JP2008204405A - Software regression test system, regression test program and regression test method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently and automatically perform a regression test by automatically extracting a portion to be preferentially tested according to a changed part so as to avoid an excessive large number of test items when performing the test. <P>SOLUTION: This regression system performs a test by storing information on test cases of the tests performed in the past and combining a dependency relation acquired by analyzing the information and an object program. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention realizes a system for performing a regression test (regression test) for confirming whether or not an unexpected influence is caused by a change of a program in software development, or a system for performing a regression test. It is related with the method of performing the program or regression test.

  For software development efficiency and automation in software development, methods, systems, etc. have been devised for automatic generation of necessary tests, discovery of duplicate tests, and optimization.

Japanese Patent Application Laid-Open No. 2004-133830 discloses that the number of executions is recorded in advance in each test case, and test cases that are affected by the correction are extracted.
Patent Document 2 discloses a program test apparatus that selects each test case from all test cases in order of importance in the function of the entire program based on information on the program.

In addition, with regard to program dependency, many types of dependency such as data dependency and control dependency, and an analysis method (Non-patent Document 1) have been proposed.
Patent Document 3 discloses an influence analysis apparatus that analyzes a variable affected by a program modification, detects a calculation path by following the execution order of the program, and analyzes the influence of the modification applied to the program.

Further, Patent Document 4 discloses investigating the range of influence when a source program is modified by examining the dependency relationship between substitution sources and substitution destinations between variables.
In the correction work (debugging) in software development, whether the correction (change) made is correct, or the correction (change) made affects other parts that have been operating normally until then. It is essential to do a regression test to see if it causes any problems. However, if the scale of the software grows and the number of test cases becomes enormous, it is impossible to re-execute all tests for each change.

Therefore, it is required to narrow down the test cases related to the change contents and perform a regression test. However, the scale of the software is large and the number of test cases is large, and it is very difficult to narrow down and execute test cases necessary for the regression test manually. Another problem is that because the software is becoming more complex, the impact of the changes cannot be fully understood and necessary test cases are overlooked.
Japanese Patent Publication No.8-1609 JP 63-116248 A Japanese Patent No. 2634011 JP-A-8-190475 Frank Tip, A Survey of Program Slicing Techniques, Journal of Programming Languages, Vol. 3, no. 3, pp. 121-189, 1995

  Therefore, the problem of the present invention is that, when a regression test is performed to confirm whether an unexpected influence has occurred due to a change in the program, the test items are not changed to an enormous amount according to the changed part. It is to automatically extract a portion to be preferentially tested, and to enable a regression test to be performed efficiently and automatically.

  In order to solve the above-described problems, in the present invention, information on test cases of tests performed in the past is stored, and the test is performed by combining the information and the dependency obtained by analyzing the target program. Regression system that performs

In other words, in order to narrow down to test cases related to the changed part,
(1) what part of the software will be executed in each test case, and (2) where in the software the impact of the changes made will be.
I have to know.

  Therefore, in the present invention, in order to realize the above (1), the statement executed when the test is performed for each test case is associated with the test case, the information is stored, and the above (2) is realized. In addition, the program to be analyzed is analyzed, dependency relationships are extracted, and the dependency relationship is used to estimate a portion that has a dependency relationship with the changed portion and is expected to change the execution result.

  Then, analyze how much dependency there is with the changed part from the number and type of dependency, and test that has been executed in the past for the part that is strongly dependent on the changed part, that is, strongly affected by the change A corresponding test case is extracted based on the correspondence information between the case and the Statement in the program, and the test is executed.

  By configuring in this way, it is possible to reliably crush a test case having a strong dependency from the changed portion, and it is possible to improve the efficiency of the regression test.

  According to the present invention, it is possible to efficiently and automatically perform a regression test in correction work (debugging) in software development. As a result, the efficiency of software development itself can be increased.

Embodiments of the present invention will be described below with reference to the drawings.
FIG. 1 shows a software regression test system 101 of the present invention.
The regression test system 101 includes a program storage unit 102, a dependency relationship storage unit 103, a user interface 104, a program analysis unit 105, a test execution unit 106, a dependency relationship analysis unit 107, a test case storage unit 108, and a statement ID storage unit 109. It consists of

The program storage unit 102 stores a source file of the test target program.
The dependency relationship storage unit 103 is a database that stores information about dependency relationships between statements in a program. For each source file, each statement has a list of affected statements and a list of affected statements.

FIG. 2 shows an outline of the dependency relationship storage unit 103. FIG. 3 shows a data structure when FIG. 2 is realized by the dependency relationship storage unit 103.
The test case storage unit 108 is a database that stores information about a program portion executed when the test case is executed. Each test case has a list of Statements in the program that is actually executed when the test case is executed.

FIG. 4 shows an outline of the test case storage unit 108. FIG. 5 shows a data structure when FIG. 4 is realized by the test case storage unit 108.
The Statement ID storage unit 109 stores a table indicating correspondence between each Statement of the test target program and an identifier (ID).

  The user interface 104 mediates between the user and the system, such as receiving an input such as a test execution instruction from a user such as a software developer, and displaying and notifying a test result.

  The program analysis unit 105 statically analyzes the test target program, extracts a dependency relationship in the program, and outputs the dependency relationship to the dependency relationship storage unit 103. Further, when the program is partially corrected or changed, the dependency relationship is extracted again for the related portion from the corrected or changed portion, and updated to the dependency relationship storage unit 103.

  The dependency relationship analysis unit 107 receives information about the modified and changed part of the program from the test execution unit 106 and uses the dependency relationship stored in the dependency relationship storage unit 103 to relate to the corrected and changed portion. A certain part is analyzed, and the result is output to the test execution unit 106.

The test execution unit 106 selects and executes a necessary test case from the test case storage unit 108 based on the result output from the dependency relationship analysis unit 107.
FIG. 6 shows a flow of processing when selecting and executing a test related to the corrected / changed part when the program is corrected / changed in the regression test system 101 shown in FIG.

  Before the execution of FIG. 6, the test target program stored in the program storage unit 102 is analyzed by the program analysis unit 105, and information related to the dependency relationship is stored in the dependency relationship storage unit 103. The test cases relating to past tests are stored in the test case storage unit 108.

  In the following description, “priority” is used as an index indicating the strength of the dependency relationship with the modified and changed Statement. The “priority” is set to increase according to the type of dependency relationship that has been changed or changed, so that the priority value can be increased. The larger the value, the lower the dependency relationship of the Statement and the lower the relationship with the changed Statement.

Further, it is assumed that the type of dependency relationship is at least one of data dependency, control dependency, and function call.
First, in S601, the test execution unit 106 determines, to the dependency relationship analysis unit 107, the modified / changed Statement in the program and the degree of strong association from the modified / changed Statement to extract the priority level. To be notified.

  In step S <b> 602, the dependency relationship analysis unit 107 extracts related statements using the dependency relationship stored in the dependency relationship storage unit 103 based on the notified modified and changed statement and the priority threshold value. I do. Details of this extraction process are shown in FIG.

In FIG. 7, a set S1, S2, and S3 of a pair of Statement and priority is prepared in S701, and all of them are emptied.
In S 702, a pair of the value 0 of the priority and the value of the modified and changed Statement in the program is entered in S 1 and S 2.

In step S703, a pair of a statement having a dependency relationship with the statement in S2 and its priority are entered in S3. Details of the processing of S703 are shown in FIG.
In FIG. 8, in S801, a set S2 ′ of State and priority pairs is prepared, and the contents of S2 are copied. In S802, it is determined whether S2 ′ is empty. If S2 ′ is empty (Yes), the process of FIG. 8 is terminated, and the process proceeds to S704 of FIG. If S2 ′ is not empty (No), the process proceeds to S803.

In step S803, a single statement / priority pair (s, p) is extracted from S2 ′.
In S804, using the dependency relationship stored in the dependency relationship storage unit 103, for the statements s1, s2, s3,... , ...

  In S805, (s1, p1), (s2, p2), (s3, p3),... Are put in S3. At this time, the Statement already in S3 is replaced when the priority value of the new pair is small, that is, when the dependency is stronger, and is discarded when the priority value is large, that is, when the dependency is weak.

In this way, by the process of FIG. 8, a pair of a State and its priority having a dependency relationship with the State included in S2 can be obtained in S3.
Then, the process proceeds to S704 in FIG. In S704, those whose priority level exceeds the threshold are removed from S3. In S705, out of the pairs in S3, those whose Statement is in S1 and whose priority value is larger in S1 than in S3 are removed from S3. That is, the weak dependency relationship is removed from S3.

  In S706, the Statement contained in S3 is entered in S1. At this time, if a pair for the same Statement already exists in S1, it is replaced with the pair in S3.

  Next, in S707, it is determined whether S3 is empty. If S3 is empty (Yes), the process returns to FIG. 6 and proceeds to S603. If S3 is not empty (No), the process proceeds to S708, and S2 is made empty. In S709, the contents of S3 are put into S2, S3 is emptied, and the process returns to S703.

  By performing the processing of FIG. 7 as described above, the result that the statement entered in S1 at the end does not exceed the priority threshold value, and is a strongly-dependent statement from the modified / changed statement is obtained. .

Then, the process proceeds to S603 in FIG. In step S <b> 603, the dependency relationship analysis unit 107 notifies the test execution unit 106 of the extracted statement that is the result of the processing in FIGS. 7 and 8. In step S604, the test execution unit 106 searches the test case storage unit 108 for a test case associated with the notified statement, and executes the found test case. At this time, the test cases that have already been executed in the regression test cycle shown in FIG. 6 are not executed. In step S605, the test case that has been executed is recorded as being performed, and the test case storage unit 108 is updated.

  Thereafter, in S606, it is determined whether or not there is a time allowance for continuing the test. If there is a surplus (Yes), the process proceeds to S607. If there is no room (No), the process proceeds to S608.

  In S607, the threshold value that determines how much of the strongly related State is extracted from the modified / changed State is increased (the value is set so as to extract up to a State having a weaker dependency), and the process returns to S601. The flow of FIG. 6 is repeated.

  In step S608, after the sufficient regression test is performed or the test time is not sufficient and the series of regression tests is completed, the dependency storage unit 108 is partially updated for the corrected portion in the program.

FIG. 9 illustrates S608 in detail.
First, in step S <b> 901, a statement that has been modified or changed by the user is designated via the user interface 104. At this time, when a statement is newly added, it can be dealt with by designating a function or source file including the added statement.

  In step S <b> 902, the program analysis unit 105 searches the dependency storage unit 103 for a statement that is directly dependent on or directly dependent on a statement at a specified location.

  Dependency extraction processing (see FIGS. 12 and 13 described later) is performed on the source file including the searched Statement, and only the corresponding entry in the dependency storage unit 103 is updated.

The regression test system of the present invention has been described in detail above with reference to FIGS.
Hereinafter, the operation of the regression test of the present invention will be described using specific examples.

Assume that the regression test system of the present invention is used in the modification work (debugging) of a sample program written in a C-like language shown in FIG.
First, for the program in FIG. 10, the program analysis unit 105 stores information about the dependency relationship in the dependency relationship storage unit 103.

  Also, the correspondence between each Statement and Statement ID in the program of FIG. 10 is assumed to be the table shown in FIG. The table shown in FIG. 11 is stored in the Statement ID storage unit 109.

The details of the process of analyzing the dependency relationship of each Statement in the program of FIG. 10 and generating information related to the dependency relationship are disclosed in Non-Patent Document 1.
In this example, data dependency, control dependency, and function call are used as the target dependency. These dependencies are created as control flow graphs that represent the state of the program as a node and transitions between statements at the time of execution as edges, and the created control flow graph is assigned and referenced to variables at each node. It can be obtained by following the path.

FIG. 12 shows a control flow graph created from the sample program of FIG.
Using this graph, for data dependence, for example, a node (i = 1) is assigned to a variable i, and a node (while (i <= max)) is referenced to a variable i. In addition, since there is a path from the node (i = 1) to the node (while (i <= max)) that does not assign to the variable i, there must be data dependency between the nodes. I understand.

  As for control dependence, for example, there is a path that leads to STOP in addition to the path from the node (while (i <= max)) to the node (fact = fact * i). You can see that it exists.

  Thus, when a control flow graph is created from a program and the dependency relationship is analyzed, a program dependency graph as shown in FIG. 13 can be obtained. In this graph, a solid line indicates data dependence, a one-dot chain line indicates control dependence, and a two-dot chain line indicates a function call.

  The program analysis unit 105 saves the program dependency relationship graph created as described above in the dependency relationship storage unit 103. That is, as described above, the program dependency relationship graph can be expressed as shown in FIG. 2, and the dependency relationship of the sample program of FIG. 10 is as shown in FIG. 14 is stored in the dependency relationship storage unit 103 in the form of the data structure shown in FIG.

  14 and 15, only the lists that are affected by the Statement ID: 0 and the Statement ID: 10 and the lists that are affected are described in the following description.

For write (int x) with Statement ID: 0,
Since the function is called from the statements of ID: 9 and 10, statement IDs: 9 and 10 are registered in the list of affected statements.

  For write (sum) of Statement ID: 10, since the function call is received from the Statement of Statement ID: 0, Statement ID: 0 is registered in the list of affected Statements. In addition, Statement IDs: 3, 6, and 9 are registered in a Statement list that is affected by data dependence from Statements of Statement IDs: 3 and 6, and control dependence from Statement of Statement ID: 9.

As described above, the dependency relationship is stored in the dependency relationship storage unit 103, and at the same time, information about the past test is stored in the test case storage unit.
Here, it is assumed that tests of test case 1, test case 2, and test case 3 have been executed as samples in the past. In the past execution, in the test case 1, the Statement IDs: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, and 11 are executed. In the test case 2, the Statement ID: 1, 2 , 3, 4, 5 and 9 are executed, and in the test case 3, no Statement is executed. In such a case, as shown in FIG. 16, the statements executed in each test case of the past test are expressed as a list, and the test case storage unit 108 in the form of the data structure shown in FIG.
Stored in

  As described above, the dependency relationships shown in FIGS. 14 and 15 are stored in the dependency relationship storage unit 103, and the test cases performed in the past tests shown in FIGS. 16 and 17 are stored in the test case storage unit 108. When information related to the relationship with the statement of the program is stored, as an example of the process of performing the regression test, the regression test when the statement “write (sum)” (Statement ID: 10) in the sample program is modified Processing will be described.

  In the regression test system of the present invention, along with the flow shown in FIG. 6, a test case for executing a strongly related statement from the statement “write (sum)” is searched and executed.

  In this example, the priority, which is an index of relevance, is set to increase by 1 every time one of data dependency, control dependency, or function call is inserted as a dependency relationship with the modified or changed Statement. Shall. That is, the smaller the priority value, the stronger the dependency.

  As shown in FIG. 18, for example, when considering the strength of dependency from “write (sum)” in the sample program of FIG. int x) ”,“ sum = 0 ”,“ sum = sum + i ”, and“ if (print) ”have a priority of 1, and“ main () ”that is related by interposing these priority 1 statements. , “I = 1”, “while (i <= max)”, “i = i + 1”, and “write (fact)” have a priority of 2.

  In FIG. 6, in S601, the test execution unit 106 sets “10” as the Statement ID modified and changed by the dependency analysis unit 107, and “10” as the priority for determining how much of the strongest dependency is extracted. “0” is notified.

In step S <b> 602, the dependency relationship analysis unit 107 extracts a statement having a dependency relationship with the statement ID: 10.
In S702 of FIG. 7, S1 = S2 = {(10, 0)} ((10, 0) represents a pair with a Statement ID: 10 and a priority value of 0), and the processing of S703 is performed. Execute.

  The processing of S703 is the processing shown in the flow of FIG. 8. In S801, S2 '= {(10, 0)}, and in S803, the State and priority pair (10, 0) is extracted from S2'. Next, S804 is executed.

  This time, we are focusing on Statement ID: 10 with priority 0, and referring to the dependency database, Statement ID: 10 has a dependency relationship with Statement ID: 0, 3, 6, 9 respectively. , Data dependency, and control dependency, the respective priorities are “0 (priority of Statement ID: 0) +1” (priority increase value when one step of function call is sandwiched) = 1 ”,“ 0 + 1 ” (Priority increase value when data dependency is sandwiched by one level) = 1 ”,“ 0 + 1 (Priority increase value when data dependency is sandwiched by one step) = 1 ”,“ 0 + 1 (Priority when control dependency is sandwiched by one step) Degree increase value) = 1 ”.

  Next, S3 is entered in S805, and S3 = {(0,1), (3,1), (6,1), (9,1)}. Thereafter, the process returns to S802. Since (10, 0) is extracted from S2 'and is empty, the process in FIG. 8 (S703 in FIG. 7) ends, and the process proceeds to S704 in FIG.

  In S704, the priority threshold is set to 0 this time, and all pairs in S3 are removed because they exceed 0, and S3 becomes empty. For this reason, no processing is performed in S705 and S706, and the processing in FIG. 7 ends in S707 (S602 in FIG. 6).

  In S602, the Statement included in S1 is set as the extracted Statement, and the process proceeds to S603. In step S <b> 603, the test execution unit 106 is notified of the statement extracted by the dependency relationship analysis unit 107.

  In S604, since S1 = {(10, 0)}, when the priority threshold is set to 0, the related Statement is only Statement ID: 10. Referring to the test case storage unit 103, since only the test case 1 passes through (sample.c: 10), the test case 1 is executed in this cycle.

After execution, the test case storage unit 103 is updated in step S605, and the executed test case is recorded.
Since the test time still exists, the process proceeds to S607, and the priority threshold is set to 1.

  In FIG. 6, in S601, the test execution unit 106 corrects and changes to the dependency analysis unit 107 “10”, and “1” as the priority for determining how much of the statement with strong dependency is extracted. To be notified.

In step S <b> 602, the dependency relationship analysis unit 107 extracts a statement having a dependency relationship with the statement ID: 10.
In S702 of FIG. 7, S1 = S2 = {(10, 0)} is set, and the processing of S703 (processing of FIG. 8) is executed.

  In S801 in FIG. 8, S2 '= {(10, 0)}, and in S803, the statement and priority pair (10, 0) is extracted from S2'. Next, S804 is executed. This time, we are focusing on Statement ID: 10 with priority 0, and referring to the dependency database, Statement ID: 10 has a dependency relationship with Statement ID: 0, 3, 6, 9 respectively. , Data dependence, data dependence, and control dependence, each priority is 0 (priority of Statement ID: 0) +1 (priority decrease value when one step of function call is sandwiched) = 1, 0 + 1 (Priority decrease value when data dependency is sandwiched by one level) = 1, 0 + 1 (Priority decrease value when data dependency is sandwiched by one step) = 1, 0 + 1 (Priority decrease value when control dependency is sandwiched by one step) = 1.

  This is entered into S3 in S805, and S3 = {(0,1), (3,1), (6,1), (9,1)}. Thereafter, the process returns to S802. Since (10, 0) is extracted from S2 'and is empty, the process shown in FIG. 8 (S703 in FIG. 7) ends. In the next step S704, the priority threshold is set to 1 this time, and since all the pairs in S3 have a priority of 1 or less, they are not removed and remain as they are. At this time, S1 = {(10,0)}, S3 = {(0,1), (3,1), (6,1), (9,1)}, and S1 and S3 have Since there is no pair for the same Statement, the pair in S3 is entered into S1 in S705 and S706, and S1 = {(10,0), (0,1), (3,1), (6, 1), (9, 1)}. Here, since S3 is not empty, the process does not end in S707. In S708 and S709, S2 = {(0,1), (3,1), (6,1), (9,1)} and again S703. A search for a statement having a dependency relationship is performed.

In S801 of FIG. 8 for the second time, S2 ′ = {(0,1), (3,1), (6,1), (9,1)}, and S803, S804, S805 for the four pairs included. I will go.
First, referring to the dependency storage unit 103 for (0, 1) and searching for a statement having a dependency relationship, the statement having the dependency relationship is a Statement ID: 10, 11 and is related by a function call. And S3 = {(10, 2), (11, 2)} is obtained as a priority pair.

  Next, when searching for (3, 1), we get {(1, 2), (6, 2), (10, 2)} as related statements and their priorities, and before (0, 1) The obtained {(10, 2), (11, 2)} is merged in S805, and S3 = {(1, 2), (6, 2), (10, 2), (11, 2)} is obtained. It is done.

  Next, when searching for (6, 1), the related Statement and its priority are {(2, 2), (3, 2), (5, 2), (8, 2), (10, 2)}. Is obtained and merged in {circle around (1)}, {6,2), (10,2), (11,2)} obtained so far, and S3 = {(1,2), (2, 2), (3, 2), (5, 2), (6, 2), (8, 2), (10, 2), (11, 2)} are obtained.

  Finally, when searching for (9,1), {(1,2), (10,2), (11,2)} are obtained as related Statements and their priorities, and the {( 1, 2), (2, 2), (3, 2), (5, 2), (6, 2), (8, 2), (10, 2), (11, 2)} and S805 Finally, S3 = {(1,2), (2,2), (3,2), (5,2), (6,2), (8,2), (10,2) , (11, 2)}, and the processing shown in FIG. 8 (S703 in FIG. 7) ends.

  In the next step S704, the priority threshold is set to 1 this time, and all the pairs in S3 are removed because the priority exceeds 1, and S3 becomes empty. For this reason, no processing is performed in S705 and S706, and the processing ends in S707 (S602 in FIG. 6).

  In S602, the Statement included in S1 is set as the extracted Statement, and the process proceeds to S603. In step S <b> 603, the test execution unit 106 is notified of the statement extracted by the dependency relationship analysis unit 107. Here, since S1 = {(10,0), (0,1), (3,1), (6,1), (9,1)}, the priority threshold is set to 1. At this time, the related Statements are Statement IDs: 0, 3, 6, 9, and 10. Referring to the test case storage unit 103, the test cases that pass through these statements are test cases 1 and 2. Since test case 1 has already been executed in the previous cycle, test case 2 is executed in this cycle.

After execution, the test case storage unit 103 is updated in step S605, and the executed test case is recorded.
In the flow as described above, a regression test is realized in which the test is repeated in order from the test case having a strong dependency relationship with the modified and changed Statement while increasing the threshold value of the priority, as long as there is time.

  Then, after a sufficient regression test is performed or the test time runs out and a series of regression tests are completed, partial update of the dependency storage unit 108 is performed in step S608 for corrections and changes in the program. .

In the partial update process of the dependency relationship storage unit 108 in S608, the program analysis unit 105 searches the dependency relationship storage unit 103 for a statement that is directly dependent on or has a direct dependency on the specified location. And the retrieved S
A dependency relationship is extracted for a source file including a statement, and only the corresponding entry in the dependency storage unit 103 is updated. The extraction of the dependency relationship is as described with reference to FIGS. 10, 12, and 13. In S608, the extracted dependency relationship is compared with the old dependency relationship as shown in FIG. 13, and only different portions are updated.

  As described above, the regression test system of the present invention has been described in detail with reference to FIGS. However, the present invention is not limited to the above-described regression system configuration, and may be realized as a software program that operates on an information processing apparatus such as a computer.

FIG. 19 shows a hardware configuration of an information processing apparatus that implements the software regression test system of the present invention.
The information processing apparatus 1900 includes a CPU 1901, a memory 1902, an input device 1903, an output device 1904, an external storage device 1905, a medium drive device 1906, and a network connection device 1907, which are connected to each other via a bus 1908.

  The memory 1902 includes, for example, a ROM (Read Only Memory), a RAM (Random Access Memory), and the like, and stores programs and data for realizing a regression test system.

The CPU 1901 implements a regression test system by executing a program using the memory 1902.
The input device 1903 is, for example, a keyboard, a pointing device, a touch panel, etc., and is used for inputting instructions and information from the user. The output device 1904 is, for example, a display or a printer, and is used for outputting an inquiry to a user of the information processing device 1900, a processing result, and the like.

  The external storage device 1905 is, for example, a magnetic disk device, an optical disk device, a magneto-optical disk device, or the like. Programs and data can be stored in the external storage device 1905, and can be loaded onto the memory 1902 and used as necessary.

  The medium driving device 1906 drives a portable recording medium 1909 and accesses the recorded contents. As the portable recording medium 1909, an arbitrary computer-readable recording medium such as a memory card, a memory stick, a flexible disk, a CD-ROM (Compact Disc Read Only Memory, an optical disk, a magneto-optical disk, a DVD (Digital Versatile Disk), etc. It is also possible to store programs and data in this portable recording medium 1909 and load them into the memory 1902 and use them if necessary.

  The network connection device 1907 communicates with an external device via an arbitrary network (line) such as a LAN or WAN, and performs data conversion accompanying the communication. If necessary, the program and data can be received from an external device and loaded into the memory 1902 for use.

A program operating on the information processing apparatus is configured to process the flows of FIGS. 6, 7, 8, and 9 using the memory 1902 of the information processing apparatus.
FIG. 20 shows a method for loading the program into the information processing apparatus when the regression test system of the present invention is realized by executing the program on the information processing apparatus.
FIG. 20A shows a method in which the information processing apparatus 2001 loads programs and data 2002 stored in an external storage device such as a hard disk of the information processing apparatus 2001.

  FIG. 20B shows a method of loading a program or data 2004 recorded on a portable storage medium such as a CD-ROM or DVD via the medium driving device of the information processing apparatus 2001. .

  FIG. 20C shows a method of loading a program and data 2003 provided by an information provider via a communication device of the information processing apparatus 2001 via a line such as a network.

  As described above, the present invention may be configured as a program for causing an information processing apparatus such as a computer to perform the same functions as the functions of the system described above. The present invention can also be configured as a computer-readable recording medium that records a program for causing an information processing apparatus such as a computer to perform the same functions as those described above. The present invention can also be configured as a computer data signal representing the above-described program embodied in a carrier wave.

  As described above, this book is configured to store information about test cases of tests performed in the past, and to perform tests by combining the dependency obtained by analyzing the information and the target program. According to the regression system of the invention, it is possible to preferentially crush test cases having a strong dependency relationship with the changed portion, and it is possible to improve the efficiency of the regression test. As a result, the efficiency of software development itself can be increased.

(Additional remark 1) In order to implement | achieve the regression test system of a software program, information processing apparatus provided with the storage means which memorize | stores information,
A Statement extraction means for extracting a Statement related to the changed Statement with reference to the dependency relationship of each Statement in the software program stored in the storage means;
The test case associated with the extracted statement is related to the extracted statement from the information about the correspondence between the test case executed in the past and each statement in the software program stored in the storage means. A test case detection means for detecting a test case;
A test execution means for executing the detected test case;
Program to function as. (1) (FIGS. 1 and 6)
(Supplementary Note 2) The Statement extraction unit extracts a Statement by selecting a priority that is defined as an index of relevance from the changed Statement that does not exceed a specified threshold.
2. The program according to claim 1, wherein the program is repeated as long as there is a time for testing the Statement extraction unit, the test case detection unit, and the test execution unit by changing the threshold value. (2) (FIGS. 1, 6, 7, and 8)
(Supplementary Note 3) The priority defined as the index of relevance is set so that the priority increases according to the type of the sandwiched dependency every time the dependency is sandwiched by one stage. The program according to appendix 2, wherein the program is calculated based on the dependency relationship. (3) (FIGS. 1, 6, 7, 8, and 18)
(Supplementary note 4) The program according to claim 1 or 2, further causing the information processing apparatus to function as test case recording means for recording the test case executed by the test execution means in the storage means. . (4)
(Supplementary Note 5) Further, the information processing apparatus further changes the statemen after the test time is over.
3. The program according to claim 2, wherein the dependency relationship changed by t is caused to function as dependency relationship update means for updating the storage means. (5) (FIGS. 1, 6, and 9)
(Supplementary note 6) The program according to supplementary note 4, wherein the test execution means does not perform a duplicate test on the test cases recorded by the test case recording means.
(Supplementary note 7) The program according to supplementary note 3, wherein the dependency relationship is at least one of data dependency, control dependency, and function call.
(Appendix 8) A system for performing a software regression test,
A statement identifier storage means for storing a correspondence table of statements and identifiers for identifying the statements in the software program;
Dependency storage means for storing the dependency of each Statement in the software program, which is described using a Statement identifier;
Test case storage means for storing information described using a Statement identifier for test cases of tests performed in the past;
Dependency analysis means for extracting a Statement related to the changed Statement based on the dependency stored in the dependency storage;
A test execution unit that searches the test case storage unit for a test case associated with the extracted statement, executes the corresponding program portion with reference to the statement identifier storage unit, and performs a test;
A system comprising: (6) (FIGS. 1 and 6)
(Additional remark 9) In the said dependence relationship analysis means, by selecting the thing whose priority defined as the parameter | index of the relationship from the said changed Statement does not exceed the designated threshold value, a Statement is extracted,
The system according to appendix 8, wherein the test is repeated by changing the threshold value as long as there is a time to test. (FIGS. 1, 6, 7, and 8)
(Supplementary Note 10) The priority defined as the index of relevance is set so that the priority increases according to the type of the sandwiched dependency every time the dependency is sandwiched by one stage. The system according to appendix 9, wherein the system is calculated based on the dependency relationship. (FIGS. 1, 6, 7, 8, and 18)
(Supplementary Note 11) The system according to Supplementary Note 8 or Supplementary Note 9, wherein the test execution unit records a test case in which the test is performed in the test case storage unit.
(Supplementary note 12) The system according to claim 9, further comprising a program analysis unit that partially updates the dependency relationship changed by the changed Statement after the test time is over.
(Supplementary note 13) The system according to supplementary note 11, wherein the test execution unit does not perform a duplicate test on a test case once performed with reference to the test case storage unit. (FIGS. 1, 6, and 9)
(Supplementary note 14) The system according to supplementary note 10, wherein the dependency relationship is at least one of data dependency, control dependency, and function call.
(Supplementary note 15) A method for executing a regression test of a software program on an information processing apparatus including a storage unit for storing information,
The information processing apparatus
Referring to the dependency relationship of each statement in the software program stored in the storage means, extract a statement related to the changed statement,
The test case associated with the extracted statement is related to the extracted statement from the information about the correspondence between the test case executed in the past and each statement in the software program stored in the storage means. Detect test cases,
Performing the detected test case;
A method characterized by that. (7) (FIGS. 1 and 6)
(Supplementary Note 16) When extracting a Statement related to the changed Statement, select the one whose priority defined as an index of relevance from the changed Statement does not exceed the specified threshold To extract the Statement,
16. The method according to claim 15, wherein the threshold value is changed, and as long as there is a test time, the extraction of the State related to the changed State, the detection of the test case, and the execution of the test case are repeated. . (FIGS. 1, 6, 7, and 8)
(Supplementary Note 17) The priority defined as the index of relevance is set so that the priority increases in accordance with the type of dependency relationship that has been inserted each time a dependency relationship such as data dependency or control dependency is inserted one step. The method according to claim 16, wherein the method is calculated based on the dependency relationship. (FIGS. 1, 6, 7, 8, and 18)
(Supplementary note 18) The method according to supplementary note 15 or supplementary note 16, wherein when the test of the searched test case is performed, the executed test case is recorded in the storage means.
(Supplementary note 19) The method according to supplementary note 16, further comprising: updating the dependency relationship changed by the changed Statement after the test time is over in the storage unit. (FIGS. 1, 6, and 9)
(Supplementary note 20) The method according to supplementary note 18, wherein when the searched test case is performed, the already executed test case recorded in the storage means is not duplicated.
(Supplementary note 21) The system according to supplementary note 17, wherein the dependency relationship is at least one of data dependency, control dependency, and function call.
(Additional remark 22) In order to implement | achieve the regression test system of a software program, the information processing apparatus provided with the storage means which memorize | stores information,
A Statement extraction means for extracting a Statement related to the changed Statement with reference to the dependency relationship of each Statement in the software program stored in the storage means;
The test case associated with the extracted statement is related to the extracted statement from the information about the correspondence between the test case executed in the past and each statement in the software program stored in the storage means. A test case detection means for detecting a test case;
A test execution means for executing the detected test case;
A computer-readable recording medium in which a program for functioning as a computer is recorded. (8) (FIGS. 1 and 6)
(Supplementary Note 23) The Statement extraction unit extracts a Statement by selecting a priority that is defined as an index of relevance from the changed Statement that does not exceed a specified threshold.
23. The computer-readable program having recorded thereon the program according to claim 22, wherein the threshold value is changed, and the statement extraction unit, the test case detection unit, and the test execution unit are repeated as long as there is time to test. Recording medium. (FIGS. 1, 6, 7, and 8)
(Supplementary Note 24) The priority defined as the index of relevance is set so that the priority increases according to the type of the sandwiched dependency every time the dependency is sandwiched by one stage. A computer-readable storage medium storing the program according to appendix 23, which is calculated based on the dependency relationship. (FIGS. 1, 6, 7, 8, and 18)
(Supplementary note 25) The program according to claim 22 or 23, further causing the information processing apparatus to function as test case recording means for recording the test case executed by the test execution means in the storage means. A computer-readable storage medium on which is recorded.
(Supplementary Note 26) Further, after the test time is over, the information processing apparatus changes the state
24. The computer-readable recording medium having recorded thereon the program according to claim 23, wherein the dependency relation changed by nt is caused to function as dependency relation update means for updating the storage means. (FIGS. 1, 6, and 9)
(Supplementary note 27) The computer-readable storage medium storing the program according to Supplementary note 25, wherein the test execution unit does not perform a duplicate test on the test case recorded by the test case recording unit.
(Supplementary note 28) The computer-readable storage medium storing the program according to supplementary note 24, wherein the dependency relationship is at least one of data dependency, control dependency, and function call.

It is a figure which shows the structure of the regression test system of this invention. 3 is a diagram illustrating an outline of a dependency relationship storage unit 103. FIG. 4 is a diagram illustrating a data structure of a dependency relationship storage unit 103. FIG. 2 is a diagram showing an outline of a test case storage unit 108. FIG. It is a figure which shows the data structure of test case storage FIG. It is a figure which shows the flow of selection and execution of a test in the software regression test system of this invention. It is a figure which shows the detailed flow of S602 of FIG. It is a figure which shows the detailed flow of S703 of FIG. It is a figure which shows the detailed flow of S608 of FIG. It is a figure which shows a sample program. It is a figure which shows the table which defines the relationship between Statement of a sample program, and Statement ID. It is a figure which shows the control flow graph of a sample program. It is a figure which shows a program dependence graph. It is a figure which shows the dependency of a sample program. It is a figure which shows the data structure in the case of storing the dependency relationship of a sample program in a dependency relationship storage part. It is a figure which shows the relationship between each test case stored in the test case storage part, and Statement. It is a figure which shows the data structure in the case of storing the relationship of FIG. 16 in a test case storage part. It is a figure explaining priority. It is a figure which shows the structure of the information processing apparatus which implement | achieves a software regression test system. It is a figure explaining the loading to the information processing apparatus of a program.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Software regression test system 102 Program storage part 103 Dependency storage part 104 User interface 105 Program analysis part 106 Test execution part 107 Dependency analysis part 108 Test case storage part 109 Statement ID storage part 1900 Information processing apparatus 1901 CPU
1902 Memory 1903 Input device 1904 Output device 1905 External storage device 1906 Medium drive device 1907 Network connection device 1908 Bus 1909 Portable recording medium 2001 Information processing device 2002 Program and data recorded in external storage device 2003 Program provided by information provider And data 2004 Programs and data recorded on portable recording media

Claims (8)

In order to realize a regression test system for a software program, an information processing apparatus including a storage unit for storing information is provided.
A Statement extraction means for extracting a Statement related to the changed Statement with reference to the dependency relationship of each Statement in the software program stored in the storage means;
The test case associated with the extracted statement is related to the extracted statement from the information about the correspondence between the test case executed in the past and each statement in the software program stored in the storage means. A test case detection means for detecting a test case;
A test execution means for executing the detected test case;
Program to function as. The Statement extraction means extracts a Statement by selecting a priority that does not exceed a specified threshold value as a relevance index from the changed Statement,
2. The program according to claim 1, wherein the program is repeated as long as there is a time for testing the Statement extraction unit, the test case detection unit, and the test execution unit by changing the threshold value.   The priority defined as the index of relevance is set so that the priority increases according to the type of dependency relationship that has been inserted each time a dependency relationship is inserted, and based on the dependency relationship. 3. The program according to claim 2, wherein the program is calculated as follows.   3. The program according to claim 1, further comprising causing the information processing apparatus to function as test case recording means for recording the test case executed by the test execution means in the storage means.   The program according to claim 2, further causing the information processing apparatus to function as a dependency update unit that updates the storage unit with a dependency changed by the changed Statement after the test time is over. A system for software regression testing,
A statement identifier storage means for storing a correspondence table of statements and identifiers for identifying the statements in the software program;
Dependency storage means for storing the dependency of each Statement in the software program, which is described using a Statement identifier;
Test case storage means for storing information described using a Statement identifier for test cases of tests performed in the past;
Dependency analysis means for extracting a Statement related to the changed Statement based on the dependency stored in the dependency storage;
A test execution unit that searches the test case storage unit for a test case associated with the extracted statement, executes the corresponding program portion with reference to the statement identifier storage unit, and performs a test;
A system comprising: A method for executing a regression test of a software program on an information processing apparatus comprising a storage means for storing information,
The information processing apparatus
Referring to the dependency relationship of each statement in the software program stored in the storage means, extract a statement related to the changed statement,
The test case associated with the extracted statement is related to the extracted statement from the information about the correspondence between the test case executed in the past and each statement in the software program stored in the storage means. Detect test cases,
Performing the detected test case;
A method characterized by that. In order to realize a regression test system for a software program, an information processing apparatus including a storage unit for storing information is provided.
A Statement extraction means for extracting a Statement related to the changed Statement with reference to the dependency relationship of each Statement in the software program stored in the storage means;
The test case associated with the extracted statement is related to the extracted statement from the information about the correspondence between the test case executed in the past and each statement in the software program stored in the storage means. A test case detection means for detecting a test case;
A test execution means for executing the detected test case;
A computer-readable recording medium in which a program for functioning as a computer is recorded.

Images