KR100871820B1 - System for assorting source error and method therefor - Google Patents

Abstract

A source error sorting system is presented to reduce false alarm rate, by selecting an item with highest possibility of true alarm reported through a bug detector. According to a source error sorting system, a sorting item search module(110) gives a corresponding basic score by comparing a bug item extracted from more than one bug detector with a class sorting item. A bug condition search module(120) gives an additional point to the bug item according to the number of bug item conditions extracted from the bug detectors. A common item search module(130) gives an additional point to the searched bug item among bug items extracted from the bug detectors. A bug selection module(140) determines a class with the highest score given based on the bug item and the scores given according thereto, as a bug.

Description

System for assorting source error and method therefor}

1 is an overall configuration of a source error screening system according to the present invention.

2 is an overall flowchart of a source error screening method according to the present invention;

3 is a detailed flowchart of a step S300 of searching for a bug item commonly reported among bug items according to the present invention and assigning an additional point.

<Description of Symbols for Main Parts of Drawings>

10: bug type classification module 110: classification item search module

120: bug occurrence condition search module 130: common item search module

140: bug screening module

The present invention relates to a source error screening system and a method thereof, and more particularly, using the result data extracted through one or more conventional bug detectors for detecting a source error, so-called 'bug', Classifying the class according to the type to give a basic score, and cross-check the error precondition of the bug and the bug check data extracted from each of the bug detectors are commonly reported By performing the process of assigning additional points to the bug items, and finally adjusting the bug items according to the given scores by class, the most likely item is the highest alarm, that is, the true alarm among the bugs reported through the bug detector. Source error screening system and method that can easily screen items and reduce false alarm rate overall .

As is well known, existing bug detectors analyze the source and find the source code where the bug may occur and report the results.

The results from a single bug detector have a relatively high false alarm and report a large number of bugs depending on the software to be analyzed. At this time, all the bugs generated in large quantities have to be analyzed as a result of False Alarm, so there was a problem that the user consumes a lot of resources.

The present invention was devised to solve the above problems, and based on the data extracted from one or more bug detectors, classify the class according to the type of bug and give a basic score, and precondition of error And Cross-check each bug check data extracted from the bug detectors to give an additional point, and finally adjust the bug items according to the given scores by class, so that the true alarm of the bugs reported through the bug detectors The present invention relates to a source error screening system and a method for easily selecting a high probability item and reducing the false alarm rate as a whole.

The present invention relates to a source error screening system, comprising: a classification item search module for comparing a bug item extracted from one or more bug detectors with a classification item for each class to give a corresponding basic score; A bug occurrence condition search module for assigning a predetermined addition point to a bug item according to a predetermined number of respective bug item occurrence conditions extracted from the bug detectors; A common item search module for searching for commonly reported items among the bug items extracted from the bug detectors, and giving an added point to the found bug items; And selecting a bug by judging the class having the highest final score as a bug based on the bug items classified through the classification item search module, the bug occurrence condition search module, and the common item search module and the scores assigned thereto. module; It includes.

In addition, a bug type classification module for classifying the bug items extracted from the bug detectors by class based on a type classification criterion of a preset bug; It characterized in that it further comprises.

In addition, the class is classified into six, characterized in that the basic score is given from 0 to 100 points according to the class.

In addition, the predetermined number of the bug item occurrence condition is characterized in that 0 to 89.

The predetermined addition points given through the common item search module may be 0 to 17 points so that the sum of the basic points and the additional points of the class does not directly exceed the points of the upper class.

Meanwhile, the present invention relates to a source error screening method using a source error screening system comprising a bug type classification module, a classification item search module, a bug occurrence condition search module, a common item search module, and a bug screening module. Comparing a bug item extracted from the one or more bug detectors with a classification item for each class to give a corresponding base score; (b) the source error screening system assigning an additional point to a bug item according to a predetermined number of respective bug item occurrence conditions extracted from the bug detectors; (c) searching for a commonly reported item among the bug items extracted from the bug detectors by the source error screening system, and assigning additional points to the found bug items; And (d) determining, by the source error screening system, the class having the highest final score assigned as a bug based on a classified bug item and a score assigned thereto. It includes.

The method may further include, before the step (a), classifying, by the source error screening system, the bug items extracted from the bug detectors by class based on a preset type of bug classification; It characterized in that it further comprises.

In addition, the step (c), (c-1) the source error screening system to search for each of the bug items extracted from the bug detectors; (c-2) the source error screening system determining whether a bug item commonly reported from the bug detectors exists; And (c-3) determining that the source error screening system adds an additional point to the detected bug item when there is a commonly reported item as a result of the determination; It characterized in that it further comprises.

The features and advantages of the present invention will become more apparent from the following detailed description based on the accompanying drawings. Prior to this, the terms or words used in the present specification and claims are defined in the technical spirit of the present invention on the basis of the principle that the inventor can appropriately define the concept of the term in order to explain his invention in the best way. It should be interpreted to mean meanings and concepts. In addition, when it is determined that the detailed description of the known function and its configuration related to the present invention may unnecessarily obscure the subject matter of the present invention, it should be noted that the detailed description is omitted.

Hereinafter, with reference to the accompanying drawings will be described in detail the present invention.

Prior to this, the process of the source error screening system and method according to the present invention is to use the bug check data of the target software extracted through the conventional first and second bug detectors in order to easily screen the bugs of the target software. It is desirable to understand.

A source error screening system according to the present invention will be described with reference to FIG. 1 is a block diagram conceptually illustrating a configuration of a source error screening system 100 according to the present invention. As illustrated, the bug type classification module 10 may include bug check data extracted from first and second bug detectors. (Hereinafter, referred to as a 'bug item') performs a function of classifying by class based on a predetermined type classification of a bug.

The classification item search module 110 compares a bug item extracted from each of the first and second bug detectors with a classification item for each class and performs a function of assigning a corresponding basic score.

The bug occurrence condition search module 120 performs a function of assigning a predetermined addition point to a bug item according to a predetermined number of respective bug item occurrence conditions extracted from the first and second bug detectors, respectively.

In particular, bugs vary from case to case and occur under dozens of conditions. Therefore, it is assumed that the number of bug occurrence conditions is easier for the bug detector to analyze, and the items having a small number of conditions are classified as bugs, and the addition points are differentiated. At this time, the addition point is set so that the sum of the base point and the addition point of the present class is less than 18 points so that the sum of the upper class does not directly exceed the score of the upper class.

The common item search module 130 searches for a commonly reported item among the bug items extracted from the first and second bug detectors, respectively, and performs a function of giving an added point to the found bug item. Specifically, the common item searching module 130 assigns 99 points to the items commonly reported from the first and second bug detectors.

For reference, examples of bugs having a common type used in a conventional bug detector are shown in Table 1 below.

First bug detector 2nd bug detector Buffer Overrun, Type Overrun 3680 [U] Indexing array with constant value that is out of bounds. 3683 Array index constant addresses the location one beyond the end of the array. 3685 [U] Indexing array with value that will definitely be out of bounds. 3688 Array index value definitely addresses the location one beyond the end of the array. Return Pointer To Local 3216 Address of local static data assigned to pointer with linkage or wider scope. 3217 Address of automatic data assigned to pointer with linkage or wider scope. 3225 Function parameter returns address of automatic data. 3230 Address of automatic data assigned to local pointer with static storage duration. 3231 Function parameter returns address of local static data. 4139 Function returns address of local static data. 4140 Function returns address of automatic data. Uninitialized Variable 3321 [U] The variable '% s'is definitely unset at this point. Missing Return Statement 0744 [U] '% s ()' has been declared with a non void result type but ends with an implicit 'return;' statement. 0745 [U] 'return;' found in '% s ()', which has been defined with a non-'void 'return type.

[Table 1] Items of bugs with common types

As described above, in order to achieve the characteristic object of the present invention that screens items which are likely to be True Alarms among the reported bugs, the bug screening module 140 according to the aspect of the present invention is classified category search module. (110), the bug occurrence condition classification module 120, the bug items classified through the common item search module 130 and the score given accordingly, preferably, the basic score initially given according to the type of bug, and the bug occurrence condition A bug item with the lowest false alarm rate, i.e. the highest final score given, based on a bug item classified according to the addition point given according to and the addition points for items commonly retrieved from the first and second bug detectors. Determines a class as a bug and selects it.

More specifically, if the false alarm rate is reduced, the false alarm rate of class A is 10% and the false alarm rate of class B is 50% .If 100 items are reported, the false alarm is (100 * 0.1). + 100 * 0.5) / 200 * 100 = 30%, and if 90 items of class A are selected and 10 items of class B are selected out of 200 items (90 * 0.1 + 10 * 0.5) / 100 * 100 = 14%, which improves the false alarm rate by 16%.

A source error screening method according to the present invention using the above-described source error screening system will be described with reference to FIGS. 2 and 3 as follows.

FIG. 2 is a flowchart illustrating a source error screening method, classifying bug items by class as shown in FIG. 10, and assigning basic scores to bug items extracted from the first and second bug detectors, respectively. (S100), a process of assigning an additional point according to the number of bug item occurrence conditions (S200), and a process of assigning an additional point to a commonly reported item among bug items extracted from the first and second bug detectors (S300). And based on the bug items classified through the bug items and the scores assigned thereto, determining and classifying the class having the highest final score as a bug (S400).

In detail, the bug type classification module 10 classifies the bug items extracted from the first and second bug detectors by class according to a preset type classification criterion of the bug (S10), and the classification item search module ( In operation 110, the bug items extracted from the first and second bug detectors are compared with the classification items for each class, and a corresponding base score is assigned (S100). The basic scores and classified bug types assigned to each class are briefly described as in the following [Table 2].

division Bug Items by Type Base score class 1 DOUBLE_FREE FREE_NON_HEAP_VARIABLE RETURN_POINTER_TO_LOCAL USE_AFTER_FREE ............ 100 class 2 BUFFER_OVERRUN * TYPE_OVERRUN * UNINITIALIZED_VARIABLE * ........... 80 class 3 NULL_POINTER_DEREFERENCE IGNORED_RETURN_VALUE ........... 60 class 4 FORMAT_STRING UNUSED_VALUE ........... 40 class 5 BUFFER_UNDERRUN SHIFT_AMOUNT_EXCEEDS_BIT_WIDTH ........... 20 class 6 Other items ........... 0

[Table 2] Class classification table by bug type

As described above, in the case of [class 1], a basic score of 100 points is given as an item to be reported unconditionally as a fatal error. In the present embodiment, the types of bugs have been specifically classified and described, but may be changed according to the user's classification criteria, and thus the present invention is not limited to the classification.

Thereafter, the bug occurrence condition search module 120 assigns a predetermined addition point to the bug item according to a predetermined number of respective bug item occurrence conditions extracted from the first and second bug detectors, respectively (S200).

For example, the number of occurrence conditions can be set to 17 points for 0 to 4, 16 points for 5 to 9, 15 points for 10 to 14, and 0 points for 85 to 89. As described above in this embodiment, the predetermined number of bug occurrence conditions is set to 0 to 89 and the addition point is set to 0 to 17 points, but the present invention is not limited thereto, and various setting changes are possible. .

As described above, the common item search module 130 searches for a commonly reported item among the bug items extracted from the first and second bug detectors, and performs an operation of assigning an additional point to the found bug item (S300). When the detailed flow is described with reference to FIG. 3, the common item search module 130 searches for bug items extracted from the first and second bug detectors (S320), and whether a commonly reported bug item exists. If it is determined whether or not (S340), and the determination result of step S340, if there is an item that is commonly reported, give a score of 99 points to the detected bug items (S360), if the determination result of step S340, The procedure of step S400 is performed.

This is to search for common types and use them to select bug items, since items commonly reported from each bug detector are likely to be true alarms.

Then, the bug screening module 140 is common from the bug item and the score given accordingly, preferably the base score initially given according to the bug type, the addition point given according to the bug occurrence condition, and the first and second bug detectors. Based on the bug items classified through the addition points for the searched items, the class having the highest final score is determined as the bug and is selected (S400).

According to the present invention described above, a bug item is reported based on the bug check data extracted through the first and second bug detectors so that the normal source error screening system 100 checks the bug of the target software. By selecting items with a high true alarm rate, you can eventually lower the false alarm rate in the class of bug items. Thus, the user can quickly prove the actual bug, and ultimately has the characteristic advantage of shortening the target software development period.

As described above and described with reference to a preferred embodiment for illustrating the technical idea of the present invention, the present invention is not limited to the configuration and operation as shown and described as described above, it is a deviation from the scope of the technical idea It will be understood by those skilled in the art that many modifications and variations can be made to the invention without departing from the scope of the invention. Accordingly, all such suitable changes and modifications and equivalents should be considered to be within the scope of the present invention.

According to the present invention as described above, there is an effect that can efficiently review the vast amount of result data extracted from the bug detector.

In addition, according to the present invention, it is possible to shorten the software development period by finding bugs of the target software in a short time.

Claims (8)

In the source error screening system, A classification item search module 110 for comparing a bug item extracted from one or more bug detectors with a classification item for each class to give a corresponding basic score; A bug occurrence condition search module 120 for assigning a predetermined addition point to a bug item according to a predetermined number of respective bug item occurrence conditions extracted from the bug detectors; A common item search module 130 for searching for commonly reported items among the bug items extracted from the bug detectors and giving an added point to the found bug items; And Based on the bug items classified through the classification item search module 110, the bug occurrence condition search module 120, and the common item search module 130 and the scores assigned thereto, the class with the highest final score is given. A bug screening module 140 for judging and determining a bug; Source error screening system comprising a. The method of claim 1, A bug type classification module (10) for classifying the bug items extracted from the bug detectors by class based on a type classification criterion of a preset bug; Source error screening system, characterized in that it further comprises. The method of claim 1, The class is classified into six, the source error screening system, characterized in that the base score is given from 0 to 100 points according to the class. The method of claim 1, The predetermined number of bug item occurrence conditions is Source error screening system, characterized in that 0 to 89. The method of claim 1, The predetermined addition point given through the common item search module is Source error screening system, characterized in that 0 to 17 points so that the sum of the basic score and the addition point of this class does not directly exceed the score of the upper class. Source error screening using a source error screening system consisting of a bug type classification module 10, a classification item search module 110, a bug occurrence condition search module 120, a common item search module 130, and a bug screening module 140. In the method, (a) the source error screening system 100 comparing a bug item extracted from one or more bug detectors with a classification item for each class and assigning a corresponding base score; (b) the source error screening system 100 assigning an additional point to a bug item according to a predetermined number of respective bug item occurrence conditions extracted from the bug detectors; (c) the source error screening system 100 searching for a commonly reported item among the bug items extracted from the bug detectors, and assigning an additional point to the found bug item; And (d) determining, by the source error screening system, the class having the highest assigned final score as a bug based on a classified bug item and a score assigned thereto; Source error screening method comprising a. The method of claim 6, Prior to step (a) above, Classifying, by the source error screening system, the bug items extracted from the bug detectors by class based on a predetermined class of bug classification; Source error screening method, characterized in that it further comprises. The method of claim 6, Step (c) is, (c-1) searching for the bug items extracted from the bug detectors by the source error screening system (100); (c-2) the source error screening system 100 determining whether a bug item commonly reported from the bug detectors exists; And (c-3) determining that the source error screening system 100 adds an additional point to the found bug item when there is a commonly reported item as a result of the determination; Source error screening method, characterized in that it further comprises.

Images