DETECTION OF ANTIPATTERNS THROUGH STATISTICAL ANALYSIS

• US 20170091073A1
• Filed: 09/30/2015
• Published: 03/30/2017
• Est. Priority Date: 09/30/2015
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Abadi, Aharon, Abadi, Moria, Tripp, Omer
• Granted Patent: US 10,055,329 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 11/3608 using formal methods, e.g. model checking, abstract interpretation theorem proving G06N5/013
• CPC Class Codes: G06F 11/3616 using software metrics
• 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 8/20 Software design
• CPC Class Codes: G06F 8/36 Software reuse
• CPC Class Codes: G06F 8/43 Checking; Contextual analysis

Abstract

A computer implemented method of detecting code antipatterns, comprising:

• 1) Receiving a code containing one or more of a plurality of code segments, each one of the one or more code segments includes one or more of a plurality of patterns.
• 2) Automatically analyzing each one of the one or more code segment to create an array of a plurality of features vectors, each one of the plurality of features vectors maps a plurality of predefined features found in one or more patterns.
• 3) Matching each one of the plurality of features vectors with a plurality of template features vectors each representing one of a plurality of antipatterns, the plurality of template features vectors is stored in an antipatterns dataset.
• 4) Determining a presence or an absence of each of the plurality of antipatterns within each of the one or more code segments according to the matching.

Claims

1. A computer implemented method of detecting code antipatterns, comprising:

receiving a code containing at least one of a plurality of code segments, each one of said at least one code segment includes at least one pattern;

automatically analyzing each one of said at least one code segment to create an array of a plurality of features vectors, each one of said plurality of features vectors maps a plurality of predefined features found in one of said at least one pattern;

matching each one of said plurality of features vectors with a plurality of template features vectors each representing one of a plurality of antipatterns, said plurality of template features vectors is stored in an antipatterns dataset; and

determining a presence or an absence of each of said plurality of antipatterns within said each at least one code segment according to said matching.

2. The method of claim 1, further comprising identifying at least one approved pattern to replace at least one of said plurality of antipatterns detected in said code segment, said at least one approved pattern exhibits a similar functional behavior as said at least one antipattern, said at least one approved pattern is retrieved from an approved patterns dataset comprising at least one approved pattern.

3. The method of claim 1, further comprising at least one of said plurality of template features vectors is created through statistical analysis of a plurality of previous patterns having functional characteristics similar to said at least one pattern, said plurality of previous patterns is stored in a previous patterns dataset.

4. The method of claim 1, further comprising at least one of said plurality of template features vectors is created using feedback of at least one other software developer who performed code review over said at least one pattern and identifies at least one of said plurality of antipatterns.

5. The method of claim 1, further comprising at least one of said plurality of template features vectors is created through analysis of a log which holds a history of at least one of a plurality of changes performed to said at least pattern in order to determine whether said at least one pattern is an antipattern.

6. The method of claim 1, further comprising at least one of said plurality of template features vectors is created through analysis of a log that holds a history of at least one of a plurality of changes made to said at least one pattern by at least two software developers to determine whether said at least one pattern is an antipattern.

7. The method of claim 1, further comprising said matching employs voting between at least one matching algorithms, each of said at least one matching algorithms employs a different method of comparing said each features vector with said plurality of template features vectors.

8. A system for detecting code antipatterns, comprising:

an interface for communicating with a user;

a program store storing a code; and

a processor coupled to the first interface and the program store for implementing the stored code, the code comprising;

code instructions for receiving a code having a plurality of code segments, each one of said plurality of code segments includes at least one pattern, said code is received according to a user input received via said interface;

code instructions for automatically analyzing each one of said plurality of code segments to create an array consisting of a plurality of features vectors, each one of said plurality of features vectors maps a plurality of predefined features found in one of said at least one pattern;

code instructions for comparing each one of said plurality of features vectors with a plurality of template features vectors each representing one of a plurality of antipatterns, said plurality of template features vectors is stored in an antipatterns dataset; and

code instructions for determining a presence or an absence of at least one of said plurality of antipatterns within said each code segments according to said compare.

9. The system of claim 8, further comprising said processor is adapted for implementing the following code instructions in said code stored:

code instructions for identifying at least one approved pattern to replace at least one of said plurality of antipatterns detected in said code segment, said at least one approved pattern exhibits a similar functional behavior as said at least one antipattern, said at least one approved pattern is retrieved from an approved patterns dataset comprising at least one approved pattern.

10. The system of claim 8, further comprising said processor is adapted for implementing the following code instructions in said code stored:

code instructions for creating at least one of said plurality of template features vectors through statistical analysis of a plurality of previous patterns having functional characteristics similar to said at least one pattern, said plurality of previous patterns is stored in a previous patterns dataset.

11. The system of claim 10, further comprising said processor is adapted for implementing the following code instructions in said code stored:

code instructions for creating said at least one template features vectors using feedback of at least one other software developer who performed code review over said at least one pattern and identifies said at least one antipattern.

12. The system of claim 10, further comprising said processor is adapted for implementing the following code instructions in said code stored:

code instructions for creating said at least one template features vectors through analysis of a log which holds a history of at least one of a plurality of changes performed to said at least one pattern in order to determine whether said at least one pattern is an antipattern.

13. The system of claim 10, further comprising said processor is adapted for implementing the following code instructions in said code stored:

code instructions for creating said at least one template features vectors through analysis of a log which holds a history of at least one of a plurality of changes made to said at least one pattern by at least two software developers to determine whether said at least one pattern is an antipattern.

14. A computer program product for detecting code antipatterns, comprising:

a non-transitory computer readable storage medium;

first program instructions to receive a code that includes at least one of a plurality of code segments, each one of said at least one code segments includes at least one pattern;

second program instructions to analyze each one of said plurality of code segments to create an array consisting of a plurality of features vectors, each one of said plurality of features vectors maps a plurality of predefined features found in one of said at least one pattern;

third program instructions to match each one of said plurality of features vectors with a plurality of template features vectors each representing one of a plurality of antipatterns, said plurality of template features vectors is stored in an antipatterns dataset; and

fourth program instructions to determine a presence or an absence of each of said plurality of antipatterns within said each at least one code segments according to said match;

wherein said first, second, third and fourth program instructions are stored on said non-transitory computer readable storage medium.

15. The computer program product of claim 14, further comprising a fifth program instruction to calculate at least one approved pattern to replace at least one of said plurality of antipatterns detected in said code segment, said at least one approved pattern exhibits a similar functional behavior as said at least one antipattern, said at least one approved pattern is retrieved from an approved patterns dataset comprising at least one approved pattern, wherein said fifth program instruction is stored on said non-transitory computer readable storage medium.

16. The computer program product of claim 14, further comprising a sixth program instruction to create at least one of said plurality of template features vectors through statistical analysis of a plurality of previous patterns having functional characteristics similar to said at least one pattern, said plurality of previous patterns is stored in a previous patterns dataset, wherein said sixth program instruction is stored on said non-transitory computer readable storage medium.

17. The computer program product of claim 16, further comprising said code instructions for creating said at least one template features vectors uses feedback of at least one other software developer who performed code review over said at least one pattern.

18. The computer program product of claim 16, further comprising said code instructions for creating said at least one template features vectors analyzes a log which holds a history of at least one of a plurality of changes performed to said at least one pattern in order to determine whether said at least one pattern is an antipattern.

19. The computer program product of claim 16, further comprising said code instructions for creating said at least one template features vectors analyzes a log which holds a history of at least one of a plurality of changes made to said at least one pattern by at least two software developers to determine whether said at least one pattern is an antipattern.

20. The computer program product of claim 14, further comprising said third program instructions employs voting between at least one matching algorithms, each of said at least one matching algorithms employs a different method of comparing said each features vector with said plurality of template features vectors.