Multi-language configuration file generation method and device, electronic equipment and storage medium

Abstract

The invention discloses a multi-language configuration file generation method and device, electronic equipment and a storage medium, and relates to the technical field of computers. The method comprises the steps that when a source language configuration file changes, configuration change information is positioned according to a version change tree of the source language configuration file and tree node structure attribute information; determining types and quantity of translation modes and translation weights of different translation modes according to a term proportion, character grammar, translation quantity and scene adjustment factors of the configuration change information, and translating the configuration change information by utilizing the selected translation mode to obtain a single translation result; determining a translation result of the configuration change information according to the single translation result and the corresponding translation weight; and updating a target language configuration file corresponding to the source language configuration file according to a translation result. According to the method, the problem of inaccurate translation results in related technologies can be solved, and the accuracy of multi-language configuration file generation is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular, to a method, device, electronic device, and computer-readable storage medium for generating multilingual configuration files. Background Art

[0002] Currently, related technologies rely on tools or models with translation capabilities to generate multilingual configuration files. Although this method can achieve automated generation of multilingual configuration files, and such translation methods have good accuracy for common languages, however, they lack the understanding of specific technical terms and sensitivity to cultural differences, resulting in inaccurate translation results or even misunderstandings in the final generated multilingual configuration files.

[0003] In view of this, improving the accuracy of multilingual configuration files is a technical problem that needs to be solved by those skilled in the art. Summary of the Invention

[0004] The present invention provides a method, device, electronic device, and computer-readable storage medium for generating multilingual configuration files, which effectively improve the accuracy of multilingual configuration files.

[0005] To solve the above technical problems, the present invention provides the following technical solutions:

[0006] On the one hand, the present invention provides a method for generating a multilingual configuration file, including:

[0007] When the source language configuration file changes, locate the configuration change information according to the version change tree of the source language configuration file and the tree node structure attribute information; the version change tree represents the difference information of the version information structure tree corresponding to the source language configuration file before and after the change; determine the types and quantities of translation methods according to the target term ratio, literal grammar, and translation volume of the configuration change information, and use the selected translation methods to translate the configuration change information to obtain a single translation result; when there are at least two translation methods, determine the translation result of the configuration change information according to the target term ratio, literal grammar, translation volume, and preset scenario adjustment factor of the configuration change information, and according to the single translation result and the corresponding translation weight; update the target language configuration file corresponding to the source language configuration file according to the translation result.

[0008] On the other hand, the present invention provides a device for generating a multilingual configuration file, including:

[0009] An incremental compilation module, configured to locate the configuration change information according to the version change tree of the source language configuration file and the tree node structure attribute information when the source language configuration file changes; the version change tree represents the difference information of the version information structure tree corresponding to the source language configuration file before and after the change.

[0010] The multi-engine fusion translation module is used to determine the types and quantities of translation methods according to the target term ratio, literal grammar, and translation volume of the configuration change information, and use the selected translation methods to translate the configuration change information to obtain a single translation result; when there are at least two translation methods, according to the target term ratio, literal grammar, translation volume, and preset scenario adjustment factor of the configuration change information, and based on the single translation result and the corresponding translation weight, determine the translation result of the configuration change information; update the target language configuration file corresponding to the source language configuration file according to the translation result.

[0011] The present invention also provides an electronic device, including a memory and a processor. When the processor executes the computer program stored in the memory, the steps of any one of the above multi-language configuration file generation methods are implemented.

[0012] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of any one of the above multi-language configuration file generation methods are implemented.

[0013] The advantages of the technical solution provided by the present invention are as follows: using a tree structure to store the version information of the language configuration file, automatically identifying the changes in the configuration file by analyzing the version change tree, realizing node-level change identification, and only translating the changed parts. By reducing redundant operations, the efficiency of generating multi-language configuration files can be effectively improved, and the time-consuming for generating multi-language configurations can be compressed from several hours / thousand files in the related art to within dozens of minutes, and the generation speed of multi-language configuration files is increased by more than 10 times. Select the most suitable translation method according to the statement complexity of the changed part of the source language configuration file, which is beneficial to improving the accuracy of the finally generated multi-language configuration file. Further, for complex situations, multiple translation methods are also supported. Use the translation capabilities that different translation methods are good at to make up for the problem that a single translation method may result in inaccurate translation due to insufficient understanding of professional terms or culture. Determine the final translation result according to different translation methods and the corresponding weights, further improving the accuracy of the multi-language configuration file.

[0014] In addition, the present invention also provides a corresponding implementation device, electronic device, and computer-readable storage medium for the multi-language configuration file generation method, further making the method more practical, and the device, electronic device, and computer-readable storage medium have corresponding advantages. Brief Description of the Drawings

[0015] To more clearly illustrate the technical solutions of the present invention or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0016] Figure 1 Schematic diagram of the hardware composition framework applicable to the multi - language configuration file generation method provided by the present invention; Figure 2 Flow chart of a multi - language configuration file generation method provided by the present invention; Figure 3 Flow chart of another multi - language configuration file generation method provided by the present invention; Figure 4 Structural framework diagram under an exemplary embodiment of the multi - language configuration file generation device provided by the present invention; Figure 5 Structural diagram of an exemplary embodiment of the electronic device provided by the present invention. Detailed implementation manners

[0017] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the following will further elaborate on the present invention in conjunction with the drawings and specific implementation manners. Among them, the terms "first", "second", "third", "fourth", etc. in the specification and the above - mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non - exclusive inclusion. The term "exemplary" means "serving as an example, embodiment, or illustration". Any embodiment described as "exemplary" here does not have to be interpreted as being superior to or better than other embodiments.

[0018] Based on the requirements of i18n (Internationalization), software needs to be able to run in different language and cultural environments. When the Chinese configuration file is iterated, synchronization problems are likely to occur between different versions, resulting in inconsistent configuration files. Therefore, it is necessary to generate multi-language configuration files. To solve the problems existing in the traditional manual maintenance of configuration files in multiple language versions by R & D personnel, related technologies use tools or models with translation capabilities to automatically generate multi-language configuration files. However, although this method can quickly generate multi-language configuration files, due to the lack of understanding of specific domain terms and sensitivity to cultural differences, the translation results of multi-language configuration files will be inaccurate and even cause misunderstandings. To further improve the accuracy of multi-language configuration files, related technologies will introduce manual secondary review. That is to say, after automatically generating multi-language configuration files, the correctness and consistency of the translation will be reviewed manually. This process will reduce the generation efficiency of multi-language configuration files and cannot guarantee high accuracy.

[0019] In view of this, the present invention uses a tree structure to store the version information of language configuration files, realizes node-level change recognition by analyzing the version change tree, and only translates the parts of the source language configuration file that have changed. The most suitable translation method is selected according to the statement complexity of the changed parts of the source language configuration file, and the translation capabilities that different translation methods are good at are used to make up for the problem of inaccurate translation results caused by insufficient understanding of the professional field or culture. The final translation result is jointly determined according to different translation methods and corresponding weights. Combining with the specific application environment architecture or specific hardware architecture on which the execution of the multi-language configuration file generation method depends, the specific application environment architecture or specific hardware architecture is described here. The following combines Figure 1 Some possible application scenarios related to the technical solution of the present invention are introduced by way of example, which may include the following content: The multi-language configuration file generation system includes a server 11 and a client 12. The server 11 and the client 12 are connected through a bus or wirelessly to achieve data transmission and interaction. The client 12 has at least a human-computer interaction module. A computer program for implementing the steps of any one of the multi-language configuration file generation methods recorded in the above embodiments is stored in the read-only memory of the server. The server 11 will pre-define multiple application programming interfaces or set up plugins of multiple translation tools with different translation capabilities. Whenever the processor of the server 11 detects a language configuration file, it will generate a corresponding version information structure tree for the language configuration file.

[0020] When the server 11 detects that the source language configuration file has changed, it determines the version change tree based on the latest version information structure tree and the second latest version information structure tree of the source language configuration file, and locates the configuration change information according to the version change tree and the structure attribute information of its upper tree nodes; it determines the types and quantities of translation methods according to the target term ratio, literal grammar, and translation volume of the configuration change information. After determining the target translation method to be selected, it calls the corresponding translation engine through the application programming interface or translates the configuration change information by calling the plug-in of the translation tool to obtain a single translation result; when there are at least two translation methods, it determines the translation result of the configuration change information according to the target term ratio, literal grammar, translation volume, and preset scenario adjustment factor of the configuration change information, and according to the single translation result and the corresponding translation weight; it updates the target language configuration file corresponding to the source language configuration file according to the translation result, thereby effectively improving the generation efficiency and generation accuracy of the multi-language configuration file. When manual review of translation accuracy is required or when user feedback on translation accuracy and satisfaction is needed, the processor of the server 11 can also generate corresponding manual review / review instructions or user feedback instructions, and then send the instructions to the user terminal 12 to display the translation result on the user interface of the human-computer interaction module, so that the accuracy of translation method selection and the accuracy of the final translation result can be continuously optimized according to historical translation records and user feedback.

[0021] It should be noted that the above application scenarios are only shown for the convenience of understanding the ideas and principles of the present invention, and the embodiments of the present invention are not limited in this regard. On the contrary, the embodiments of the present invention can be applied to any applicable scenario. After introducing the technical solutions of the present invention, the various non-limiting embodiments of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. First, please refer to Figure 2 , Figure 2 which is a schematic flowchart of a method for generating a multi-language configuration file provided in this embodiment. This embodiment may include the following content:

[0022] S201: When the source language configuration file changes, locate the configuration change information according to the version change tree of the source language configuration file and the structure attribute information of the tree nodes.

[0023] Among them, the source language configuration file refers to the configuration file in which the configuration information has changed. This configuration file is written based on the source language, such as Chinese. In order to run in other different language and cultural environments, it is necessary to correspondingly update the configuration files in other languages. In this embodiment, the generation process of the language configuration file under this application is described by taking the generation of a language configuration file as an example. For the sake of convenience of description, the language configuration file to be newly generated is defined as the target language configuration file, and the target language configuration file is written based on the target language, such as English. The version change tree refers to the change situation of the version information structure tree recording the source language configuration file before and after the change. The tree node structure attribute information refers to all the attribute information of the changed node, indicating the position of the node in the tree. By determining the position of the node in the tree, the change tracking at the node level can be identified. The configuration change information refers to the different parts of the source language configuration file before and after the change. In this step, by comparing the tree structure and the node attributes, it is determined whether there are changes and the changed nodes, and then the changed data is determined.

[0024] S202: Determine the types and quantities of translation methods according to the target term ratio, literal grammar and translation volume of the configuration change information.

[0025] The translation capabilities of different translation methods vary. A translation method is a way of translating through a certain type of translation tool, translation engine, or translation plugin. For example, Translation Engine A can respond quickly, which is suitable for simple sentences or scenarios where a better generation rate of language configuration files is required. Translation Engine B translates professional terms accurately, and Translation Engine C has a better effect on matching professional terms and culture, but it responds slowly. Based on this, the present invention determines the complexity of sentence translation based on the proportion of terms, the grammar of the text, and the translation volume, and selects the most suitable translation method for it according to the complexity of sentence translation. The translation method includes the type of the selected translation tool, translation engine, or translation plugin and which translation methods are selected for translation. The target term proportion refers to the proportion of professional terms to the total number of words. The target term refers to a noun, phrase, or short sentence considered to be a professional term in the current translation. The text grammar is used to indicate whether the sentence structure corresponding to the configuration change information is complex, such as whether there are nested structures and the number of nested structures. Nested structures include nested object clauses, subject clauses, adverbial clauses, attributive clauses, conditional clauses, reason clauses, etc. For example, "The administrator (subject) found (predicate) that the running status of the RAID (Redundant Arrays of Independent Disks) 5 array was abnormal (the SVO (subject-verb-object) structure of the object)". The main clause is "The administrator found X", and the nested object clause is "The running status of the RAID 5 array was abnormal". The more nested structures there are, the more complex the sentence is. Any existing method can be used to determine the number of nested structures in the current translation sentence. For example, the benepar_zh model (model name) can be used to obtain the number of nested levels. The translation volume refers to the number of words that need to be translated. For example, some specified words do not need to be translated and some words are prohibited from being translated. The number of words that need to be translated or the number of words that do not need to be translated can be counted to measure the concept of translation volume.

[0026] S203: Translate the configuration change information using the selected translation method to obtain a single translation result. When there are at least two translation methods, according to the target term proportion, text grammar, translation volume, and preset scenario adjustment factor of the configuration change information, and based on the single translation result and the corresponding translation weight, determine the translation result of the configuration change information.

[0027] After determining the number of types of translation method selections in S202, according to the structure of the source language configuration file, parse and extract the configuration change information line by line as the content to be translated, and call the corresponding translation tool or translation engine or translation plugin to translate the configuration change information or pass it to the translation engine to obtain the corresponding translation result. For the convenience of distinction, it is defined as a single translation result. When the translation method selected in S202 is one kind, determine the translation result of the configuration change information according to the single translation result. When the translation method selected in S202 is multiple kinds, it is also necessary to further determine the weights of different translation methods. The weights also need to be determined according to the translation complexity of the configuration change information. Considering the translation capabilities of different translation methods for different translation scenarios, a preset scenario adjustment factor can be further set. The preset scenario adjustment factor can be used to adjust the term ratio and / or the sentence structure complexity. Finally, jointly determine the final translation result according to the single translation results of different translation methods and the corresponding weights.

[0028] S204: Update the target language configuration file corresponding to the source language configuration file according to the translation result.

[0029] The translation result in this step refers to the translation result corresponding to the configuration change information finally determined through the above steps. If the configuration change information is an addition, the translation result can be directly added to the corresponding position of the target language configuration file. If the configuration change information is a deletion, the original old content can be deleted at the corresponding position of the target language configuration file; if the configuration change information is a modification, the translation result can replace or overwrite the old content, or the content that has changed in the target language configuration file can be deleted first, and then the translation result can be copied to the target language configuration file, so as to only update the part of the target language configuration file that has changed, effectively improving the generation efficiency of the multi-language configuration file.

[0030] In the technical solution provided in this embodiment, the version information of the language configuration file is stored using a tree structure. By analyzing the version change tree, the changes in the configuration file can be automatically identified, achieving node-level change identification, and only translating the parts that have changed. By reducing redundant operations, the generation efficiency of the multi-language configuration file can be effectively improved, reducing the time required to generate the multi-language configuration from several hours / thousands of files in the related art to within dozens of minutes, and the generation speed of the multi-language configuration file is increased by more than 10 times. Select the most suitable translation method according to the sentence complexity of the changed part of the source language configuration file, which is beneficial to improving the accuracy of the finally generated multi-language configuration file. Further, for high-complexity situations, multiple translation methods are also supported. Utilize the translation capabilities that different translation methods are good at to make up for the problem that a single translation method may lead to inaccurate translation results due to insufficient understanding of the professional field or culture. Jointly determine the final translation result according to different translation methods and the corresponding weights, further improving the accuracy of the multi-language configuration file.

[0031] In the above embodiments, there is no limitation on how to locate the configuration change information according to the version change tree and the tree node structure attribute information of the source language configuration file. The present invention also gives a schematic implementation method, which may include the following contents:

[0032] Use an abstract syntax tree to construct a version information structure tree of the source language configuration file; obtain the new version information structure tree generated by the source language configuration file at the current moment and the old version information structure tree generated at the previous moment of the current moment; calculate the attribute hash values of each node of the new version information structure tree and the old version information structure tree respectively, and determine the changed nodes by comparing the attribute hash values of each node of the new version information structure tree and the old version information structure tree.

[0033] In this embodiment, an abstract syntax tree is used to represent the data structure of the source program code corresponding to the language configuration file. For the sake of convenience of description, it is defined as a version information structure tree. Whenever the source program code structure corresponding to the language configuration file changes, the corresponding version information structure tree also changes. When the source language configuration file is updated, such as adding a hard disk / mechanical hard disk node, by comparing the two version information structure trees, that is, the new version information structure tree and the old version information structure tree, the content that needs to change can be quickly located, so as to quickly determine the content that needs to be translated, such as change type: add node, path: hard disk / mechanical hard disk node, content: "mechanical hard disk status: ${status}". The nodes in the version information structure tree have structure attributes, that is, their positions in the tree. Calculate the hash values of all the attributes of the nodes, and the addition / modification / removal of the nodes can be identified by comparing the hash values of the nodes.

[0034] As can be seen from the above, by analyzing the abstract syntax tree of the source language configuration file, changes in the source language configuration file can be automatically identified like a version comparison tool by translating the source code. By comparing the hash values, the added or modified configuration items can be accurately identified for the added or modified nodes, and then only the changed parts are translated, without re-translating the entire file, reducing redundant operations and improving the generation efficiency and accuracy of the language configuration file.

[0035] Furthermore, this embodiment also gives a construction method of the version information structure tree, which may include the following contents:

[0036] Obtain each file fragment of the source language configuration file, and determine the root node, sub-nodes, and leaf nodes of the version information structure tree corresponding to the source language configuration file according to the ownership relationship and hierarchical relationship between different objects included in each file fragment; for each sub-node, generate corresponding text translation tags and / or parameter placeholder tags according to the node attribute information of the current sub-node, and generate corresponding structural relationship tags according to the hierarchical relationship and connection relationship between the current sub-node and other sub-nodes, and the connection relationship between the current sub-node and each leaf node; for each leaf node, generate corresponding text translation tags and / or parameter placeholder tags according to the node attribute information of the current leaf node, and generate corresponding structural relationship tags according to the hierarchical relationship between the current leaf node and other leaf nodes, and the connection relationship between the current leaf node and each sub-node; traverse each sub-node and each leaf node, and automatically record the position information of the target node with text translation tags in the version information structure tree.

[0037] In this embodiment, a file fragment refers to a code fragment of a language configuration file, such as JSON (JavaScript Object Notation, object notation). File fragments can be used to describe the hardware of a server, such as processor type, number of processors, hard disk, network card, power supply, operating system, such as the current operating system name, version number, installation date, patch list and patch date, software environment, such as various databases, various server types, application version, relevant configuration parameter information, such as network configuration, storage configuration, system parameters, virtualization environment, etc. Different file fragments describe different objects, and the objects include entity hardware or software environment or parameter values in the above content. According to the structure of the server, there are hierarchical relationships, parallel relationships, ownership relationships, etc. between these objects. The present invention uses a tree structure to represent, and the tree structure includes a root node, branch nodes, and leaf nodes. The root node is the top node of the tree, and all other nodes start from the root node. A branch node refers to a node other than a leaf node, and a branch node has at least one child node. A leaf node refers to a node without a child node and is located at the end of the tree. The connection relationship between nodes represents the relationship between corresponding data. Based on this, the present invention determines the root node, sub-nodes, and leaf nodes of the version information structure tree through the association relationship between different objects included in each file fragment. For example, the JSON fragment is: , correspondingly, ROOT (root node) is disk (hard disk), and there are 2 branch nodes, one is solid state drive (ssd), and the other is raid.

[0038] After the tree structure is constructed, in order to accurately locate the changed content, corresponding tags are also defined for different nodes as the structural attribute information of the tree nodes. The text translation tag is used to identify the corresponding text segment as the text content to be translated participating in the translation. When a node carries this type of tag, it is an ordinary text node of the regular text that needs to be translated. For example, "available space" will automatically record its position in the tree, such as the path: disk / ssd / text. The parameter placeholder tag is used to identify the corresponding text segment as the parameter symbol content that does not need to participate in the translation. When a node carries this type of tag, it is a parameter placeholder node, marked with a special symbol (such as ), and the system will lock these parameters from participating in the translation to avoid incorrect modification. The structure relationship tag is used to identify the parent-child relationship and sibling relationship between nodes. For example, raid is a child node of disk, ensuring that the translated file maintains the original structure. Based on the three types of tags, the version information structure tree after the conversion of the above JSON segment can be expressed as: .

[0039] As can be seen from the above, in this embodiment, by modeling the file structure with three types of tags, it is possible to lock the parameters that do not participate in the translation, avoid incorrect modification, which is beneficial to improving the translation accuracy. Through the structure relationship tag, it can ensure that the translated file maintains the original structure, effectively improving the translation accuracy and the generation accuracy of the multilingual configuration file.

[0040] Based on the above embodiment, in order to further improve the accuracy of the translation result, the present invention also provides the following embodiments, which may include the following content: Pre-construct a cultural adaptation library, a terminology library, and a corresponding terminology lock; during the process of translating the configuration change information using the selected target translation method, if it is identified based on the terminology lock that the current word to be translated is in the prohibited translation list, then no translation is required; if it is identified based on the terminology lock that the current word to be translated is in the review list, then a match is made in the terminology library according to the preset matching rules, and the target translation word obtained from the match is sent to the pre-bound client for review; if the match in the terminology library fails, the current word will be translated using the target translation method, and the target translation word obtained from the translation is sent to the pre-bound client for review; if it is identified based on the terminology lock that the current word to be translated is in the cultural adaptation list, then the translation is performed based on the cultural adaptation library.

[0041] Among them, the target translation method is the currently selected translation method. The term library can be used to determine the number of target terms. The term library can be constructed in advance according to different application scenarios. The term library includes multiple standard words, and each standard word includes its corresponding general expressions in different translation languages. When a word that does not appear in the term library is translated for the first time, such as "SSD" → "solid state drive", it can be automatically recorded in the term library. When translating, the matching items are retrieved from the term library according to the preset matching rules. The preset matching rules include the exact matching rule and the context matching rule. For example, when "SSD" is encountered again, according to the exact matching rule, "SSD" can directly retrieve "solid state drive", and according to the context matching rule, such as "SSD performance" → "solid state drive performance". The term lock constructs a three-level term protection mechanism, as shown in Table 1 below, which can include a prohibited translation list, a review list, and a cultural adaptation list. The rules for different lists are different. The cultural adaptation library includes at least parameter formats, parameter symbols, and professional terms that match the target translation language. For example, if the target translation language is English, for the parameter format, it can include the date format. For example, the source language parameter format: "2025 / 03 / 07" → the target language parameter format "07 / 03 / 2025", and the parameter symbol such as the currency symbol. For example, the source language parameter format: "¥100" → the target language parameter format "€8.50". Professional terms refer to the concepts of the same name in different cultures.

[0042] Table 1 Example Table of Term Lock

[0043] As can be seen from the above, in this embodiment, the accuracy of the translation result is ensured through the combination of the word library lock, the three-level term protection mechanism, and cultural adaptation, which is beneficial to improving the adaptation of the translation result to different professional fields and different cultural scenarios and improving the generation accuracy of the multilingual configuration file.

[0044] The above embodiments do not make any restrictions on how to integrate multiple translation results. The present invention also provides an implementation method for determining the final translation result in the scenario of multiple translation methods, which may include the following content:

[0045] Determine the corresponding translation benchmark weights according to the translation performance of different translation methods; determine the translation scenario weights according to the target term density, literal grammar, translation volume, and preset statement complexity adjustment factor of the configuration change information; under the translation weight constraint conditions, adjust the translation benchmark weights corresponding to different translation methods according to the translation scenario weights to obtain the translation weights corresponding to different translation methods; select the maximum similarity from the similarities between the translation results corresponding to every two translation methods, and determine the multi-translation method difference degree according to the difference between the maximum similarity and the preset value; when the multi-translation method difference degree is less than the first preset difference threshold, use the single translation result of the translation method corresponding to the maximum value among the translation weights as the translation result of the configuration change information; when the multi-translation method difference degree is greater than or equal to the first preset difference threshold and less than or equal to the second preset difference threshold, fuse the single translation results of each translation method and use the fused translation as the translation result of the configuration change information. When the multi-translation method difference degree is greater than the second preset difference threshold, trigger manual review and make a mark.

[0046] In this embodiment, the translation benchmark weights are the basic weights of different translation methods. For example, the translation benchmark weight W1 of translation method A with strong generality is 0.55, the translation benchmark weight W2 of translation method B with professional term advantages is 0.45, and the translation benchmark weight W3 of translation method C with response speed advantages is 0.25. The translation scenario weights are the weights determined based on the current translation situation. The target term density is determined according to the total number of target terms, the total number of words in the configuration change information, and the preset term adjustment factor. Exemplarily, the target term density = total number of target terms / total number of words × preset term adjustment factor α, and the preset term adjustment factor α can be flexibly selected according to the actual situation. For example, α = 0.3, and the preset statement complexity β can be flexibly selected according to the actual situation. For example, β = 0.2. Correspondingly, the translation scenario weight = 1 + α × term density + β × (target term proportion + statement complexity + translation volume). Exemplarily, the target term proportion + statement complexity + translation volume jointly determine the translation complexity of the configuration change information. The statement complexity can be represented by the nesting number, and the translation volume can be represented by the proportion of special symbols that do not need to be translated. In this embodiment, the translation complexity C of the configuration change information can also be calculated by calling the relational expression C = C1 × (number of target terms / total number of words) + C2 × number of nesting layers + C3 × proportion of special symbols. C1, C2, and C3 are the adjustment coefficients of the corresponding items and can be flexibly selected according to the actual situation. The present invention does not make any limitation on this. For example, C1 = 0.4, C2 = 0.3, C3 = 0.3. Correspondingly, the translation complexity C of the configuration change information can be calculated based on C = 0.4 × (number of target terms / total number of words) + 0.3 × number of nesting layers + 0.3 × proportion of special symbols. The sentence cost such as 、 、 、 、 , , the numerical value of the translation complexity C can quantitatively represent the complexity. For a sentence with a subject-predicate-object structure, such as "The SSD has excellent performance", which belongs to a simple sentence, the value of the translation complexity C is between 0 and 0.3. For a sentence with a modifier structure based on the subject-predicate-object, such as "The SSD performance under extreme temperatures", which belongs to a regular sentence, the value of the translation complexity C is between 0.3 and 0.6. For a sentence with a nested structure, such as "When the RAID5 array operates in the degraded mode", which belongs to a complex sentence, the value of the translation complexity C is above 0.6. Different values of the translation complexity C represent different translation complexities, and different translation methods are determined according to different translation complexities, as shown in Table 2.

[0047] Table 2 Translation Complexity and Translation Method Selection

[0048] Among them, the target term number / total word number is the proportion of the target term. The translation weight constraint conditions refer to the upper and lower limits of the weights of different translation methods and the total weight of all translation methods. For example, the upper limit of the weight of a translation method can be 0.55, and the lower limit can be 0.15, and the sum of the total weights of all translation methods remains 1. In a simple way, the translation weight can be directly obtained by multiplying the translation scenario weight and the translation benchmark weight. For example, the translation weight = translation benchmark weight × (1 + α × term density + β × C). When the adjusted translation weight exceeds the upper limit, the upper limit value is taken; when it is lower than the lower limit, the lower limit value is taken.

[0049] In this embodiment, different fusion methods are selected according to the difference degrees of different translation methods. First, the maximum similarity is selected from the similarities of the corresponding translation results of every two translation methods. For example, taking the selected translation methods including three translation methods A, B, and C as an example, using any similarity calculation method, such as using Jaro-Winkler (the name of the string similarity calculation method) for similarity calculation, which can be represented by Sim, calculate the similarity values of the single translation results output by every two translation methods, Sim(translation method A, translation method B), Sim(translation method A, translation method C), Sim(translation method B, translation method C), and then select the maximum similarity value from them, that is, max(Sim(translation method A, translation method B), Sim(translation method A, translation method C), Sim(translation method B, translation method C)). The difference degree of multiple translation methods is determined according to the difference between the maximum similarity and the preset value M. The preset value can be taken as 1, 1.5, 2, 3, etc. The present invention does not make any limitation on this. Correspondingly, the difference degree D of multiple translation methods = M - max(Sim(translation method A, translation method B), Sim(translation method A, translation method C), Sim(translation method B, translation method C)). The first preset difference threshold and the second preset difference threshold can be flexibly selected according to the actual situation. Taking 0.3 and 0.7 as an example, D < 0.3: Automatically select the translation result output by the translation method with the highest weight. 0.3 ≤ D ≤ 0.7: Integrate the single translation results of the selected translation methods to generate a fused translation as a more efficient method. The configuration change information is divided into multiple segments, and for the translation results of different segments, the currently optimal translation results can be selected from all the selected translation methods. A pre-set evaluation criterion can be set. For example, for the same segment, if the translations are quite different from each other, then the one with the highest weight is taken as the optimal, or according to the type to which the segment belongs, select the result of the translation method that has an advantage in translating this type, and then integrate the optimal translation segments of each segment into the final translation result. D > 0.7 indicates that the accuracy of the translation method may not be high, then trigger manual review and mark the controversial points, such as marking as an error, as Figure 3 shown

[0050] In this embodiment, the weights of different translation methods are adjusted in combination with different scenarios, so that the translation accuracy is more in line with the current scenario, which can improve the translation accuracy. Fusing according to the difference degrees of different translation methods preferentially further improves the accuracy of the translation result, which is beneficial to improving the accuracy of the multilingual configuration file.

[0051] In order to further improve the accuracy of the translation result, after obtaining the translation result in the above embodiment, translation verification and translation feedback can also be carried out, which can include the following contents:

[0052] Perform grammar checking, semantic checking, and cultural difference checking on the translation result. If the translation result passes all of the grammar checking, semantic checking, and cultural difference checking, add the translation result and the configuration items of the corresponding source language configuration file as candidate standard terms to the candidate term library. If the translation result does not pass all of the grammar checking, semantic checking, and cultural difference checking, send the translation result to the client to trigger manual review, and update or add the standard terms based on the manually corrected result and the configuration items of the corresponding source language configuration file to the term library. When the preset manual review conditions are met, such as when the number of candidate standard terms in the candidate term library exceeds the preset threshold, send the candidate term library to the client to trigger manual review, and update or add the standard terms based on the manually reviewed and corrected result and the configuration items of the corresponding source language configuration file to the term library.

[0053] Among them, grammar checking is used to check whether the translation result conforms to the grammar rules of the target language, and semantic checking is used to check whether the target terms in the translation result are semantically consistent with the original text. The process of cultural difference checking: Obtain the cultural adaptation library corresponding to the target translation language; Use Unicode CLDR v43 (a data repository for languages and regions) to query the translation result data in real time, verify the translation result based on the cultural adaptation library, and correct the parameter formats, parameter symbols, or professional terms that do not conform to the target translation language. For example, if the SSD in the translation result is translated as hard disk, and the context is storage and device, the manually corrected result or the manually reviewed result is solid state drive, add it to the term library, and when "SSD" appears again and the context contains "storage" or "device", replace it with "solid state drive".

[0054] As can be seen from the above, in this embodiment, by performing multiple checks on the translation result, and through the checking of professional terms and cultural differences, it is beneficial to improve the accuracy of professional translation, solve the problem of misunderstandings caused by cultural differences between different languages, and improve the accuracy of the multilingual configuration file. At the same time, expand and correct the term library through manual review and review results, further improve the accuracy of a single translation result, and improve the accuracy of the multilingual configuration file.

[0055] Based on the above embodiments, in order to further improve the accuracy of the translation result, the present invention also provides various implementation methods for determining the final translation result according to the verification result, which may include the following:

[0056] As an exemplary verification implementation method, a language benchmark threshold is determined according to the difference degree between the target translation language and the source language corresponding to the source language configuration file; a scenario threshold is determined according to the target term density, literal grammar, translation volume and a preset statement complexity adjustment factor of the configuration change information; a semantic weight is determined according to at least one target sentence component and the corresponding component weight in the configuration change information; a back-translation accuracy recognition condition is determined according to the scenario threshold and the language benchmark threshold; the translation result is back-translated from the target translation language back to the source language based on the semantic weight to obtain back-translation configuration information; when the difference between the back-translation configuration information and the configuration change information meets the back-translation accuracy recognition condition, the target language configuration file corresponding to the source language configuration file is updated according to the translation result.

[0057] In this embodiment, different language benchmark thresholds are set according to the grammatical difference degree between the target translation language and the source language. Taking the source language as Chinese, as shown in Table 3, the language benchmark threshold for Japanese can take a value greater than 82%, the language benchmark threshold for Arabic can take a value greater than 78%, and the language benchmark threshold for English can take a value greater than 85%.

[0058] Table 3 Language Benchmark Threshold Table

[0059] In this embodiment, the scenario threshold is a threshold set according to different scenarios. The target term density is determined according to the target term ratio and a preset term adjustment factor. The scenario threshold can be expressed as, for example, α×target term density + β×translation complexity C. The translation complexity C can be calculated according to the method described in the above embodiment. Among them, α and β can take different values, such as 0.1 and 0.05. Correspondingly, as a simple representation method, the minimum value of the back-translation accuracy recognition condition is the sum of the two, that is, the back-translation accuracy recognition condition is greater than or equal to the language benchmark threshold + the scenario threshold. A dependency syntactic analyzer can be used to identify the sentence components in the translation result, such as the subject, predicate, and object. The subject and object, as noun phrases NP, are given a certain weight, the predicate VP, as a verb phrase, is given a certain weight, and the modifier is given the lowest weight. For example, the semantic weight = 0.6×NP + 0.3×VP + 0.1×modifier.

[0060] As another verification method, verification can be performed through structural consistency. In the source version information structure tree of the source language configuration file, the source root node and the source modification relationship tree are determined; in the target version information structure tree of the updated target language configuration file, the target root node and the target modification relationship tree are determined; according to the source root node, the source modification relationship tree, the target root node and the target modification relationship tree, the tree structure similarity is determined; when the tree structure similarity meets the preset similarity condition, the updated target language configuration file is used as the latest language configuration file of the target translation language at the current moment.

[0061] In this embodiment, the ROOT nodes of the source language and the target language are identified, and the modified relationship tree is compared. The modified relationship formula includes a left subtree and a right subtree. The preset similarity condition can be: the tree structure similarity is at least greater than 90±△, where △ represents the fluctuation value. When the tree structure similarity does not meet the preset similarity condition, the different nodes are marked, and manual review is initiated. At the same time, correction suggestions are generated, such as adjusting the word order, adjusting according to the manual correction suggestions, and updating the target language configuration file based on the adjusted translation result as the translation result.

[0062] As can be seen from the above, this embodiment accurately identifies the translation results with incorrect word translations from the reverse translation verification, and can accurately identify the translation results with problems in word order through the node tree alignment verification, improving the accuracy of the translation results, facilitating the timely adjustment of the translation results, and being beneficial to improving the accuracy and efficiency of the multi-language configuration file.

[0063] The present invention also provides a corresponding device for the multi-language configuration file generation method, further making the method more practical. Among them, the device can be described from the perspective of functional modules and the perspective of hardware. The multi-language configuration file generation device provided by the present invention will be introduced below. This device is used to implement the multi-language configuration file generation method provided by the present invention. The following description will specifically introduce the functions of each program module in this embodiment. The multi-language configuration file generation device described below can be correspondingly referred to the multi-language configuration file generation method described above.

[0064] From the perspective of functional modules, see Figure 4 , Figure 4 is a structural diagram of the multi-language configuration file generation device provided in this embodiment in a specific implementation manner. The device may include:

[0065] The incremental compilation module 401 is used to locate the configuration change information according to the version change tree of the source language configuration file and the tree node structure attribute information when the source language configuration file changes.

[0066] The multi-engine fusion translation module 402 is used to determine the types and quantities of translation methods according to the target term ratio, literal grammar, and translation volume of the configuration change information, and use the selected translation methods to translate the configuration change information to obtain a single translation result; when there are at least two translation methods, according to the target term ratio, literal grammar, translation volume, and preset scenario adjustment factor of the configuration change information, and according to the single translation result and the corresponding translation weight, determine the translation result of the configuration change information; update the target language configuration file corresponding to the source language configuration file according to the translation result.

[0067] Exemplarily, in some embodiments of this embodiment, the above incremental compilation module 401 can also be used to: construct a version information structure tree of the source language configuration file using an abstract syntax tree; obtain the new version information structure tree generated by the source language configuration file at the current moment and the old version information structure tree generated at the previous moment of the current moment; calculate the attribute hash values of each node of the new version information structure tree and the old version information structure tree respectively, and determine the changed nodes by comparing the attribute hash values of each node of the new version information structure tree and the old version information structure tree.

[0068] Exemplarily, in some other embodiments of this embodiment, the above multi-engine fusion translation module 402 can also be used to: determine a language benchmark threshold according to the difference degree between the target translation language and the source language corresponding to the source language configuration file; determine a scenario threshold according to the target term density, literal grammar, translation volume and a preset statement complexity adjustment factor of the configuration change information; the target term density is determined according to the target term ratio and a preset term adjustment factor; determine a semantic weight according to at least one target sentence component and the corresponding component weight in the configuration change information; determine a back-translation accuracy recognition condition according to the scenario threshold and the language benchmark threshold; re-translate the translation result from the target translation language back to the source language based on the semantic weight to obtain back-translation configuration information; when the difference between the back-translation configuration information and the configuration change information meets the back-translation accuracy recognition condition, update the target language configuration file corresponding to the source language configuration file according to the translation result.

[0069] Exemplarily, in some other embodiments of this embodiment, the above multi-engine fusion translation module 402 can also be used to: determine a source root node and a source modification relationship tree in the source version information structure tree of the source language configuration file; determine a target root node and a target modification relationship tree in the target version information structure tree of the updated target language configuration file; determine a tree structure similarity according to the source root node, the source modification relationship tree, the target root node and the target modification relationship tree; when the tree structure similarity meets a preset similarity condition, use the updated target language configuration file as the latest language configuration file of the target translation language at the current moment.

[0070] Exemplarily, in some other embodiments of the present embodiment, the above multi-engine fusion translation module 402 can also be used to: determine corresponding translation benchmark weights according to the translation performance of different translation methods; determine translation scenario weights according to the target term density, literal grammar, translation volume, and preset statement complexity adjustment factor of the configuration change information; wherein, the professional term density is determined according to the total number of professional terms, the total number of words in the configuration change information, and a preset correction coefficient; under the translation weight constraint conditions, adjust the translation benchmark weights corresponding to different translation methods according to the translation scenario weights to obtain the translation weights corresponding to different translation methods; select the maximum similarity from the similarities between the translation results corresponding to every two translation methods, and determine the multi-translation method difference degree according to the difference between the maximum similarity and a preset value; when the multi-translation method difference degree is less than the first preset difference threshold, use the single translation result of the translation method corresponding to the maximum value among the translation weights as the translation result of the configuration change information; when the multi-translation method difference degree is greater than or equal to the first preset difference threshold and less than or equal to the second preset difference threshold, fuse the single translation results of each translation method, and use the fused translation as the translation result of the configuration change information.

[0071] Exemplarily, in some embodiments of the present embodiment, the above incremental compilation module 401 may further include a tree construction module, and this module can be used to: obtain each file fragment of the source language configuration file, and determine the root node, molecular nodes, and leaf nodes of the version information structure tree corresponding to the source language configuration file according to the ownership relationship and hierarchical relationship between different objects included in each file fragment; for each molecular node, generate corresponding literal translation tags and / or parameter placeholder tags according to the node attribute information of the current molecular node, and generate corresponding structure relationship tags according to the hierarchical relationship and connection relationship between the current molecular node and other molecular nodes, and the connection relationship between the current molecular node and each leaf node; wherein, the literal translation tag is used to identify that the corresponding text fragment is the text content to be translated participating in the translation, the parameter placeholder tag is used to identify that the corresponding text fragment is the parameter symbol content that does not need to participate in the translation, and the structure relationship tag is used to identify the parent-child relationship and sibling relationship between nodes; for each leaf node, generate corresponding literal translation tags and / or parameter placeholder tags according to the node attribute information of the current leaf node, and generate corresponding structure relationship tags according to the hierarchical relationship between the current leaf node and other leaf nodes, and the connection relationship between the current leaf node and each molecular node; traverse each molecular node and each leaf node, and automatically record the position information of the target nodes with literal translation tags in the version information structure tree.

[0072] Exemplarily, in some other embodiments of this embodiment, the above multi-engine fusion translation module 402 can also be used to: pre-construct a cultural adaptation library, a terminology library, and corresponding terminology locks; the cultural adaptation library at least includes parameter formats, parameter symbols, and professional terms that match the target translation language; the terminology locks include a prohibited translation list, a review list, and a cultural adaptation list; the terminology library includes multiple standard words, and each standard word includes its general expression in different translation languages; during the process of translating the configuration change information using the selected target translation method, if it is identified based on the terminology lock that the current word to be translated is in the prohibited translation list, then no translation is required; if it is identified based on the terminology lock that the current word to be translated is in the review list, then a match is made in the terminology library according to the preset matching rules, and the target translation word obtained by the match is sent to the pre-bound client for review; if the match in the terminology library fails, the current word will be translated using the target translation method, and the target translation word obtained by the translation is sent to the pre-bound client for review; if it is identified based on the terminology lock that the current word to be translated is in the cultural adaptation list, then the translation is performed based on the cultural adaptation library.

[0073] The multi-language configuration file generation device mentioned above is described from the perspective of functional modules. Further, the present invention also provides an electronic device, which is described from the perspective of hardware. Figure 5 FIG. 5 is a schematic structural diagram of an electronic device provided in an embodiment of the present invention. The electronic device includes a memory 501 and a processor 502. A computer program is stored in the memory 501, and the processor 502 is configured to run the computer program to execute the steps in any of the above embodiments of the multi-language configuration file generation method.

[0074] An embodiment of the present application also provides a computer-readable storage medium. A computer program is stored in the computer-readable storage medium, where the computer program is configured to execute the steps in any of the above embodiments of the multi-language configuration file generation method when running.

[0075] In an exemplary embodiment, the above computer-readable storage medium may include, but is not limited to: USB flash drives, read-only memories (ROM for short), random access memories (RAM for short), external hard drives, magnetic disks, or optical discs, etc., various media that can store computer programs.

[0076] An embodiment of the present application also provides a computer program product. The above computer program product includes a computer program, and when the computer program is executed by a processor, it implements the steps in any of the above embodiments of the multi-language configuration file generation method.

[0077] Embodiments of the present application also provide another computer program product, including a non-volatile computer-readable storage medium storing a computer program, which when executed by a processor implements the steps in any of the above-described embodiments of the multi-language configuration file generation method.

[0078] The above has introduced in detail a multi-language configuration file generation method, apparatus, electronic device, and computer-readable storage medium provided by the present invention. Each embodiment in this specification is described in a progressive manner, and the key point of each embodiment is the difference from other embodiments. The same or similar parts between the embodiments can be referred to each other. Whether the units and algorithm steps of each example described in the disclosed embodiments are executed in the form of electronic hardware or computer software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, and such implementation should not be considered to exceed the scope of the present invention. Without departing from the principle of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the protection scope of the present invention.

Claims

1. A method for generating a multilingual configuration file, characterized in that, Including: When the source language configuration file changes, locate the configuration change information according to the version change tree and tree node structure attribute information of the source language configuration file; The version change tree represents the difference information of the version information structure tree corresponding to the source language configuration file before and after the change; According to the target term ratio, literal grammar and translation volume of the configuration change information, determine the types and quantities of translation methods, and use the selected translation methods to translate the configuration change information to obtain a single translation result; When there are at least two translation methods, determine the translation weights of different translation methods according to the target term ratio, literal grammar, translation volume and preset scenario adjustment factor of the configuration change information, and determine the translation result of the configuration change information according to the single translation result and the corresponding translation weights; Update the target language configuration file corresponding to the source language configuration file according to the translation result.

2. The method for generating a multilingual configuration file according to claim 1, wherein The locating the configuration change information according to the version change tree and tree node structure attribute information of the source language configuration file includes: Construct the version information structure tree of the source language configuration file by using the abstract syntax tree; Obtain the new version information structure tree generated by the source language configuration file at the current moment and the old version information structure tree generated at the previous moment of the current moment; Calculate the attribute hash values of each node of the new version information structure tree and the old version information structure tree respectively, and determine the changed nodes by comparing the attribute hash values of each node of the new version information structure tree and the old version information structure tree.

3. The multi-language profile generation method according to claim 1, wherein The updating the target language configuration file corresponding to the source language configuration file according to the translation result includes: Determine the language benchmark threshold according to the difference degree between the target translation language and the source language corresponding to the source language configuration file; Determine the scenario threshold according to the target term density, literal grammar, translation volume and preset statement complexity adjustment factor of the configuration change information; the target term density is determined according to the target term ratio and the preset term adjustment factor; Determine the semantic weight according to at least one target sentence component and the corresponding component weight in the configuration change information; Determine the back translation accuracy recognition condition according to the scenario threshold and the language benchmark threshold; Translate the translation result back to the source language from the target translation language based on the semantic weight to obtain the back translation configuration information; When the difference between the back translation configuration information and the configuration change information meets the back translation accuracy recognition condition, update the target language configuration file corresponding to the source language configuration file according to the translation result.

4. The multi-language profile generation method according to claim 1, wherein The updating the target language configuration file corresponding to the source language configuration file according to the translation result includes: Determine the source root node and the source modification relationship tree in the source version information structure tree of the source language configuration file; Determine the target root node and the target modification relationship tree in the target version information structure tree of the updated target language configuration file; Determine the tree structure similarity according to the source root node, the source modification relationship tree, the target root node and the target modification relationship tree. When the tree structure similarity meets the preset similarity condition, the updated target language configuration file is used as the latest language configuration file of the target translation language at the current moment.

5. The multi-language configuration file generation method according to claim 1, characterized in that Determining the translation result of the configuration change information according to the single translation result and the corresponding translation weight includes: Determining the corresponding translation benchmark weight according to the translation performance of different translation methods; Determining the translation scenario weight according to the target term density, literal grammar, translation volume, and preset statement complexity adjustment factor of the configuration change information; Under the translation weight constraint condition, adjusting the translation benchmark weights corresponding to different translation methods according to the translation scenario weight to obtain the translation weights corresponding to different translation methods; Selecting the maximum similarity from the similarities between the translation results corresponding to every two translation methods, and determining the multi-translation method difference degree according to the difference between the maximum similarity and the preset value; When the multi-translation method difference degree is less than the first preset difference threshold, using the single translation result of the translation method corresponding to the maximum value among the translation weights as the translation result of the configuration change information; When the multi-translation method difference degree is greater than or equal to the first preset difference threshold and less than or equal to the second preset difference threshold, fusing the single translation results of each translation method, and using the fused translation as the translation result of the configuration change information.

6. The multi-language profile generation method according to claim 1, wherein Before locating the configuration change information according to the version change tree and node structure attribute information corresponding to the source language configuration file, it includes: Obtaining each file segment of the source language configuration file, and determining the root node, molecular nodes, and leaf nodes of the version information structure tree corresponding to the source language configuration file according to the ownership relationship and hierarchical relationship between different objects included in each file segment; For each molecular node, generating corresponding literal translation labels and / or parameter placeholder labels according to the node attribute information of the current molecular node, and generating corresponding structure relationship labels according to the hierarchical relationship and connection relationship between the current molecular node and other molecular nodes, and the connection relationship between the current molecular node and each leaf node; wherein, the literal translation label is used to identify that the corresponding text segment is the text content to be translated participating in the translation, the parameter placeholder label is used to identify that the corresponding text segment is the parameter symbol content that does not need to participate in the translation, and the structure relationship label is used to identify the parent-child relationship and sibling relationship between nodes; For each leaf node, generating corresponding literal translation labels and / or parameter placeholder labels according to the node attribute information of the current leaf node, and generating corresponding structure relationship labels according to the hierarchical relationship between the current leaf node and other leaf nodes, and the connection relationship between the current leaf node and each molecular node; Traversing each molecular node and each leaf node, and automatically recording the position information of the target node with a literal translation label in the version information structure tree.

7. The multi-language profile generation method according to any one of claims 1 to 6, characterized in that Translating the configuration change information using the selected translation method to obtain a single translation result, including: Pre-build a cultural adaptation library, a terminology library, and corresponding terminology locks; the cultural adaptation library at least includes parameter formats, parameter symbols, and professional terms that match the target translation language; the terminology locks include a no-translation list, a review list, and a cultural adaptation list; the terminology library includes multiple standard words, and each standard word includes its general expressions corresponding to different translation languages; During the process of translating the configuration change information using the selected target translation method, if it is identified based on the terminology lock that the current word to be translated is in the no-translation list, then no translation is required; If it is identified based on the terminology lock that the current word to be translated is in the review list, then a match is made in the terminology library according to the preset matching rules, and the target translation word obtained from the match is sent to the pre-bound client for review; if the match in the terminology library fails, the current word will be translated using the target translation method, and the translated target translation word will be sent to the pre-bound client for review; If it is identified based on the terminology lock that the current word to be translated is in the cultural adaptation list, then the translation is performed based on the cultural adaptation library.

8. A multilingual configuration file generation device, characterized in that, It includes: An incremental compilation module, which is used to locate configuration change information according to the version change tree and tree node structure attribute information of the source language configuration file when the source language configuration file changes; the version change tree represents the difference information of the version information structure tree corresponding to the source language configuration file before and after the change; A multi-engine fusion translation module, which is used to determine the types and quantities of translation methods according to the target term ratio, text grammar, and translation volume of the configuration change information, and use the selected translation method to translate the configuration change information to obtain a single translation result; when there are at least two translation methods, according to the target term ratio, text grammar, translation volume, and preset scenario adjustment factor of the configuration change information, and according to the single translation result and the corresponding translation weight, determine the translation result of the configuration change information; Update the target language configuration file corresponding to the source language configuration file according to the translation result.

9. An electronic device, characterized in that, It includes: A memory, which is used to store computer programs; A processor, which is used to implement the steps of the multi-language configuration file generation method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is executed by the processor, it implements the steps of the multi-language configuration file generation method according to any one of claims 1 to 7.

Images