Methods, systems and computer-readable media for detecting a partial commit

Patent metadata

Field	Value
Publication number	US-9785430-B2
Application number	US-201414198878-A
Country	US
Kind code	B2
Filing date	Mar 6, 2014
Priority date	Mar 21, 2013
Publication date	Oct 10, 2017
Grant date	Oct 10, 2017

Abstract

The present invention provides a method and system for detecting a partial commit of software. A dependency information of the software is extracted from a version history and a bug database. A dimensional matrix containing a set of commit, and relationship information with a set of files with each commit is created from the dependency information. A centrality matrix is computed by performing a first set of matrix transformations on the dimensional matrix. A set of missing files of a partial commit, is identified by performing a second set of matrix transformations on the centrality matrix and a file vector, the file vector including a file dependency information of the partial commit.

First claim

What is claimed:

1. A software version analysis device comprising a memory having programmed instructions stored thereon and at least one processor coupled to the memory and configured to execute the stored programmed instructions to:
extract historical dependency information from a version repository and a bug database for a software application;
generate a dimensional matrix based on the historical dependency information, wherein the dimensional matrix comprises at least an indication of a set of commits and an indication of a set of files included in each of the commits;
generate a centrality matrix from the dimensional matrix based on one or more matrix transformations performed on the dimensional matrix;
obtain file dependency information of a received another commit associated with the software application, the another commit identifying a plurality of files to be committed;
determine when the another commit is a partial commit having one or more missing files, which were not included in the plurality of files to be committed, based on one or more matrix transformations performed using the file dependency information and the centrality matrix; and
determine a set of plausible files for the another commit, generate a ranking of the set of plausible files according to a probability of occurrence in the another commit, determine the missing files based on the ranking or a predetermined threshold, and update a knowledge database or the predetermined threshold, when the determining indicates that the another commit is a partial commit.

2. The software version analysis device of claim 1 , wherein the dependency information comprises one or more of an indication of one or more transactions performed on the software application, a set of bugs in the software application, or fix information available for the set of bugs.

3. The software version analysis device of claim 1 , wherein the centrality matrix is a right singular matrix and the one or more matrix transformations comprises a singular vector decomposition of the dimensional matrix.

4. The software version analysis device of claim 1 , wherein the file dependency information comprises a vector comprising one or more of an author of the commit, a set of committed files, or a type of modification.

5. The software version analysis device of claim 1 , wherein the missing files are represented as a weighted vector, wherein a weight of each of the missing files in the weighted vector indicates a probability that the missing file should have been included in the plurality of files to be committed of the another commit.

6. The software version analysis device of claim 1 , wherein the processor is further configured to execute the stored programmed instructions to:
receive feedback on an inconsistency on the missing files; and
update the knowledge database or the predetermined threshold based on the feedback.

7. A method of detecting software application partial commits implemented by one or more software version analysis devices, the method comprising:
extracting, historical dependency information from a version repository and a bug database for a software application;
generating a dimensional matrix based on the historical dependency information, wherein the dimensional matrix comprises at least an indication of a set of commits and an indication of a set of files included in each of the commits;
generating a centrality matrix from the dimensional matrix based on one or more matrix transformations performed on the dimensional matrix;
obtaining file dependency information of a received another commit associated with the software application, the another commit identifying a plurality of files to be committed;
determining when the another commit is a partial commit having one or more missing files, which were not included in the plurality of files to be committed, based on one or more matrix transformations performed using the file dependency information and the centrality matrix; and
determining a set of plausible files for the another commit, generating a ranking of the set of plausible files according to a probability of occurrence in the another commit, determining the missing files based on the ranking or a predetermined threshold, and updating a knowledge database or the predetermined threshold, when the determining indicates that the another commit is a partial commit.

8. The method of claim 7 , wherein the dependency information comprises one or more of an indication of one or more transactions performed on the software application, a set of bugs in the software application, or fix information available for the set of bugs.

9. The method of claim 7 , wherein the centrality matrix is a right singular matrix and the one or more matrix transformations comprises a singular vector decomposition of the dimensional matrix.

10. The method of claim 7 , wherein the file dependency information comprises a vector comprising one or more of an author of the commit, a set of committed files, or a type of modification.

11. The method of claim 7 , wherein the missing files are represented as a weighted vector, wherein a weight of each of the missing files in the weighted vector indicates a probability that the missing file should have been included in the plurality of files to be committed of the another commit.

12. The method of claim 7 , further comprising:
receiving feedback on an inconsistency on the missing files; and
updating the knowledge database or the predetermined threshold based on the feedback.

13. A non-transitory computer readable medium having stored thereon instructions for detecting software application partial commits comprising machine executable code which when executed by at least one processor, causes the processor to:
extract historical dependency information from a version repository and a bug database for a software application;
generate a dimensional matrix based on the historical dependency information, wherein the dimensional matrix comprises at least an indication of a set of commits and an indication of a set of files included in each of the commits;
generate a centrality matrix from the dimensional matrix based on one or more matrix transformations performed on the dimensional matrix;
obtain file dependency information of a received another commit associated with the software application, the another commit identifying a plurality of files to be committed;
determine when the another commit is a partial commit having one or more missing files, which were not included in the plurality of files to be committed, based on one or more matrix transformations performed using the file dependency information and the centrality matrix; and
determine a set of plausible files for the another commit, generate a ranking of the set of plausible files according to a probability of occurrence in the another commit, determine the missing files based on the ranking or a predetermined threshold, and update a knowledge database or the predetermined threshold, when the determining indicates that the another commit is a partial commit.

14. The non-transitory computer readable medium of claim 13 , wherein the dependency information comprises one or more of an indication of one or more transactions performed on the software application, a set of bugs in the software application, or fix information available for the set of bugs.

15. The non-transitory computer readable medium of claim 13 , wherein the centrality matrix is a right singular matrix and the one or more matrix transformations comprises a singular vector decomposition of the dimensional matrix.

Assignees

• Infosys Ltd

Inventors

• Viswanathan Sree Aurovindh
• Rama Girish Maskeri

Classifications

• G06F8/65

  Updates (security arrangements therefor G06F21/57) · CPC title

• G06F8/71 Primary

  Version control (security arrangements therefor G06F21/57); Configuration management · CPC title

• G06F8/72

  Code refactoring · CPC title