CN116204445A - Test case generation method, device, code testing method, equipment and medium - Google Patents

Abstract

The present invention relates to the field of software testing, and in particular, to a test case generating method, a device, a code testing method, a device, and a medium. The test case generation method comprises the following steps: the combination logic is used as a target condition to cover test points and a condition tree is generated by using a preset rule; constructing a population of genetic algorithm with each individual as an initial test case; node interception is carried out on the condition tree to obtain effective nodes and an initial weight value is given; taking individuals of the population as input of codes to be tested and running the codes to be tested on a verification platform; updating the weight value of the effective node according to the coverage times and the initial weight value of the effective node collected from the verification platform; calculating the fitness of individuals in the population according to the updated weight value and the coverage state of the effective nodes collected from the verification platform; and adjusting individuals of the population based on the fitness until a preset output condition is met, and outputting. The scheme of the invention greatly reduces the number of redundant test cases, improves the convergence rate and improves the test efficiency.

Description

Test case generation method, device, code test method, equipment and medium

technical field

The invention relates to the field of software testing, in particular to a test case generation method, device, code testing method, equipment and media.

Background technique

Code coverage is a method to indirectly measure software quality by calculating the proportion of the source code executed during the test to the total source code. It potentially guarantees the quality of the actual product while ensuring the quality of the test. Based on this, it can find places in the program that have not been tested by the test case, and further create new test cases to increase the coverage. Ideally, the code coverage test is expected to achieve a 100% coverage rate, but this is usually difficult to achieve, especially for large-scale projects with a large amount of code and complex logic, the coverage rate of conditions, paths, etc. is even difficult to reach 90%. Very harsh trigger conditions, conventional random tests run for hours, or even days, cannot cover the corresponding test points.

In order to reduce the code testing cycle, improve code coverage, and reduce the technical requirements for code testers, the industry currently introduces intelligent search algorithms to automatically generate code coverage test cases. However, the commonly used test case generation method in the industry is based on random algorithm or constrained random algorithm. This open-loop method does not adjust the test case generation strategy based on the code state, and then generates a large number of redundant test cases, resulting in low test efficiency. Therefore, it is urgent needs improvement.

Contents of the invention

In view of this, it is necessary to provide a test case generation method, device, code testing method, equipment and medium for the above technical problems.

According to a first aspect of the present invention, a method for generating a test case is provided, the method for generating a test case includes:

Analyzing the code to be tested to use combinational logic as a target condition coverage test point, and using preset rules to generate a condition tree for the target condition coverage test point;

Construct a population of genetic algorithms, wherein each individual in the population is an initial test case of the code to be tested;

Performing node interception on the conditional tree to obtain valid nodes, and assigning initial weight values to the valid nodes;

Using the individuals of the population as the input of the code to be tested, and running the code to be tested on the verification platform;

updating the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value;

calculating the fitness of individuals in the population according to the update weight value and the coverage status of the valid nodes collected from the verification platform;

The individuals of the population are adjusted based on the fitness until the preset output conditions are met, and the adjusted test cases are output.

In some embodiments, the preset rules include at least one of the following:

Taking the combined logic corresponding to the target condition coverage test point as the root node;

Processing the combination logic based on the preset sub-item segmentation rules, using the combination logic itself as a parent node, and using the child items of the combination logic of the parent node as child nodes;

For branch conditions, the branch that causes the variable change of this node is listed as a child node;

When the node is a code input parameter, the output of a code invisible component, or the value change of a node variable has no logical relationship with other variables, the node is a leaf node and will not be decomposed;

The minimum number of nodes passed by each node to reach the root node is taken as the layer number of the node.

In some embodiments, the preset subitem segmentation rules include at least one of the following:

Segmentation rule 1. Convert logical operators to obtain combinational logic including only AND relation or OR relation;

Segmentation rule 2, setting the priority of the AND relationship, or the relationship, to follow the design language rules of the code to be tested;

Segmentation rule 3. There is only a single form of logical relationship between sub-items, and each sub-item corresponds to a sub-node;

Segmentation rule 4. Before decomposing sub-items, it is necessary to simplify the combination logic according to the logical operation rules to ensure that each item does not contain logical operators;

Segmentation rule 5. For the logical relationship of the root node, after conversion by segmentation rule 1, segmentation rule 2, and segmentation rule 4, the OR relationship is changed to an AND relationship;

Segmentation rule 6. If the simplified logical relation contains both an OR relation and an AND relation, first decompose the sub-nodes of the low-priority logical relation, and then decompose the high-priority sub-nodes from the generated sub-nodes .

In some embodiments, the preset output condition includes that there is an individual in the current population whose test can cover the target condition coverage test point, and the number of iterations of the current population reaches the preset number of iterations.

In some embodiments, the step of outputting the adjusted test case when the individual of the population is adjusted based on the fitness until the preset output condition is met includes:

In response to the presence of an individual test in the current population that can cover the target condition coverage test point or the number of iterations in the current population reaches the preset number of iterations, then stop the iteration and output the test case corresponding to the individual or when the preset number of iterations is reached The test cases corresponding to the individuals in the population.

In some embodiments, the step of adjusting the individuals of the population based on the fitness until the preset output condition is met, and outputting the adjusted test case further includes:

In response to the fact that none of the individual tests in the current population can cover the target condition coverage test point and the number of iterations of the current population does not reach the preset number of iterations, the population change operation is performed on the individuals in the current population based on the fitness to generate the next generation population .

In some embodiments, the step of adjusting the individuals of the population based on the fitness until the preset output condition is met, and outputting the adjusted test case further includes:

In response to generating the next-generation population, return to the step of using the individuals of the population as the input of the code to be tested and running the code to be tested on the verification platform.

In some embodiments, the population changing operation includes at least one of selection, crossover and mutation operations.

In some embodiments, the step of performing node interception on the conditional tree to obtain valid nodes includes:

Select all nodes whose node layer number is less than or equal to the preset intercepted node layer number from the generated condition tree as valid nodes.

In some embodiments, the step of assigning initial weight values to the effective nodes includes:

Assign an initial weight value to each valid node according to the following formula 1:

公式一；

formula one;

表示节点的层号，/>

表示节点在层中的索引号，/>

表示预设截取节点层数，

表示第/>

层中的节点总数，/>

表示逻辑属性系对应的值，/>

表示第/>

层、第/>

个节点的初始权重值。in,

Indicates the layer number of the node, />

Indicates the index number of the node in the layer, />

Indicates the number of preset interception node layers,

Indicates the first />

the total number of nodes in the layer, />

Indicates the value corresponding to the logical attribute system, />

Indicates the first />

Layer, No. />

The initial weight value of each node.

In some embodiments, the value selection rules corresponding to logical attributes are as follows:

；If the logical attribute of the node is "and", then

;

等于此节点对应的父节点所包括的子节点个数。If the logical attribute of the node is "or", then

Equal to the number of child nodes included in the parent node corresponding to this node.

In some embodiments, the step of updating the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value includes:

Update the weight value for each valid node according to the following formula two;

公式二；

formula two;

表示种群进化的代数且/>

大于等于1，/>

表示第/>

层、第/>

个节点在使用第/>

代种群进行测试时被覆盖的次数，/>

表示种群中个体总数量，/>

表示第/>

次迭代时第/>

层、第/>

个节点的更新权重值。in,

represents the algebra of population evolution and />

greater than or equal to 1, />

Indicates the first />

Layer, No. />

node is using the />

The number of times the generation population is covered when testing, />

Indicates the total number of individuals in the population, />

Indicates the first />

At iteration />

Layer, No. />

The updated weight value of each node.

In some embodiments, the step of calculating the fitness of individuals in the population according to the updated weight value and the coverage status of the valid nodes collected from the verification platform includes:

Calculate the fitness of each individual in the population according to the following formula three;

公式三；

formula three;

表示第/>

代种群中，第/>

个个体的适应度，/>

表示使用第/>

代种群的第/>

个个体测试时，第/>

层的第/>

个节点的覆盖状态对应的值。in,

Indicates the first />

In the generation population, the />

The fitness of an individual, />

Indicates the use of the />

No./> of generation population

When an individual is tested, the first />

Layer />

The value corresponding to the coverage state of a node.

In some embodiments, the value selection rules corresponding to the coverage state of a node are as follows:

值为1；If the node is covered then

value is 1;

值为0。If the node is not covered then

The value is 0.

In some embodiments, the code under test is used to verify the DUT.

In some embodiments, the DUT includes memory chips, processors, and SoC chips.

In some embodiments, each test case of the code under test includes a plurality of input parameters.

According to a second aspect of the present invention, a test case generation device is provided, the test case generation device includes:

A generating module, which is used to analyze the code to be tested to cover the test point with combinational logic as the target condition, and use preset rules to generate a condition tree for the target condition coverage test point;

A building block for constructing a population of genetic algorithms, wherein each individual in the population is an initial test case of the code to be tested;

An intercepting module, configured to perform node interception on the conditional tree to obtain valid nodes, and assign initial weight values to the valid nodes;

An operation module, configured to use the individual of the population as the input of the code to be tested, and run the code to be tested on the verification platform;

An updating module, configured to update the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value;

A calculation module, configured to calculate the fitness of individuals in the population according to the update weight value and the coverage status of the valid nodes collected from the verification platform;

An output module, configured to adjust the individuals of the population based on the fitness until a preset output condition is met and output an adjusted test case.

According to the third aspect of the present invention, the present invention provides a code testing method, which uses the test cases output by the test case generation method described above to perform tests on codes.

According to a fourth aspect of the present invention, a computing device is also provided, and the computing device includes:

at least one processor; and

The memory stores a computing device program that can run on the processor, and the processor executes the aforementioned test case generation method when executing the program.

According to the fifth aspect of the present invention, there is also provided a computing device readable storage medium, the computing device readable storage medium stores a computing device program, and when the computing device program is executed by a processor, the aforementioned test case generation method is executed.

The above-mentioned test case generation method generates a condition tree for the target condition coverage test point by analyzing the code to be tested and using preset rules, intercepts the required effective nodes from the condition tree, and generates an initial weight value for each effective node, and then In the population iteration process of the genetic algorithm, it is updated according to the coverage times of effective nodes and the initial weight value, and then the population is adjusted according to the updated weight value to calculate the adaptability of the population individual, and the coverage test can be generated by continuous iteration to cover the target condition The number of redundant test cases is greatly reduced, and the automatic adjustment of fitness according to the node hit probability can significantly speed up the iterative process and improve the convergence speed, so that the generation of test cases no longer depends on the technical level of code testers, reducing the cost of testing. The required labor cost has significantly improved the test efficiency.

In addition, the present invention also provides a test case generation device, a code testing method, a computing device, and a computing device-readable storage medium, which can also achieve the above technical effects, and will not be repeated here.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention, and those skilled in the art can obtain other embodiments according to these drawings without any creative effort.

Fig. 1 is the flowchart of a kind of test case generating method that an embodiment of the present invention provides;

Fig. 2 adopts the flow chart that the method of the present invention automatically generates test case that Fig. 2 provides for another embodiment of the present invention;

Fig. 3 is an example diagram of a conditional tree corresponding to a combination logic in the code to be tested provided by an embodiment of the present invention;

4 is a schematic structural diagram of a test case generation device provided by another embodiment of the present invention;

FIG. 5 is an internal structure diagram of a computing device in another embodiment of the present invention.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

It should be noted that all expressions using "first" and "second" in the embodiments of the present invention are to distinguish two entities with the same name but different parameters or parameters that are not the same, see "first" and "second" It is only for the convenience of expression, and should not be construed as a limitation on the embodiments of the present invention, which will not be described one by one in the subsequent embodiments.

In one embodiment, please refer to shown in Figure 1, the present invention provides a kind of test case generation method 100, specifically, described test case generation method comprises the following steps:

Step 101, analyzing the code to be tested to use combinational logic as the target condition coverage test point, and using preset rules to generate a condition tree for the target condition coverage test point;

In this embodiment, analyzing the code to be tested refers to analyzing the structure and logic of the code text, specifically, analyzing the program structure and logic based on the syntax of the code programming language. Combination logic includes at least two or more logic operations. Combination logic can be a combination of the same logic operations or a combination of different logic operations. For example, a condition coverage test point includes three AND operations and two OR operations. The preset rules are pre-set by testers or users to split and associate combination logic to form a node relationship tree including a root node and several child nodes.

Step 102, constructing a genetic algorithm population, wherein each individual in the population is an initial test case of the code to be tested;

In this embodiment, the genetic algorithm is an evolutionary iterative search algorithm. After the initial population is generated, the population individuals are selected, crossed and mutated according to specific rules according to the performance (also called fitness) of the individuals in the population to generate new Population, on the basis of inheriting the excellent genes of the original population, the new population maintains the diversity and evolutionary possibility of the population, and finally generates the optimal population through continuous iteration. In order to improve the performance of genetic algorithm, a series of improved algorithms are proposed, such as improved genetic algorithm, adaptive genetic algorithm, fuzzy genetic algorithm and so on. Among them, the adaptive genetic algorithm can improve the convergence accuracy and convergence speed of the algorithm through the adaptive adjustment of the genetic parameters, and the adaptive genetic algorithm is preferably used to construct the population. The initial test case refers to the test case that cannot cover the target condition coverage test point that has not been iterated in this search. The initial test case can be a randomly generated test case defined by the tester or a test case generated by the existing constraint method , of course, the initial test case can also be the final test case that failed to cover the target condition coverage test point after the previous genetic algorithm search.

Step 103, performing node interception on the conditional tree to obtain effective nodes, and assigning initial weight values to the effective nodes;

In this embodiment, node interception refers to selecting some nodes from the conditional tree, for example, intercepting nodes close to the root node in the conditional tree, or intercepting nodes at a specific depth in the conditional tree, and so on. Weight value refers to the measure of the importance or influence of a given set of data in solving a certain problem. There are many ways to calculate the weight value, and the specific method depends on the actual situation. For example, existing weight calculation methods can be used, such as relative weight method, percentage method, grade method, score method, linear weighting method or a combination of multiple methods.

Step 104, using the individuals of the population as the input of the code to be tested, and running the code to be tested on the verification platform;

In this embodiment, the verification platform refers to a tool that can run the code to be tested and collect, analyze and count the running process results of the code. This example does not limit the version and type of the verification platform, and only needs to have corresponding analysis and statistics functions That's it.

Step 105, updating the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value;

Step 106, calculating the fitness of individuals in the population according to the update weight value and the coverage status of the valid nodes collected from the verification platform;

Step 107 adjusts the individuals of the population based on the fitness until the adjusted test case is output when the preset output condition is satisfied.

In this embodiment, adjusting the population individuals refers to changing the population to form new individuals, and then forming a new population from each new individual, and returning to step 104 to implement population iteration after forming the new population.

The above-mentioned test case generation method generates a condition tree for the target condition coverage test point by analyzing the code to be tested and using preset rules, intercepts the required effective nodes from the condition tree, and generates an initial weight value for each effective node, and then In the population iteration process of the genetic algorithm, it is updated according to the coverage times of effective nodes and the initial weight value, and then the population is adjusted according to the updated weight value to calculate the adaptability of the population individual, and the coverage test can be generated by continuous iteration to cover the target condition The number of redundant test cases is greatly reduced, and the automatic adjustment of fitness according to the node hit probability can significantly speed up the iterative process and improve the convergence speed, so that the generation of test cases no longer depends on the technical level of code testers, reducing the cost of testing. The required labor cost has significantly improved the test efficiency.

In some embodiments, the preset rules include at least one of the following:

Taking the combined logic corresponding to the target condition coverage test point as the root node;

Processing the combination logic based on the preset sub-item segmentation rules, using the combination logic itself as a parent node, and using the child items of the combination logic of the parent node as child nodes;

For branch conditions, the branch that causes the variable change of this node is listed as a child node;

When the node is a code input parameter, the output of a code invisible component, or the value change of a node variable has no logical relationship with other variables, the node is a leaf node and will not be decomposed;

The minimum number of nodes passed by each node to reach the root node is taken as the layer number of the node.

In some embodiments, the preset subitem segmentation rules include at least one of the following:

Segmentation rule 1. Convert logical operators to obtain combinational logic including only AND relation or OR relation;

Segmentation rule 2, setting the priority of the AND relationship, or the relationship, to follow the design language rules of the code to be tested;

Segmentation rule 3. There is only a single form of logical relationship between sub-items, and each sub-item corresponds to a sub-node;

Segmentation rule 4. Before decomposing sub-items, it is necessary to simplify the combination logic according to the logical operation rules to ensure that each item does not contain logical operators;

Segmentation rule 5. For the logical relationship of the root node, after conversion by segmentation rule 1, segmentation rule 2, and segmentation rule 4, the OR relationship is changed to an AND relationship;

Segmentation rule 6. If the simplified logical relation contains both an OR relation and an AND relation, first decompose the sub-nodes of the low-priority logical relation, and then decompose the high-priority sub-nodes from the generated sub-nodes .

In some embodiments, the preset output condition includes that there is an individual in the current population whose test can cover the target condition coverage test point, and the number of iterations of the current population reaches the preset number of iterations.

In this embodiment, by setting two conditions for stopping iterations, excessive invalid iterations can be effectively avoided, and stable execution of the genetic algorithm can be ensured.

In this embodiment, the aforementioned step 107 is to adjust the individuals of the population based on the fitness until the preset output conditions are met, and output the adjusted test cases, including:

In response to the presence of an individual test in the current population that can cover the target condition coverage test point or the number of iterations in the current population reaches the preset number of iterations, then stop the iteration and output the test case corresponding to the individual or when the preset number of iterations is reached The test cases corresponding to the individuals in the population.

In some embodiments, the aforementioned step 107, adjusting the individuals of the population based on the fitness until the preset output conditions are met, outputting the adjusted test cases further includes:

In response to the fact that none of the individual tests in the current population can cover the target condition coverage test point and the number of iterations of the current population does not reach the preset number of iterations, the population change operation is performed on the individuals in the current population based on the fitness to generate the next generation population .

In some embodiments, the aforementioned step 107, adjusting the individuals of the population based on the fitness until the preset output conditions are met, outputting the adjusted test cases further includes:

In response to generating the next-generation population, return to the step of using the individuals of the population as the input of the code to be tested and running the code to be tested on the verification platform.

In some embodiments, the population changing operation includes at least one of selection, crossover and mutation operations.

It should be noted that, in the specific implementation process, it is preferable to use the above three operations in combination, for example, the selection operation is performed on the population individuals first, then the crossover operation is performed, and finally the mutation operation is performed to obtain a new population.

In some embodiments, performing node interception on the conditional tree in step 103 to obtain valid nodes includes:

Select all nodes whose node layer number is less than or equal to the preset intercepted node layer number from the generated condition tree as valid nodes.

In some embodiments, assigning initial weight values to the effective nodes in the aforementioned step 103 includes:

Assign an initial weight value to each valid node according to the following formula 1:

公式一；

formula one;

表示节点的层号，/>

表示节点在层中的索引号，/>

表示预设截取节点层数，

表示第/>

层中的节点总数，/>

表示逻辑属性系对应的值，/>

表示第/>

层、第/>

个节点的初始权重值。in,

Indicates the layer number of the node, />

Indicates the index number of the node in the layer, />

Indicates the number of preset interception node layers,

Indicates the first />

the total number of nodes in the layer, />

Indicates the value corresponding to the logical attribute system, />

Indicates the first />

Layer, No. />

The initial weight value of each node.

In some embodiments, the value selection rules corresponding to logical attributes are as follows:

；If the logical attribute of the node is "and", then

;

等于此节点对应的父节点所包括的子节点个数。If the logical attribute of the node is "or", then

Equal to the number of child nodes included in the parent node corresponding to this node.

In some embodiments, the aforementioned step 105, updating the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value includes:

Update the weight value for each valid node according to the following formula two;

公式二；

formula two;

表示种群进化的代数且/>

大于等于1，/>

表示第/>

层、第/>

个节点在使用第/>

代种群进行测试时被覆盖的次数，/>

表示种群中个体总数量，/>

表示第/>

次迭代时第/>

层、第/>

个节点的更新权重值。in,

represents the algebra of population evolution and />

greater than or equal to 1, />

Indicates the first />

Layer, No. />

node is using the />

The number of times the generation population is covered when testing, />

Indicates the total number of individuals in the population, />

Indicates the first />

At iteration />

Layer, No. />

The updated weight value of each node.

In some embodiments, the aforementioned step 106, calculating the fitness of individuals in the population according to the updated weight value and the coverage status of the valid nodes collected from the verification platform, includes:

Calculate the fitness of each individual in the population according to the following formula three;

公式三；

formula three;

表示第/>

代种群中，第/>

个个体的适应度，/>

表示使用第/>

代种群的第/>

个个体测试时，第/>

层的第/>

个节点的覆盖状态对应的值。in,

Indicates the first />

In the generation population, the />

The fitness of an individual, />

Indicates the use of the />

No./> of generation population

When an individual is tested, the first />

Layer />

The value corresponding to the coverage state of a node.

In some embodiments, the value selection rules corresponding to the coverage state of a node are as follows:

值为1；If the node is covered then

value is 1;

值为0。If the node is not covered then

The value is 0.

In some embodiments, the code under test is used to verify the DUT. Wherein, DUT is the abbreviation of Design under Test. Preferably, the DUT can be a memory chip, a processor, a SoC chip, etc. according to functions.

In some embodiments, each test case of the code under test includes a plurality of input parameters.

In yet another embodiment, in order to facilitate the understanding of the solution of the present invention, please refer to the following for various software test application scenarios as shown in Figure 2. This embodiment takes a specific combination logic in a test case as an example to describe in detail the present invention for Code conditional coverage testing provides another test case generation method. Specifically, the conditional tree for the target coverage test point is generated through code text analysis and certain rules, and the nodes with the required number of layers are intercepted, and the adaptation of each node is generated according to the algorithm. degree coefficient, and then according to the adaptive genetic algorithm and the designed fitness function, through continuous iteration to generate test cases that can cover the target condition coverage test points, in the process, the fitness function is automatically adjusted according to the node hit probability to speed up the iteration process. The specific implementation process is as follows:

Step 1: Analyze the code to generate a conditional tree; analyze the program structure and logic according to the syntax of the code programming language, and use specific rules to generate a conditional tree. The rules for generating a conditional tree are as follows:

(1) The root node is the target condition coverage test point, and the root node is the combination logic.

(2) For combinatorial logic, the parent node is the combinatorial logic itself, and the child node is the child of the parent node's combinatorial logic. The sub-item division rules are as follows:

Segmentation rule 1, logical operators include: and (&&), or (||), not (!), only processing && and || operators, ! Expressions are converted, such as !(a==1) into (a!=1);

Segmentation rule 2, the priority of && and || follows the corresponding programming language rules;

Segmentation rule 3. There is only a single form of logical relationship between sub-items. Each sub-item corresponds to a sub-node. In addition to sub-items, sub-nodes also include logical relationship attributes. If the logical relationship between sub-items in this layer is ||, the logic The relationship attribute is "or", if the logical relationship between sub-items in this layer is &&, then the logical relationship attribute is "and";

Segmentation rule 4. Before decomposing sub-items, the combination logic needs to be simplified according to the logical operation rules to ensure that each item does not contain logical operators;

Segmentation rule 5, for the root node, after the logic relation expression is converted by segmentation rule 1, segmentation rule 2, and segmentation rule 4, the || relation;

Segmentation rule six, if the simplified logical relation contains both || and &&, decompose the low-priority logical relation sub-nodes first, and then decompose the high-priority sub-nodes from the generated sub-nodes;

For example, for the parent node combination logic (a>1)&&((b==5)&&(c!=7))||((d<7)&&(!(e!=1))), according to the split Rule 1 first turns into (a>1)&&((b==5)&&(c!=7))||((d<7)&&(e==1));

If the current node is not the root node, turn it into (a>1)&&(b==5)&&(c!=7)||(d<7)&&(e==1) according to the splitting rule; according to the splitting rule Third, the || operator has low priority and can be decomposed into two layers of sub-formulas. The first layer is (a>1)&&(b==5)&&(c!=7) and (d<7)&&(e ==1), the logical relationship attributes are all "or", the second layer is (a>1), (b==5), (c!=7), (d<7), (e==1) , the logical relationship attributes are all "and".

If the current node is the root node, according to the splitting rule five turns into (a>1)&&(b==5)&&(c!=7)&&(d<7)&&(e==1), then the child node only There is a layer (a>1), (b==5), (c!=7), (d<7), (e==1), and the logical relationship attribute is "and".

(3) For the branch condition, the branch that causes the variable change of this node is listed as a child node, and the logical relationship attributes are all "or";

For example, for a node (a>1), the node variable is a, if the state variable state, when state==1 and state==3 will cause a to change, then the child node of (a>1) is (state== 1) and (state==3), the logical relationship attributes are both "or".

(4) When a node is a code input parameter, an output of an invisible component of code, or a value change of a node variable has no logical relationship with other variables, the node is a leaf node and will not be decomposed.

(5) Each child node is set as the layer number of the node according to the minimum number of nodes passed by to reach the root node.

The conditional tree generated based on the above rules is shown in Figure 3. Each ellipse represents a node, and the target relational expression (conditional relational expression or conditional, logical mixed relational expression) is marked inside, starting from the current node and ending at the parent node with an arrow The logical relationship attributes of the current node are marked on the line segment.

Step 2, intercept corresponding nodes from the conditional tree according to the configuration of interception layers. The configuration parameter is the number of node layers to be intercepted, which is used to determine which nodes can be used to calculate the fitness, and the selected nodes are called effective nodes. The effective node selection rule is: when the node layer number is less than or equal to the configured layer number, the node is a valid node

Step 3: Assign initial weight values to each valid node according to the algorithm. The initial weight value of a valid node is set according to the following rules:

公式一；

formula one;

The meaning of the parameters in formula 1 is explained as follows:

表示节点的层号；

Indicates the layer number of the node;

表示节点在层中的索引号，不妨以图3中第一层节点为例，该层包括四个节点，从左到右的四个节点其索引号可以分别记作1至4；

Indicates the index number of the node in the layer. Take the first layer node in Figure 3 as an example. This layer includes four nodes. The index numbers of the four nodes from left to right can be recorded as 1 to 4 respectively;

表示预设截取节点层数；

Indicates the number of preset interception node layers;

表示第/>

层中的节点总数；

Indicates the first />

the total number of nodes in the layer;

表示逻辑属性系对应的值，若节点的逻辑属性为“与”，则/>

；若节点的逻辑属性为“或”，则/>

等于此节点对应的父节点所包括的子节点个数；

Indicates the value corresponding to the logical attribute. If the logical attribute of the node is "and", then />

; If the logical attribute of the node is "or", then />

Equal to the number of child nodes included in the parent node corresponding to this node;

表示第/>

层、第/>

个节点的初始权重值。

Indicates the first />

Layer, No. />

The initial weight value of each node.

Step 4, refresh the node weight. The node weight is used to calculate the fitness of the individual in the genetic algorithm. The weight of each effective node is continuously optimized as the population evolves to weaken the influence of non-key nodes on the fitness of the individual, and then make the population increase the coverage of key nodes. evolution. Node weights are updated as follows:

公式二；

formula two;

The meaning of the parameters in formula 2 is explained as follows:

表示种群进化的代数且/>

大于等于1；

represents the algebra of population evolution and />

greater than or equal to 1;

表示第/>

层、第/>

个节点在使用第/>

代种群进行测试时被覆盖的次数；

Indicates the first />

Layer, No. />

node is using the />

The number of times covered when testing on behalf of the population;

表示种群中个体总数量；

Indicates the total number of individuals in the population;

表示第/>

次迭代时第/>

层、第/>

个节点的更新权重值。

Indicates the first />

At iteration />

Layer, No. />

The updated weight value of each node.

Step five, genetic parameter setting. Set parameters such as the crossover rate and mutation rate of the genetic algorithm.

Step six, population initialization. Genetic algorithm chromosomes are generated by binary encoding of input parameters, and populations are generated by constrained random algorithms.

Step seven, run the verification platform and the code to be tested. Decode individual chromosomes to generate input parameters in test cases, input these parameters to the verification platform, and run the verification platform and the code under test.

Step eight, collecting node coverage information. After testing the test cases corresponding to all individuals in the population, count the coverage of each effective node, including whether it is covered in each test case and the number of times it is covered in the current generation population.

Step 9, calculate the individual fitness of the population. The individual fitness is calculated according to the following formula:

公式三；

formula three;

The meaning of the parameters in Formula 3 is explained as follows:

表示第/>

代种群中，第/>

个个体的适应度

Indicates the first />

In the generation population, the />

individual fitness

表示使用第/>

代种群的第/>

个个体测试时，第/>

层的第/>

个节点的覆盖状态对应的值，若节点被覆盖则/>

值为1，若节点未被覆盖则/>

值为0。

Indicates the use of the />

No./> of generation population

When an individual is tested, the first />

Layer />

The value corresponding to the coverage state of a node, if the node is covered then />

The value is 1, if the node is not covered then />

The value is 0.

Step ten, population individual analysis. Mainly analyze whether the condition coverage test point is reached. If the target condition coverage test point is reached or the maximum allowed evolution algebra is reached, the search process will be stopped. If the target condition coverage test point is reached, the test case will be output. If the target condition coverage test point is not reached, the search will fail and the operating parameters need to be adjusted. , search again; if the target condition coverage test point is not reached and the maximum allowable evolution algebra is not reached, the selection, crossover and mutation operations of the traditional genetic algorithm are performed to generate a new population, and the process from step 4 to step 10 is repeated .

A method for generating test cases in this embodiment has at least the following beneficial technical effects:

(1) Propose the construction strategy of the conditional tree, design the adaptive weight generation and adjustment rules based on the conditional tree and node coverage information, and then determine the corresponding fitness function to optimize the convergence speed of the algorithm.

(2) Based on the adaptive genetic algorithm, the initial population is generated in a constrained random manner, and the fitness evaluation of the test cases generated by decoding the individual population is performed according to the operating information. Based on the evaluation results, an optimized new population is generated through selection, crossover, and mutation operations .

(3) Continuously optimize fitness calculation parameters according to code coverage information, increase the weight of key (bottleneck) conditions, and accelerate algorithm convergence.

(4) The code test process information is considered in the process of generating new test cases, which greatly reduces the proportion of redundant test cases, and the adaptive fitness function is used to further reduce this proportion.

(5) It only needs to manually configure the running parameters, and the test case generation process does not require manual intervention, and the conditional node does not touch the interior of the invisible component of the code, and is not affected by it, which significantly reduces the difficulty and complexity of test case generation and reduces labor costs .

In some embodiments, as shown in FIG. 4, the present invention also provides a test case generation device 200, the test case generation device includes:

The generation module 201 is used to analyze the code to be tested to cover the test point with the combinational logic as the target condition, and use preset rules to generate a condition tree for the target condition coverage test point;

A construction module 202, configured to construct a population of genetic algorithms, wherein each individual in the population is an initial test case of the code to be tested;

An intercepting module 203, configured to perform node interception on the conditional tree to obtain valid nodes, and assign initial weight values to the valid nodes;

The running module 204 is used to use the individuals of the population as the input of the code to be tested, and run the code to be tested on the verification platform;

An update module 205, configured to update the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value;

A calculation module 206, configured to calculate the fitness of individuals in the population according to the update weight value and the coverage status of the valid nodes collected from the verification platform;

The output module 207 is configured to adjust the individuals of the population based on the fitness until the preset output condition is met and output the adjusted test cases.

The above-mentioned test case generation device generates a condition tree for the target condition coverage test point by analyzing the code to be tested and using preset rules, intercepts the required effective nodes from the condition tree, and generates an initial weight value for each effective node, and then In the population iteration process of the genetic algorithm, it is updated according to the coverage times of effective nodes and the initial weight value, and then the population is adjusted according to the updated weight value to calculate the adaptability of the population individual, and the coverage test can be generated by continuous iteration to cover the target condition The number of redundant test cases is greatly reduced, and the automatic adjustment of fitness according to the node hit probability can significantly speed up the iterative process and improve the convergence speed, so that the generation of test cases no longer depends on the technical level of code testers, reducing the cost of testing. The required labor cost has significantly improved the test efficiency.

It should be noted that, for specific limitations on the test case generation device, reference may be made to the above-mentioned limitations on the test case generation method, which will not be repeated here. Each module in the above-mentioned test case generation device can be fully or partially realized by software, hardware and combinations thereof. The above-mentioned modules can be embedded in or independent of the processor in the computing device in the form of hardware, and can also be stored in the memory of the computing device in the form of software, so that the processor can call and execute the corresponding operations of the above-mentioned modules.

According to another aspect of the present invention, a kind of code test method is provided, described code test method adopts the test case output of test case generation method described in the above embodiment to code execution test, and described test case generation method comprises:

Analyzing the code to be tested to use combinational logic as a target condition coverage test point, and using preset rules to generate a condition tree for the target condition coverage test point;

Construct a population of genetic algorithms, wherein each individual in the population is an initial test case of the code to be tested;

Performing node interception on the conditional tree to obtain valid nodes, and assigning initial weight values to the valid nodes;

Using the individuals of the population as the input of the code to be tested, and running the code to be tested on the verification platform;

updating the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value;

calculating the fitness of individuals in the population according to the update weight value and the coverage status of the valid nodes collected from the verification platform;

The individuals of the population are adjusted based on the fitness until the preset output conditions are met, and the adjusted test cases are output.

According to another aspect of the present invention, a computing device is provided, which may be a server, and its internal structure diagram is shown in FIG. 5 . The computing device includes a processor, memory, network interface, and database connected by a system bus. Wherein, the processor of the computing device is used to provide computing and control capabilities. The memory of the computing device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computing device program, and a database. The internal memory provides an environment for the operation of the operating system and computing device programs in the non-volatile storage medium. The computing device's database is used to store data. The network interface of the computing device is used to communicate with external terminals via a network connection. When the computing device program is executed by the processor, the test case generation method described above is realized. Specifically, the test case generation method includes the following steps:

Analyzing the code to be tested to use combinational logic as a target condition coverage test point, and using preset rules to generate a condition tree for the target condition coverage test point;

Constructing a population of genetic algorithms, wherein each individual in the population is an initial test case of the code to be tested;

Performing node interception on the conditional tree to obtain valid nodes, and assigning initial weight values to the valid nodes;

Using the individuals of the population as the input of the code to be tested, and running the code to be tested on the verification platform;

updating the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value;

calculating the fitness of individuals in the population according to the update weight value and the coverage status of the valid nodes collected from the verification platform;

The individuals of the population are adjusted based on the fitness until the preset output conditions are met, and the adjusted test cases are output.

According to yet another aspect of the present invention, a computing device readable storage medium is provided, on which a computing device program is stored, and when the computing device program is executed by a processor, the method for generating test cases described above is implemented, specifically, including Perform the following steps:

Analyzing the code to be tested to use combinational logic as a target condition coverage test point, and using preset rules to generate a condition tree for the target condition coverage test point;

Constructing a population of genetic algorithms, wherein each individual in the population is an initial test case of the code to be tested;

Performing node interception on the conditional tree to obtain valid nodes, and assigning initial weight values to the valid nodes;

Using the individuals of the population as the input of the code to be tested, and running the code to be tested on the verification platform;

updating the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value;

calculating the fitness of individuals in the population according to the update weight value and the coverage status of the valid nodes collected from the verification platform;

The individuals of the population are adjusted based on the fitness until the preset output conditions are met, and the adjusted test cases are output.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be realized by instructing related hardware through a computing device program, and the computing device program can be stored in a non-volatile computing device. When the program of the computing device is read from the storage medium, it may include the processes of the embodiments of the above-mentioned methods when executed. Wherein, any references to memory, storage, database or other media used in the various embodiments provided in the present application may include non-volatile and/or volatile memory. Nonvolatile memory can include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Chain Synchlink DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The technical features of the above embodiments can be combined arbitrarily. To make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be It is considered to be within the range described in this specification.

The above-mentioned embodiments only represent several implementation modes of the present application, and the description thereof is relatively specific and detailed, but it should not be construed as limiting the scope of the patent for the invention. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the scope of protection of the patent application should be based on the appended claims.

Claims (21)

1. A test case generation method is characterized in that, the test case generation method comprises: Analyzing the code to be tested to use combinational logic as a target condition coverage test point, and using preset rules to generate a condition tree for the target condition coverage test point; Construct a population of genetic algorithms, wherein each individual in the population is an initial test case of the code to be tested; Performing node interception on the conditional tree to obtain valid nodes, and assigning initial weight values to the valid nodes; Using the individuals of the population as the input of the code to be tested, and running the code to be tested on the verification platform; updating the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value; calculating the fitness of individuals in the population according to the update weight value and the coverage status of the valid nodes collected from the verification platform; The individuals of the population are adjusted based on the fitness until the preset output conditions are met, and the adjusted test cases are output. 2. The test case generation method according to claim 1, wherein the preset rules include at least one of the following: Taking the combined logic corresponding to the target condition coverage test point as the root node; Carry out sub-item segmentation for the combination logic based on preset sub-item segmentation rules, so that the combination logic itself is used as a parent node, and the sub-items of the combination logic of the parent node are used as child nodes; For branch conditions, the branch that causes the variable change of this node is listed as a child node; When the node is a code input parameter, the output of a code invisible component, or the value change of a node variable has no logical relationship with other variables, the node is a leaf node and will not be decomposed; The minimum number of nodes passed by each node to reach the root node is taken as the layer number of the node. 3. The test case generation method according to claim 2, wherein the preset sub-item segmentation rules include at least one of the following: Segmentation rule 1. Convert logical operators to obtain combinational logic including only AND relation or OR relation; Segmentation rule 2, setting the priority of the AND relationship, or the relationship, to follow the design language rules of the code to be tested; Segmentation rule 3. There is only a single form of logical relationship between sub-items, and each sub-item corresponds to a sub-node; Segmentation rule 4. Before decomposing sub-items, it is necessary to simplify the combination logic according to the logical operation rules to ensure that each item does not contain logical operators; Segmentation rule 5. For the logical relationship of the root node, after conversion by segmentation rule 1, segmentation rule 2, and segmentation rule 4, the OR relationship is changed to an AND relationship; Segmentation rule 6. If the simplified logical relation contains both an OR relation and an AND relation, first decompose the sub-nodes of the low-priority logical relation, and then decompose the high-priority sub-nodes from the generated sub-nodes . 4. The test case generation method according to claim 3, wherein the preset output conditions include that there is a certain individual test in the current population that can cover the target condition coverage test point, and the number of iterations of the current population reaches the preset number of iterations. 5. The test case generation method according to claim 4, wherein the step of adjusting the individual of the population based on the fitness until the preset output condition is met, the step of outputting the adjusted test case includes : In response to the presence of an individual test in the current population that can cover the target condition coverage test point or the number of iterations in the current population reaches the preset number of iterations, then stop the iteration and output the test case corresponding to the individual or when the preset number of iterations is reached The test cases corresponding to the individuals in the population. 6. test case generating method according to claim 4, is characterized in that, described based on described fitness, the step of outputting the adjusted test case when the individual of described population is adjusted until satisfying preset output condition, also include: In response to the fact that none of the individual tests in the current population can cover the target condition coverage test point and the number of iterations of the current population does not reach the preset number of iterations, the population change operation is performed on the individuals in the current population based on the fitness to generate the next generation population . 7. test case generating method according to claim 6, is characterized in that, the step of outputting the adjusted test case when the individual of described population is adjusted until satisfying preset output condition based on described fitness, also include: In response to generating the next-generation population, return to the step of using the individuals of the population as the input of the code to be tested and running the code to be tested on the verification platform. 8. The test case generation method according to claim 7, wherein the population change operation includes at least one of selection, crossover and mutation operations. 9. test case generation method according to claim 7, is characterized in that, described conditional tree is carried out node interception to obtain the step of valid node, comprising: Select all nodes whose node layer number is less than or equal to the preset intercepted node layer number from the generated condition tree as valid nodes. 10. test case generating method according to claim 7, is characterized in that, described step of giving initial weight value for described effective node, comprises: Assign an initial weight value to each valid node according to the following formula 1:

formula one; in,

Indicates the layer number of the node, />

Indicates the index number of the node in the layer, />

Indicates the number of preset interception node layers, />

Indicates the first />

the total number of nodes in the layer, />

Indicates the value corresponding to the logical attribute system, />

Indicates the first />

Layer, No. />

The initial weight value of each node. 11. The test case generation method according to claim 10, characterized in that, the value rules corresponding to the logic attribute are as follows: If the logical attribute of the node is "and", then

; If the logical attribute of the node is "or", then

Equal to the number of child nodes included in the parent node corresponding to this node. 12. test case generation method according to claim 11, is characterized in that, described according to the coverage number of times and described initial weight value of the described valid node that collects from described verification platform to the weight value of described valid node Steps for updating to obtain updated weight values include: Update the weight value for each valid node according to the following formula two;

formula two; in,

represents the algebra of population evolution and />

greater than or equal to 1, />

Indicates the first />

Layer, No. />

node is using the />

The number of times the generation population is covered when testing, />

Indicates the total number of individuals in the population, />

Indicates the first />

At iteration />

Layer, No. />

The updated weight value of each node. 13. test case generation method according to claim 12, is characterized in that, described according to described update weight value and the coverage status of the described effective node that collects from described verification platform, calculate the fitness of individual in the population steps, including: Calculate the fitness of each individual in the population according to the following formula three;

formula three; in,

Indicates the first />

In the generation population, the />

The fitness of an individual, />

Indicates the use of the />

generation population

When an individual is tested, the first />

Layer />

The value corresponding to the coverage state of a node. 14. The test case generation method according to claim 13, characterized in that, the value value rules corresponding to the coverage state of the node are as follows: If the node is covered then

value is 1; If the node is not covered then

The value is 0. 15. The test case generation method according to claim 1, wherein the code to be tested is used to verify the DUT. 16. The test case generation method according to claim 15, wherein the DUT comprises a memory chip, a processor, and a SoC chip. 17. The test case generation method according to claim 1, wherein each test case of the code to be tested includes a plurality of input parameters. 18. A test case generating device, characterized in that, the test case generating device comprises: A generating module, which is used to analyze the code to be tested to cover the test point with combinational logic as the target condition, and use preset rules to generate a condition tree for the target condition coverage test point; A building block for constructing a population of genetic algorithms, wherein each individual in the population is an initial test case of the code to be tested; An intercepting module, configured to perform node interception on the conditional tree to obtain valid nodes, and assign initial weight values to the valid nodes; An operation module, configured to use the individual of the population as the input of the code to be tested, and run the code to be tested on the verification platform; An update module, configured to update the weight value of the valid node according to the coverage times of the valid node collected from the verification platform and the initial weight value to obtain an updated weight value; A calculation module, configured to calculate the fitness of individuals in the population according to the update weight value and the coverage status of the valid nodes collected from the verification platform; An output module, configured to adjust the individuals of the population based on the fitness until a preset output condition is met and output an adjusted test case. 19. A code testing method, characterized in that the code testing method uses the test case output by the test case generation method according to any one of claims 1-17 to execute the test on the code. 20. A computing device, comprising: at least one processor; and A memory, the memory stores a computing device program that can run in the processor, and the processor executes the test case generation method described in any one of claims 1-17 when executing the program. 21. A computing device-readable storage medium, the computing device-readable storage medium stores a computing device program, characterized in that, when the computing device program is executed by a processor, it executes any one of claims 1-17. Test case generation method.

Images