CN109271308A - Class based on search integrates the generation method of initial population in cycle tests problem - Google Patents
Class based on search integrates the generation method of initial population in cycle tests problem Download PDFInfo
- Publication number
- CN109271308A CN109271308A CN201710578788.8A CN201710578788A CN109271308A CN 109271308 A CN109271308 A CN 109271308A CN 201710578788 A CN201710578788 A CN 201710578788A CN 109271308 A CN109271308 A CN 109271308A
- Authority
- CN
- China
- Prior art keywords
- class
- tree
- initial population
- cycle tests
- sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3672—Test management
- G06F11/3688—Test management for test execution, e.g. scheduling of test suites
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Class based on search integrates the generation method of initial population in cycle tests problem.It has been firstly introduced into a constraint condition, that is, when testing the class of certain sequence, has not allowed to break the strong dependence between class.Then propose a kind of generation method in this constraint condition lower class cycle tests, including analysis system obtain relying on by force matrix, by matrix transposition and traverse to construct multiway tree or multiway tree forest, the tree that can merge merged, the class that completion does not traverse so obtain complete forest, by every one tree in random sequence level traversal forest to obtain class testing sequence；The present invention solves to a certain extent when solving class using intelligent algorithm (such as genetic algorithm and particle swarm algorithm) and integrating cycle tests, due to initial population ideal adaptation angle value is lower, total quality is poor and the problem of influence algorithm the convergence speed and optimizing result.The class testing sequence generated by this method is not lost into randomness as initial population, while total quality increases, strengthens the optimizing ability of simple generic algorithm and simple particle swarm algorithm.
Description
Technical field
The invention belongs to software testing technology fields, especially integrate cycle tests generation technique in the class based on search
In, a kind of generation method of the initial population (class testing sequence) under certain constraint condition.
Background technique
The determination of the integrated cycle tests of class is one of the key difficulties in object-oriented software integration testing.Object-oriented is soft
Part has the characteristics that encapsulation, inheritance and polymorphism, so its measuring technology and traditional area procedure-oriented software You Hen great
Not.Testing cost caused by time that class testing ordinal relation is found to software defect and design test pile, accordingly, it is determined that class
Integration testing sequence become integration testing in an important research problem.
During object-oriented software test, the development cost of test pile is often very high, so determining that class is integrated
The top priority of cycle tests is exactly to reduce the development cost of test pile.Based on the method for graph theory due to itself limitation, nothing
Method is suitable for large program.And when determining class testing sequence, need to consider to influence the various factors of test pile, therefore,
The problem of solving class testing sequence can be converted into optimization problem.Current existing method includes genetic algorithm (Genetic
Algorithm, GA) and particle swarm optimization algorithm (Particle Swarm Optimization, PSO) etc., they belong to people
Work intelligent algorithm.Such methods in order to indicate construction test pile cost, often using the overall complexity of sequence as fitness
Function, by generating a certain number of individual composition initial populations at random, a class testing sequence is as an individual.It is right later
Population carries out a series of evolutional operation, after certain algebra of evolving, the final optimal solution obtained under fitness function.
Currently, simple genetic algorithm and particle swarm algorithm are all to generate initial population, initial population by random device
The fitness of individual is lower, restricts convergence speed of the algorithm to a certain extent, causes to be unable to get in defined evolutionary generation
More preferably solve.Therefore, simple evolution algorithm is not able to satisfy actual needs gradually.
Summary of the invention
It is lower in order to solve existing method initial population individual adaptation degree, lead to the problem that optimizing is ineffective, this hair
It is bright to provide a kind of generation method of the class testing sequence (population at individual) under constraint condition.That is, introducing a kind of constraint condition pair
Initial population is pre-processed, constraint condition are as follows: when breaking that dependence generates class testing sequence between class, is not allowed to break between class
Strong dependence.Firstly, determining the dependence type between class；Then, the developing algorithm for proposing multiway tree, exist by force according to
Root node and leaf node of the target class and source class for the relationship of relying respectively as multiway tree；Finally, carrying out level time to multiway tree
It goes through, generates with randomness and meet the individual of constraint condition.Method executes a primary generation individual, so, execute number
It is identical as population scale.Wherein, population scale is generally twice to three times of class number.Thus individual institute's group that method generates
At initial population, do not lose randomness, while total quality increases, simple generic algorithm and simple population can be reinforced
The optimizing ability of algorithm.Method the following steps are included:
(1) it indicates for convenience, all classes of examining system is numbered from 1.
(2) strong dependence matrix M, M [i, j]=1 between class are obtained by static analysis indicates that class i relies on by force class j, if M [i,
J]=there is no strong dependences for 0 expression, two classes.
(3) it by matrix M transposition and traverses.Transposed matrix M ' is obtained first, and the purpose of transposition is that strong dependence is converted
Indicate that class i relies on by force class j at the front-rear position relationship in class testing sequence, such as M [i, j]=1, in order to meet constraint, in the sequence
Class j need to be before class i, and the matrix M ' after transposition can more intuitively indicate this context.M [i, j]=1 become M ' [j,
I]=1, indicate class j before class i.Then transposed matrix is traversed, M ' [j, i]=1 is worked as, then using class j as root node, class i conduct
Leaf node constructs a multiway tree.
(4) tree that can merge is merged, some class is root node in one tree, and such conduct in another one tree
Leaf node thinks that two trees can merge at this time.If there is no such class, then original state is kept.
(5) class each not traversed so far is obtained separately as one tree comprising all classes several trees
Forest.
(6) it is pressed by each tree in random sequence successively level traversal forest when some node has multiple child nodes
Random sequence traversal.
The obtained class testing sequence ensure that great randomness while meeting constraint condition.As
Class testing sequence reduces the overall complexity of individual, improves the total quality of population as initial population individual.
Detailed description of the invention
Fig. 1 is flow chart of the invention.
Fig. 2 is strong dependence matrix.
Fig. 3 is transposed matrix.
Fig. 4 is multiway tree.
Fig. 5 is the multiway tree after merging.
Fig. 6 is forest.
Detailed description of the invention
With reference to the accompanying drawings and examples to the detailed description of the invention.
In Fig. 1, the present invention includes that analysis system is relied on by force in the generation method of constraint condition lower class cycle tests
Matrix, by matrix transposition and traverse with construct multiway tree, merge the class that multiway tree, completion do not traverse obtain complete forest,
By every one tree in random sequence level traversal forest, traversing result is class testing sequence, and its step are as follows:
(1) analysis system obtains relying on matrix by force: an existing system, contains 10 classes, number 1-10 altogether.Fig. 2 is analysis
Strong dependence the matrix M, M [i, j]=1 obtained afterwards indicates that class i relies on by force class j, and being worth indicates that strong dependence is not present in two classes for 0.
In Fig. 2, M [1,3]=M [2,3]=M [4,9]=M [5,3]=M [6,5]=M [8,5]=1 illustrates class 1,2, and the last 5 relies on class 3, class semi-finals
Class 9, class 6 are relied on, the Final 8 relies on class 5.
(2) Fig. 3 is before matrix M ', M ' [i, j]=1 after transposition indicates that class i needs to come class j in the sequence.It is constraining
Under the conditions of, before class 3 will come class 1,2,5 in sequence, before class 5 will come class 6,8, before class 9 will come class 4, except this it
Relative position is any between outer class, and so there is no need to break strong dependence.Then, transposed matrix M ' is traversed, if M ' [i, j]
=1, then using i as root node, j constructs multiway tree as leaf node.Fig. 4 is the multiway tree constructed after Ergodic Matrices.
(3) in Fig. 4, class 5 is not only used as root node on one tree, but also leaf node is used as on another one tree, so
Class 5 is the node that can merge, and two trees are merged, and eliminates duplicate node.What Fig. 5 was indicated is the multiway tree after merging.
(4) after having merged multiway tree, if each class is separately as one there is also the class not traversed
Tree is just constituted named comprising 10 class 4 entitled trees 3, tree 9, tree 7, tree 10(with root node in this way) tree forest.
(5) finally, traversing 4 trees according to random sequence level, if some node has multiple child nodes, also by random
Order traversal.In this way, random sequence can be obtained, such as: 9-4-3-2-5-8-6-1-7-10 or 10-3-1-5-6-8-2-7-9-4
Deng.Verified, the class testing sequence tested does not break the strong dependence between class.
Claims (6)
1. the class based on search integrates the generation method of initial population in cycle tests problem it is characterized in that, being firstly introduced into one
Kind constraint condition: when testing the class of a certain sequence, do not allow to break the strong dependence between class.Then, the structure of multiway tree is proposed
Build algorithm.Finally traversal multiway tree obtains class sequence.This method mainly include the following steps include: 1) analysis system obtain by force according to
Rely matrix；2) by matrix transposition and traverse to construct multiway tree；3) merge multiway tree；4) using each class not traversed as
One tree and multiway tree form forest；5) class testing sequence is obtained by every one tree in random sequence level traversal forest.
2. the class according to claim 1 based on search integrates the generation method of initial population in cycle tests problem,
It is characterized in that, in step 1), analysis system obtains relying on matrix by force, if matrix is M, M [i, j]=1 indicates that class i relies on by force class
J, being worth indicates that strong dependence is not present in two classes for 0.
3. the class according to claim 1 based on search integrates the generation method of initial population in cycle tests problem,
It is characterized in that, in step 2, by matrix transposition and traverses to construct multiway tree, if the matrix after transposition is M ', M ' [i, j]=1
Before indicating that class i needs to come class j in the sequence.Ergodic Matrices M ' later, when working as M ' [i, j]=1, using i as root node, j makees
For leaf node, multiway tree is constructed.
4. the class according to claim 1 based on search integrates the generation method of initial population in cycle tests problem,
It is characterized in that, in step 3), when merging multiway tree, if some class is in two trees respectively as root node and leaf section
Point then merges identical node in this two trees, that is, realizes the merging of multiway tree.
5. the class according to claim 1 based on search integrates the generation method of initial population in cycle tests problem,
It is characterized in that, in step 4), if, using each class as one tree, obtained with step 2 there is also the class not traversed
The tree composition forest arrived.
6. the class according to claim 1 based on search integrates the generation method of initial population in cycle tests problem,
It is characterized in that, in step 5), by random sequence, level traversal successively is carried out to every one tree in forest, if certain node has
Multiple child nodes are also traversed by random sequence.The node traversed is sequentially output and obtains class testing sequence, is obtained in this way
The existing randomness of class testing sequence, and meet constraint condition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710578788.8A CN109271308A (en) | 2017-07-17 | 2017-07-17 | Class based on search integrates the generation method of initial population in cycle tests problem |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710578788.8A CN109271308A (en) | 2017-07-17 | 2017-07-17 | Class based on search integrates the generation method of initial population in cycle tests problem |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109271308A true CN109271308A (en) | 2019-01-25 |
Family
ID=65147663
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710578788.8A Pending CN109271308A (en) | 2017-07-17 | 2017-07-17 | Class based on search integrates the generation method of initial population in cycle tests problem |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109271308A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111880795A (en) * | 2020-07-29 | 2020-11-03 | 中国银联股份有限公司 | Front-end interface generation method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070162894A1 (en) * | 2006-01-11 | 2007-07-12 | Archivas, Inc. | Method of and system for dynamic automated test case generation and execution |
| CN102937933A (en) * | 2012-11-14 | 2013-02-20 | 中国矿业大学 | Class test sequence determining method based on testing level |
-
2017
- 2017-07-17 CN CN201710578788.8A patent/CN109271308A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070162894A1 (en) * | 2006-01-11 | 2007-07-12 | Archivas, Inc. | Method of and system for dynamic automated test case generation and execution |
| CN102937933A (en) * | 2012-11-14 | 2013-02-20 | 中国矿业大学 | Class test sequence determining method based on testing level |
Non-Patent Citations (1)
| Title |
|---|
| 张艳梅: "一种基于动态依赖关系的类集成测试方法", 《计算机学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111880795A (en) * | 2020-07-29 | 2020-11-03 | 中国银联股份有限公司 | Front-end interface generation method and device |
| CN111880795B (en) * | 2020-07-29 | 2024-03-12 | 中国银联股份有限公司 | Front-end interface generation method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103257921B (en) | Improved random forest algorithm based system and method for software fault prediction | |
| CN105095491B (en) | Process model restorative procedure based on Petri network basic structure | |
| CN103870260B (en) | The method and system of business interface exploitation | |
| CN103559129B (en) | Statistical regression test data generating method based on genetic algorithm | |
| CN109407654A (en) | A kind of non-linear causality analysis method of industrial data based on sparse depth neural network | |
| CN107423217B (en) | Black box fuzzy test method and system based on variation tree | |
| CN103605605B (en) | Based on the be dominant high-order mutation testing method of relation and two-stage genetic algorithm of statement | |
| CN105718368A (en) | Software testing data amplification method | |
| CN110825615A (en) | Software defect prediction method and system based on network embedding | |
| CN104360947B (en) | Multi-process deadlock detection method based on petri network | |
| CN110175085A (en) | A kind of Hadoop system abnormal cause diagnostic method and device using map analysis | |
| CN110942098A (en) | Power supply service quality analysis method based on Bayesian pruning decision tree | |
| CN109271308A (en) | Class based on search integrates the generation method of initial population in cycle tests problem | |
| CN103678123A (en) | Method applied to recognizing vulnerabilities of single-event soft errors in processor systems | |
| Tomašević et al. | Ergodic behaviour of a multi-type growth-fragmentation process modelling the mycelial network of a filamentous fungus | |
| CN104750610B (en) | Message-passing parallel program variant reduction method based on degree of being dominant | |
| CN105528296B (en) | A kind of class cluster test method of object-oriented software | |
| CN109783361A (en) | The method and apparatus for determining code quality | |
| CN104615679A (en) | Multi-agent data mining method based on artificial immunity network | |
| CN103729297A (en) | Test case generation method based on hierarchic genetic algorithm | |
| Fell et al. | An algebra of unbounded operators | |
| CN104461861B (en) | Path test data generation method based on EFSM models | |
| CN104199649B (en) | The path method for decomposing of interactive information between a kind of process for father and son | |
| CN104731695B (en) | A kind of unit test system and method for supporting the input of table driving bottom | |
| CN109190204A (en) | A kind of complex mechanical product module partition method based on complex network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190125 |