CN110162335A - Code refactoring method, apparatus, computer equipment and medium - Google Patents
Code refactoring method, apparatus, computer equipment and medium Download PDFInfo
- Publication number
- CN110162335A CN110162335A CN201910345048.9A CN201910345048A CN110162335A CN 110162335 A CN110162335 A CN 110162335A CN 201910345048 A CN201910345048 A CN 201910345048A CN 110162335 A CN110162335 A CN 110162335A
- Authority
- CN
- China
- Prior art keywords
- code
- sample
- code data
- sorted
- type
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000013515 script Methods 0.000 claims abstract description 63
- 238000013145 classification model Methods 0.000 claims abstract description 56
- 238000000605 extraction Methods 0.000 claims abstract description 51
- 238000003860 storage Methods 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims description 62
- 238000003066 decision tree Methods 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 16
- 238000012216 screening Methods 0.000 claims description 11
- 238000001514 detection method Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 description 40
- 238000010586 diagram Methods 0.000 description 10
- 238000012545 processing Methods 0.000 description 8
- 238000004422 calculation algorithm Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 239000000203 mixture Substances 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000005538 encapsulation Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000007115 recruitment Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 238000005194 fractionation Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/70—Software maintenance or management
- G06F8/72—Code refactoring
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a code reconstructing method, device, computer equipment and storage mediums, first by obtaining code data to be sorted, treat classification code data and carry out feature extraction, obtain the sample characteristics to be sorted of code data to be sorted；Then the sample characteristics to be sorted of code data to be sorted are input in preset code classification model, obtain the object code reconstruct type of code data to be sorted；Type is reconstructed further according to the object code of code data to be sorted, obtains corresponding object reconstruction script；Classification code data are finally treated using object reconstruction script and carry out code refactoring；The efficiency for not only increasing code refactoring also achieves the automatic classification and reconstruct of the code data to be sorted to different code reconstruct type.
Description
Technical field
The present invention relates to intelligent decision field more particularly to a kind of code refactoring method, apparatus, computer equipment and storage
Medium.
Background technique
With the fast development of computer technology, the ease for use and response speed of software product have been increasingly becoming influence user
One of an important factor for usage experience.And the ease for use and response speed of software product be in its product code quality it is close
Cut it is relevant, if the product bad for some code architectures wants the current technology development of response and the update of software product is changed
Generation, just need to code structure to its inside repeatedly modified, as modification number increases, the code modified can undoubtedly be produced
Raw some mistakes, gradually decrease, therefore just need that code is reconstructed so as to cause the operational efficiency of product.Weight is treated at present
The method that structure code is reconstructed is mainly by some senior engineers or architect, first by reading source code manually
Mode go discovery to need the code that reconstructs, multiple manual reconstruct is then carried out to the code that need to be reconstructed again.However works as and face
It is a variety of different types of to reconfiguration code, engineer or frame are perhaps generally required when the quantity of reconfiguration code is huger
Structure teacher takes a substantial amount of time and energy, goes to treat reconfiguration code according to different code refactoring methods and is reconstructed, this is not only
The difficulty for increasing code refactoring is also greatly reduced the efficiency of code refactoring.
Summary of the invention
The embodiment of the present invention provides a kind of code refactoring method, apparatus, computer equipment and storage medium, to solve code
The inefficient problem of reconstruct.
A kind of code refactoring method, comprising:
Code data to be sorted is obtained, feature extraction is carried out to the code data to be sorted, obtains the generation to be sorted
The sample characteristics to be sorted of code data；
The sample characteristics to be sorted of the code data to be sorted are input in preset code classification model, institute is obtained
State the object code reconstruct type of code data to be sorted；
Type is reconstructed according to the object code of the code data to be sorted, obtains corresponding object reconstruction script；
Code refactoring is carried out to the code data to be sorted using the object reconstruction script.
A kind of code refactoring device, comprising:
Sample characteristics extraction module to be sorted, for obtaining code data to be sorted, to the code data to be sorted into
Row feature extraction obtains the sample characteristics to be sorted of the code data to be sorted；
Input module, for the sample characteristics to be sorted of the code data to be sorted to be input to preset code classification
In model, the object code reconstruct type of the code data to be sorted is obtained；
Object reconstruction script obtains module, for reconstructing class according to the object code of the code data to be sorted
Type obtains corresponding object reconstruction script；
Code refactoring module, for carrying out code weight to the code data to be sorted using the object reconstruction script
Structure.
A kind of computer equipment, including memory, processor and storage are in the memory and can be in the processing
The computer program run on device, the processor realize above-mentioned code refactoring method when executing the computer program.
A kind of computer readable storage medium, the computer-readable recording medium storage have computer program, the meter
Calculation machine program realizes above-mentioned code refactoring method when being executed by processor.
In above-mentioned code refactoring method, apparatus, computer equipment and storage medium, by obtaining code data to be sorted,
It treats classification code data and carries out feature extraction, obtain the sample characteristics to be sorted of code data to be sorted；It then will be to be sorted
The sample characteristics to be sorted of code data are input in preset code classification model, obtain the target generation of code data to be sorted
Code reconstruct type；Type is reconstructed further according to the object code of code data to be sorted, obtains corresponding object reconstruction script；Finally
Classification code data are treated using object reconstruction script and carry out code refactoring；The efficiency for not only increasing code refactoring, is also realized
It automatic classification to the code data to be sorted of different code reconstruct type and reconstructs.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below by institute in the description to the embodiment of the present invention
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the invention
Example, for those of ordinary skill in the art, without any creative labor, can also be according to these attached drawings
Obtain other attached drawings.
Fig. 1 is an application environment schematic diagram of code refactoring method in one embodiment of the invention；
Fig. 2 is an exemplary diagram of code refactoring method in one embodiment of the invention；
Fig. 3 is another exemplary diagram of code refactoring method in one embodiment of the invention；
Fig. 4 is another exemplary diagram of code refactoring method in one embodiment of the invention；
Fig. 5 is another exemplary diagram of code refactoring method in one embodiment of the invention；
Fig. 6 is another exemplary diagram of code refactoring method in one embodiment of the invention；
Fig. 7 is a functional block diagram of code refactoring device in one embodiment of the invention；
Fig. 8 is another functional block diagram of code refactoring device in one embodiment of the invention；
Fig. 9 is another functional block diagram of code refactoring device in one embodiment of the invention；
Figure 10 is a schematic diagram of computer equipment in one embodiment of the invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are some of the embodiments of the present invention, instead of all the embodiments.Based on this hair
Embodiment in bright, every other implementation obtained by those of ordinary skill in the art without making creative efforts
Example, shall fall within the protection scope of the present invention.
Code refactoring method provided in an embodiment of the present invention, the code refactoring method can apply ring using as shown in Figure 1
In border.Specifically, which applies in code refactoring system, which includes as shown in Figure 1
Client and server-side, client are communicated with server-side by network, for solving the problems, such as the inefficient of code refactoring.
Wherein, client is also known as user terminal, refers to corresponding with server-side, provides the program of local service for client.Client can
It is mounted on but is not limited to various personal computers, laptop, smart phone, tablet computer and portable wearable device
On.Server-side can be realized with the server cluster of the either multiple server compositions of independent server.
In one embodiment, as shown in Fig. 2, providing a kind of code refactoring method, the service in Fig. 1 is applied in this way
It is illustrated, includes the following steps: for end
S10: obtaining code data to be sorted, treats classification code data and carries out feature extraction, obtains code number to be sorted
According to sample characteristics to be sorted.
Wherein, code data to be sorted refers to the data of the reconstruct classification of type of pending code.Optionally, the generation to be sorted
Code data can be the code data directly acquired from the code database of server-side；It is also possible in the code text of client
The code data pre-saved, or the local code data for directly uploading or being sent to client.Client waits for this point
Category code data are sent to server-side, and server-side gets code data to be sorted.
Wherein, sample characteristics to be sorted refer to one group of characteristic for reflecting code data to be sorted in structure.It is to be sorted
The sample characteristics to be sorted of code data may include: lines of code, code redundancies, member variable, function numbers and function
At least one of in parameter.Specifically, treat the progress feature extraction of classification code data can be by obtaining from the database of server-side
Feature extraction script compiled in advance is taken, classification code data are then treated using corresponding feature extraction script and carry out feature
It extracts.
Preferably, classification code data progress feature can also be treated by using code instrumentation SourceMonitor to mention
It takes.Wherein, SourceMonitor tool be it is a can be to the generation that multilingual (C++, C#, VB, net, Java and HTML) writes
The tool that code is measured, and for different language, the code metric value exported is different.It specifically, as long as will be wait divide
Category code data are input in code instrumentation SourceMonitor, and code instrumentation SourceMonitor can be exported directly wait divide
Sample characteristics to be sorted corresponding to category code data obtain the sample characteristics to be sorted of code data to be sorted.Specifically, generation
Code line number can pass through<Lines>tag extraction in code instrumentation SourceMonitor；Code redundancies can pass through code instrumentation
<Repetive rate>tag extraction in SourceMonitor；Member variable can pass through code instrumentation SourceMonitor
In<Variable>tag extraction；Similarly, other sample characteristics to be sorted of code data to be sorted can pass through generation respectively
Other labels in code tool SourceMonitor extract.
S20: the sample characteristics to be sorted of code data to be sorted are input in preset code classification model, obtain to
The object code of classification code data reconstructs type.
Wherein, code classification model is also referred to as Decision-Tree Classifier Model, refers to the characteristic of division based on decision tree, can treat point
The model that category code data are classified automatically.Specifically, decision tree refer to it is known it is various happen probability on the basis of, lead to
It crosses and constitutes decision tree to seek the probability that the desired value of net present value (NPV) is more than or equal to zero, it belongs to a kind of tree, wherein each
Internal node indicates the test on an attribute, and each branch represents a test output, and each leaf node represents a kind of classification.
Wherein, decision tree belongs to a kind of supervision study.Supervision study refer to by give a pile sample, each sample have one group of attribute and
One classification, and these classifications be it is pre-determined, then through overfitting emerging object can be provided correctly point
A kind of method of class.
Classified automatically since code classification model can treat classification code data.Therefore, as long as by code to be sorted
The sample characteristics to be sorted of data are input in the code classification model, can directly export the target generation of code data to be sorted
Code reconstruct type obtains the object code reconstruct type of the code data to be sorted.Wherein, code refactoring type fingering line code
Reconstruct classification of type after resulting typonym, object code reconstruct type refers to that code data to be sorted carries out the reconstruct of code
Resulting typonym after classification of type.In the present embodiment, the sample characteristics to be sorted of code data to be sorted are input to
After being trained in code classification model, the object code of obtained code data to be sorted reconstruct type can be with are as follows: too long letter
The excessive type of several classes of types, class, function parameter cross polymorphic type, duplicated code type, the direct access type of class members's variable or condition
Any one in expression formula repeat type.Specifically, too long type function refers to the generation that the line number of function in code is more than threshold value
Code；The excessive type of class refers to the code that function member variable and function numbers in code are more than threshold value；Function parameter is crossed polymorphic type and is referred to
Function parameter is more than the code of threshold value in code: duplicated code type refers to the code that duplicated code block in code is more than threshold value；Item
Part expression formula repeat type refers to that in condition discriminant function, the code block there are repetition judgement is more than the code of threshold value；Class members
Variables access type refers to through the function of class the code for accessing member variable.
S30: type is reconstructed according to the object code of code data to be sorted, obtains corresponding object reconstruction script.
Wherein, object reconstruction script refer in advance it is compiled, can be to the code number to be sorted of different target code refactoring type
According to the script for carrying out code refactoring.It is to be appreciated that if the object code reconstruct type for obtaining code data to be sorted is different,
The object reconstruction script of corresponding acquisition is also different.In this step, the database of server-side is previously stored with several different generations
Reconstruct script corresponding to code reconstruct type.
Specifically, type is reconstructed according to the object code of code data to be sorted, obtains corresponding object reconstruction script, it can
To include: the object code reconstruct type for obtaining code data to be sorted, then by canonical matching method, by the object code weight
Structure type is matched one by one with the script of each reconstruct script in database mark；Finally class will be reconstructed with the object code
The reconstruct script of type successful match, is determined as object reconstruction script, wherein script mark refers to the one kind for distinguishing different reconstruct scripts
Mark.Such as: the script of object reconstruction script corresponding to too long type function is identified as too long function；Duplicated code type institute
The script of corresponding object reconstruction script is identified as duplicated code；The script mark of object reconstruction script corresponding to the excessive type of class
Know is that class is excessive.
S40: classification code data are treated using object reconstruction script and carry out code refactoring.
Since object reconstruction script has the function of that treating classification code data carries out code refactoring, in step S30
After getting corresponding object reconstruction script, the object reconstruction script can be directlyed adopt and treat classification code data progress code weight
Structure.Optionally, treat classification code data carry out code refactoring can be with are as follows: first treat classification code data split to obtain it is more
Then a subfunction is packaged the subfunction after fractionation to obtain encapsulation subfunction；Finally by the multiple sub- letters of encapsulation of calling
Several modes realizes code refactoring.
The present embodiment treats classification code data and carries out feature extraction, obtain wait divide by obtaining code data to be sorted
The sample characteristics to be sorted of category code data；Then the sample characteristics to be sorted of code data to be sorted are input to preset generation
In code disaggregated model, the object code reconstruct type of code data to be sorted is obtained；Finally according to the mesh of code data to be sorted
Code refactoring type is marked, obtains corresponding object reconstruction script, and the progress of classification code data is treated using object reconstruction script
Code refactoring；To realize automatic classification and reconstruct to the code data to be sorted of different target code refactoring type.
In one embodiment, as shown in figure 3, it is preset being input to the sample characteristics to be sorted of code data to be sorted
In code classification model, before obtaining the object code reconstruct type of code data to be sorted, code refactoring method further include:
S201: obtaining N number of sample code data, and each sample code data include corresponding reconstruction of the original code type, N
For positive integer.
Wherein, sample code data refer to the sample data that code architecture is not good enough, need to be reconstructed, and N is positive integer.Per the same
This code data includes corresponding reconstruction of the original code type, and reconstruction of the original code type refers to that sample code data carry out code
Resulting typonym after reconstruct classification of type.I.e. each sample code data are all to have been subjected to processing classification and difference in advance
The code data of corresponding reconstruction of the original code type on mark.Specifically, the reconstruction of the original code type of sample code data
It may include: too long type function, the excessive type of class, that function parameter crosses polymorphic type, duplicated code type, class members's variable is direct
At least one of in access type and conditional expression repeat type.
Optionally, sample code data can be and get source code from open source system in advance, then to the source of acquisition
Code is resulting after being detected, identified and being classified.Wherein, open source system can for Tomcat system, ArgoUML system or
Apache Ant system etc..Be also possible to pre-save in the code text of client has been subjected to the code number of processing classification
According to；The either local code data for having been subjected to processing classification for directly uploading or being sent to client.Client is by the sample
Code data is sent to server-side, and server-side can get sample code data.
S202: carrying out feature extraction to each sample code data, and the training sample for obtaining each sample code data is special
Sign.
Wherein, training sample feature refers to one group characteristic of the reflected sample code data in structure.Sample code number
According to training sample feature may include: in lines of code, code redundancies, member variable, function numbers and function parameter
At least one of.Specifically, feature extraction is carried out to each sample code data, obtains the training sample of each sample code data
The method and detailed process of feature treat classification code data with above-mentioned steps S10 and carry out feature extraction, obtain code to be sorted
The method of the sample characteristics to be sorted of data is identical with detailed process, does not do redundancy herein and repeats.
S203: according to the training sample feature of each sample code data and corresponding reconstruction of the original code type to decision
Tree-model is trained, and obtains code classification model.
Wherein, code classification model refers to the training sample feature by a large amount of sample code data and corresponding original generation
Code reconstruct type is trained rear generated, can treat the model that classification code is classified.Specifically, by each sample generation
Code data training sample feature and corresponding reconstruction of the original code type be input in decision-tree model, then by using
C4.5 algorithm is trained decision-tree model, that is, produces the code classification model after training.Wherein, C4.5 algorithm is a system
Column are used in the algorithm in the classification problem of machine learning and data mining.The target of C4.5 algorithm is supervised learning, gives one
Data set, wherein each tuple can be described with one group of attribute value, each tuple belongs in the classification of a mutual exclusion
Certain is a kind of.C4.5 algorithm can find a dependence value to the mapping relations of classification, and this mapping can be used for by study
The entity unknown to new classification is classified.
For the present embodiment by obtaining N number of sample code data, each sample code data include corresponding source code weight
Structure type, N are positive integer；Then feature extraction is carried out to each sample code data, obtains the instruction of each sample code data
Practice sample characteristics；Finally fought to the finish according to the training sample feature of each sample code data and corresponding reconstruction of the original code type
Plan tree-model is trained, and obtains code classification model；It ensure that the accuracy by the arrived code classification model of training.
In one embodiment, as shown in figure 4, according to the training sample feature of each sample code data and corresponding original
Code refactoring type is trained decision-tree model, obtains code classification model, comprising:
S2031: the training sample feature of each sample code data and corresponding reconstruction of the original code type are formed into sample
Eigen collection, and sample characteristics collection is divided into sample training collection and sample verifying collection.
Wherein, sample characteristics collection refers to the sample data for training decision-tree model, and sample characteristics collection includes each sample
The training sample feature of code data and corresponding reconstruction of the original code type.Specifically, sample characteristics collection is by several samples
Eigen composition data set, each sample characteristics all include the sample code data training sample feature and with the sample generation
The reconstruction of the original code type of code data.It is to be appreciated that the training sample feature and corresponding original of each sample code data
Beginning code refactoring type is associated.
Sample training collection refers to the data set for establishing code classification model.Sample verifying collection refers to after verifying foundation
The data set of the effect of code classification model.Specifically, sample characteristics collection is divided into sample training collection and sample verifying collection can
It is divided using random division or the method for crosscheck；The ratio value of sample training collection and sample verifying collection can be with after division
Are as follows: sample training collection: sample verifies collection=6:4, sample training collection: sample verifies collection=7:3 or sample training collection: sample is tested
Demonstrate,prove collection=7.5:2.5 etc..Preferably, in order to improve the precision of code classification model, in this step, the sample characteristics that will acquire
The 75% of collection is used as sample training collection, for being trained to model；It is tested using the 25% of the sample characteristics collection of acquisition as sample
Card collection, for carrying out recruitment evaluation to the model for completing training.
S2032: sample training collection being input in decision-tree model and is trained, and obtains preliminary classification model.
Specifically, it is made of due to sample training collection training sample feature and corresponding reconstruction of the original code type,
Therefore the training sample feature of sample training concentration and corresponding reconstruction of the original code type need to be input in decision-tree model,
And the decision-tree model is trained by using C4.5 algorithm, preliminary classification model can be obtained.Optionally, training sample
Feature may include: at least one in lines of code, code redundancies, member variable, function numbers and function parameter.It is original
Code refactoring type includes: that too long type function, the excessive type of class, function parameter cross polymorphic type, duplicated code type, class members
At least one of in the direct access type of variable and conditional expression repeat type result.
S2033: preliminary classification model is verified using sample verifying collection, is verified accuracy rate.
Wherein, verifying accuracy rate refers to using sample verifying collection to institute after the preliminary classification model progress recruitment evaluation after training
The result obtained.Specifically, carrying out verifying to preliminary classification model using sample verifying collection includes: the instruction for first concentrating sample verifying
Practice sample characteristics to be input in the preliminary classification model, then the reconstruction of the original code type after being trained is tested the sample
Reconstruction of the original code type after card collection training is matched one by one with corresponding preset reconstruction of the original code type, last basis
Matching result is verified accuracy rate.
Illustratively, if it includes 10 sample code data that the sample verifying got, which is concentrated, by 10 sample codes
After the sample training feature of data is input to preliminary classification model, have 7 in the reconstruction of the original code type after obtaining 10 training
Reconstruction of the original code type after a training matches with corresponding preset reconstruction of the original code type, original after 3 training
Code refactoring type does not match that it is 0.7 that the verifying accuracy rate, which is then calculated, with corresponding preset reconstruction of the original code type.
S2034: if verifying accuracy rate is greater than or equal to default accuracy rate, preliminary classification model is determined as code classification
Model.
Wherein, default accuracy rate refers in advance to resulting result after preliminary classification model progress initial assessment.Specifically, will
The verifying accuracy rate according to obtained in step S233 is compared with default accuracy rate, if the verifying accuracy rate is greater than or equal in advance
If accuracy rate, then corresponding preliminary classification model is determined as code classification model；If the verifying accuracy rate is less than default accurate
Rate then needs to optimize processing to the preliminary classification model.Specifically, optimizing processing to preliminary classification model can be by making
It is realized with minimum optimization loss function.A loss function is determined first, is then iterated to the loss function, it is minimum
Change the loss function, until convergence realizes the optimization processing to preliminary classification model to obtain minimum optimization loss function.
The present embodiment is by by the training sample feature of each sample code data and corresponding reconstruction of the original code type
Sample characteristics collection is formed, and sample characteristics collection is divided into sample training collection and sample verifying collection；Then sample training collection is defeated
Enter and be trained into decision-tree model, obtains preliminary classification model, and using sample verifying collection to affiliated preliminary classification model
It is verified, is verified accuracy rate；If affiliated verifying accuracy rate is greater than or equal to default accuracy rate, by preliminary classification model
It is determined as code classification model；Further ensure the accuracy of the code classification model of foundation.
In one embodiment, as shown in figure 5, treating classification code data carries out feature extraction, code number to be sorted is obtained
According to sample characteristics to be sorted, specifically include:
S101: characteristic parameter collection is obtained, characteristic parameter collection includes M parameter identification, and M is positive integer.
Wherein, characteristic parameter collection refers to the characteristic set for presetting the code data to be sorted that need to be extracted.Feature ginseng
Manifold may include: at least one in lines of code, code redundancies, member variable, function numbers and function parameter.Specifically
Ground, characteristic parameter collection include M parameter identification, and M is positive integer.Parameter identification refers to the one kind assigned for each characteristic parameter
Identification number.Such as: the parameter identification of lines of code can be Lines；The parameter identification of code redundancies can be Repetive
rate；The parameter identification of member variable can be Variable；The parameter identification of function numbers can be Number；Function parameter
Parameter identification can be Parameter.
S102: corresponding feature extraction script is obtained according to each parameter identification.
Wherein, feature extraction script, which refers to, can directly treat the text that classification code data carry out feature extraction.In this implementation
In example, feature extraction script is database that is compiled in advance and being stored in server-side, therefore according to the parameter identification of acquisition
Can corresponding feature extraction script directly be obtained from the database of server-side.Such as: according to the parameter identification Lines of lines of code
Corresponding feature extraction script<Lines>can be got from the database of server-side；According to the parameter identification of code redundancies
Repetive rate can get corresponding feature extraction script<Repetive rate>from the database of server-side；According at
The parameter identification Variable of member's variable can get corresponding feature extraction script<Variable>from the database of server-side；
Corresponding feature extraction script < Number can be got from the database of server-side according to the parameter identification Number of function numbers
>；Corresponding feature extraction script can be got from the database of server-side according to the parameter identification Parameter of function parameter
<Parameter>。
S103: classification code data are treated using feature extraction script and carry out feature extraction, obtain code data to be sorted
Sample characteristics to be sorted.
Wherein, sample characteristics to be sorted refer to one group of characteristic for reflecting code data to be sorted in structure.Specifically,
The sample characteristics to be sorted of code data to be sorted may include: lines of code, code redundancies, member variable, function numbers
And function parameter.Specifically, the direct feature extraction of classification code data progress is treated since each feature extraction script all has
Function, therefore can directly by the feature extraction script obtained according to step S102 treat classification code data carry out feature mention
It takes, obtains the sample characteristics to be sorted of code data to be sorted.
For the present embodiment by obtaining characteristic parameter collection, characteristic parameter collection includes M parameter identification；Then according to each parameter
Mark obtains corresponding feature extraction script, finally treats classification code data using feature extraction script and carries out feature extraction,
Obtain the sample characteristics to be sorted of code data to be sorted；Further ensure the sample to be sorted of the code data to be sorted of acquisition
The accuracy and validity of eigen.
In one embodiment, as shown in fig. 6, before obtaining N number of sample code data, code refactoring method further include:
S2011: L source code data are obtained, L is positive integer.
Wherein, source code data refer to the code data without any processing.Optionally, L source code data can be with
It is the source code obtained from a certain code system, or from open source system such as Tomcat, ArgoUML or Apache Ant
The source code of acquisition, L are positive integer.In this step, L source code data of acquisition may be according to different programming languages
What speech was compiled；Such as: source code data can be by java language compilation's or be by c++ language compilation,
It can also be through python language compilation etc..
S2012: code detection is carried out to each source code data, obtains the programming language class of each source code data
Type.
Wherein, programming language type refers to a kind of abstract concept for describing the compiler language of different source code data.It is optional
Ground, the programming language type of source code data may include: java language, c++ language, python language and SSH language etc..
Specifically, the automatic inspection of code can be realized by the detection instrument of program language by carrying out code detection to each source code data
It surveys, as long as each source code data are input in the detection instrument of program language, each source code number can be obtained
According to corresponding programming language type.
Preferably, code detection is being carried out to each source code data, is obtaining the programming language of each source code data
Say type after, can also the programming language type according to corresponding to each source code data, assign each source code data
Corresponding type identification.Wherein, type identification refers to for distinguishing programming language type corresponding to different source code data not
A kind of same mark.Type identification can be indicated with any one in number, capitalization or big minuscule.Such as: it assigns
The corresponding type identification of source code data compiled with java language is a；Assign the source code compiled with c++ language
The corresponding type identification of data is b；Assigning the corresponding type identification of source code data compiled with python language is c；
Assigning the corresponding type identification of source code data compiled with SSH language is d.
S2013: according to preset screening sample information, each source code data are sieved according to programming language type
Choosing, obtains sample code data.
Wherein, screening sample information refers to a kind of regular or requirement for being screened to source code data.Specifically,
The programming language type as corresponding to different source code data may be different, according to preset screening sample information,
Screening according to programming language type to each source code data can be by canonical matching method, by each source code data
Corresponding programming language type is matched with preset screening sample information, finally by the source code data of successful match
As sample code data.
Illustratively, if preset screening sample information is to filter out the code number that programming language type is java language
According to then by canonical matching method, by programming language type corresponding to each source code data and preset screening sample letter
After breath is matched, resulting sample code data are all the code datas due to java language compilation.
For the present embodiment by obtaining L source code data, L is positive integer；Then each source code data are carried out
Code detection obtains the programming language type of each source code data；Finally according to preset screening sample information, according to volume
Journey language form screens each source code data, obtains sample code data；It not only ensure that sample code data
Accuracy, further improve the accuracy of code classification model.
It should be understood that the size of the serial number of each step is not meant that the order of the execution order in above-described embodiment, each process
Execution sequence should be determined by its function and internal logic, the implementation process without coping with the embodiment of the present invention constitutes any limit
It is fixed.
In one embodiment, a kind of device is provided, code refactoring method one in the code refactoring device and above-described embodiment
One is corresponding.As shown in fig. 7, the code refactoring device includes sample characteristics extraction module 10 to be sorted, input module 20, target weight
Structure script obtains module 30 and code refactoring module 40.Detailed description are as follows for each functional module:
Sample characteristics extraction module 10 to be sorted treats the progress of classification code data for obtaining code data to be sorted
Feature extraction obtains the sample characteristics to be sorted of code data to be sorted；
Input module 20, for the sample characteristics to be sorted of code data to be sorted to be input to preset code classification mould
In type, the object code reconstruct type of code data to be sorted is obtained；
Object reconstruction script obtains module 30, for reconstructing type according to the object code of code data to be sorted, obtains
Corresponding object reconstruction script；
Code refactoring module 40 carries out code refactoring for treating classification code data using object reconstruction script.
Preferably, as shown in figure 8, code refactoring device further include:
Sample code data acquisition module 201, for obtaining N number of sample code data, each sample code data include
Corresponding reconstruction of the original code type, N are positive integer；
Training sample characteristic extracting module 202 is obtained for carrying out feature extraction to each sample code data per the same
The training sample feature of this code data；
Model training module 203, for according to each sample code data training sample feature and corresponding original generation
Code reconstruct type is trained decision-tree model, obtains code classification model.
Preferably, as shown in figure 9, model training module 203 includes:
Component units 2031, for by the training sample feature of each sample code data and corresponding reconstruction of the original code
Type forms sample characteristics collection, and sample characteristics collection is divided into sample training collection and sample verifying collection；
Input unit 2032, is trained for sample training collection to be input in decision-tree model, obtains preliminary classification
Model；
Authentication unit 2033 is verified accuracy rate for verifying using sample verifying collection to preliminary classification model；
Determination unit 2034 is used for when verifying accuracy rate more than or equal to default accuracy rate, and preliminary classification model is true
It is set to code classification model.
Preferably, characteristic extracting module 10 includes:
Characteristic parameter collection acquiring unit, for obtaining characteristic parameter collection, characteristic parameter collection includes M parameter identification, and M is positive
Integer；
Feature extraction script acquiring unit, for obtaining corresponding feature extraction script according to each parameter identification；
Sample characteristics extraction unit to be sorted is mentioned for treating classification code data progress feature using feature extraction script
It takes, obtains the sample characteristics to be sorted of code data to be sorted.
Preferably, code refactoring device further include:
Source code data acquisition module, for obtaining L source code data, L is positive integer；
Code detection module obtains each source code data for carrying out code detection to each source code data
Programming language type；
Screening module is used for according to preset screening sample information, according to programming language type to each source code number
According to being screened, sample code data are obtained.
Specific about code refactoring device limits the restriction that may refer to above for code refactoring method, herein not
It repeats again.Modules in above-mentioned code refactoring device can be realized fully or partially through software, hardware and combinations thereof.On
Stating each module can be embedded in the form of hardware or independently of in the processor in computer equipment, can also store in a software form
In memory in computer equipment, the corresponding operation of the above modules is executed in order to which processor calls.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 10.The computer equipment include by system bus connect processor, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program and data
Library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The database of machine equipment is used to store the data used in the code refactoring method in above-described embodiment.The computer equipment
Network interface is used to communicate with external terminal by network connection.To realize one kind when the computer program is executed by processor
Code refactoring method.
In one embodiment, a kind of computer equipment is provided, including memory, processor and storage are on a memory
And the computer program that can be run on a processor, processor realize code refactoring in above-described embodiment when executing computer program
Method.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program realizes code refactoring method in above-described embodiment when being executed by processor.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
It is apparent to those skilled in the art that for convenience of description and succinctly, only with above-mentioned each function
Can unit, module division progress for example, in practical application, can according to need and by above-mentioned function distribution by different
Functional unit, module are completed, i.e., the internal structure of described device is divided into different functional unit or module, more than completing
The all or part of function of description.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations；Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features；And these are modified
Or replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution should all
It is included within protection scope of the present invention.
Claims (10)
1. a kind of code refactoring method characterized by comprising
Code data to be sorted is obtained, feature extraction is carried out to the code data to be sorted, obtains the code number to be sorted
According to sample characteristics to be sorted；
The sample characteristics to be sorted of the code data to be sorted are input in preset code classification model, obtain it is described to
The object code of classification code data reconstructs type；
Type is reconstructed according to the object code of the code data to be sorted, obtains corresponding object reconstruction script；
Code refactoring is carried out to the code data to be sorted using the object reconstruction script.
2. code refactoring method as described in claim 1, which is characterized in that it is described by the code data to be sorted to
Classification samples feature is input in preset code classification model, obtains the object code reconstruct class of the code data to be sorted
Before type, the code refactoring method further include:
N number of sample code data are obtained, each sample code data include corresponding reconstruction of the original code type, and N is positive
Integer；
Feature extraction is carried out to each sample code data, the training sample for obtaining each sample code data is special
Sign；
According to the training sample feature of each sample code data and the corresponding reconstruction of the original code type to decision
Tree-model is trained, and obtains code classification model.
3. code refactoring method as claimed in claim 2, which is characterized in that described according to each sample code data
Training sample feature and the corresponding reconstruction of the original code type are trained decision-tree model, obtain code classification mould
Type, comprising:
The training sample feature of each sample code data and the corresponding reconstruction of the original code type are formed into sample
Feature set, and the sample characteristics collection is divided into sample training collection and sample verifying collection；
The sample training collection is input in decision-tree model and is trained, preliminary classification model is obtained；
The preliminary classification model is verified using sample verifying collection, is verified accuracy rate；
If the verifying accuracy rate is greater than or equal to default accuracy rate, the preliminary classification model is determined as code classification mould
Type.
4. code refactoring method as described in claim 1, which is characterized in that described to carry out spy to the code data to be sorted
Sign is extracted, and the sample characteristics to be sorted of the code data to be sorted are obtained, comprising:
Characteristic parameter collection is obtained, the characteristic parameter collection includes M parameter identification, and M is positive integer；
Corresponding feature extraction script is obtained according to each parameter identification；
Feature extraction is carried out to the code data to be sorted using the feature extraction script, obtains the code number to be sorted
According to sample characteristics to be sorted.
5. code refactoring method as claimed in claim 2, which is characterized in that before the N number of sample code data of acquisition,
The code refactoring method further include:
L source code data are obtained, L is positive integer；
Code detection is carried out to each source code data, obtains the programming language class of each source code data
Type；
According to preset screening sample information, each source code data are sieved according to the programming language type
Choosing, obtains sample code data.
6. a kind of code refactoring device characterized by comprising
Sample characteristics extraction module to be sorted carries out the code data to be sorted special for obtaining code data to be sorted
Sign is extracted, and the sample characteristics to be sorted of the code data to be sorted are obtained；
Input module, for the sample characteristics to be sorted of the code data to be sorted to be input to preset code classification model
In, obtain the object code reconstruct type of the code data to be sorted；
Object reconstruction script obtains module, for reconstructing type according to the object code of the code data to be sorted, obtains
Take corresponding object reconstruction script；
Code refactoring module, for carrying out code refactoring to the code data to be sorted using the object reconstruction script.
7. code refactoring device as claimed in claim 6, which is characterized in that the code refactoring device further include:
Sample code data acquisition module, for obtaining N number of sample code data, each sample code data include corresponding to
Reconstruction of the original code type, N is positive integer；
Training sample characteristic extracting module obtains each described for carrying out feature extraction to each sample code data
The training sample feature of sample code data；
Model training module, for according to each sample code data training sample feature and the corresponding original generation
Code reconstruct type is trained decision-tree model, obtains code classification model.
8. code refactoring device as claimed in claim 7, which is characterized in that the model training module includes:
Component units, for by the training sample feature of each sample code data and the corresponding reconstruction of the original code
Type forms sample characteristics collection, and the sample characteristics collection is divided into sample training collection and sample verifying collection；
Input unit is trained for the sample training collection to be input in decision-tree model, obtains preliminary classification model；
Authentication unit is verified accuracy rate for verifying using sample verifying collection to the preliminary classification model；
Determination unit is used for when the verifying accuracy rate is greater than or equal to default accuracy rate, and the preliminary classification model is true
It is set to code classification model.
9. a kind of computer equipment, including memory, processor and storage are in the memory and can be in the processor
The computer program of upper operation, which is characterized in that the processor realized when executing the computer program as claim 1 to
Any one of 5 code refactoring methods.
10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In realization code refactoring method as described in any one of claim 1 to 5 when the computer program is executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910345048.9A CN110162335A (en) | 2019-04-26 | 2019-04-26 | Code refactoring method, apparatus, computer equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910345048.9A CN110162335A (en) | 2019-04-26 | 2019-04-26 | Code refactoring method, apparatus, computer equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110162335A true CN110162335A (en) | 2019-08-23 |
Family
ID=67640146
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910345048.9A Pending CN110162335A (en) | 2019-04-26 | 2019-04-26 | Code refactoring method, apparatus, computer equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110162335A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111625275A (en) * | 2020-04-29 | 2020-09-04 | 贝壳技术有限公司 | Code reconstruction planning method and device, storage medium and electronic equipment |
| CN112947993A (en) * | 2019-12-31 | 2021-06-11 | 深圳市明源云链互联网科技有限公司 | Method and device for reconstructing system framework, electronic equipment and storage medium |
| CN113238796A (en) * | 2021-05-17 | 2021-08-10 | 北京京东振世信息技术有限公司 | Code reconstruction method, device, equipment and storage medium |
| CN114327415A (en) * | 2022-03-17 | 2022-04-12 | 武汉天喻信息产业股份有限公司 | Compiling method and device for compiling java file |
| WO2022121724A1 (en) * | 2020-12-07 | 2022-06-16 | 华为技术有限公司 | Data processing apparatus and method |
| WO2023155487A1 (en) * | 2022-02-18 | 2023-08-24 | 华为云计算技术有限公司 | Code refactoring method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107220180A (en) * | 2017-06-08 | 2017-09-29 | 电子科技大学 | A kind of code classification method based on neutral net language model |
| CN108446115A (en) * | 2018-03-12 | 2018-08-24 | 中国银行股份有限公司 | A kind of method and device of code reuse |
| CN108897572A (en) * | 2018-07-19 | 2018-11-27 | 北京理工大学 | A kind of complicated type reconstructing method based on variable association tree |
-
2019
- 2019-04-26 CN CN201910345048.9A patent/CN110162335A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107220180A (en) * | 2017-06-08 | 2017-09-29 | 电子科技大学 | A kind of code classification method based on neutral net language model |
| CN108446115A (en) * | 2018-03-12 | 2018-08-24 | 中国银行股份有限公司 | A kind of method and device of code reuse |
| CN108897572A (en) * | 2018-07-19 | 2018-11-27 | 北京理工大学 | A kind of complicated type reconstructing method based on variable association tree |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112947993A (en) * | 2019-12-31 | 2021-06-11 | 深圳市明源云链互联网科技有限公司 | Method and device for reconstructing system framework, electronic equipment and storage medium |
| CN112947993B (en) * | 2019-12-31 | 2021-12-07 | 深圳市明源云链互联网科技有限公司 | Method and device for reconstructing system framework, electronic equipment and storage medium |
| CN111625275A (en) * | 2020-04-29 | 2020-09-04 | 贝壳技术有限公司 | Code reconstruction planning method and device, storage medium and electronic equipment |
| CN111625275B (en) * | 2020-04-29 | 2023-10-20 | 贝壳技术有限公司 | Code reconstruction planning method and device, storage medium and electronic equipment |
| WO2022121724A1 (en) * | 2020-12-07 | 2022-06-16 | 华为技术有限公司 | Data processing apparatus and method |
| CN113238796A (en) * | 2021-05-17 | 2021-08-10 | 北京京东振世信息技术有限公司 | Code reconstruction method, device, equipment and storage medium |
| WO2023155487A1 (en) * | 2022-02-18 | 2023-08-24 | 华为云计算技术有限公司 | Code refactoring method and device |
| CN114327415A (en) * | 2022-03-17 | 2022-04-12 | 武汉天喻信息产业股份有限公司 | Compiling method and device for compiling java file |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110162335A (en) | Code refactoring method, apparatus, computer equipment and medium | |
| Davison et al. | PyNN: a common interface for neuronal network simulators | |
| Li et al. | Software defect prediction via convolutional neural network | |
| US11681925B2 (en) | Techniques for creating, analyzing, and modifying neural networks | |
| Rajpal et al. | Not all bytes are equal: Neural byte sieve for fuzzing | |
| CN106599922B (en) | Transfer learning method and system for large-scale data calibration | |
| CN105184160B (en) | A kind of method of the Android phone platform application program malicious act detection based on API object reference relational graphs | |
| CN109829155A (en) | Determination method, automatic scoring method, apparatus, equipment and the medium of keyword | |
| US20180373986A1 (en) | Machine learning using dynamic multilayer perceptrons | |
| US9984336B2 (en) | Classification rule sets creation and application to decision making | |
| CN110021439A (en) | Medical data classification method, device and computer equipment based on machine learning | |
| Gupta et al. | Neural attribution for semantic bug-localization in student programs | |
| CN106537333A (en) | Systems and methods for a database of software artifacts | |
| EP3332320A1 (en) | Model integration tool | |
| US20210141801A1 (en) | String Parsed Categoric Encodings for Machine Learning | |
| CN108681746A (en) | A kind of image-recognizing method, device, electronic equipment and computer-readable medium | |
| CN109753653A (en) | Entity name recognition methods, device, computer equipment and storage medium | |
| US11640539B2 (en) | Techniques for visualizing the operation of neural networks using samples of training data | |
| US20180357201A1 (en) | Ability-providing-data generation apparatus | |
| CN108255702A (en) | A kind of test case creation method, apparatus, equipment and storage medium | |
| CN109828750A (en) | Auto-configuration data buries method, apparatus, electronic equipment and storage medium a little | |
| CN110502224A (en) | Interface analogy method, device and computer equipment based on HTTP request | |
| CN108897572A (en) | A kind of complicated type reconstructing method based on variable association tree | |
| CN110162972A (en) | A kind of UAF leak detection method based on sentence combined coding deep neural network | |
| CN111210356B (en) | Medical insurance data analysis method and device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |