Simple Translate App with OpenAI gpt api
Create a Chinese index JSON file that stores all translated texts, and check if a corresponding translation exists by reading this file
/**
* core
**/
type Lang = string;
interface IndexJsonFile {
[copyWriting: string]?: {
lang: Lang;
index: number;
}[];
}
/**
* Relationship between application and copywriting
**/
interface AppCWsFile {
[app: string]?: keyof IndexJsonFile[];
}Considering the performance limitations of Steno, split specific languages into files less than 100MB, with the split rule as {language}-{index}.json
type CN = string;
type I18n = string;
/**
* file structure
*
* fileName: {I18n}-{index}.json
*
**/
interface CopyWritingFile {
[text: CN]?: I18n;
}graph LR
Input(Input the list of text to be translated)--> Check(Check if the text to be translated exists)
Check --> Exist(Exists) --> Ignore[Ignore]
Check --> NotExist(Not exists) --> GPT(AI Translation)
GPT --> LangChain(Generate the request text based on the pre-built Prompt) --> Parser(Parse the API response format)
Parser --> Write[Write to JSON]
graph LR
Input(Input the application name)--> GetCWs(Get all copywritings under the current application)
GetCWs --> GetCWsFile(Get the corresponding storage file of all copywritings)
GetCWsFile --> Set(Deduplicate based on language and index)
Set --> GetFileHandler(Get the handler of the corresponding file) --> Temp(Create a temporary folder)
Temp --> Group(Generate I18n files based on language)
Group --> Group
Group --> Send(Send files under Temp back to the client)
Send --> End[Clean up Temp]