KR20010075872A - Method for automatic generating call graph for software maintenance - Google Patents

Abstract

The present invention relates to a method for automatically generating a call graph for software maintenance, comprising: a first step of a translation unit extracting a media language from a source program and storing the media language in a media language storage unit; A second step of extracting unit extracting function information necessary for generating a call graph from the intermediate language and storing it in a reverse engineering integrated information file; A third step of automatically generating a call graph information model node and edge information of an input graph, which is a call information of a function, from the reverse engineering integrated information file and storing the call graph information model storage unit in a call graph information model storage unit; A fourth step of determining, by the call graph direction graph generator, the level of each node using the node and edge information of the input graph which is the function call information, and identifying the type of the edge according to the relationship between the node and the adjacent node; A fifth step of the call graph direction graph generation unit determining an order of nodes at the level of each node; A sixth step of generating, by the call graph direction graph generator, absolute coordinates of the node and the edges; And a seventh step of automatically generating a call graph by using the coordinates of the real node and the edges generated by the call graph graphic interface unit in the direction graph, and used in a computer-aided software engineering tool.

Description

How to automatically create call graph for software maintenance {METHOD FOR AUTOMATIC GENERATING CALL GRAPH FOR SOFTWARE MAINTENANCE}

The present invention relates to a method for automatically generating a call graph for software maintenance, and in particular, a direction graph for easily understanding and analyzing the function call relations of source programs written in COBOL, C, C ++, JAVA, etc. A method of automatically generating a call graph arranged hierarchically and a computer-readable recording medium recording a program for realizing the method.

In general, a graph automatic generation method for software maintenance requires an abstract structure and an algorithm for automatically generating the structure, and the graph generation algorithm is classified into various types according to the type of graph and the application field of the graph. .

First, depending on the type of graph, the cycle directed graph, the cycleless graph, the cycle undirected graph, the cycleless direction graph, the tree and planar graphs, etc. It is divided into software engineering, computer network and petri net according to the application of the graph. In particular, in the field of software engineering, Computer Aided Software Engineering (CASE), which provides abstract structures of different representation techniques, for example, data-flow diagram, structure chart, flow chart. (flowchart), state transition diagram, entity relationship diagram, and so on.

The computer-aided software engineering tools provide a user interface that allows the user to create and process graphs directly, and the creation of call graphs can produce layouts that are slightly improved than drawing on real paper.

However, the computer-aided software engineering tool as described above, in the case of complex source code, it takes much time and effort to analyze the control structure of the source code and generate the call graph, even if the user generates the call graph. Since it is a graph having positive orientation, there is a problem of poor readability.

In order to solve the above problems, a method for automatically hierarchically arranging call graphs as direction graphs has been developed and used.

The conventional call graph automatic generation method arranges the type of the selected graph (plane graph, direction graph, tree graph, etc.), and generates a specific type of arrangement (plane arrangement, hierarchical arrangement, etc.).

The hierarchical layout method of the direction graph is used to automatically generate the bicycle direction graph, and is particularly useful for the call graph generation in the software engineering field. The generation of the bicyclic directional graph analyzes source programs of a programming language to be maintained, stores nodes and edges which are graph data dependent on the programming language, and then uses a transitive closure algorithm for the stored graph data. Automatic placement can be achieved by applying a barycenter algorithm or the like.

However, as described above, the conventional method for automatically generating a call graph extracts all the information of the syntax step from source programs written in various program languages and stores them in a form dependent on a specific language, thereby extending the program language to be analyzed. Each time, a new storage structure, that is, a structure in which analysis information of various programming languages can be integrated and stored in a single grammar form that is easily extensible, a cycle call graph occurs, and a lattice type Positional placement is not considered and there is a problem that an unbalanced positional arrangement can be created.

Accordingly, the present invention has been made to solve the above problems, and by automatically generating the function call information of the source code in a hierarchical direction graph to support the user to easily understand the program structure, the complexity of each function And a computer-readable recording medium recording a method for automatically generating a call graph for software maintenance that can easily analyze source code to be maintained by providing measurement information such as a structure diagram and a program for realizing the method. The purpose is to provide.

1 is a configuration diagram of an embodiment of an automatic call graph generation system to which the present invention is applied.

2 is a flowchart illustrating an embodiment of a method for automatically generating a call graph for software maintenance according to the present invention.

FIG. 3 is a flow diagram of one embodiment of a subroutine for determining the level of a node in FIG.

4 is a detailed flowchart of one embodiment of a subroutine for determining the above-weighted order of call graph layout in FIG.

FIG. 5 is a flow diagram of one embodiment of a subroutine for determining the bottom-weighted order of call graph layout in FIG.

FIG. 6 is a detailed flowchart of one embodiment of a subroutine for determining the above-weighted position of the call graph layout in FIG.

FIG. 7 is a flow diagram of one embodiment of a subroutine for determining the bottom-weight center position of the call graph layout in FIG.

8 is a detailed flowchart of an embodiment of a subroutine for determining the up-lattice-weighted position of the call graph layout in FIG.

9 is a detailed flowchart illustrating a subroutine for generating coordinates of nodes X and Y of the call graph layout of FIG. 2.

FIG. 10 is a detailed flowchart illustrating a subroutine for generating coordinates of the edges X and Y of the call graph layout of FIG. 2.

* Explanation of symbols for the main parts of the drawings

10: monitor 11: printer

12: call graph automatic generation system 13: source code analysis system

14: source program 15: intermediate language storage unit

16: reverse engineering integrated information file 20: call graph graphical interface

21: call graph automatic generation unit 30: call graph direction graph generation unit

31: call graph information model storage unit 40: translation unit

41: extraction unit

According to an aspect of the present invention, there is provided a method for generating a call graph for software maintenance, comprising: a first step of a translation unit extracting a media language from a source program and storing the media language in a media language storage unit; A second step of extracting unit extracting function information necessary for generating a call graph from the intermediate language stored in the intermediate language storage unit and storing the extracted function information in a reverse engineering integrated information file; A third step in which an automatic call graph generation unit searches node and edge information of an input graph, which is call information of a function necessary for generating a call graph, from the reverse engineering integrated information file from a media language and stores it in a call graph information model storage unit; The call graph direction graph generator determines the level of each node by using the node and edge information of the input graph which is the function call information stored in the call graph information model storage unit, and determines the level of the edge according to the relationship of the node adjacent to the corresponding node. A fourth step of identifying the type; A fifth step of the call graph direction graph generation unit determining an order of nodes at the level of each node; A sixth step of generating, by the call graph direction graph generator, absolute coordinates of the node and the edges; And a seventh step of automatically generating a call graph by using the coordinates of the real node and the edges generated by the call graph graphic interface unit as the direction graph.

In addition, the present invention, in the automatic call graph generation system having a microprocessor, in the call graph generation method for software maintenance, the first function that the translation unit extracts the intermediate language from the source program and stores it in the intermediate language storage unit; ; A second function of an extracting unit extracting function information necessary for generating a call graph from an intermediate language stored in the intermediate language storage unit and storing the function information in a reverse engineering integrated information file; A third function of automatically generating a call graph information model storing unit and edge information of an input graph which is call information of a function required for generating a call graph from the reverse engineering integrated information file from an intermediate language, and storing them in a call graph information model storage unit; The call graph direction graph generator determines the level of each node by using the node and edge information of the input graph which is the function call information stored in the call graph information model storage unit, and determines the level of the edge according to the relationship of the node adjacent to the corresponding node. A fourth function of identifying a kind; A fifth function of the call graph direction graph generator determining an order of nodes at the level of each node; A sixth function of generating, by the call graph direction graph generator, absolute coordinates of the node and the edges; And a computer-readable recording medium having recorded thereon a program for realizing a seventh function of automatically generating a call graph using coordinates of real nodes and edges generated by the call graph graphic interface unit in the direction graph.

The above objects, features and advantages will become more apparent from the following detailed description taken in conjunction with the accompanying drawings. Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a configuration diagram of an embodiment of an automatic call graph generation system for software maintenance to which the present invention is applied.

As shown in the figure, a call graph automatic generation system for software maintenance to which the present invention is applied includes a call graph automatic generation system 12 and a source code analysis system 13, and the source code analysis system. 13, the source program 14 written in COBOL, C, C ++, JAVA, etc. is extracted by the translation unit 40 to extract the intermediate language, which is an independent intermediate language, and the intermediate language storage unit 15. ), And extracts only the function information necessary for generating the call graph in the stored intermediate language from the extracting unit 41 and stores it in the reverse engineering integrated information file 16.

In addition, the call graph automatic generation system 12 includes a call graph automatic generation unit 21 and a call graph graphic interface unit 20. The call graph automatic generation unit 21 includes a call graph information model storage unit 31 and a call graph direction graph generation unit 30.

The call graph automatic generation unit 21 searches node and edge information of an input graph, which is a function call information necessary for generating a call graph, from the reverse engineering integrated information file 16 from the media language and stores the call graph information model storage unit ( 31), the call graph direction graph generation unit 30 generates a call graph direction graph using node and edge information of the input graph stored in the call graph information model storage unit 31, and the call. The graph graphic interface unit 20 automatically generates a call graph and a call tree by using the X and Y coordinates of the real nodes and edges generated by the call graph direction graph and monitors 10 or the printer 11. Will output

2 is a flowchart illustrating an embodiment of a method for automatically generating a call graph according to the present invention. First, the translation unit 40 extracts an intermediate language from the source program 14 to the intermediate language storage unit 15. Store (100), the extraction unit 41, the function information for generating a call graph from the intermediate language stored in the intermediate language storage unit 15 is stored in the reverse engineering information integrated file 16, and the reverse engineering information integrated file 102, after searching the sentence of the 'n' line of the information of the nodes and the edges of the input graph, the function call information required for generating the call graph stored in (16) in the call graph information model storage unit 31 (102) In the call graph direction graph generation unit 30, the level algorithm of each node is identified using the level algorithm (104), and the order-based algorithm is used to determine the top-weight center order of the call graph layout. 106), down-weight Determine the seam order (108).

Then, the positions of the top-weight center, the bottom-weight center, and the top-lattice-weight center of the call graph layout are determined (110 to 114), and after generating nodes X and Y coordinates of the call graph layout (116). And generating coordinates of the edges X and Y (118), and using the generated coordinates of the nodes and the edges, the call graph graphic interface unit 20 sends the call graph and the call tree to the monitor 10 and the printer 11. Output

3 to 10 are a series of processes for generating a call graph performed by the call graph direction graph generator 30.

FIG. 3 is a detailed flowchart illustrating a subroutine for determining a level of a node in FIG. 2. The depth first search is extended to arrange all nodes hierarchically at each level and to define the types of edges. Identification process.

First, ordinal i of the first visited node among all nodes is initialized to '0' (202), and ordinal i is initialized to '1' (203).

Next, it is determined whether ordinal i is smaller than n, the size of the searched node (203).

The determination result 203 returns if ordinal i is greater than or equal to n, the size of the searched node, and if ordinal i is less than n, the size of the searched node, it is determined whether the node of ordinal i has been visited ( 204).

In step 204, if the node of ordinal i is visited, i is increased by 1 (207), and then the process of determining whether ordinal i is smaller than n, the size of the searched node, for the next node visit (203). If the node of ordinal i is not visited, '1' is assigned to the level of ordinal i (205), and it is determined whether the node adjacent to the node of ordinal i is null (206).

The determination result 206, if the node adjacent to the node of ordinal i is null, performs step 207 of increasing i by 1, and ordinal i arrives from the node of ordinal i if the node adjacent to the node is not null. It visits all possible nodes (208) and determines whether or not it visited node j adjacent to the node of ordinal number i (209).

As a result of the determination (209), if the node j adjacent to the ordinal i is not visited, a tree edge is generated between the node ordinal i and the adjacent node j (210), and i level + is applied to the adjacent node j. A level of 1 is generated (211), and a process (209) is performed to determine whether a visit has been made to a node j adjacent to the node of ordinal i.

As a result of the determination (209), if a visit is made to the node j adjacent to the node of ordinal i, it is determined whether the level of the node of ordinal i is smaller than the level of the adjacent node j (212).

The determination result 212, if the level of the node of ordinal i is less than the level of the adjacent node j, generates a forward edge between the node of ordinal i and the adjacent node j (213), the forward edge (forward) A dummy node and a dummy edge are generated for the edge (214), and a process (209) is performed to determine whether a visit has been made to the node j adjacent to the ordinal node i.

As a result of the determination 212, if the level of the ordinal node i is not smaller than the level of the adjacent node j, it is determined whether the level of the ordinal node i is greater than the level of the adjacent node j (215).

As a result of the determination 215, if the level of the node of ordinal i is greater than the level of adjacent node j, a back edge is generated between the node of ordinal i and the adjacent node j (216), and the back edge After generating a dummy node and a dummy edge for the node (217), a process (209) is performed to determine whether a visit is made to the node j adjacent to the ordinal node i.

In the determination result 215, if the level of the ordinal node is not greater than the level of the adjacent node j, a tree edge is generated between the ordinal node i and the adjacent node j and the forward edge between the ancestor nodes of the node j. generate a forward edge (218), modify the level by adding 1 to the level of adjacent node j of the ordinal node i (219), and add a dummy node and a dummy node to the forward edge. After generating a dummy edge (220), a process (209) is performed to determine whether a visit has been made to a node j adjacent to the ordinal node i.

4 and 5 are a process of determining the order of all nodes on each level by reducing intersections between edges in order to provide high quality readability and traceability, wherein up-weight center and down-weight center are utilized. The procedure is carried out after repeating Figure 3 and 4, and then repeating Figure 3.

4 is a detailed flowchart of an embodiment of a subroutine for determining the above-weighted order of the call graph layout in FIG. 2.

First, search n levels arranged hierarchically (301), initialize ordinal i of the first level to '0' (302), and determine whether ordinal i is smaller than n, the size of the searched level. (303).

In the determination result 303, if ordinal i is smaller than n, which is the magnitude of the searched level, the nodes of ordinal i th level are weighted in ascending order of ordinal number 304, and each node of the i + 1 th level is assigned to its node. The center of gravity is calculated as the weighted average of ancestors as follows (305).

Upper-weight center of each node = sum of weights of ancestors / number of ancestors

Sorting the nodes of ordinal i + 1th level in ascending order by the measured weight (306), increasing ordinal i by 1 (307), and determining whether i is less than n for the next level visit Perform 303.

The determination result 303 returns if ordinal i is greater than or equal to n, which is the magnitude of the searched level.

FIG. 5 is a detailed flowchart of an exemplary embodiment of a subroutine for determining the down-weight centering order of the call graph layout in FIG. 2.

First, search for n levels arranged hierarchically (401), initialize the ordinal of the last level to n (402), and determine whether ordinal i of the last level is greater than '0', the size of the searched level. (403).

In the determination result 403, if ordinal i of the last level is greater than '0' which is the size of the searched level, the nodes of ordinal i th level of the last level are weighted in ascending ordinal order (404), and ordinal i- Each node of the first level is computed as below-weight center as the weighted average of its descendants as follows (405).

Bottom-weight center of each node = sum of weights of descendants / number of descendants

After arranging each node of the ordinal i-1 level in ascending order by the measured weight (406), i is decreased by 1 (407), and it is determined whether i is greater than '0' (403). do.

The determination result 403 returns if the ordinal i of the last level is less than or equal to '0' which is the size of the searched level.

6 to 8 are heuristic processes for offsetting the spacing between levels and the spacing between nodes as steps of a location algorithm.

FIG. 6 is a detailed flowchart of an exemplary embodiment of a subroutine for determining a position of the center of gravity of the call graph layout in FIG. 2.

First, search for n levels in which the order of nodes in each level is determined (501), initialize the ordinal i of the first level to '0' (502), and then ordinal i is less than n, the size of the searched level. Determine (503).

In the determination result 503, if i is less than n, the nodes of ordinal i th level are weighted in ascending ordinal order (504), and the nodes of i + 1 th level are weighted in ascending ordinal order, After visiting m nodes to compute the position of the node in the above-weighted method (505), initializing ordinal j of the first node to '0' (506), ordinal j is the i + 1th level It is determined whether the number of nodes is less than m (507).

If i is greater than or equal to n, the determination result 503 is returned.

In the determination result 507, if j is less than m, the above-weight center of the j node is calculated as follows, and the weight of the measured j node is stored (509).

Center of gravity of each node = sum of weights of ancestors / number of ancestors

In operation 510, it is determined whether the weight of the j node is greater than the weight of the j + 1 node, which is the next node.

In the determination result 507, if j is greater than or equal to m, i is increased by 1, and then a process 503 of determining whether i is less than n is performed for the next level of visit.

In the determination result 510, if the weight of j node is greater than the weight of j + 1 node, the weight of j node +1 is sequentially added to the weights assigned to each node from the j + 1 node to the last node m-1. In addition, after modifying and storing the weights of the nodes (511), increasing j by 1 (512), and determining whether j is less than m for a visit of the next node (507).

As a result of the determination 510, if the weight of the j node is less than or equal to the weight of the j + 1 node, the process of increasing j by 1 is performed 512.

FIG. 7 is a detailed flowchart illustrating an embodiment of a subroutine for determining a bottom-weighted position of the call graph layout in FIG. 2.

First, we search for n levels rearranged by the above-weight centers (601), initialize the last ordinal i to n (602), and then determine whether ordinal i is greater than the last level '0'. Determine (603).

In the determination result 603, if i is greater than 0, m nodes are visited (604) to calculate the position of nodes in the ordinal i-1 th level by the below-weighted method (604), and the ordinal number of the first visited node After initializing j to '0' (605), it is determined whether j is less than m (606), and if i is less than or equal to 0, it is returned.

In the determination result 606, if j is smaller than m, the bottom-weight center of the j node is calculated as follows, and the weight of the j node is stored according to the measured weight (608).

Down-weight center of each node = sum of weights of descendants / number of descendants

If the weight of j node is greater than the weight of j + 1 node (609), if the weight of j node is greater than the weight of j + 1 node, from node j-1 to node m-1, which is the last node, The weight of each node is sequentially added to the weights assigned to each node, and the weights of the nodes are modified and stored (610), and j is increased by 1 for the next node visit (611), where j is greater than m. If the weight of the j node is less than or equal to the weight of the j + 1 node, the process of increasing j by 1 is performed (611), and then the next node is visited. A process 606 of determining whether j is less than m is performed.

If j is greater than or equal to m, the determination result 606 increases i by 1 for a next level visit (607), and determines whether i is greater than 0 (603).

8 is a detailed flowchart of a subroutine for determining the up-lattice-weighted position of the call graph layout in FIG. 2, wherein the lattice structure (only a node having two or more ancestors of its own is found in a balanced layout). It is a process to generate.

First, search for n levels relocated by the down-weight center (701), initialize ordinal i of the second level to 1 (702), and determine whether ordinal i is less than n, the magnitude of the searched level. (703).

In the determination result 703, if ordinal i is smaller than n, which is the size of the searched level, nodes of the lattice structure of ordinal i th level are searched to calculate the position of the node in the above-lattice-weighted method. After visiting m nodes of the total number (704), initializing the ordinal j of the first node to 0 (705), determining whether ordinal j is less than m, the total number of nodes of the i th level (706), and determining If the result 703, ordinal i, is greater than or equal to n, the magnitude of the level searched, it is returned.

As a result of the determination (706), if ordinal j is greater than or equal to m, the total number of nodes of the i-th level, the value of ordinal i is increased by one (707), and then the level of the ordinal i is searched for the next level visit. If the ordinal j is smaller than m, the total number of nodes of the i-th level, it is determined whether a lattice structure exists in the j node (708).

If the lattice structure does not exist in the j node, the determination result 708 increases the ordinal j by 1 (712), and then determines whether the ordinal j is smaller than m, the total number of nodes of the i th level, for the next node to visit. If the lattice structure is present in the j node, the lattice-weight center of the j node is calculated as follows, and the weight of the j node is stored according to the measured weight (709).

Upper-lattice-weight center of each node = sum of weights of ancestors / number of ancestors

In operation 710, it is determined whether the existing weight of the j node is smaller than the measured weight of the j node.

In the determination result 710, if the existing weight of j node is smaller than the measured weight of j node, the weight is corrected by adding all descendant nodes of j node by the weight difference of j (711), and the ordinal j is increased by one. In operation 712, a process 706 is performed to determine whether ordinal j is smaller than m, the total number of nodes of the i-th level, for the next node to visit.

In the determination result 710, if the existing weight of the j node is greater than or equal to the measured weight of the j node, the ordinal j is increased by 1 (712), and the ordinal j is the total number of nodes of the i th level for the visit of the next node. A process 706 of determining whether the number is smaller than m is performed.

9 and 10 are a process of outputting a call graph, which is a process of automatically generating a call graph which is a hierarchical direction graph by determining coordinates of X and Y of a real node and an edge.

FIG. 9 is a detailed flowchart illustrating a subroutine for generating coordinates of nodes X and Y of the call graph layout of FIG. 2.

First, search for n levels at which nodes are located at each level (801), initialize ordinal i of the first level to '0' (802), and then ordinal i is less than n, the size of the level found. Determine (803).

In the determination result 803, if ordinal i is smaller than n, the size of the searched level, m nodes are visited to calculate coordinates for outputting positions of nodes in the i-th level (804), and the first After initializing the ordinal j of the node to '0' (805), it is determined whether the ordinal j is smaller than m, which is the total number of the i-th level (806).

As a result of the determination 803, if ordinal i is greater than or equal to n, which is the magnitude of the searched level, it is returned.

In the determination result 803, if ordinal j is greater than or equal to m, which is the total number of i-th levels, the ordinal i is decreased by 1 (807), and then it is determined whether ordinal i is smaller than n, the size of the searched level. Perform process 803.

In the determination result 803, if ordinal j is less than m, which is the total number of i-th levels, it is determined whether ordinal j is a pseudo node (808), and if j is a pseudo node, the j pseudo node is transformed into an intersection point of an absolute edge. And the intersection is calculated by calculating the X and Y coordinates according to the offsets between the levels and the nodes (809), then increasing j by 1 (812), and is j less than m for the next node visit? In operation 803, if j is not the parent node, the j node X coordinate and the width position are calculated by the gap offset between the nodes (810), and the Y node of the j node by the gap offset between the levels. After calculating the coordinates and the height position (811), j is increased by 1 (812).

10 is a detailed flowchart illustrating a subroutine for generating the coordinates of the edges X and Y of the call graph layout in FIG. 2, and shows the flow chart of the edges X and Y coordinates of the direction graph layout. This is a process for outputting a call graph to the monitor 10 or the printer 11 by generating coordinates of X and Y.

First, n-1 level pairs with edges at n levels are searched (901), the ordinal i of the first level is initialized to 1 (902), and then the magnitude of the level at which the ordinal i of the first level is searched. It is determined whether is smaller than n-1 (903).

In the determination result 903, if the ordinal i of the first level is greater than or equal to n-1, which is the size of the searched level, is returned, and if the ordinal i of the first level is less than n-1, the size of the searched level, In order to calculate coordinates for outputting positions of edges between the i th level and the i + 1 th level, m edges are visited (904), and the ordinal j of the first node is initialized to '0' (905). ).

Then, it is determined whether ordinal j is smaller than m, the total number of nodes of the i th level (906), and if ordinal j is greater than or equal to m, the total number of nodes of the i th level, the ordinal i of the first level is reduced by one. Thereafter, in operation 907, a process 903 of determining whether ordinal i of the first level is smaller than n-1, which is the size of the searched level, is performed, and ordinal j of the first node m is the total number of nodes of the i th level. If smaller, it is determined whether the trunk of ordinal j is a dummy trunk (908).

In the determination result 908, if the trunk of ordinal j is a dummy trunk, the dummy trunk of ordinal j is replaced with an absolute trunk (909), and the X coordinate of the node j trunk is determined by the X coordinate and the width position of the node at the i th level. Calculate the coordinates, calculate the Y coordinate of the node j edge by the node's Y coordinate and the height position (910), and then calculate the X coordinate of the j edge by the node's X coordinate and the width position at the i + 1th level. After calculating the Y coordinate of the j trunk by the Y coordinate and the height position of the node (911), the ordinal j is increased by 1 (912), and determining whether j is less than m for the next node to visit. Perform 906.

In the determination result 908, if the edge of ordinal j is not a dummy trunk, the X coordinate of the j edge is calculated based on the X coordinate and the width position of the node at the i th level, and the ordinal number is determined by the Y coordinate and the height position of the node. A process 910 of calculating the Y coordinate of the j edge is performed.

The present invention described above is capable of various substitutions, modifications, and changes without departing from the spirit of the present invention for those skilled in the art to which the present invention pertains. It is not limited to the drawing.

As described above, the present invention provides a program maintainer with an easy-to-read and traceable high-quality call graph to easily provide the program maintainer with the understanding and analysis of the source program structure, and using each function call information. By providing measurement information such as complexity and structure diagram, it is easily used to analyze the source code to be maintained, thereby minimizing the time, effort, and cost required for the maintenance work of the program. There is an effect that can be automatically generated.

Claims (3)

In the call graph generation method for software maintenance, A first step of the translation unit extracting the intermediate language from the source program and storing the intermediate language in the intermediate language storage unit; A second step of extracting unit extracting function information necessary for generating a call graph from the intermediate language stored in the intermediate language storage unit and storing the extracted function information in a reverse engineering integrated information file; A third step in which an automatic call graph generation unit searches node and edge information of an input graph, which is call information of a function necessary for generating a call graph, from the reverse engineering integrated information file from a media language and stores it in a call graph information model storage unit; The call graph direction graph generator determines the level of each node by using the node and edge information of the input graph which is the function call information stored in the call graph information model storage unit, and determines the level of the edge according to the relationship of the node adjacent to the corresponding node. A fourth step of identifying the type; A fifth step of the call graph direction graph generation unit determining an order of nodes at the level of each node; A sixth step of generating, by the call graph direction graph generator, absolute coordinates of the node and the edges; And A seventh step of automatically generating a call graph by using the coordinates of the real nodes and edges generated by the call graph graphical interface as the direction graph. Automatic call graph generation method for software maintenance including a. The method of claim 1, The fourth step, An eighth step of visiting all nodes reachable from a given node; A ninth step of generating a tree edge and assigning a level if a node adjacent to the predetermined node has been visited; A tenth step of generating a forward edge according to a relationship between a level value of a node adjacent to the predetermined node, and generating a dummy node and a dummy edge with respect to the forward edge; An eleventh step of generating a back edge according to a relationship between a level value of a node adjacent to the predetermined node, and generating a dummy node and a dummy edge for the back edge; And After generating a tree edge between the predetermined node and the adjacent node according to the relationship between the level value of the node and the adjacent node, generates a forward edge between the ancestor nodes of the adjacent node, and modify the level value of the adjacent node And a twelfth step of generating a dummy node and a dummy edge for the forward edge. In the automatic call graph generation system with a microprocessor, A first function of the translation unit extracting the intermediate language from the source program and storing the intermediate language in the intermediate language storage unit; A second function of an extracting unit extracting function information necessary for generating a call graph from an intermediate language stored in the intermediate language storage unit and storing the function information in a reverse engineering integrated information file; A third function of automatically generating a call graph information model storing unit and edge information of an input graph which is call information of a function required for generating a call graph from the reverse engineering integrated information file from an intermediate language, and storing them in a call graph information model storage unit; The call graph direction graph generator determines the level of each node using node and edge information of the input graph, which is the function call information stored in the call graph information model storage unit, and determines the A fourth function of identifying a kind; A fifth function of the call graph direction graph generator determining an order of nodes at the level of each node; A sixth function of generating, by the call graph direction graph generator, absolute coordinates of the node and the edges; And The seventh function of the call graph graphical interface automatically generates a call graph using the coordinates of the actual nodes and edges generated by the direction graph. A computer-readable recording medium having recorded thereon a program for realizing this.