CN115309634A - Micro-service extraction method, system, medium, equipment and information processing terminal - Google Patents

Abstract

The invention belongs to the technical field of micro-service extraction, and discloses a micro-service extraction method, a system, a medium, equipment and an information processing terminal, which are used for carrying out hierarchical division on a source code; acquiring a time sequence diagram of each method of a control layer through reverse engineering, and acquiring class diagrams of a physical layer, a database access layer and other layers; modeling a display service function of the time sequence diagram; carrying out implicit business function modeling on the class diagram; extracting candidate micro-services of the source code based on the business function model through spectral clustering; evaluating the quality of the candidate micro-service; and the candidate micro-services are visualized in a graph structure form, and an adjusting function is provided for the framework personnel. The method uses the spectral clustering algorithm to extract the micro-services, and achieves the aims of high cohesion inside the micro-services and low coupling among the micro-services; the extraction of the micro-service after the business function modeling is carried out on the source code greatly reduces the use threshold of the method, and reduces the time cost and the labor cost of extracting the monomer architecture to the micro-service.

Description

A microservice extraction method, system, medium, device and information processing terminal

technical field

The invention belongs to the technical field of microservice extraction, and in particular relates to a microservice extraction method, system, medium, equipment and information processing terminal.

Background technique

Migrating a system with a monolithic architecture to a microservice architecture can better solve the problems faced by a system based on a monolithic architecture. The task of microservice extraction is mainly to reasonably extract classes with the same business function from the single-architecture system to form a group of candidate microservices. Compared with traditional manual extraction and identification methods, existing microservice extraction methods realize automatic or semi-automatic identification of microservices, but the key input for determining the quality of candidate microservices still depends on the generation of professionals and the use of methods The threshold is high, and the automatic extraction method is difficult. At present, the methods of automatic extraction are mainly divided into static and dynamic methods, which have the following characteristics: 1. The method of extracting microservices based on program running and execution traces depends on the design of test cases, and the test cases must be familiar with the system software designers. As a result, legacy systems often have the problems of complex business, high turnover of developers, and incomplete documentation. Therefore, the design of test cases is difficult to completely cover all business functions, and the incomplete design of test cases will lead to the coverage of the proposed microservice class. The rate is low, so the threshold for using this method is high. 2. The method of extracting microservices based on source code generally uses the similarity of code as the basis for extracting microservices, which cannot fully reflect the business relationship. The above determines the scalability of microservices, so the use of this method has great hidden dangers. The existence of the above problems has led to the difficulty of automatic extraction of microservices.

Through the above analysis, the existing problems and defects of the existing technology are: the method of extracting microservices based on the program execution trajectory depends on the design of the test case, the test case must be generated by a software designer who is familiar with the system, and the legacy system often has complex business , the problem of high turnover of developers and incomplete documentation, so it is difficult to design test cases to completely cover all business functions, and incomplete design of test cases will lead to low coverage of the proposed microservice class, so this method uses The threshold is high. The microservice extraction method based on the source code generally uses code similarity as the basis for microservice extraction, which cannot fully reflect the business relationship. The quality of the extracted microservices is poor, and the usability is poor, and the quality of the microservices will be determined to a large extent. The scalability of microservices, so the applicability of this method is not high.

Contents of the invention

Aiming at the problems existing in the prior art, the present invention provides a microservice extraction method, system, medium, equipment, and information processing terminal, and particularly relates to a microservice extraction method, system, medium, and equipment based on the source code of a single system and information processing terminals.

The present invention is realized in this way, a microservice extraction method, the microservice extraction method includes: generating a class diagram and a sequence diagram according to the source code; extracting business functions based on the class diagram and the sequence diagram; using a fast spectral clustering algorithm to The extracted business functions are extracted from microservices, and finally the quality of the generated candidate microservices is evaluated and visualized.

Divide the source code into layers; obtain the sequence diagram of each method of the control layer through reverse engineering, and obtain the class diagrams of the entity layer, database access layer and other layers; model the display business function of the sequence diagram; Business function modeling; extract candidate microservices of source code based on business function model through spectral clustering; evaluate the quality of candidate microservices; finally visualize candidate microservices in the form of graph structure, providing adjustment functions for architects.

Further, the microservice extraction method includes the following steps:

Step 1: Divide the source code into layers, obtain the sequence diagram of the control layer through reverse engineering, and obtain the class diagrams of the entity layer, database access layer and other layers; this is because irregular source code writing will affect the extraction of microservices.

Step 2: Model the display business function of the sequence diagram, and model the implicit business function of the class diagram; because the sequence diagram can reflect the dynamic relationship between classes, the analysis of the class diagram mainly considers that the domain-driven model is the current micro Service splitting is a popular idea. Through class diagram analysis, domain relationships can be mined to supplement some business relationships that cannot be reflected in sequence diagrams.

Step 3: Extract candidate microservices of the source code based on the business function model through spectral clustering, because spectral clustering regards the identification of microservices as graph clustering, and extracts microservices with high cohesion and low coupling.

Step 4: Evaluate the quality of the candidate microservices, visualize the candidate microservices in the form of a graph structure and provide adjustment functions; use the poor quality microservices to improve the legacy system, the scalability is poor, and the microservices cannot be fully reflected Taking advantage of the architecture, evaluating the quality of microservices can avoid this situation.

Further, in the first step, the source code of the single legacy system is divided into layers, and the control layer, business layer, database access layer, entity layer and other layers in the source code are identified.

According to the source code of the single system program, the sequence diagram and class diagram of the system are generated through reverse engineering:

(1) Use the SequenceDiagram tool to generate a sequence diagram for each method in the control layer of the source code, where the control layer of the source code refers to the class that interacts with the user interface display;

(2) Use the PlantUML Parser tool to generate class diagrams for the database access layer, entity layer, and other layers for each class file of the source code.

Further, in the second step, by modeling the sequence diagram to display business functions, a calling relationship mapping table between classes is obtained, specifically including:

(1) Among multiple sequence diagram files, read a sequence diagram file for analysis, and when the sequence diagram file is empty, end the display business function modeling process;

(2) Count the number of calls f _ij between two classes C _i and C _j , and add 1 to f _ij every time there is a call relationship between C _i and C _j ;

(3) The number of calls between the two classes C _i and C _j is stored in the Map structure, where the key is a character string formed by splicing the class names of the two classes with "_", and the value is the number of calls f _ij ;

(4) After all the sequence diagram files are parsed, the model Map is output.

Model the implicit business function according to the class diagram, and obtain the semantic similarity relationship matrix between classes:

(1) In multiple class diagram files, read two class diagrams for analysis, and when the class diagram file is empty, end the implicit business function modeling process;

(2) Calculate the semantic similarity S _ij between the two class graphs C _i and C _j ; make a word bag dictionary according to the input text information, and match the words in the text with the keys in the word bag to obtain the corpus; Initialize the tf-idf transformation model, obtain the converted corpus corpus_tfidf, use the Lsi model to train the corpus_tfidf corpus, and calculate the similarity of the sparse matrix; the format conversion converts the word segmentation list that needs to find similarity into a corpus doc_test_vec, and obtains the similarity of the text.

Further, in the third step, after obtaining the business function model of the source code, perform clustering using a spectral clustering algorithm to obtain candidate microservices.

(1) Construct a similarity matrix, the formula is as follows:

Among them, W _ij represents the similarity matrix, if there is a call relationship between two classes i and j, then w _ij = map _ij otherwise w _ij = S _ij ; D _ij is a diagonal matrix, and the value on the diagonal is in the W matrix The sum of the corresponding row or column;

(2) Construct the Laplacian matrix L, and normalize L, the formula is as follows:

L=D-W;

(3) Decompose the eigenvalues of the Laplacian matrix, use the Lanczos method to accelerate the decomposition process, obtain the first k smallest eigenvalues and corresponding eigenvectors, and finally form a k-dimensional eigenmatrix F;

(4) Use Kmeans to cluster the k-dimensional feature matrix F; use AFK-MC ² to select the initial cluster center, randomly select an initial center sample c ₁ ; calculate the proposal distribution q(x) of all data sets , randomly select a data point from q(x) and calculate the distance dx; use Markov chain Monte Carlo to sample a sequence of length m, and take the last k-1 as the center point C={C1,C2, ...,Ck};

Use A-means to reduce the time for K-means algorithm to assign data points to cluster C; calculate the distance from each data point _xi to all cluster centers, select the centroid C _k closest to itself, and calculate the The isometric index α _i , the formula is as follows:

α _i =abs(||i-μ ₁ || ² -||i-μ ₂ || ² );

公式如下：Calculate the improvement threshold

The formula is as follows:

时，x_i在轮迭代中不会移动，将x_i固定分配给簇C_k，不再参与距离计算和重新分配。when

When _xi does not move in round iterations, _xi is fixedly assigned to cluster C _k , and no longer participates in distance calculation and reassignment.

After spectral clustering, a set of clustering results is obtained, and each cluster is a microservice candidate.

Further, the quality assessment of the extracted candidate microservices in the step 4 includes:

Using the modular evaluation method, the higher the modular evaluation value, the lower the coupling degree between the microservices, and the higher the aggregation degree inside the microservice. The calculation formula is as follows:

Among them, N is the number of microservices, N _i is the number of classes in microservices, W _i is the weight sum of internal edges of microservice i, W _{i, j} is the edge between microservice i and microservice j weight and .

Visualize the extracted candidate microservices, display the microservices of each class file and the calling relationship between classes based on Echarts in the form of a graph structure, and drag the vertices to manually adjust the microservices to which the class belongs.

Another object of the present invention is to provide a microservice extraction system applying the microservice extraction method, the microservice extraction system comprising:

The layer division module is used to divide the source code into layers, and obtain the sequence diagram of the control layer through reverse engineering, and obtain the class diagram of the entity layer, database access layer and other layers;

The business function modeling module is used to model the display business function of the sequence diagram and model the implicit business function of the class diagram;

The candidate microservice extraction module is used to extract the candidate microservices of the source code based on the business function model through spectral clustering;

The quality evaluation module is used to evaluate the quality of candidate microservices, visualize the candidate microservices in the form of graph structure and provide adjustment functions.

Another object of the present invention is to provide a computer device, the computer device includes a memory and a processor, the memory stores a computer program, when the computer program is executed by the processor, the processor executes the The steps of the microservice extraction method described above.

Another object of the present invention is to provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the microservice extraction method.

Another object of the present invention is to provide an information data processing terminal, which is used to implement the microservice extraction system.

Combining the above-mentioned technical solutions and technical problems to be solved, please analyze the advantages and positive effects of the technical solutions to be protected by the present invention from the following aspects:

First, in view of the technical problems existing in the above-mentioned prior art and the difficulty of solving the problems, closely combine the technical solution to be protected in the present invention and the results and data in the research and development process, etc., to analyze in detail and profoundly how to solve the technical solution of the present invention Technical problems, some creative technical effects brought about after solving the problems. The specific description is as follows:

The microservice extraction method provided by the present invention performs business function modeling based on source code, and solves the problem of poor modularity of the method for extracting microservices directly based on source code similarity; uses spectral clustering algorithm to extract microservices, Better achieve the goal of high cohesion within microservices and low coupling between microservices; extracting microservices after business function modeling of source code greatly reduces the threshold for using methods, and architects do not need to understand The legacy system can obtain candidate microservices, which realizes the automation of microservice extraction, and reduces the time and labor costs of extracting microservices from single architectures.

The present invention solves the problem that the generation of the key input part of the microservice extraction technology requires the participation of professionals and the threshold for using the method is high by proposing a method for modeling business functions based on source codes. This invention allows architects to quickly obtain candidate microservices only by automatically generating sequence diagrams and class diagrams through reverse engineering when extracting microservices from legacy systems with monolithic architectures. There is no need to analyze the business functions of the legacy system, which lowers the threshold for using the microservice extraction method. At the same time, the invention visualizes the extracted microservices in the form of a graph, which is convenient for architects to intuitively observe and adjust the extraction results.

Second, regarding the technical solution as a whole or from the perspective of a product, the technical effects and advantages of the technical solution to be protected by the present invention are specifically described as follows:

The microservice extraction method provided by the present invention is dedicated to using the source code of a single legacy system to automatically obtain business function information to extract microservices, reducing the threshold for using the microservice extraction method. The invention helps architects to quickly extract microservices from single legacy systems, making it unnecessary to understand the business information of single legacy systems, lowering the threshold for using the method, and facilitating technicians to obtain high-cohesion and low-coupling microservices faster , to reduce the cost and risk of migrating from a monolithic architecture to a microservice architecture.

Third, as an auxiliary evidence of the inventiveness of the claims of the present invention, it is also reflected in the following important aspects:

(1) The technical scheme of the present invention fills up the technical gap in the industry at home and abroad:

The current mainstream microservice extraction technology has the following problems: 1. The method of extracting microservices based on program execution traces depends on the design of test cases. Test cases must be generated by software designers who are familiar with the system, and legacy systems often have complex businesses. , the problem of high turnover of developers and incomplete documentation, so it is difficult to design test cases to completely cover all business functions, and incomplete design of test cases will lead to low coverage of the proposed microservice class, so this method uses The threshold is high. 2. The method of extracting microservices based on source code generally uses the similarity of code as the basis for extracting microservices, which cannot fully reflect the business relationship. The above determines the scalability of microservices, so the use of this method has great hidden dangers. The existence of the above problems has led to the difficulty of automatic extraction of microservices.

Ben Fangming proposed a technology for extracting business functions based on source codes, using reverse engineering to generate sequence diagrams from source codes, and then obtain calling relationships reflecting program business functions. While the quality of microservices can be improved, the participation of experts is not required, which reduces the difficulty of automatic extraction of microservices.

(2) The technical solution of the present invention solves the technical problem that people have always been eager to solve but failed to achieve success:

Microservice extraction technology has been committed to researching a method that can automatically split a single system into microservices, and the obtained microservices have good functional modularity, are easy to expand, develop and maintain, and can solve the problem of difficult maintenance of single systems. and extended questions. But the current microservice abstraction methods cannot solve the above problems at the same time.

The method for extracting business functions based on source codes in the present invention solves the current automatic acquisition of program business functions, which needs to be analyzed based on the running track of the program, which cannot be completely automated without the design of test cases by experts; overcomes the technical problem of The technical problem of poor quality microservices extracted by the source code method.

Description of drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings that need to be used in the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For Those of ordinary skill in the art can also obtain other drawings based on these drawings without making creative efforts.

Fig. 1 is a flow chart of a microservice extraction method provided by an embodiment of the present invention;

Fig. 2 is a flow chart of display business function modeling provided by an embodiment of the present invention;

FIG. 3 is a flowchart of implicit business function modeling provided by an embodiment of the present invention;

FIG. 4 is a result diagram of visualization of candidate microservices provided by an embodiment of the present invention.

Detailed ways

In order to make the object, technical solution and advantages of the present invention more clear, the present invention will be further described in detail below in conjunction with the examples. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

Aiming at the problems existing in the prior art, the present invention provides a microservice extraction method, system, medium, equipment and information processing terminal. The present invention will be described in detail below in conjunction with the accompanying drawings.

1. Explain the embodiment. In order to make those skilled in the art fully understand how to implement the present invention, this part is an explanatory embodiment for explaining the technical solution of the claims.

As shown in Figure 1, the microservice extraction method provided by the embodiment of the present invention includes the following steps:

S101, performing hierarchical division on the source code;

S102, obtaining the sequence diagram and class diagram of the control layer through reverse engineering;

S103, displaying business function modeling on the sequence diagram;

S104, performing implicit business function modeling on the class diagram;

S105, extract candidate microservices of the source code based on the business function model through spectral clustering;

S106, evaluating the quality of the candidate microservices;

S107. Visualize the candidate microservices in the form of a graph structure and provide an adjustment function.

As a preferred embodiment, the microservice extraction method provided in the embodiment of the present invention specifically includes the following steps:

S1: Divide the source code of the single legacy system into layers, as shown in Figure 2, identify the control layer, business layer, database access layer, entity layer and other layers in the source code. Generally follow the source code of the development specification, the same layer is located under the same package, and the suffix of the class is the same. The control layer is generally located under the package of the classes that directly receive user requests and respond to users, such as Controller, Action, or View. The database access layer is generally located under Dao or a package that directly interacts with the database. The service layer is generally located under the package that interacts with the control layer and the database layer such as Service. Classes in the physical layer are generally used to encapsulate data and perform data transmission. Other classes mainly involve project configuration classes and tool classes.

S2: Generate sequence diagrams and class diagrams through reverse engineering. Specifically, this step includes the following sub-steps:

S21: Use the SequenceDiagram tool to generate a sequence diagram for each method in the control layer of the source code, where the control layer of the source code refers to the class that interacts with the user interface display;

S22: Use the PlantUMLParser tool to generate a class diagram for each class file of the source code.

S3: Perform display service function modeling according to the sequence diagram. FIG. 2 is a flow chart of display service function modeling provided by an embodiment of the present invention. The detailed process is as follows:

S31: Firstly, among multiple sequence diagram files, read a sequence diagram file for analysis, and when the sequence diagram file is empty, end the display business function modeling process;

S32: Count the number of calls f _ij between two classes C _i and C _j , if there is no call relationship between C _i and C _j , add 1 to f _ij ;

S33: Store the number of calls between the two classes C _i and C _j in the Map structure, where the key is a string formed by splicing the class names of the two classes with "_", and the value is the number of calls f _ij ;

S34: After the analysis of all the sequence diagram files is completed, the model Map is output.

S4: Perform implicit business function modeling according to the class diagram. FIG. 3 is a flowchart of the implicit function modeling provided by the embodiment of the present invention. The detailed process is as follows:

S41: Firstly, in multiple class diagram files, read two class diagrams for analysis, and when the class diagram file is empty, end the implicit business function modeling process;

S42: Calculate the semantic similarity S _ij between the two class diagrams C _i and C _j . First make a word bag dictionary according to the input text information, and then match the words in the text with the key in the word bag to get the corpus corpus, initialize the tf-idf transformation model, get the converted corpus corpus_tfidf, and use the corpus_tfidf corpus to use the Lsi model Perform training, calculate the similarity of the sparse matrix, and convert the word segmentation list that needs to find the similarity into a corpus doc_test_vec to obtain the similarity of the text.

S5: Use spectral clustering to perform clustering to obtain candidate microservices. The detailed process is as follows:

S51: Construct a similarity matrix, the formula is as follows, W _ij represents a similarity matrix, if there is a calling relationship between two classes i and j, then w _ij = map _ij otherwise w _ij = S _ij ; D _ij is a diagonal matrix, diagonal The value above is the sum of the corresponding row or column in the W matrix.

S52: Construct a Laplacian matrix L, and normalize L, the formula is as follows:

L=D-W

S53: Perform eigenvalue decomposition on the Laplacian matrix, use the Lanczos method to accelerate the decomposition process, obtain the first k smallest eigenvalues and corresponding eigenvectors, and finally form a k-dimensional feature matrix F.

S54: Perform clustering on the k-dimensional feature matrix F using Kmeans. First, use AFK-MC ² to select the initial cluster center, randomly select an initial center sample c1, and then calculate the proposal distribution q(x) of all data sets (as shown in formula 5), randomly select from q(x) Extract a data point and calculate the distance dx (as shown in formula 6); finally use Markov chain Monte Carlo (Metropolis-Hastings) to sample a sequence of length m, and take the last k-1 as the center point C= {C1,C2,...,Ck}.

d _x ＝d(x,C _i-1 ) ²

Use A-means to reduce the time it takes for the K-means algorithm to assign data points to cluster C. Calculate the distance from each data point x _i to all cluster centers, select the centroid C _k closest to itself, and calculate the isometric index α _i of this point, the formula is as follows:

α _i ＝abs(||i-μ ₁ || ² -||i-μ ₂ || ² )

公式如下：Calculate the improvement threshold

The formula is as follows:

时，认为xi在此轮迭代中将不会移动，将x_i固定分配给簇C_k，不再参与距离计算和重新分配。when

When , it is considered that xi will not move in this round of iterations, and _xi is fixedly allocated to cluster C _k , and no longer participates in distance calculation and reallocation.

After spectral clustering in the above steps, a set of clustering results will be obtained, and each cluster is a microservice candidate.

S6: Evaluate the quality of the extracted candidate microservices. Using the modular evaluation method, the higher the modular evaluation value, the lower the coupling degree between the microservices, and the higher the aggregation degree within the microservice. The formula is as follows:

Among them, N is the number of microservices, N _i is the number of classes in microservices, W _i is the weight sum of internal edges of microservice i, W _{i, j} is the edge between microservice i and microservice j weight and .

S7: Visualize the candidate microservices extracted by this method, and display the microservices to which each class belongs in the form of a graph structure based on Echarts. As shown in Figure 4, the vertices of the graph represent class files, the colors of the vertices represent the microservices they belong to, the vertices with the same color belong to the same microservice, and the edges represent the calling relationship between classes.

The microservice extraction system provided by the embodiment of the present invention includes:

The layer division module is used to divide the source code into layers, and obtain the sequence diagram of the control layer through reverse engineering, and obtain the class diagram of the entity layer, database access layer and other layers;

The business function modeling module is used to model the display business function of the sequence diagram and model the implicit business function of the class diagram;

The candidate microservice extraction module is used to extract the candidate microservices of the source code based on the business function model through spectral clustering;

The quality evaluation module is used to evaluate the quality of candidate microservices, visualize the candidate microservices in the form of graph structure and provide adjustment functions.

2. Application examples. In order to prove the creativity and technical value of the technical solution of the present invention, this part is the application example of the claimed technical solution on specific products or related technologies.

The technical solution in this aspect will be described in detail below in combination with the microservice extraction embodiment of the application program SpringBlog.

The first step is to pull the source code of the SpringBlog application on github and analyze the hierarchical relationship of the source code. Identify the control layer, business layer, database access layer, entity layer, and other layers in the source code. The class of the control layer ends with the suffix Controller, which is a class that directly receives user requests and responds to users. The database access layer is generally located under repositories, the package that interacts with the database. The service layer is located under the package that interacts with the control layer and the database layer, such as Service. Classes in the physical layer are generally used to encapsulate data and perform data transmission, and are mainly located under the models and forms packages. Other classes mainly involve project configuration classes and tool classes.

The second step is to use tools to obtain the timing diagram of the control layer and the class diagrams of other layers for the control layer of SpringBlog.

The third step is to model the display business function according to the sequence diagram. Among multiple sequence diagram files, read a sequence diagram file for analysis, and when the sequence diagram file is empty, end the display business function modeling process; output the model Map. Count the number of calls f _ij between two classes C _i and C _j , if there is no call relationship between C _i and C _j , add 1 to f _ij ; count the number of calls between two classes C _i and C _j Stored in the Map structure, where the key is a string composed of the class names of the two classes spliced with "_", and the value is the number of calls f _ij ;

The fourth step is to model the implicit business function according to the class diagram. Calculate the semantic similarity S _ij between two class diagrams Ci and Cj. First make a word bag dictionary according to the input text information, and then match the words in the text with the key in the word bag to get the corpus corpus, initialize the tf-idf transformation model, get the converted corpus corpus_tfidf, and use the corpus_tfidf corpus to use the Lsi model Perform training, calculate the similarity of the sparse matrix, and convert the word segmentation list that needs to find the similarity into a corpus doc_test_vec to obtain the similarity of the text.

The fifth step is to use spectral clustering for clustering to obtain candidate microservices.

The sixth step is to use the modular evaluation method. The higher the modular evaluation value, the lower the coupling degree between the microservices, and the higher the aggregation degree within the microservice. The formula is as follows:

Among them, N is the number of microservices, N _i is the number of classes in microservices, W _i is the weight sum of internal edges of microservice i, W _{i, j} is the edge between microservice i and microservice j weight and .

The seventh step is to visualize the candidate microservices extracted by this method.

3. Evidence of the relevant effects of the embodiment. The embodiment of the present invention has achieved some positive effects in the process of research and development or use, and indeed has great advantages compared with the prior art. The following content is described in conjunction with the data and charts of the test process.

Compared with the traditional source code-based microservice extraction method, the microservice extraction method proposed in this paper has indeed greatly improved the module quality. The following table shows the quality evaluation of the two methods for SpringBlog application extraction microservices.

Among them, MEMSC is the MQ value of the microservice extracted by the traditional method, and OUR is the MQ value of the microservice obtained by the method proposed in this paper. K is the splitting granularity of microservices. We split the SpringBlog application into 6 microservices, 7 microservices, 8 microservices, 9 microservices, and 10 microservices, and count the module quality MQ of microservices . The value of MQ represents the quality of the microservice. The higher the value, the better the modularity of the microservice, and the more the microservice meets the standard of high cohesion and low coupling.

It can be seen from observation that the microservice extraction method proposed in this aspect has higher microservice quality than the traditional microservice extraction method under five different split granularities. Especially in the case of split granularity k=9, the quality of microservices has been greatly improved.

It should be noted that the embodiments of the present invention can be realized by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in memory and executed by a suitable instruction execution system such as a microprocessor or specially designed hardware. Those of ordinary skill in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and/or contained in processor control code, for example, on a carrier medium such as a magnetic disk, CD or DVD-ROM, such as a read-only memory Such code is provided on a programmable memory (firmware) or on a data carrier such as an optical or electronic signal carrier. The device and its modules of the present invention can be realized by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., It can also be realized by software executed by various types of processors, or by a combination of the above hardware circuits and software such as firmware.

The above is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Anyone familiar with the technical field within the technical scope disclosed in the present invention, whoever is within the spirit and principles of the present invention Any modifications, equivalent replacements and improvements made within shall fall within the protection scope of the present invention.

Claims (10)

1. A micro-service extraction method is characterized by comprising the following steps:

performing hierarchical division on a source code; acquiring a time sequence diagram of each method of a control layer through reverse engineering, and acquiring class diagrams of a physical layer, a database access layer and other layers; modeling a display service function of the time sequence diagram; carrying out implicit business function modeling on the class diagram; extracting candidate micro-services of the source code based on the business function model through spectral clustering; evaluating the quality of the candidate micro-service; and finally, visualizing the candidate micro-services in a graph structure form to provide an adjusting function for the framework personnel.

2. A microservice extraction process as claimed in claim 1, characterised in that it comprises the following steps:

step one, performing hierarchical division on a source code, acquiring a timing diagram of a control layer through reverse engineering, and acquiring class diagrams of a physical layer, a database access layer and other layers;

step two, performing display service function modeling on the time sequence diagram, and performing implicit service function modeling on the class diagram;

extracting candidate micro-services of the source codes based on the business function model through spectral clustering;

and step four, evaluating the quality of the candidate micro-service, visualizing the candidate micro-service in a graph structure form and providing an adjusting function.

3. The microservice extraction method of claim 2, wherein in the first step, the source code of the single legacy system is hierarchically divided, and a control layer, a business layer, a database access layer, a physical layer and other layers in the source code are identified;

generating a time sequence diagram and a class diagram of the system through reverse engineering according to the source code of the single system program:

(1) Generating a timing graph for each method in a control layer of the source code using a SequenceDiagram tool, wherein the control layer of the source code refers to a class that interacts with a user interface display in charge;

(2) Class diagrams for the database access layer and the physical and other layers are generated for each class file of the source code using the PlantUML Parser tool.

4. The microservice extraction method of claim 2, wherein in the second step, the obtaining of the call relation mapping table between classes by performing display service function modeling on the timing diagram specifically comprises:

(1) Reading one time sequence diagram file from a plurality of time sequence diagram files for analysis, and ending the business function modeling process when the time sequence diagram file is empty;

(2) Counting the calling times fij between the two classes Ci and Cj, and adding 1 to fij when the calling relation between Ci and Cj occurs once;

(3) Storing the calling times between the two classes Ci and Cj in a Map structure, wherein the key is a character string formed by splicing the class names of the two classes by using a _, and the value is the calling times fij;

(4) After all the time sequence diagram files are analyzed, outputting a model Map;

carrying out implicit business function modeling according to the class diagram to obtain a semantic similarity relation matrix between classes:

(1) Reading two class diagrams from a plurality of class diagram files for analysis, and ending the implicit business function modeling process when the class diagram files are empty;

(2) Calculating semantic similarity Sij between the two class diagrams Ci and Cj; making a bag dictionary according to input text information, and matching the words in the text with keys in the bag to obtain a corpus; initializing a tf-idf transformation model to obtain a converted corpus coprus _ tfidf, training a coprus _ tfidf corpus by using an Lsi model, and calculating the sparse matrix similarity; and converting the format by using a word segmentation list needing to find the similarity as a corpus doc _ test _ vec to obtain the similarity of the text.

5. The microservice extraction method of claim 2, wherein in the third step, after obtaining the service function model of the source code, clustering is performed by using a spectral clustering algorithm to obtain candidate microservices;

(1) Constructing a similarity matrix, wherein the formula is as follows:

wherein Wij represents a similarity matrix, if a calling relationship exists between two classes i and j, wij = mapij, and vice versa; dij is a diagonal matrix, and the value on the diagonal is the sum of corresponding rows or columns in the W matrix;

(2) Constructing a Laplace matrix L and normalizing the L, wherein the formula is as follows:

L＝D-W；

(3) Decomposing eigenvalues of the Laplacian matrix, accelerating the decomposition process by using a Lanczos method to obtain the first k minimum eigenvalues and corresponding eigenvectors, and finally forming a k-dimensional eigenvalue matrix F;

(4) Clustering the k-dimensional feature matrix F by using Kmeans; selecting initial clustering centers by using AFK-MC2, and randomly selecting and extracting an initial center sample c ₁ (ii) a Calculating a proposal distribution q (x) of all data sets, randomly extracting a data point from q (x) and calculating a distance dx; sampling a sequence with the length of m by using Markov chain Monte Carlo, and taking the last k-1 sequences as central points C = { C1, C2,.., ck };

using A-means to reduce the time for the K-means algorithm to assign data points to cluster C; calculate each data point x _i Selecting the closest centroid C to the distance from the cluster center _k Calculating the equidistant index alpha of said points _i The formula is as follows:

α _i ＝abs(||i-μ ₁ || ² -||i-μ ₂ || ² )；

calculating an improvement threshold

The formula is as follows:

when in use

When x is _i Do not move in round iterations, x _i Fixed allocation to cluster C _k No longer participating in distance calculation and reallocation;

after spectral clustering, a group of clustering results is obtained, and each clustering cluster is a micro service candidate.

6. The microservice extraction method of claim 2, wherein the quality assessment of the extracted candidate microservices at step four comprises:

using a modularization evaluation mode, the higher the modularization evaluation value is, the lower the coupling degree between the micro services is, and the higher the polymerization degree inside the micro service is, and the calculation formula is as follows:

where N is the number of microservices, N _i Is the number of classes within the microservice, W _i Is the weighted sum, W, between the internal edges of the microservice i _i，j Is the sum of the weights of the edges between micro service i and micro service j;

and visualizing the extracted candidate micro-services, displaying the micro-services to which the class files belong and the calling relation among the classes in a graph structure form based on Echarts, and dragging a vertex to manually adjust the micro-services to which the classes belong.

7. A microservice extraction system to which the microservice extraction method of any of claims 1 to 6 is applied, the microservice extraction system comprising:

the hierarchical division module is used for hierarchically dividing the source code, acquiring a timing chart of the control layer through reverse engineering, and acquiring class diagrams of a physical layer, a database access layer and other layers;

the business function modeling module is used for performing display business function modeling on the time sequence diagram and performing implicit business function modeling on the class diagram;

the candidate micro-service extraction module is used for extracting candidate micro-services of the source codes based on the business function model through spectral clustering;

and the quality evaluation module is used for evaluating the quality of the candidate micro-service, visualizing the candidate micro-service in a graph structure form and providing an adjusting function.

8. A computer arrangement, characterized in that the computer arrangement comprises a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to carry out the steps of the microservice extraction method according to any of the claims 1-6.

9. A computer-readable storage medium, storing a computer program which, when executed by a processor, causes the processor to carry out the steps of the microservice extraction method according to any of claims 1 to 6.

10. An information data processing terminal characterized by being configured to implement the micro-service extraction system of claim 7.

Images