MODULE GENERATING APPARATUS, MODULE GENERATING METHOD, AND COMPUTER PRODUCT

• US 20090307672A1
• Filed: 08/18/2009
• Published: 12/10/2009
• Est. Priority Date: 02/27/2007
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Yamashita, Koichiro
• Granted Patent: US 8,286,150 B2
• Time in Patent Office: Days
• US Class Current: 717/155
• CPC Class Codes: G06F 8/00 Arrangements for software engineering testing or debugging G06F11/36; administrative, planning or organisation aspects of software project management G06Q10/06
• CPC Class Codes: G06F 8/42 Syntactic analysis
• CPC Class Codes: G06F 8/44 Encoding

Abstract

A computer is caused to function as a parsing unit, a macroblocking analyzing unit, a junction-node restructuring unit, an identical portion merging/restructuring unit, a similar portion merging/restructuring unit, and an intermediate language restructuring unit. The parsing unit performs syntax analysis of a source code. The macroblocking analyzing unit segments the program written in the source code into blocks and appends a virtual portion representing a unique number in a statement, to a number for identifying a variable for the statement in each block to virtualize a calculation pattern. The junction-node restructuring unit extracts a node directly related to a subroutine block. The identical portion merging/restructuring unit merges pre-processing together and post-processing together for a subroutine called up at a multiple portions in the program. The similar portion merging/restructuring unit integrates subroutines having similar structures into a related subroutine.

Claims

1. A module generating apparatus comprising:

a parsing unit that performs syntax analysis of a source code for a program and outputs information concerning the syntax analysis;

a macroblocking analyzing unit that based on syntax analysis information output from the parsing unit, segments the program into a plurality of blocks through structural analysis of the program and outputs block structure information obtained as a result of the segmentation and appendant information for virtualizing a calculation pattern of each block;

a junction-node restructuring unit that based on the block structure information and the appendant information output from the macroblocking analyzing unit, focuses on one block, defined as a subroutine, extracts a node directly related to the block of focus by tracing back a processing sequence to a node that is in an ancestor block upstream in the processing sequence from the block of focus and from which the block of focus is reached as a result of sequential processing, and by tracing the processing sequence to a node that is in a descendant block downstream in the processing sequence from the block of focus and that is processed after the block of focus;

an identical portion merging/restructuring unit that, with respect to a plurality of portions that use the block of focus extracted from the program by the junction-node restructuring unit, compares nodes in ancestor blocks of the block of focus to merge the ancestor blocks, compares nodes in descendant blocks of the block of focus to merge the descendant blocks, and outputs merged-block information obtained as a result of merging the ancestor blocks and merging the descendant blocks;

a similar portion merging/restructuring unit that with respect to a second block of focus extracted from the program by the junction-node restructuring unit, compares the merged-block information to define, as subroutines of an identical type, a plurality of blocks of focus that have similar or identical merged block information, further merges similar or identical merged block information, and outputs abstracted code information obtained as a result of the merging; and

an intermediate language restructuring unit that outputs information of a commonized component, based on the abstracted code information and the syntax analysis information.

2. The module generating apparatus according to claim 1, wherein

the macroblocking analyzing unit appends, to a number identifying a variable for a statement in each block, a virtual portion representing a number unique in the statement, the virtual portion being appended as the appendant information and to express the number for identifying the variable by a combination of a lower portion representing a number unique in the source code entirely and an upper portion made up of the virtual portion.

3. The module generating apparatus according to claim 2, wherein

the identical portion merging/restructuring unit performs comparison between nodes in the ancestor block and between nodes in the descendant block by identifying a variable based on the virtual portion.

4. The module generating apparatus according to claim 2, wherein

the similar portion merging/restructuring unit compares the merged block information by identifying a variable based on the virtual portion.

5. A module generating method comprising:

parsing involving syntax analysis of a source code for a program and obtaining information concerning the syntax analysis;

macroblocking-analyzing involving segmentation of the program into a plurality of blocks through structural analysis of the program based on syntax analysis information obtained from the parsing, obtaining block structure information as a result of the segmentation, and obtaining appendant information for virtualizing a calculation pattern of each block;

commonizing a component, based on the block structure information and the appendant information.

6. The module generating method according to claim 5, wherein

the commonizing further involves junction-node-restructuring that is based on the block structure information and the appendant information and involves focusing on one block defined as a subroutine, extracting a node directly related to the block of focus by tracing back a processing sequence to a node that is in an ancestor block upstream in the processing sequence from the block of focus and from which the block of focus is reached as a result of sequential processing, and by tracing the processing sequence to a node that is in a descendant block downstream in the processing sequence from the block of focus and that is processed after the block of focus.

7. The module generating method according to claim 6, wherein

the commonizing further involves identical-portion-merging that, with respect to a plurality of portions that use the block of focus extracted from the program at the junction-node-restructuring, involves comparing nodes in ancestor blocks of the block of focus, merging the ancestor blocks, comparing nodes in descendant blocks of the block of focus, merging the descendant blocks, and obtaining merged-block information as a result of the merging of the ancestor blocks and of the descendant blocks.

8. The module generating method according to claim 7, wherein

the commonizing further involves similar-portion-merging that, with respect to a second block of focus extracted from the program at the junction-node restructuring, involves comparing the merged-block information to define, as subroutines of an identical type, a plurality of blocks of focus that have similar or identical merged block information, further merging similar or identical merged block information, and obtaining abstracted code information as a result of the merging.

9. The module generating method according to claim 8, wherein

the commonizing further involves intermediate-language-restructuring that obtains information concerning a commonized component, based on the abstracted code information and the syntax analysis information.

10. The module generating method according to claim 9, wherein

the macroblocking-analyzing further involves appending, to a number identifying a variable for a statement in each block, a virtual portion representing a number unique in the statement, the virtual portion being appended as the appendant information and to express the number for identifying the variable by a combination of a lower portion representing a number unique in the source code entirely and an upper portion made up of the virtual portion.

11. The module generating method according to claim 10, wherein

the identical-portion-merging further involves comparison between nodes in the ancestor block and between nodes in the descendant block by identifying a variable based on the virtual portion.

12. The module generating apparatus according to claim 10, wherein

the similar-portion-merging further involves comparing the merged block information by identifying a variable based on the virtual portion.

13. A computer-readable recording medium storing therein a module generating program that causes a computer to execute:

parsing involving syntax analysis of a source code for a program and obtaining information concerning the syntax analysis;

macroblocking-analyzing involving segmentation of the program into a plurality of blocks through structural analysis of the program based on syntax analysis information obtained from the parsing, obtaining block structure information as a result of the segmentation, and obtaining appendant information for virtualizing a calculation pattern of each block;

junction-node-restructuring based on the block structure information and the appendant information and involving, focusing on one block defined as a subroutine, extracting a node directly related to the block of focus by tracing back a processing sequence to a node that is in an ancestor block upstream in the processing sequence from the block of focus and from which the block of focus is reached as a result of sequential processing, and by tracing the processing sequence to a node that is in a descendant block downstream in the processing sequence from the block of focus and that is processed after the block of focus;

identical-portion-merging that, with respect to a plurality of portions that use the block of focus extracted from the program at the junction-node-restructuring, involves comparing nodes in ancestor blocks of the block of focus, merging the ancestor blocks, comparing nodes in descendant blocks of the block of focus, merging the descendant blocks, and obtaining merged-block information as a result of the merging of the ancestor blocks and of the descendant blocks;

similar-portion-merging that, with respect to a second block of focus extracted from the program at the junction-node restructuring, involves comparing the merged-block information to define, as subroutines of an identical type, a plurality of blocks of focus that have similar or identical merged block information, further merging similar or identical merged block information, and obtaining abstracted code information as a result of the merging; and

intermediate-language-restructuring that obtains information concerning a commonized component, based on the abstracted code information and the syntax analysis information.

14. The computer-readable recording medium according to claim 13, wherein

the macroblocking-analyzing further involves appending, to a number identifying a variable for a statement in each block, a virtual portion representing a number unique in the statement, the virtual portion being appended as the appendant information and to express the number for identifying the variable by a combination of a lower portion representing a number unique in the source code entirely and an upper portion made up of the virtual portion.

15. The computer-readable recording medium according to claim 14, wherein

the identical-portion-merging further involves comparison between nodes in the ancestor block and between nodes in the descendant block by identifying a variable based on the virtual portion.

16. The computer-readable recording medium according to claim 14, wherein

the similar-portion-merging further involves comparing the merged block information by identifying a variable based on the virtual portion.