|
Software Patent Abstract
A software testing system uses a graph traversal algorithm to explore
a model simulating a software product in order to identify errors
in the software product. The model employs a Petri's net construct
for maintaining state and governing transitions. In particular,
the model mediates between a test driver and the software product.
The model-based approach is usable both to validate the design of
the software and verify the implementation of that design. Using
the Petri net model, the test space is bounded.
Software Patent Claims
We claim:
1. A system for error-checking a software product comprising: at
least one processor and a computer-readable storage medium coupled
to the at-least one processor; a model of the software product defining
a model space comprising a set of states of the model, the model
having places, transitions, and tokens, wherein a transition is
operable to pass at least one token from at least one place, and
wherein the model of the software product comprises components automatically
extracted from the software product; and a simulation engine for
manipulating the software product and the model of the software
product to cause an occurrence of an event in the software product
and to cause an occurrence of a corresponding transition in the
model of the software product, and to compare a state of the software
product after occurrence of the event with the state of the model
of the software product after occurrence of the corresponding transition,
and to log errors found in the software product.
2. The system according to claim 1, wherein logged errors are determined
when the state of the software product after occurrence of the event
differs from the state of the model of the software product after
occurrence of the corresponding transition.
3. The system according to claim 1, further comprising a graph
traversal algorithm set, in accordance with which the software product
and the model of the software product are manipulated by the simulation
engine.
4. The system according to claim 1, further comprising a model
editor for creating and manipulating the model of the software product.
5. The system according to claim 4, wherein the model editor comprises
a visual user interface having therein visual model components that
are selectable and mutable by a user.
6. The system according to claim 5, wherein the visual user interface
of the model editor further comprises a window having model units,
each model unit comprising model components, wherein selection of
a model unit causes the components of the selected unit to be displayed
in a separate window as visual model components.
7. The system according to claim 1, wherein the components automatically
extracted from the software product include a component corresponding
to a method of a class in the software product and a component corresponding
to an attribute of the class in the software product.
8. A computer-implemented method of producing error-checked software
code comprising: producing a body of software code to be error-checked;
creating a model of the desired functionality of the software code
to be error-checked, wherein the model comprises at least one Petri
net, and wherein creating a model of the desired functionality of
the software code to be error-checked comprises extracting information
from the software code to be error-checked; and dynamically determining
whether there is a difference in behavior of the software code to
be error-checked as compared to the behavior of the model, whereby
if there is such a difference in behavior, the difference is logged
as an error found in the software code to be error-checked.
9. The method according to claim 8, further comprising modifying
the software to eliminate the logged error.
10. The method according to claim 8, wherein the model of the desired
functionality of the software code to be error-checked is derived
from a specification for the software code to be error-checked.
11. The method according to claim 8 wherein extracting information
from the software code to be error-checked comprises extracting
an attribute and a method of a class of the software code to be
error-checked, wherein the attribute and method are used as a state
and a transition respectively in the model.
12. The method according to claim 8, wherein the model defines
a model space consisting of a set of states of the model, and wherein
dynamically determining whether there is a difference in behavior
of the software code to be error-checked as compared to the behavior
of the model comprises receiving a graph traversal algorithm and
traversing the model space pursuant to the received graph traversal
algorithm.
13. A computer-readable storage medium having stored thereon computer-executable
instructions for performing the method according to claim 8.
14. A computer-implemented method for testing a software product
to identify at least one error in the software product comprising:
receiving input by the software product from a simulation engine;
in response to said input, causing the occurrence of an event in
the software product, whereby a state of the software product after
the occurrence of said event is comparable with a state of a corresponding
model of the software product to identify the at least one error
found in the software product, wherein the model comprises components
automatically extracted from the software product; and providing
data associated with the at least one error for review.
Mobile Phone Patent Description
TECHNICAL FIELD
This invention relates generally to computer software development
and, more particularly, relates to automated model-based testing
of computer software using an extended Petri's net formulation.
BACKGROUND OF THE INVENTION
Computers and computerized devices have become increasingly common
and increasingly powerful in today's world, enabling automation
of tasks previously performed by humans, and in addition performing
or facilitating some tasks that were previously impossible. However,
as powerful as computers are, the process of creating the software
typically used by computers is still a very human-oriented process.
Typically, a designer first creates a specification that describes
the desired functionality of a piece of software. Subsequently,
a developer writes computer code, typically in source code form,
to instruct the computer as what operations to perform, actions
to take, etc. Once the source code is complete, computer-readable
code may be created therefrom. As will be appreciated by those of
skill in the art, this process of converting source code into a
computer-readable form is often referred to as "compiling."
Once the compiled version of the code has been produced, the first
stage of development is over, but the software is typically not
yet ready to be released to the public or other intended consumer.
Rather, a testing process is now initiated to identify and correct
any bugs, errors, or flaws in the software. At this stage in development,
the types of problems that are addressed are typically coding or
implementation errors, rather than errors in usability, or other
similar errors that pertain more to the overall design of the software
as embodied in the specification.
The testing process is an expensive and time-consuming process.
Typically an individual called a "tester" tests specific
use scenarios with respect to the software. This may be done in
a number of ways. One way is for the tester to actually operate
the software in the same manner that an end-user would, trying out
many different types of operations, keystrokes, menu selections,
etc. During this process, the tester observes the behavior of the
software, as executed by the computer, to identify any odd or incorrect
behavior. The tester reports any problems to the developers, who
in turn typically will rewrite a portion of the source code to correct
the problem. It will be appreciated that this method of testing
is expensive and time consuming. Moreover, the test space, or the
group of all actions and combinations thereof that the tester could
possibly test, is enormous. Given the complexity of present software
products, the test space is essentially unbounded, i.e. due to time
or other constraints, a human tester will typically not be able
to test every action or combination of actions that is in the test
space, and moreover may not even be aware of many or most such actions
or combination of actions.
Another method often used to test software involves the production
of test code. In particular, an individual familiar with the target
software and its intended operation can write test code that will
manipulate the software in the same way that might result from actual
use of the software by an end user. Typically, the test code functions
at a fairly low level, interfacing directly with APIs rather than
interfacing with the system from the high level at which an actual
user would operate. Although this approach is common, and may be
preferable over other existing types of testing, it does possess
a number of rather serious shortcomings. For example, a primary
weakness of this method is that, like the previously discussed manual
testing method, it is bounded by human creativity and time constraints
in defining and traversing usage scenarios.
SUMMARY OF THE INVENTION
Embodiments of the invention are directed to an automated graph
traversal algorithm for identifying errors in computer code using
a model. The model employs a Petri's net construct for maintaining
state and for governing transitions in the model of a target body
of software. In particular, the model of a software body is extracted
and acts as a mediator between a test driver and an implementation
of the software body. The model-based approach is usable both to
validate the design of the software and to verify the implementation
of that design. In contrast to prior methods that test a selection
of use scenarios without testing or even identifying the remaining
scenarios, the graph traversal algorithm according to an embodiment
of the invention is usable to provide deterministic coverage on
scenarios whether it tests all or only some. In this way, the test
space is bounded rather than fairly unlimited and unmanageable.
Generally, embodiments of the invention provide a number of other
benefits as well, including hiding unnecessary complexity while
preserving and using the relevant details, and providing improved
efficiency and cost effectiveness in code testing by testing many
more use scenarios without requiring additional effort or time from
a human tester.
The steps exercised in testing software in accordance with many
embodiments of the invention include first building a model of the
software using the design specification for the software. The model
may be a state machine that embodies state and also supports the
passing of tokens to model sequential operations, numerical bounds,
etc. In an embodiment of the invention, a tester extracts states,
guards, and transitions from the specification and codes them via
the model. Methods and attributes can be obtained through system
reflection, and used to build configuration files for the test driver.
Next, the model is linked to the implementation of the design. Finally,
an event generator uses the model to simulate operation of the software.
In particular the event generator uses graph theory to traverse
the model, exploring millions of test situations. The event generator
may also test specified scenarios. As a final step, the event generator
reports bugs and errors to allow developers or other personnel to
take corrective action.
The modular nature of the code checking process also provides robustness
with respect to code changes. For example, if 5% of the target code
changes, there will be an approximately 5% impact on the model.
In an embodiment, any changes that render the model syntactically
incorrect are flagged for correction. Furthermore, in embodiments
of the invention, a tester may choose to ignore certain areas of
the model.
In an embodiment of the invention, a collection of graph traversal
algorithms is employed to represent test sequence scenarios. The
first graph traversal algorithm is based on random exploration of
the graph resulting from the model. As transitions occur, objects
may change state and tokens can move from state to state. A key
that is used as a seed controls the randomness of the graph traversal,
with the two simulations using the same seed yielding the same exploration
path. Transitions with respect to states are weighted to give the
probability of occurrence for each transition. The weights may also
be used to forbid one or more transitions, such as private transitions
or transitions to be skipped at that time due to incompleteness
or otherwise.
Additional features and advantages of the invention will be made
apparent from the following detailed description of illustrative
embodiments that proceeds with reference to the accompanying figures.
BRIEF DESCRIPTION OF THE DRAWINGS
While the appended claims set forth the features of the present
invention with particularity, the invention, together with its objects
and advantages, may be best understood from the following detailed
description taken in conjunction with the accompanying drawings
of which:
FIG. 1 is a block diagram generally illustrating an exemplary computer
system usable in an implementation according to an embodiment of
the present invention;
FIG. 2A is a state diagram embodying an example Petri net having
states and employing tokens to allow or disallow transitions;
FIG. 2B is a state diagram embodying an example Petri net having
states and employing tokens to allow or disallow transitions, wherein
the Petri net is in another state;
FIG. 2C is a state diagram embodying an example Petri net having
states and employing tokens to allow or disallow transitions, wherein
the Petri net is in yet another state;
FIG. 3 is an architectural diagram of a software error-checking
system in accordance with an embodiment of the invention;
FIG. 4 is a relational diagram illustrating the relationship of
a class in the software product to the simulation model;
FIG. 5A is a user interface illustration showing an interface to
a model editor;
FIG. 5B is a user interface illustration showing an interface to
a model editor, wherein a first menu item has been selected;
FIG. 5C is a user interface illustration showing an interface to
a model editor, wherein a second menu item has been selected;
FIG. 5D is a user interface illustration showing an interface to
a model editor, wherein a third menu item has been selected;
FIG. 6A is a user interface illustration showing an interface to
a model editor, wherein a project window is displayed;
FIG. 6B is a user interface illustration showing an interface to
a model editor, wherein a project window and a hierarchically related
component window are both displayed; and
FIG. 7 is a flow chart illustrating steps taken in accordance with
an embodiment of the invention to produce error-checked software.
DETAILED DESCRIPTION OF THE INVENTION
Turning to the drawings, wherein like reference numerals refer
to like elements, the invention is described hereinafter in the
context of a computing environment. Although it is not required
for practicing the invention, the invention is described as it is
implemented by computer-executable instructions, such as program
modules, that are executed by a personal computer (PC) or other
computer. Generally, program modules include routines, programs,
objects, components, data structures and the like that perform particular
tasks or implement particular abstract data types.
The invention may be implemented in computer system configurations
other than a PC. For example, hand-held devices, multi-processor
systems, microprocessor-based or programmable consumer electronics,
network PCs, minicomputers, mainframe computers and the like may
be used in implementations of the invention. 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 memory storage
devices.
Although the invention may be incorporated into many types of computing
environments as suggested above, the following detailed description
of the invention is set forth in the context of an exemplary general-purpose
computing device in the form of a computer 20. Thus, before describing
the invention in detail, the exemplary computing environment in
which the invention may be implemented is described in connection
with FIG. 1.
The computer 20 includes a processing unit 21, a system memory
22, and a system bus 23 that couples various system components including
the system memory to the processing unit 21. The system bus 23 may
be any of several types of bus structures including a memory bus
or memory controller, a peripheral bus, and a local bus using any
of a variety of bus architectures. The system memory includes read
only memory (ROM) 24 and random access memory (RAM) 25. A basic
input/output system (BIOS) 26, containing the basic routines that
help to transfer information between elements within the PC 20,
such as during start-up, is stored in ROM 24. The PC 20 further
includes a hard disk drive 27 for reading from and writing to a
hard disk 60, a magnetic disk drive 28 for reading from or writing
to a removable magnetic disk 29, and an optical disk drive 30 for
reading from or writing to a removable optical disk 31 such as a
CD ROM or other optical media.
The hard disk drive 27, magnetic disk drive 28, and optical disk
drive 30 are connected to the system bus 23 by a hard disk drive
interface 32, a magnetic disk drive interface 33, and an optical
disk drive interface 34, respectively. The drives and their associated
computer-readable media provide nonvolatile storage of computer
readable instructions, data structures, program modules and other
data for the computer 20. Although the exemplary environment described
herein employs a hard disk 60, a removable magnetic disk 29, and
a removable optical disk 31, it will be appreciated by those skilled
in the art that other types of computer readable media which can
store data that is accessible by a computer, such as magnetic cassettes,
flash memory cards, digital video disks, Bernoulli cartridges, random
access memories, read only memories, memory keys, memory sticks,
and the like may also be used in the exemplary operating environment.
A number of program modules may be stored on the hard disk 60,
magnetic disk 29, optical disk 31, ROM 24 or RAM 25, including an
operating system 35, one or more applications programs 36, other
program modules 37, and program data 38. A user may enter commands
and information into the computer 20 through input devices such
as a keyboard 40 and a pointing device 42. Other input devices (not
shown) may include a microphone, joystick, game pad, satellite dish,
scanner, or the like. These and other input devices are often connected
to the processing unit 21 through a serial port interface 46 that
is coupled to the system bus, but may be connected by other interfaces,
such as a parallel port, game port or a universal serial bus (USB).
A monitor 47 or other type of display device is also connected to
the system bus 23 via an interface, such as a video adapter 48.
In addition to the monitor, computers typically include other peripheral
output devices, not shown, such as speakers and printers.
The computer 20 may operate in a networked environment using logical
connections to one or more remote computers, such as a remote computer
49. The remote computer 49 may be a PC, 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
20, although only a memory storage device 50 has been illustrated
in FIG. 1. The logical connections depicted in FIG. 1 include a
local area network (LAN) 51 and a wide area network (WAN) 52. Such
networking environments are commonplace in offices, enterprise-wide
computer networks, intranets and the Internet.
When used in a LAN networking environment, the computer 20 is connected
to the local network 51 through a network interface or adapter 53.
When used in a WAN networking environment, the computer 20 typically
includes a modem 54 or other means for establishing communications
over the WAN 52. The modem 54, which may be internal or external,
is connected to the system bus 23 via the serial port interface
46. In a networked environment, program modules depicted relative
to the computer 20, or portions thereof, may be stored in the remote
memory storage device. 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.
In the description that follows, the invention will be described
with reference to acts and symbolic representations of operations
that are performed by one or more computers, unless indicated otherwise.
As such, it will be understood that such acts and operations, which
are at times referred to as being computer-executed, include the
manipulation by the processing unit of the computer of electrical
signals representing data in a structured form. This manipulation
transforms the data or maintains it at locations in the memory system
of the computer, which reconfigures or otherwise alters the operation
of the computer in a manner well understood by those skilled in
the art. The data structures where data is maintained are physical
locations of the memory that have particular properties defined
by the format of the data. However, while the invention is being
described in the foregoing context, it is not meant to be limiting
as those of skill in the art will appreciate that various of the
acts and operations described hereinafter may also be implemented
in hardware.
In accordance with one embodiment of the invention, a Petri net
formulation is used to model a software application or other target
software body under test. The concept of a Petri net was first introduced
in the dissertation Kommunikation mit Automaten, submitted in 1962
by Carl Adam Petri to the faculty of Mathematics and Physics at
the Technische Universitaat Darmstadt, Germany. In overview a Petri
net is a graphical and mathematical modeling tool that is characterized
by a number of elements including places, tokens at places (and
passable between places), transitions, and arcs that interconnect
places and transitions.
Arcs that connect places to transitions are referred to as "input
arcs," while arcs that connect transitions to places are referred
to as "output arcs." In other words, an input arc originates
at a place and terminates at a transition, while an output arc originates
at a transition and terminates at a place. A Petri net usable within
an embodiment of the invention is not limited to these specific
arcs, and it will be appreciated by those of skill in the art that
there are a number of other arcs that may be used additionally or
alternatively.
The transitions are the active elements of the model, representing
actions that can occur in the system. The occurrence of such an
action is referred to as a "firing" of the associated
transition. The state of the Petri net at a given time is represented
in the number and types of tokens that reside in each place. Thus,
the firing of a transition changes the state of the system. The
state, i.e. the locations of the tokens, is also sometimes referred
to as the marking of the system.
A transition cannot occur unless that transition is "enabled."
In general, a transition is enabled once the predetermined criteria
governing the transition are met. For example, one possible criterion
is that a certain number, or at minimum a certain number, of tokens
must reside at the input to the transition, i.e. at the place at
the other end of the input arc. Additionally or alternatively, it
may be required that a given number of tokens must or must not reside
at the output from the transition. When the transition fires or
occurs, it decrements, by one or more, the number of tokens at its
input places and increments the number of tokens at its output places.
It will be understood herein that the term "fire" simply
connotes occurrence of the specified event or transition without
any further limitations intended. Each arc may have a cardinality
or order, and the number of tokens passed during a transition firing
is a function of the order of the relevant arc. When multiple transitions
are simultaneously enabled, the concept of transition priority is
useful. In particular, an enabled transition with a certain priority
will not fire until after any other enabled transition having a
higher priority has fired.
In order to more realistically model a system under test, a transition
delay may be used to temporarily inhibit enabled transitions that
would otherwise fire. For example, certain types of transitions
may be delayed by a predetermined amount to account for real-world
timing, and/or a random or pseudo-random delay may be applied to
some transitions to model real-world variability or uncertainty.
A simplified example Petri net is shown in FIGS. 2A-2C. In particular,
the Petri net 201 comprises a ring structure with two places 203,
205, and two transitions, 207, 209. Transition 207 has input and
output arcs 211, 213 connected to places 203 and 205 respectively.
Transition 209 has input and output arcs 215, 217 connected to places
205 and 203 respectively. Although any combination of token rules,
arc order and transition priorities are possible, the example Petri
net shown in FIGS. 2A-2C has arc order of 1 (i.e. one token is transferred
per transition firing), and the precondition for firing of either
transition 207, 209 is that a token is waiting at the input place
203, 205. Each transition that is enabled is circumscribed by a
dashed lined. Thus in FIG. 2A, transition 207 is enabled, while
transition 209 is not enabled. After firing of transition 207, the
state of the Petri net is as shown in FIG. 2B. A token has been
passed from one place 203 to the other 205. Both transitions 207,
209 are now enabled, since the precondition is met for both, i.e.
a token awaits at the respective input place 203, 205 for each.
From this state, a firing of transition 209 would result in a restoration
of the original state as shown in FIG. 2A. If instead transition
207 fires, the resultant state is shown in FIG. 2C. It can be seen
that the tokens now both reside at place 205. As a result, transition
209 is enabled, but transition 207 is not.
In keeping with the invention, operation of a software body, such
as an application or other software, is simulated via a Petri net
that is directed via an event driven generator engine using graph
traversal algorithms. An overview of the architecture of the error
correction system in keeping with an embodiment of the invention
will be discussed with reference to FIG. 3. The system 301 comprises
primarily a simulation model 303, a model editor 305, a graph traversal
algorithm set 307, and an event-driven generator engine 309, and
includes or interfaces with a software specification 311 and a corresponding
software implementation 313 under test. The software specification
311 is used by testing personnel to develop the simulation model
303 using the model editor 305. The software implementation 313
should typically conform to the specification 311 as well, since
the specification 311 is a first step in development that is used
by developers in creating a corresponding implementation. The graph
traversal algorithm set 307 is created by the event-driven generator
engine 309, which uses the algorithms of the graph traversal algorithm
set 307 to explore the simulation model 303 to test the software
implementation 313, as will be discussed in greater detail hereinafter.
FIG. 4 illustrates in greater detail an aspect of the relationship
between the simulation model 303 and the software implementation
313, referred to in FIG. 4 by the reference numerals 403 and 413
respectively. The software implementation comprises objects, each
of which has a class 421. Each class 421 in turn comprises at least
one method 423 and at least one attribute 425. The methods 423 can
be viewed as potential actions, and hence each method is likened
to a transition 427 in the simulation model 403. Similarly, the
collection of the current conditions of each attribute 425 represents
the state of the system. Hence, each attribute 425 is likened to
a state 429 in the simulation model 403. The method of extracting
the class methods 423 and attributes 425 to build the simulation
model 403 is not critical.
A human operator may perform the extraction based on the software
specification 311, or an automatic extraction of some or all of
this information based on an actual implementation 313 may be executed
instead or in addition. It is the behavior of the collective system
embodied in the software implementation 313 that is to be tested
for conformance to the desired behavior as embodied in the software
specification 311. Hence the actual behavior of the simulation model
403 is still preferably coded into the model 403 based on the software
specification 311. For example, based on the software specification
311, a transition from a given marking should result in a particular
state of the simulation model 403. The model is thus preferably
constructed in accordance with this constraint rather than in accordance
with the actual behavior of the actual implementation 313. In fact,
there is no requirement that the software implementation 313 be
complete at the time that the simulation model 403 is constructed.
For example, it may be desired to test the function and coherence
of the system described by the specification prior to constructing
an actual implementation based on that specification.
An example user interface for the model editor 305 is shown in
greater detail in FIGS. 5A-5D. FIG. 5A shows the interface 501 having
unselected menu items "File" 503, "Windows"
505, and "Fire" 507. Generally, the "File" menu
item 503 is selectable to present an array of file-related commands
and options, as will be discussed in greater detail with respect
to FIG. 5B. The "Windows" menu item 505 is selectable
to present an array of commands and options related to windows and
views, as will be discussed in greater detail with respect to FIG.
5C. Finally, the "Fire" menu item 507 is selectable to
present an array of commands and options related to the operation
of the simulation, as will be discussed in greater detail with respect
to FIG. 5D.
FIG. 5B illustrates the user interface 501 of FIG. 5A wherein the
user has selected the "File" menu item 503. Pursuant to
the user selection, a drop-down menu is presented having therein
a number of options. The options illustrated in the example of FIG.
5B include a "New" option 511, an "Open" option
513, a "Save" option 515, a "Properties" option
517, and an "Exit" option 519. The "New" option
511 is usable by the operator to create a new project. Preferably,
the user specifies both a name and location for the file corresponding
to the project, as well as a name and location for the implementation
to be tested by the new model. In addition, a seed value is used
in certain embodiments of the invention, and the user may either
specify a seed value when creating a new project or a default seed
value, such as 1234, may be used. The seed insures that graph traversal
algorithm generators will produce the same sequence of random values
in order to insure reproduction of the exploration path. Using a
different seed would usually produce a different traversal of the
state space. A random generator is used for specific graph traversal
algorithm requiring random data, however for fully deterministic
graph traversal algorithms, a random seed value is not required.
The "Open" option 513 is usable by the operator to open
a previously created project by browsing for and selecting such
a project. The "Save" option 515 is usable by the operator
to save the current project. If the project has not been named and
a memory storage location specified for the project prior to selection
of the "Save" option 515, the user is preferably prompted
to enter a name and location upon selection of the "Save"
option 515. The "Properties" option 517 allows a user
to view and modify the properties of the project, e.g. the implementation
version being tested, the seed value, and so on. Finally, the "Exit"
option 519 allows a user to exit the model editor program.
FIG. 5C illustrates the user interface 501 of FIG. 5A wherein the
user has selected the "Windows" menu item 505. As with
the "File" menu item 503, a drop-down menu is presented
pursuant to the user selection of the "Windows" menu item
505, having therein a number of options. As illustrated in the example
of FIG. 5C, the options include a "States Tree" option
521, a "Class List" option 523, a "Cascade"
option 525, a "Tile Horizontal" option 527, and a "Tile
Vertical" option 529. The "States Tree" option 521
is usable by the operator to display a tree of states of the model.
The "Class List" option 523 can be selected to provide
a list of available classes in the assembly under test. Preferably
extended selection is available, and each class has only one component
model corresponding to it. The "Cascade" option 525 simply
acts to cascade the model editor windows. Each of the "Tile
Horizontal" option 527, and the "Tile Vertical" option
529 causes the windows associated with the model editor to be arranged
in the respective specified manner.
FIG. 5D illustrates the user interface 501 of FIG. 5A wherein the
user has selected the "Fire" menu item 507. As with the
previously described menu items, a drop-down menu is presented pursuant
to the user selection of the "Fire" menu item 507, the
drop-down menu containing a number of options. As illustrated in
the example of FIG. 5D, the options include a "Run" option
531, a "Cover+" option 533, and a "Stop" option
535. Selection of the "Run" option 531 acts to start the
simulation pursuant to the model. Preferably a recompilation of
the assembly under test and the model are performed at this time
prior to running to assure inclusion of any changes to the code
or the model. If any errors are found, a list of errors is displayed.
If there are no errors, the simulation is run. Selection of the
"Cover+" option 533 causes a graph traversal algorithm
to be used that ensures traversal of all reachable paths and unreachable
paths are reported. The simulation ends when there are no further
paths to traverse. Finally, selection of the "Stop" option
535 causes the simulation to be stopped and causes a report of the
simulation results to be released, such as to a predetermined directory
or other location.
FIGS. 6A and B show other user interface aspects of the model editor
305. In particular, FIG. 6A shows a project window 603 within model
editor window 601. The project window 603 corresponds to a project
entitled "Project 1" as indicated in its title bar. Within
the project window 603 there are icons associated with classes within
the model of Project 1. In the example of FIG. 6A, the icons include
"Task" 605, "Job" 607, and "Task Row"
609. By right-clicking an icon 605, 607, 609 or by otherwise indicating
their selection, a new window is opened for display of the model
for the associated class. The user is then able to define or edit
the model, which preferably corresponds to a Petri net.
For example, in the example user interface shown in FIG. 6B, the
user has right-clicked the icon 605 for the "Task" class,
and a new window 611 entitled "Task" has been opened.
Within window 611, the states 613, 615, 617, 619 and transitions
614, 616, 618, 620 of the model are shown and can be edited or defined.
Transitions 614, 616, 618, 620 (arcs) between states can be created
via dragging and dropping of one state onto another or otherwise.
Once the transitions 614, 616, 618, 620 have been created, a method
corresponding to each transition 614, 616, 618, 620 may be defined
or edited by right-clicking the transition 614, 616, 618, 620.
In an embodiment of the invention, the formulation of the model
is an extension of the Petri's net. In order to minimize state explosion,
activation tokens carry a hidden state. The token hidden state is
preferably used to define the state of the model itself through
logical expressions in state and transitions (e.g., guard/post).
This feature insures that all relevant details are included in the
simulation without impairing the readability of the model.
In testing the assembly under test, the event driven generator
engine 309 stimulates the firing of transitions in the assembly
under test in accordance with a graph traversal algorithm from the
graph traversal algorithm 307 and observes the resultant state of
the program. The collection of data regarding the state of the program
may be accomplished via system reflection. To minimize the amount
of data collected, it is preferable to collect such data only after
the system has stabilized after a firing. The set of occurrences
resulting from a fired event can be referred to as a single transaction.
The event driven generator engine 309 compares the list of objects
that have been modified and their states with the same properties
in the model, which has been identically stimulated, and produces
an analysis based thereon.
Initially, the event driven generator engine 309 checks that only
the objects that should have changed according to the model have
changed in the system under test during the transaction. A transaction
is given a status. In an embodiment of the invention, the status
of a transaction may be one of "Success," "No Operation,"
"Bug," and "Unknown."
A status of "Success" indicates that no object was modified
other than those that were expected to be modified during the transaction.
A status of "No Operation" indicates that the event driven
generator engine 309 fired an event which could have revealed a
route to a disallowed transition (i.e. a "sneak path"),
and the result was that the system was robust and that no object
was modified other than those that were expected to be modified.
A status of "Bug" indicates that either a defect or a
sneak path was found as a result of the transaction. When a bug
is indicated, the event driven generator engine 309 may prune the
sequence of transactions and launch a debugger on the transaction
prior to the one in which the bug was found. Finally, a status of
"Unknown" indicates that a bug occurred but that the event
driven generator engine 309 is unable to know whether the error
occurred in an object during the transaction or instead occurred
in an external resource. For example, a network time-out would be
viewed by the event driven generator engine 309 in embodiment of
the invention as an error occurring in an external resource.
A flow chart of a process of developing error-checked software
according to an embodiment of the invention is hereinafter described
with reference to FIG. 7. Initially, a specification for the software
of interest is developed at step 701. The specification is a description
of the manner in which the software should perform once written.
As discussed above, this step is performed by personnel, such as
management personnel, and the personnel developing the specification
may or may not overlap with the development personnel that will
eventually write the software. At step 703, which may be accomplished
in parallel with step 701, development personnel write code that
corresponds to the specification. There may be alterations made
to the specification while the software is being written, as issues
arise and are resolved.
At this stage in the process of FIG. 7, a software specification
has been created and a software product that ostensibly meets that
specification has been created as well. In the following steps,
the software code is checked and refined. In particular, at step
705, a tester, who may or may not be a developer, constructs a Petri
net model of the software product based on the specification. As
discussed above, some of the work of constructing the model may
be automated, such as via extraction of class definitions from the
software product. Once the model has been developed, the tester
uses an event driven generator engine to test the software via the
model. In particular, at step 707, the event driven generator engine
constructs a set of graph traversal algorithms that can be used
to define different routes of exploration through the space defined
by the model. For example, a graph traversal algorithm may establish
an initial state of the system and may then cause specified transitions
to be fired. Random exploration of the model space is also possible.
Transitions with respect to states are preferably weighted to give
the probability of occurrence for each transition, with the weights
also being usable to forbid one or more transitions, such as private
transitions or transitions known to be error-prone.
At step 709, the event driven generator engine uses a graph traversal
algorithm to explore the model space, checking that the state of
the software implementation is consistent with the corresponding
state of the model. Note that in embodiments of the invention, a
tester may choose to ignore certain areas of the model. As part
of this process, the event driven generator engine preferably also
maintains a record of defects found while exploring the model space.
For example, an error would be evident when after a particular transition,
or transaction of transitions, the state of the affected components
in the model is different from the state of the affected components
in the software product, or when the identity of at least one of
the affected components in the model or software product is different
from the identity of every affected component in the other of the
model or software product.
At step 711 the tester determines whether any substantial defects
have been reported. The inquiry as to whether any substantial defects
were reported may consist of an inquiry as to whether any defect
at all has been reported, or may be instead an inquiry as to whether
any defect greater than an acceptable severity has been reported.
If no substantial defects were reported then the software product
is approved for release and the process terminates at step 713.
If instead it is determined at step 711 that one or more substantial
defects were reported, then the process continues to step 715 for
further development of the software to correct some or all of the
reported errors. At this step, the goal in an embodiment of the
invention is to modify the software to correspond to the software
specification as was originally intended. After the software has
been redeveloped to correct the reported error or errors, it may
be tested again as at step 709 or may be released. Generally, if
it is tested again, it will not be necessary to reconstruct the
model and the graph traversal algorithm set before checking for
errors.
It will be appreciated that an improved software testing method
and system has been described, the method and system employing a
Petri net model for simulation of a software product. In view of
the many possible embodiments to which the principles of this invention
may be applied, it should be recognized that the embodiments described
herein with respect to the drawing figures are meant to be illustrative
only and should not be taken as limiting the scope of invention.
For example, those of skill in the art will recognize that some
elements of the illustrated embodiments shown in software may be
implemented in hardware and vice versa or that the illustrated embodiments
can be modified in arrangement and detail without departing from
the spirit of the invention. Therefore, the invention as described
herein contemplates all such embodiments as may come within the
scope of the following claims and equivalents thereof. |