implemented from the logical level. In design, from strategic and tactics to determine how to meet functional needs and quality requirements. One step is reflected in design layer model, especially static view, state machine diagram, and interactive view. Comparison: analysis, design, implementation, and expansion. See the modeling step (Stages of Modeling), develop process. 111. Design Time refers to the (situation) that appears in the design activity of the software development process. Comparison: Analysis time. See Modeling Time, Stages Of Modeling 112. Destroy Destroy Destroy an object and reclaims its resources. It is usually a clear activity or other activities, constraints or waste recovery results. See Destruction 113. Destruction (Destruction) destroyed an object and recovered its resources. Eliminating a constituent object will result in eliminating its components. Eliminating an object without eliminating an object that is generally associated or aggregated, but the link containing the object will be canceled. See Composition, FINAL State, Instantiation. The representation of the destruction of the destruction of the process is implemented, see cooperation drawings and sequence diagrams (Figures 13-162). In the sequential diagram, the painting large × indicator is destroyed (Figure 13-92) in the sequence diagram. Its location is where the object is destroyed, or the object is self-ended. The information of the destroyed object can be "Destroy" with the structure type "DESTROY". In the cooperation diagram, the object on the object is used to indicate that the object is destroyed in the interaction. If the object is generated in interaction and is destroyed, you can use constraint {Transient}. 114. Development Process To produce a series of parallel ordered work in order to produce software with control, reusable way, as well as related guidance. The purpose of the software development process is to ensure the function and system reliability required for the completed system. See the modeling step (stages of model, discussion UML is a modeling language, not a process. Its purpose is to describe the model, and the model can be implemented in different development processes. To standardize, the results of describing development work are more important than explaining the development process. Because there are a lot of good modeling methods, and the final user of the model does not have to know its construction process. Therefore, UML supports multiple processes. For more information on interaction, use case trigger, technical process, see [Jacobson-99]. Modeling steps and development phases The modeling steps are suitable for iterative development processes, including initial, refinement, constructing, and transition. In the development of an application, these processes are processed, but each step contains one or more iterations. In one iteration, each model element travels according to the step from the analysis to the configuration. Each element has a corresponding step. Although the development phase and modeling steps are not synchronized, there is associated between the two. In the early development phase and early iterations in each stage, the early modeling steps were more emphasized. MUT 13-93 of Figure 13-93 During the development phase, Figures 13-93 show a contrast of the degree of weight in the subsequent step and iterative process.
In the early days, mainly focusing on the analysis; establishing a design and implementation element process model for designing and implementation; all elements are completed in the construction and transition. 115. Diagram Model Element Set Graphic Representation. Usually arc (association) and vertices (other model elements) are commonly connected to each other. UML support class diagram, object map, case example, sequential diagram, collaboration diagram, state diagram, activity diagram, component graph, and extended diagram. See Background Information, Font Usage, HyperLink, Keyword, Label, Package, PRH, representation element, feature Property List. The semantic map is not a semantic element, but a semantic model element. The meaning of the element is not affected by the representation. The figure is included in the package. The diagram and complex symbols of the representation of the method are composed of paths connected by the path. Information Topology is represented without relying on the size of the symbol, the location (there are some exceptions, such as the order of time axes). There are three important visual relationships: linkage (usually two-dimensional graphics); container (closed graphic containing two-dimensional symbol); visible accessories ("close to" another icon of another icon). They are the mapping of nodes and links in the representation of the language. UML drawing on two dimensional surface. Some content is a three-dimensional projection of a two-dimensional graphic, such as a pipe. But they are still considered a two-dimensional log. In the future, 3D Brocks may be applied to desktop computers, but there is still not popular. There are 4 charts in the UML representation: chart, two-dimensional graphics, path, string. The icon is a graphic that is fixed, the shape is fixed. It cannot be expanded and does not contain content. The icon can appear in other icons, the endpoint of the path, or as a stand-alone icon. For example, the aggregated symbol (diamond), navigation (arrow), ternary (beef eye), object destruction (large x) is an icon. Two-dimensional symbols have different sizes, which can extend and contain other content (such as strings or other symbols). Many of them have the same size or different points. The path ended in the symbol boundary indicates the path connected to the two-dimensional symbol. For example, a class symbol (rectangle), state (rounded rectangle), comment (with a rectangle with folding). The path is a line, or the curved symbols connected to the end point. In concept, the path is a single topology entity, but it can have multiple symbols indicating the message. The message cannot be separated from the path. There are many other graphics symbols (no flirtat) on both ends of the path. The path can have end points - the icon that limits the path to the path in the path-end order. For example, associated symbols (solid lines), generalized symbols (solid lines with triangular icons), dependent symbols (dotted lines) are paths. The string represents different information in a "unsatisfactory" form. UML assumes that each usage of the string can become a model information through the corresponding syntax analysis. Equisitions have their own syntax, such as attributes, operations and conversions. These syntax can be identified by tools. The string can be the content of the point, is the element in the list (when the location of the location is transmitted information), the label of the symbol or path, or the independent element in the figure. For example, the representation and constraints of the class name, the conversion label, and multiplex are string. 116. Direct Class's most complete explanation of an object class. Inheritance, multiple inheritance, multiple inheritance. Semantic an object can be an instance of multiple classes - if it is an instance of a class, it is an instance of the ancestors of such a class.
The direct class is the most complete and detailed explanation of the object. Objects are direct instances of direct classes, is an indirect instance of direct class ancestors. Objects cannot make an instance of the progeny (according to the definition). If the system allows multiple classifications, there is no direct class of the full description object. Objects can be a combined direct instance of multiple classes. As long as these subcates of these classes do not describe the same object, the object is a direct instance of all classes that contains it. There is no ancestral relationship between the direct columns of the object. If an object is generated in the instance of the class, the object is a direct instance of such a class. 117. Direct Instance Direct instance is an instance (such as an object) that is described in detail by a given class. Expressed as "Object O is a direct instance of class C". Here C is the direct class of O. See Direct Class. 118. Discriminator selects a child's pseudo attribute from a group in the generalization relationship. All subsequins a given nature of the dual-intranet, corresponding to other implied nature of the dual-all. It represents a special dimension. See Generalization, PowerType, PseudoAttribute. Semantics sometimes, model elements can be characterized based on different properties. Each nature represents a special independent orthogonal dimension. For example, the vehicle can be characterized according to the power (gasoline engine, rocket engine, wind, animal, human), or by the applicable range (terrestrial, water, aerial, outer space). The description is a name for explaining dimensions. An object can be used from multiple dimensions, but they must be expressed as specific examples. The generalization relationship can have a discrimination, which is a string representing the dimension of a child with a parent. All the specializations obtained with the same discriminant name to a doubt of parents constitute a group, each group is a separate special dimension. A complete set of discriminant names represents the dimension of the specialty parents. The example must be an instance of subclass of each special group. For example, vehicle must have both power and jurisdiction. Each discriminator represents a parental abstraction property, which is characterized by a child with a biopiarcharge. There are multiple dimensions with multiple dimensions, all dimensions must be specific elements. Therefore, the child in a discriminant group is abstract. Each of them is just a partial description of the parents, focusing on a certain nature and ignores others. For example, subclasses of vehicles focusing on power ignore jurisdiction. The specific element requires simultaneous specialization of all dimensions. It can be achieved by multiple inheritance of the specific element of each child to each dimension, or by multiple categories of the instance of children to each dimension. The description is still abstract before consisting of all the discriminant. For example, actual vehicles must be powered and jurisdiction. The water-powered water transport is a sailboat; there is no name, but this example is present in the fairy tale. The default discriminating identifier indicates "empty" discriminating, "empty" discriminating is still considered legal ("default" discriminating). This convention allows common non-judicial cases. If all the generalization paths are not discriminated, all the children are in the same discrimination; in other words, all the specialization belongs to a discrimination, which is the same semantic with the case where there is no discrimination. Structure Each Charge (generalized) arc has a discriminant string, which can be an empty string. The discriminant is its parent's pseudo property, which must be unique in its parent attributes and associated roles. The domain of pseudo attributes is a subclass set. The same discriminant name is allowed to appear multiple times in different children and parents, indicating that these children belong to the same part. The representation of the representation can be expressed as a text tag on the general arrow (Fig. 13-94). If the two generalized arcs with the same discrimination are shared, the discrimination can be marked on the arrow. Examples Examples 13-94 show a specialization of EMPLOYEE in two dimensions: Employees Status and Locality. Each dimension has the range of values represented by subclasses. However, the generated examples of modulus should be satisfied in both dimensions.
For example, LIAISON Class Supervisor and Expatriate. The description of the single dimension is abstract before two dimensions in multiple inheritance subclasses. Figure 13-94 Disjudier 119. Disjoint Substate An sub-state that cannot be existing in other sub-states in the same composition state. Contrast: concurrency state. See complex transformation, composite state 120. Distribution Unit is located in an operating system process or some objects or components that exist in the processor, which can be generated as runtime or aggregation . It is a design concept in the configuration view. 121. Dynamic classification A generalized semantic variable, which can transform types or roles. Contrast: static class. See multiple classification semantics in many programming languages, the objects cannot be changed after the object instantiation will change the class. This static class constraint simplifies the implementation of the compiler, but is not necessary to logically. For example, an object implemented as a circle under a static class is only a circle without stretching on an x-axis. Under the dynamic class, it can be stretched on one axis into an ellipse. Two types of elements can be used in the UML model. Although the difference is important to implement, different options have little effect on the model. The same class must be defined in two ways, but its support can be different in both cases. Dynamic Concurrency is an active state that represents multiple concurrent execution. See the Activity Graph. 122. Dynamic View The Description and Implementation of the Model Element Different Time The Static Structure in the Static View is different. The dynamic view is a composition term, which includes examples, state machine diagrams, active views, and interactive views. 123. The second step in the elaboration software development process, start the system design, development and testing the architecture. During this time, the architectural part of the analysis view and the design view is completed. If the executable prototype is constructed, the partial implementation view is also completed. See the Development Process 124. Element Compositions the atoms of the model. This book is said to be an element that can be used in the UML model - ie model elements that express semantics information and expression elements of model elements with graphics. See Figure (Diagram), the meaning of model elements is quite wide, and there is no specific semantic. All elements can have the following features: 125. Tagged value any elements can have 0 or more tag values. The label is the name of the numeric meaning. The label in the UML is not fixed, which can carry any information that is meaningful to the modeling or tool. Standard Elements Documentation126. Entry Action The action performed when it enters a state. See Exit Action, running to completion (State Machine), Transition. A state of semantics can be selected with an inlet action. No matter how to enter this state, the inlet action is performed before the activity of the external state or the activity of the internal state. The entrance movement cannot be crossed in any way. Whether it is independent state or nested state activities, the entry action must be performed first. Perform order. If a two states with an active conversion link include an inlet action and an outlet action, execution order is: Export activity of the source state and its external state (excluding the source status and external state of the target state), then execute Conversion activities. Finally, the entry activity is performed from the external state of the target state (outer priority) until the entry activity of the target state is performed. Figure 13-117 The conversion of Figure 13-11 has such an activity.
Indicates that the method is encoded according to the internal conversion grammatical code, with virtual event name Entry (entry is a reserved word, cannot be used as the name of the actual event). Entry / Action Sequence has an entry activity in each state, which can be a sequence so that it will not lose generality. Discussing portions and export activities are not required in semantics (entry activity can be subordinate to all entry states), but they implement the state package, so that the external use is separated from the internal structure. They allow definitions of initialization and final event, but don't worry that it will be slightly. For exceptions, it is especially useful because it defines the activity even when an exception occurs. The entrance activity helps to complete the initialization work that must be completed when entering a state. The state of the variable with accumulation information can be initialized. For example, support the user interface with a keypad to enter an account or a phone number, you want to clear the last number before entering. The entrance activity is also used to assign temporary storage spaces to the status. Usually, export activities are used with entrance activities. The entrance activity is allocated, and the export activity is released. Even if external conversion occurs, resources will be released. This is a good way to process exceptions and user errors. User-level errors will trigger advanced conversions and skip nested states, but the nested state can release space before losing control. 127. Enumeration enumeration is a data structure, which constitutes a literal value with name. Typically, the name of the enumeration name and its literal value is declared. See Classifier, Data Type Semantic Enumeration is a user-defined data type, which includes the ordered sequence of names and enumerated literal names. Each literal name is a value within the enumeration range - a predefined instance of this data type. For example, RGB color = {red, green, blue}. The Boolean data type is a predefined enumeration type with a literal value of True and False. The indication method is expressed as a rectangle. In the above points, there is a keyword "enum" on the enumeration name (Figure 13-95). The second point is a list of enumerated strip values. The third point (if present) is the operation on this type. They must be inquiry, so they don't have to declare anything. 128. Event (event) occupies an explanation of the appearance of a certain period of time and space. STATE MACHINE, Transition, Trigger Semantic Figure 13-95 Declaration of the Enumeration In the state machine, an event can trigger a state transition. The event is transmitted from the generator generator by the parameter table (which can be empty) to the recipient. The time that occurs is the implicit parameter of each event. Other parameters are defined in the event. The appearance of an event (instance) has a parameter (actual value) corresponding to each event parameter. The value of these parameters can be referenced by the action triggered by the event. There are four events: Call Event receives a request to activate an action. The desired result is that the recipient of the event triggers a conversion to perform the corresponding operation. The parameters of the event are operational pointers, the parameters of the operation and the return pointer. After the conversion is completed, the caller reclaims the control. Change Event meets the Boolean condition in an expression, no parameters. This event implies a continuous test for control conditions. When the condition is true when the condition is true, the event will happen. In an event, the time that the condition is satisfied is often related to the occurrence of other events, so it usually does not require a timing query. Signal Event Receives a signal that is a specific name-based entity that communicates between objects. The signal has a clear parameter list. It is given to another or some objects by an object. Broadcasting can be considered to give signals for all objects. But in practice, the effects of the two may vary. The sender clarifies the signal of the signal when signaling, and the signal to the object may trigger a zero or one conversion thereof. The signal is a means of asynchronous communication between objects. Synchronous communication, two asynchronous signals are required, one in each communication direction.
The signal can be generally described, and the subsection generated from the parent signal inherits the parameters of the parent signal, and it can also increase its own parameters. The child signal satisfies the trigger of its ancestors. Time event satisfies a time expression, after entering a state, after a certain period of time; or get an event after an absolute time. Note that whether it is a time period or an absolute time, you can use a realistic clock or a virtual internal clock (different internal clocks with different objects). The representation of different objects is shown. Standard Elements Creating (Destroy) 129. Exception Basic Execution Mechanism The signal caused by failure behavior. See a composite state, signal (Signal). Semantic exceptions typically implicitly generated by implementation mechanisms corresponding to failure. It can be treated as signals sent to active objects or process activities, resulting in abnormal execution. The state of the object can take corresponding actions to exceptions. This includes ignoring the current process, jumps to somewhere in the execution, performing an action without transforming status, or ignoring events. The function to the high-level state conversion makes exceptions flexibly and strong. Because the exception is the signal, it has a parameter table. The parameter value is set by the actuator (such as an operating system) found. These parameters can be read in handling exceptions. In most languages, the operation of the abnormal operation can control it or forward it. The representation is distinguished and exception in the configuration type "exception". The event name in the state machine does not mark the structure. 130. Exit Action Exit Active Activity. Entry action, running to completion (State Machine), Transition Semantic Status You can choose to have an export activity. Whether in any way to exit this state, after the completion of the activities of the internal state or the attached event, export activities must be performed before the activity of the external state. Export activities cannot be skipped in any way. Be sure to be executed before the status is lost. Inlets and export activities are not required in semantics (entry activity can be subordinate to all entry states), but they achieve the state package, thereby separating the external use with the internal structure. They allow definitions of initialization and final event, but don't worry that it will be slightly. For exceptions, it is especially useful because it defines the activities that are both ever appear. Indicates that the export activity is encoded in accordance with internal conversion, with virtual event name exit (exit is a reserved word, cannot be used as the name of the actual event). EXIT / Active Sequence Each state has only one exit activity, which can be a sequence so that it will not lose generality. Discussing export activities can accomplish the necessary cleanup activities when introducing the state. Its maximum role is to release the temporary storage area and other resources allocated during the state execution. Usually, export activities are used with entrance activities. The entrance activity is allocated, and the export activity is released. Even thirten often, resources are still released. 131. Drake (export) In the context of the package, it can be accessed outside its namespace by adjusting the visibility of the package. Unlike access and import, they are accessible to external elements to be accessed in the package. See Access (IMPORT), Visibility Semantic Pack By adjusting the visibility level of elements, allowing other packages to see this element, thereby implementing (for packets to import the element are public Protecting its own son). Discussing an element to see another elements of the same level must meet two conditions. The package containing this element should be set to a common visible. In addition, the packets that reference the target elements must be accessible or imported into the package containing the target element. Both are must. 132. Expression (Expression) can be translated by a given language. Many expressions generate a certain value after translation, other implementation of a particular action. For example, expressions "(7 5 * 3)" calculate an integer.
Semantic expressions define a statement, which can be used by some instances, values, or operations in context. The expression does not modify its calculated environment. The expression contains a string and the name of the language used. The expression contains the name and expression of the language used by the translation (in accordance with the syntax code of the design language). Assume that the language can be translated. Providing a translator is a task for modeling tools. The language used can be constraint - illustrative language, such as OCL; can be a programming language, such as C , SmallTalk; or human language. Of course, the expression of natural language writing cannot be automatically calculated by the tool, but must have someone to handle it. Different subclasses of expressions produce different types of values, with Boolean expressions, object integration expressions, time expressions, process expressions. Expression in the UML model can be actually active, constraint, guardians, and more. The representation of the method is a string that is expressed in the specified language. The syntax of the expression is identified by the tool or syntax analyzer. It is believed that the analyzer can identify these strings and get the corresponding value; or identify the semantic structure to understand the expression. For example, the Boolean expression can calculate the TRUE or FALSE value. Modeling tools should know the language used, which is generally not indicated. Expression itself usually has shown its own meaning. Example Self.cost If there is no control condition, it is considered to be True. If the control condition of the use case is met, the extension will be performed. If you reference multiple locations in the base case, it is possible to expand any one of them. If the control condition has been true, the extension can be executed multiple times. Each piece of the extended use case performs the same number of times. If the number of execution must be restricted, the corresponding control condition can be used. The control condition of the extended relationship is calculated when the instance of the execution of the syndrome is performed. If the condition is TRUE or empty, the extension is performed. In many cases, the condition includes the occurrence of an event or the value required for the extension case, such as a signal that starts an extended. Conditions may depend on the state of the use case instance, including the attribute value of the base case. If the event does not occur or the condition is false, the expansion segment of the use case is not started. After the expansion segment is executed, the use case instance returns the basis, and continues to perform downward from just now. If the control condition is met, other inserted extensions can be performed immediately. If the extension point references a plurality of locations, any of the control conditions may be satisfied. At any position in the sequence, the conditions can be true. If there is a plurality of insertion sequences in an extension case, all extended segments will be executed as long as the control condition is satisfied in the first extension. The expansion conditions are no longer re-judging for subsequent. Subsequent, inserts the use case instance to reach the corresponding position. Between the two extensions, the use case example returns to perform a base case. Once start, you must complete all pieces. Note: Usually use cases are unseasive state machines (syntax), not executable processes. Because control conditions may require an emergence of external events. The use of an example is implemented as a class, and it needs to be converted to a clear control structure. As the realization of the grammar needs to be converted to a more effective but more embarrassing executable. Note: Base and extension are related, an extension can be the basis for further expansion. There is no difficulty in this, and the above rules are still used - insertion is nested. For example, it is assumed that the use case b extends the use case A at executing the extended point X; use case C and expands the use case B at Y points. (Figure 13-96). A The instance of A is executed after the extension X, start execution case B. After the instance reaches the Y point, start execution case C. C After the execution is complete, return to use B, after the extension B is executed, then execute the return case A. This is the same as nested process calls or other nested structures. Figure 13-96 Nested Extension Figure 13-97 Nested relationship representation from the extension example to blew a dashed arrow, pointing to the group arrow as a branches, with a keyword "extend". After the keyword, the extension name can be listed in parentheses. Figure 13-97 is an extension of an extended relationship, and Figure 13-98 is a sequence of behavioral sequences of the use case. Discussing expansion, including, generalization relationships are increasing to the original case. They have a lot of common, but in fact, they are very convenient. Table 13-1 compares for three strokes. Figures 13-98 The behavioral sequence list 13-1 uses an example relationship 134. Extension Point references a naming mark in a sequence in a sequence, in which additional behavior can be inserted in these locations. The declaration of the extension point opens the possibility of use case expansion. The insertion piece is a behavioral sequence of the extended use case (use case with the propeller associated with the base case). Extended relationships include extended point name lists that explain where the extension is inserted from inserting its behavior. See Extension, Use Case. The semantic extension has a name and references one or several positions in the sequence of routine. Extended points can reference a location between the two behavior steps of the use case. In addition, it can reference a set of discrete locations. It is also possible to reference a region in the behavioral sequence (not exceeding all positions between the sequence between two steps). The position is a state in the state of the use case, or other descriptions in the corresponding component in the manner - between two statements in the statement list or two messages in the interaction. The extended relationship contains optional control conditions and extension points list, and the number of extensions in the list is equal to the number of expansion pieces.
