Automatic test script generation method and system based on deep learning

Abstract

The invention discloses an automatic test script generation method and system based on deep learning, and belongs to the field of data communication of a data center, and the system comprises a test script receiving module, a test script management module, a user interaction unit, a parameter association mapping module and an automatic test script generation module. The test script automatic generation module specifically comprises a system management module, a test case classification module, a test instruction generation module, a test script generation module and a model training and updating module; according to the method, mapping from the test step text to the script code is realized, the defects of poor flexibility, weak expandability and low script generation efficiency of a template matching method are overcome, the workload of testers is greatly reduced, the generated result can be examined, corrected and confirmed by the testers, and the safety, reliability and accuracy of the test script are ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of automatic test script generation, and in particular, relates to a method and system for automatic test script generation based on deep learning. Background Art

[0002] As project structures and services become increasingly complex, testing will become an important part of project quality control. Massive request interfaces and business scenarios require testers to perform corresponding regression verification to ensure the normal operation of services. Manual testing that relied on testers in the early days has become a thing of the past, and automated testing has gradually become the mainstream.

[0003] However, automated testing requires a lot of time to write the relevant automated testing code. Not only that, the writing of automated testing code also relies on testers to learn the corresponding technical syntax, which places high demands on testers. When the product code is changed, the test script often needs to be rewritten or updated, which increases a lot of development investment.

[0004] Based on the current situation that the preparation time for automated testing is long and the threshold for writing use cases is high, there is an urgent need for a method that can parse the interface description language, provide simple and easy-to-use interface operations, cooperate with the tester's business logic, customize test cases, and generate automated test scripts. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method and system for automatically generating test scripts based on deep learning in response to the shortcomings of the background technology.

[0006] The present invention adopts the following technical solutions to solve the above technical problems:

[0007] A test script automatic generation system based on deep learning, comprising a test script receiving module, a test script management module, a user interaction unit, a parameter association mapping module, and a test script automatic generation module;

[0008] A test script receiving module, used for generating an interface description language according to the identified interface related parameters;

[0009] A test script management module, used for storing the interface description language generated according to the identified interface related parameters;

[0010] A user interaction unit, used to perform a standard analysis on the interface description language to obtain an interaction interface, and to set business parameters and issue a preview request through the interaction interface;

[0011] The parameter association mapping module is used to establish parameter association mapping, verify the preview results through the set parameter constraints, and generate test scripts;

[0012] A test script automatic generation module is used to generate test scripts, which specifically includes a system management module, a test case classification module, a test instruction generation module, a test script generation module, and a model training and update module;

[0013] Among them, the test case classification module is used to analyze the interface description language generated according to the identified interface related parameters to obtain the corresponding test subsystem and test subsystem knowledge base;

[0014] A test instruction generation module is used to analyze the interface description language and convert it into corresponding test instructions;

[0015] The test script generation module is used to convert the test instructions output by the test instruction generation module into script codes, and display them to the user, perform standard parsing to obtain an interactive interface, and set business parameters through the interactive interface and issue a preview request;

[0016] The model training and updating module is used to train and update the network model used.

[0017] As a further preferred solution of a deep learning-based test script automatic generation system of the present invention, the test script receiving module includes a parameter annotation module and an interface parsing module;

[0018] The parameter annotation module is used to identify interface related parameters in the form of annotations;

[0019] The interface parsing module is used to parse the annotations in the code to automatically generate an interface description language.

[0020] As a further preferred solution of a test script automatic generation system based on deep learning of the present invention, the parameter association mapping refers to associating different description languages ​​with consistent actual logic; the establishment of parameter association mapping includes: converting different description languages ​​into parameters under the template logical relationship and comparing them, and establishing parameter association mapping if the logic is the same.

[0021] As a further preferred solution of a deep learning-based test script automatic generation system of the present invention, the model training and updating module includes a parsing module, a detection model, a direction classifier, and a recognition model;

[0022] Wherein, the parsing module is used to generate an interface description language according to the identified interface related parameters;

[0023] A detection model is used to perform a standard analysis on the interface description language to obtain an interactive interface, and to crop multiple text foreground areas of the interactive interface into multiple quadrilateral text line images through various connected domain algorithms;

[0024] A direction classifier is used to classify the directions of the quadrilateral text line images, calculate the true rotation angle of the quadrilateral text line images, and transform the quadrilateral text line images into positive text line rectangular images through perspective transformation;

[0025] The recognition model is used to perform OCR recognition on text line rectangular images, and at the same time, combine the business corpus TopK to obtain the text content, and combine the foreground area coordinates and text content to extract text from the test script image file and output editable formatted text content.

[0026] As a further preferred solution of a test script automatic generation system based on deep learning of the present invention, the system management module includes an acquisition control unit, a data processing and RAM read-write module unit, a port control unit, a synchronous clock control unit, a command deframing unit, a FIFO data cache unit, a Flash data storage control unit, and a data reading unit. The acquisition control unit, port control unit, synchronous clock control unit, command deframing unit, FIFO data cache unit, Flash data storage control unit, and data reading unit are respectively connected to the data processing and RAM read-write module unit.

[0027] A method for automatically generating a test script system based on deep learning, specifically comprising the following steps;

[0028] Step 1, generating an interface description language according to the identified interface related parameters;

[0029] Step 2, storing the interface description language generated according to the identified interface related parameters;

[0030] Step 3, parsing the interface description language to obtain an interactive interface, setting business parameters through the interactive interface and issuing a preview request;

[0031] Step 4: Establish parameter association mapping, verify the preview results through the set parameter constraints, and generate a test script;

[0032] Step 5, analyze the interface description language generated according to the identified interface-related parameters, obtain the corresponding test subsystem and test subsystem knowledge base, analyze the interface description language, convert it into corresponding test instructions, convert the test instructions output by the test instruction generation module into script code, and display it to the user, perform standard parsing to obtain an interactive interface, and at the same time set business parameters through the interactive interface and issue a preview request to train and update the network model used.

[0033] As a further preferred solution of the method for automatically generating test scripts based on deep learning of the present invention, the network model used is trained and updated, which specifically includes the following steps:

[0034] Step 5.1, generating an interface description language according to the identified interface related parameters through a parsing module;

[0035] Step 5.2, the interface description language is parsed in a standardized manner by a detection model to obtain an interactive interface, and multiple text foreground areas of the interactive interface are cropped into multiple quadrilateral text line images by various connected domain algorithms;

[0036] Step 5.3, classify the direction of the quadrilateral text line image through a direction classifier, calculate the true rotation angle of the quadrilateral text line image, and transform the quadrilateral text line image into a positive text line rectangular image through perspective transformation;

[0037] Assume that the coordinates of the vertices of the quadrilateral are P = [(x1, y1), (x2, y2), (x3, y3), (x4, y4)], and the true angle is θ;

[0038] Calculate the angle transformation matrix MaT_θ, the specific calculation is as follows:

[0039]

[0040] Calculate the four-point coordinates P′ after angle transformation. The specific calculation is as follows:

[0041]

[0042] P′=[(x′1,y′1),(x′2,y′2),(x′3,y′3),(x′4,y′4)]

[0043] Calculate the perspective transformation matrix MaT_T. The specific calculation is as follows:

[0044] w=max(x′1,x′2,x′3,x′4)-min(x′1,x′2,x′3,x′4)

[0045] h=max(y′1,y′2,y′3,y′4)-min(y′1,y′2,y′3,y′4)

[0046] P m =[(0,0),(w,0),(w,h),(0,h)]

[0047] MaT_T=PT(P′,P m )

[0048] Among them, w and h represent the width and height of the quadrilateral, P m Represents the coordinates of the quadrilateral after affine transformation, and PT represents the affine matrix calculation;

[0049] Calculate the mixed transformation matrix MaT_M, the specific calculation is as follows:

[0050] MaT_M=MaT_θ×MaT_T

[0051] To correct a quadrilateral image into a regular text row rectangular image, it is necessary to first perform an affine transformation with MaT_θ and then perform a perspective transformation with MaT_T. The quadrilateral image only needs to be matrix multiplied with MaT_M to calculate the corresponding regular text row rectangular image.

[0052] Step 5.4, perform OCR recognition on the text line rectangle image through the recognition model, and combine it with the business corpus TopK to obtain the text content, and combine the foreground area coordinates and text content to extract text from the test script image file and output editable formatted text content.

[0053] As a further preferred solution of the method for automatically generating test scripts based on deep learning of the present invention, the generation of the interface description language includes the following steps: identifying interface-related parameters in the form of annotations; and automatically generating the interface description language by parsing the annotations in the code.

[0054] As a further preferred solution of the method for automatically generating test scripts based on deep learning of the present invention, the interface description language includes at least: json language, YAML language and WSDL language.

[0055] As a further preferred solution of the method for automatically generating test scripts based on deep learning of the present invention, obtaining the interactive interface includes: parsing the interface specification through a description language parser.

[0056] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0057] 1. The present invention discloses a method and system for automatically generating test scripts based on deep learning, which generates an interface description language according to the interface-related parameters of the identification; stores the interface description language generated according to the interface-related parameters of the identification; performs standard analysis on the interface description language to obtain an interactive interface, and sets business parameters through the interactive interface and issues a preview request; establishes a parameter association mapping, verifies the preview result through the set parameter constraints, and generates a test script; analyzes the interface description language generated according to the interface-related parameters of the identification to obtain a corresponding test subsystem and a test subsystem knowledge base, analyzes the interface description language, converts it into a corresponding test instruction, converts the test instruction output by the test instruction generation module into a script code, and displays it to the user, performs standard analysis to obtain an interactive interface, and sets business parameters through the interactive interface and issues a preview request, and trains and updates the network model used; the present invention realizes the mapping of test step text to script code, solves the shortcomings of poor flexibility, weak scalability, and low script generation efficiency of the template matching method, greatly reduces the workload of testers, and the generated results can be used for testers to review, correct and confirm, ensuring the security, reliability and accuracy of the test script;

[0058] 2. The present invention obtains an interactive interface through standardized analysis of the interface description language, thereby realizing simple and easy-to-use interface operation; by setting business parameters and establishing associated mappings, it is possible to avoid spending a lot of time on writing related codes, and the writing of codes does not depend on the tester's understanding of other writing languages; through the automatic generation method of test scripts, when the product code is updated or the description language is different, the situation of rewriting or updating the test script is avoided, thereby reducing development investment. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0060] Figure 1 It is a schematic diagram of a deep learning-based test script automatic generation system of the present invention;

[0061] Figure 2 It is a structural principle diagram of the test script automatic generation module of the present invention;

[0062] Figure 3 It is a schematic diagram of the test script receiving module of the present invention;

[0063] Figure 4It is a schematic diagram of the model training and updating module of the present invention;

[0064] Figure 5 It is a structural principle diagram of the system management module of the present invention;

[0065] Figure 6 It is a flow chart of a method for automatically generating test scripts based on deep learning of the present invention. DETAILED DESCRIPTION

[0066] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings:

[0067] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The present invention is described in detail below based on the drawings and preferred embodiments, and the purpose and effect of the present invention will become clearer. It should be understood that the specific embodiments described here are only used to explain the present invention and are not used to limit the present invention.

[0068] A deep learning-based test script automatic generation system, such as Figure 1 As shown, specifically including:

[0069] A test script receiving module, used for generating an interface description language according to the identified interface related parameters;

[0070] A test script management module, used for storing the interface description language generated according to the identified interface related parameters;

[0071] A user interaction unit, used to perform a standard analysis on the interface description language to obtain an interaction interface, and to set business parameters and issue a preview request through the interaction interface;

[0072] The parameter association mapping module is used to establish parameter association mapping, verify the preview results through the set parameter constraints, and generate test scripts;

[0073] The test script automatic generation module is used to generate test scripts, such as Figure 2 As shown, it specifically includes a system management module, a test case classification module, a test instruction generation module, a test script generation module, and a model training and updating module;

[0074] Among them, the test case classification module is used to analyze the interface description language generated according to the identified interface related parameters to obtain the corresponding test subsystem and test subsystem knowledge base;

[0075] A test instruction generation module is used to analyze the interface description language and convert it into corresponding test instructions;

[0076] The test script generation module is used to convert the test instructions output by the test instruction generation module into script codes, and display them to the user, perform standard parsing to obtain an interactive interface, and set business parameters through the interactive interface and issue a preview request;

[0077] The model training and updating module is used to train and update the network model used.

[0078] like Figure 3 As shown, the test script receiving module includes a parameter annotation module and an interface parsing module;

[0079] The parameter annotation module is used to identify interface related parameters in the form of annotations;

[0080] The interface parsing module is used to parse the annotations in the code to automatically generate an interface description language.

[0081] The parameter association mapping refers to associating different description languages ​​with consistent actual logic; the establishment of parameter association mapping includes: converting different description languages ​​into parameters under the template logic relationship and comparing them, and establishing parameter association mapping if the logic is the same.

[0082] like Figure 4 As shown, the model training and updating module includes a parsing module, a detection model, a direction classifier, and a recognition model;

[0083] Wherein, the parsing module is used to generate an interface description language according to the identified interface related parameters;

[0084] A detection model is used to perform a standard analysis on the interface description language to obtain an interactive interface, and to crop multiple text foreground areas of the interactive interface into multiple quadrilateral text line images through various connected domain algorithms;

[0085] A direction classifier is used to classify the directions of the quadrilateral text line images, calculate the true rotation angle of the quadrilateral text line images, and transform the quadrilateral text line images into positive text line rectangular images through perspective transformation;

[0086] The recognition model is used to perform OCR recognition on text line rectangular images, and at the same time, combine the business corpus TopK to obtain the text content, and combine the foreground area coordinates and text content to extract text from the test script image file and output editable formatted text content.

[0087] like Figure 5As shown, the system management module includes an acquisition control unit, a data processing and RAM read-write module unit, a port control unit, a synchronous clock control unit, a command deframing unit, a FIFO data cache unit, a Flash data storage control unit, and a data reading unit. The acquisition control unit, the port control unit, the synchronous clock control unit, the command deframing unit, the FIFO data cache unit, the Flash data storage control unit, and the data reading unit are respectively connected to the data processing and RAM read-write module unit.

[0088] A method for automatically generating a test script system based on deep learning, such as Figure 6 As shown, the specific steps include the following:

[0089] Step 1, generating an interface description language according to the identified interface related parameters;

[0090] Step 2, storing the interface description language generated according to the identified interface related parameters;

[0091] Step 3, parsing the interface description language to obtain an interactive interface, setting business parameters through the interactive interface and issuing a preview request;

[0092] Step 4: Establish parameter association mapping, verify the preview results through the set parameter constraints, and generate a test script;

[0093] Step 5, analyze the interface description language generated according to the identified interface-related parameters, obtain the corresponding test subsystem and test subsystem knowledge base, analyze the interface description language, convert it into corresponding test instructions, convert the test instructions output by the test instruction generation module into script code, and display it to the user, perform standard parsing to obtain an interactive interface, and at the same time set business parameters through the interactive interface and issue a preview request to train and update the network model used.

[0094] Training and updating the network model used includes the following steps:

[0095] Step 5.1, generating an interface description language according to the identified interface related parameters through a parsing module;

[0096] Step 5.2, the interface description language is parsed in a standardized manner by a detection model to obtain an interactive interface, and multiple text foreground areas of the interactive interface are cropped into multiple quadrilateral text line images by various connected domain algorithms;

[0097] Step 5.3, classify the direction of the quadrilateral text line image through the direction classifier, calculate the true rotation angle of the quadrilateral text line image, and transform the quadrilateral text line image into a positive text line rectangular image through perspective transformation;

[0098] Assume that the coordinates of the vertices of the quadrilateral are P = [(x1, y1), (x2, y2), (x3, y3), (x4, y4)], and the true angle is θ;

[0099] Calculate the angle transformation matrix MaT_θ, the specific calculation is as follows:

[0100]

[0101] Calculate the four-point coordinates P′ after angle transformation. The specific calculation is as follows:

[0102]

[0103] P′=[(x′1,y′1),(x′2,y′2),(x′3,y′3),(x′4,y′4)]

[0104] Calculate the perspective transformation matrix MaT_T. The specific calculation is as follows:

[0105] w=max(x′1,x′2,x′3,x′4)-min(x′1,x′2,x′3,x′4)

[0106] h=max(y′1,y′2,y′3,y′4)-min(y′1,y′2,y′3,y′4)

[0107] P m =[(0,0),(w,0),(w,h),(0,h)]

[0108] MaT_T=PT(P′,P m )

[0109] Among them, w and h represent the width and height of the quadrilateral, P m Represents the coordinates of the quadrilateral after affine transformation, and PT represents the affine matrix calculation;

[0110] Calculate the mixed transformation matrix MaT_M, the specific calculation is as follows:

[0111] MaT_M=MaT_θ×MaT_T

[0112] To correct a quadrilateral image into a regular text row rectangular image, it is necessary to first perform an affine transformation with MaT_θ and then perform a perspective transformation with MaT_T. The quadrilateral image only needs to be matrix multiplied with MaT_M to calculate the corresponding regular text row rectangular image.

[0113] Step 5.4, perform OCR recognition on the text line rectangle image through the recognition model, and combine it with the business corpus TopK to obtain the text content, and combine the foreground area coordinates and text content to extract text from the test script image file and output editable formatted text content.

[0114] The generation of the interface description language comprises the following steps: identifying interface related parameters in the form of annotations; and automatically generating the interface description language by parsing the annotations in the code.

[0115] The interface description language includes at least: json language, YAML language and WSDL language.

[0116] The acquisition of the interactive interface includes: parsing the interface specification by a description language parser

[0117] The present invention discloses a method and system for automatically generating test scripts based on deep learning, which generates an interface description language according to the interface-related parameters of the identification; stores the interface description language generated according to the interface-related parameters of the identification; performs standard parsing on the interface description language to obtain an interactive interface, and sets business parameters through the interactive interface and issues a preview request; establishes a parameter association mapping, verifies the preview result through the set parameter constraints, and generates a test script; analyzes the interface description language generated according to the interface-related parameters of the identification to obtain a corresponding test subsystem and a test subsystem knowledge base, analyzes the interface description language, converts it into a corresponding test instruction, and performs a parameter association mapping on the preview result. The test instructions output by the test instruction generation module are converted into script codes and displayed to the user. The interactive interface is obtained by performing specification parsing. At the same time, business parameters are set through the interactive interface and a preview request is issued to train and update the network model used. The invention automatically determines the test subsystem corresponding to the current test case, obtains the corresponding subsystem knowledge base, and realizes the mapping of the test step text to the script code, which solves the shortcomings of the template matching method, such as poor flexibility, weak scalability, and low script generation efficiency, greatly reduces the workload of testers, and the generated results can be used by testers for review, correction and confirmation, ensuring the security, reliability and accuracy of the test script.

[0118] The present invention obtains an interactive interface through standardized analysis of the interface description language, thereby realizing simple and easy-to-use interface operation; by setting business parameters and establishing associated mappings, it is realized that there is no need to spend a lot of time on writing related codes, and the writing of codes does not depend on the tester's understanding of other writing languages; through the automatic generation method of test scripts, when the product code is updated or the description language is different, the situation of rewriting or updating the test script is avoided, thereby reducing development investment.

[0119] Those skilled in the art can understand that the above are only preferred examples of the invention and are not intended to limit the invention. Although the invention is described in detail with reference to the above examples, those skilled in the art can still modify the technical solutions recorded in the above examples or replace some of the technical features with equivalents. Any modification, equivalent replacement, etc. made within the spirit and principle of the invention should be included in the protection scope of the invention. All technical features in this embodiment can be freely combined according to actual needs.

[0120] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A deep learning-based test script automatic generation system, characterized by: It includes a test script receiving module, a test script management module, a user interaction unit, a parameter association mapping module, and a test script automatic generation module; Wherein, the test script receiving module is used to generate an interface description language according to the identified interface related parameters; A test script management module, used for storing the interface description language generated according to the identified interface related parameters; A user interaction unit, used to perform a standard analysis on the interface description language to obtain an interaction interface, and to set business parameters and issue a preview request through the interaction interface; The parameter association mapping module is used to establish parameter association mapping, verify the preview results through the set parameter constraints, and generate test scripts; A test script automatic generation module is used to generate test scripts, which specifically includes a system management module, a test case classification module, a test instruction generation module, a test script generation module, and a model training and update module; Among them, the test case classification module is used to analyze the interface description language generated according to the identified interface related parameters to obtain the corresponding test subsystem and test subsystem knowledge base; A test instruction generation module is used to analyze the interface description language and convert it into corresponding test instructions; The test script generation module is used to convert the test instructions output by the test instruction generation module into script codes, and display them to the user, perform standard parsing to obtain an interactive interface, and set business parameters through the interactive interface and issue a preview request; The model training and updating module is used to train and update the network model used.

2. The deep learning-based test script automatic generation system according to claim 1, characterized in that: The test script receiving module includes a parameter annotation module and an interface parsing module; The parameter annotation module is used to identify interface related parameters in the form of annotations; The interface parsing module is used to parse the annotations in the code to automatically generate an interface description language.

3. The deep learning-based test script automatic generation system according to claim 1, characterized in that: The parameter association mapping refers to associating different description languages ​​with consistent actual logic; the establishment of parameter association mapping includes: converting different description languages ​​into parameters under the template logic relationship and comparing them, and establishing parameter association mapping if the logic is the same.

4. The deep learning-based test script automatic generation system according to claim 1, characterized in that: The model training and updating module includes parsing module, detection model, direction classifier, and recognition model; Wherein, the parsing module is used to generate an interface description language according to the identified interface related parameters; A detection model is used to perform a standard analysis on the interface description language to obtain an interactive interface, and to crop multiple text foreground areas of the interactive interface into multiple quadrilateral text line images through various connected domain algorithms; A direction classifier is used to classify the directions of the quadrilateral text line images, calculate the true rotation angle of the quadrilateral text line images, and transform the quadrilateral text line images into positive text line rectangular images through perspective transformation; The recognition model is used to perform OCR recognition on text line rectangular images, and at the same time, combine the business corpus TopK to obtain the text content, and combine the foreground area coordinates and text content to extract text from the test script image file and output editable formatted text content.

5. The deep learning-based test script automatic generation system according to claim 1, characterized in that: The system management module includes an acquisition control unit, a data processing and RAM read-write module unit, a port control unit, a synchronous clock control unit, a command deframing unit, a FIFO data cache unit, a Flash data storage control unit, and a data reading unit. The acquisition control unit, the port control unit, the synchronous clock control unit, the command deframing unit, the FIFO data cache unit, the Flash data storage control unit, and the data reading unit are respectively connected to the data processing and RAM read-write module unit.

6. A method for automatically generating a test script based on deep learning according to any one of claims 1 to 5, characterized in that: The specific steps include: Step 1, generating an interface description language according to the identified interface related parameters; Step 2, storing the interface description language generated according to the identified interface related parameters; Step 3, parsing the interface description language to obtain an interactive interface, setting business parameters through the interactive interface and issuing a preview request; Step 4: Establish parameter association mapping, verify the preview results through the set parameter constraints, and generate a test script; Step 5, analyze the interface description language generated according to the identified interface-related parameters, obtain the corresponding test subsystem and test subsystem knowledge base, analyze the interface description language, convert it into corresponding test instructions, convert the test instructions output by the test instruction generation module into script code, and display it to the user, perform standard parsing to obtain an interactive interface, and at the same time set business parameters through the interactive interface and issue a preview request to train and update the network model used.

7. The method for automatically generating test scripts based on deep learning according to claim 6, characterized in that: Training and updating the network model used includes the following steps: Step 5.1, generating an interface description language according to the identified interface related parameters through a parsing module; Step 5.2, the interface description language is parsed in a standardized manner by a detection model to obtain an interactive interface, and multiple text foreground areas of the interactive interface are cropped into multiple quadrilateral text line images by various connected domain algorithms; Step 5.3, classify the direction of the quadrilateral text line image through a direction classifier, calculate the true rotation angle of the quadrilateral text line image, and transform the quadrilateral text line image into a positive text line rectangular image through perspective transformation; Assume that the coordinates of the vertices of the quadrilateral are P = [(x1, y1), (x2, y2), (x3, y3), (x4, y4)], and the true angle is θ; Calculate the angle transformation matrix MaT_θ, the specific calculation is as follows: Calculate the four-point coordinates P′ after angle transformation. The specific calculation is as follows: P′=[(x′1,y′1),(x′2,y′2),(x′3,y′3),(x′4,y′4)] Calculate the perspective transformation matrix MaT_T. The specific calculation is as follows: w=max(x′1,x′2,x′3,x′4)-min(x′1,x′2,x′3,x′4) h=max(y′1,y′2,y′3,y′4)-min(y′1,y′2,y′3,y′4) P m =[(0,0),(w,0),(w,h),(0,h)] MaT_T=PT(P′,P m ) Among them, w and h represent the width and height of the quadrilateral, P m Represents the coordinates of the quadrilateral after affine transformation, and PT represents the affine matrix calculation; Calculate the mixed transformation matrix MaT_M, the specific calculation is as follows: MaT_M=MaT_θ×MaT_T To correct a quadrilateral image into a regular text row rectangular image, it is necessary to first perform an affine transformation with MaT_θ and then perform a perspective transformation with MaT_T. The quadrilateral image only needs to be matrix multiplied with MaT_M to calculate the corresponding regular text row rectangular image. Step 5.4, perform OCR recognition on the text line rectangle image through the recognition model, and combine it with the business corpus TopK to obtain the text content, and combine the foreground area coordinates and text content to extract text from the test script image file and output editable formatted text content.

8. The method for automatically generating test scripts based on deep learning according to claim 6, characterized in that: The generation of the interface description language comprises the following steps: identifying interface related parameters in the form of annotations; and automatically generating the interface description language by parsing the annotations in the code.

9. The method for automatically generating test scripts based on deep learning according to claim 6, characterized in that: The interface description language includes at least: json language, YAML language and WSDL language.

10. The method for automatically generating test scripts based on deep learning according to claim 6, characterized in that: The acquisition of the interactive interface includes: parsing the interface specification by a description language parser.

Images