使用GARbro 提取游戏的
- s文件放入到script目录中
- ogg文件文件放入 voice 中
Last active
October 3, 2022 06:45
-
-
Save kasuganosora/25f4df2b6f520c2981389117fbeb13cb to your computer and use it in GitHub Desktop.
提取近月少女的礼仪的语音和脚本变成vits的训练语料
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| package main | |
| import ( | |
| "fmt" | |
| "github.com/faiface/beep" | |
| "github.com/faiface/beep/vorbis" | |
| "github.com/faiface/beep/wav" | |
| "golang.org/x/text/encoding/japanese" | |
| "golang.org/x/text/transform" | |
| "io" | |
| "log" | |
| "math/rand" | |
| "os" | |
| "path/filepath" | |
| "regexp" | |
| "strconv" | |
| "strings" | |
| ) | |
| func main() { | |
| if len(os.Args) < 2 { | |
| log.Println("请输入角色名, 多个角色用 _ 分割") | |
| return | |
| } | |
| characterName := os.Args[1] | |
| characterId := -1 | |
| if len(os.Args) > 2 { | |
| characterId, _ = strconv.Atoi(os.Args[2]) | |
| } | |
| // 当前目录下面创建DUMMY_WAV文件夹 | |
| pwd, _ := os.Getwd() | |
| // 当前文件夹新建 output 文件夹 | |
| outputDir := filepath.Join(pwd, "output") | |
| if _, err := os.Stat(outputDir); os.IsNotExist(err) { | |
| _ = os.Mkdir(outputDir, 0755) | |
| } | |
| dummyWavDir := filepath.Join(outputDir, "DUMMY_WAV") | |
| if _, err := os.Stat(dummyWavDir); os.IsNotExist(err) { | |
| _ = os.Mkdir(dummyWavDir, 0755) | |
| } | |
| scripts := getScripts(characterName) | |
| log.Println("共有", len(scripts), "条台词, 现在进行音频转换并且写入文件列表...") | |
| hashMap := make(map[string]bool) // 去除重复的台词 | |
| trainFilelist := make([]string, 0) | |
| testFilelist := make([]string, 0) | |
| for _, script := range scripts { | |
| if _, ok := hashMap[script.Text]; ok { | |
| continue | |
| } | |
| hashMap[script.Text] = true | |
| oggFileName := filepath.Join(pwd, "voice", script.Voice+".ogg") | |
| wavFileName := filepath.Join(dummyWavDir, script.Voice+".wav") | |
| if err := oggToWav(oggFileName, wavFileName); err != nil { | |
| log.Println(err) | |
| continue | |
| } | |
| var fileLine string | |
| if characterId == -1 { | |
| fileLine = fmt.Sprintf("DUMMY_WAV/%s.wav|%s", script.Voice, script.Text) | |
| } else { | |
| fileLine = fmt.Sprintf("DUMMY_WAV/%s.wav|%d|%s", script.Voice, characterId, script.Text) | |
| } | |
| // 随机80%的数据用于训练, 20%的数据用于测试 | |
| randNum := rand.Intn(100) | |
| if randNum < 80 { | |
| trainFilelist = append(trainFilelist, fileLine) | |
| } else { | |
| testFilelist = append(testFilelist, fileLine) | |
| } | |
| } | |
| if err := saveToFile(filepath.Join(outputDir, characterName+"_train_filelist.txt"), strings.Join(trainFilelist, "\n")); err != nil { | |
| panic(err) | |
| } | |
| if err := saveToFile(filepath.Join(outputDir, characterName+"_test_filelist.txt"), strings.Join(testFilelist, "\n")); err != nil { | |
| panic(err) | |
| } | |
| } | |
| // Script 台词 | |
| type Script struct { | |
| Text string | |
| Voice string | |
| } | |
| // getScripts 获取台词 | |
| // 遍历 script 文件夹下的s文件, 获取对应角色的所有台词 | |
| func getScripts(characterName string) (ret []Script) { | |
| ret = make([]Script, 0) | |
| // 遍历 script 文件夹下的s文件 | |
| pwd, _ := os.Getwd() | |
| scriptDir := filepath.Join(pwd, "script") | |
| charactersName := strings.Split(characterName, "_") | |
| err := filepath.Walk(scriptDir, func(path string, info os.FileInfo, err error) error { | |
| if err != nil || info.IsDir() { | |
| return nil | |
| } | |
| if !strings.HasSuffix(info.Name(), ".s") { | |
| return nil | |
| } | |
| // 打开文件 | |
| f, err := os.Open(path) | |
| if err != nil { | |
| return err | |
| } | |
| log.Println("open file: ", path) | |
| defer f.Close() | |
| // 编码转换 从JIS转换为UTF-8 | |
| reader := transform.NewReader(f, japanese.ShiftJIS.NewDecoder()) | |
| content, err := io.ReadAll(reader) | |
| if err != nil { | |
| return err | |
| } | |
| // 用正则表达式匹配出所有的台词 | |
| // ％v_yuu0182 | |
| // 【大蔵遊星】 | |
| // 「これでいい……？」 | |
| // ^message,show:true | |
| // ^music01,file:BGM42 | |
| re := regexp.MustCompile(`％(.+?)\s+【(.+?)】\s+「([^」]+)」`) | |
| matches := re.FindAllStringSubmatch(string(content), -1) | |
| for _, match := range matches { | |
| if len(match) < 3 { | |
| continue | |
| } | |
| if strContains(match[2], charactersName) { | |
| s := Script{ | |
| Text: DBC2SBC(strings.TrimSpace(match[3])), | |
| Voice: strings.TrimSpace(match[1]), | |
| } | |
| if s.Voice == "" { | |
| continue | |
| } | |
| // 如果 去掉 …… 后的台词为空, 则不添加 | |
| if strings.TrimSpace(strings.ReplaceAll(s.Text, "……", "")) == "" { | |
| continue | |
| } | |
| ret = append(ret, s) | |
| fmt.Printf("%s: %s\n", s.Text, s.Voice) | |
| } | |
| } | |
| return nil | |
| }) | |
| if err != nil { | |
| log.Println(err) | |
| return | |
| } | |
| return | |
| } | |
| func strContains(str string, substr []string) bool { | |
| str = strings.TrimSpace(str) | |
| for _, s := range substr { | |
| if str == s { | |
| return true | |
| } | |
| } | |
| return false | |
| } | |
| func DBC2SBC(s string) string { | |
| var strLst []string | |
| for _, i := range s { | |
| insideCode := i | |
| if insideCode == 12288 { | |
| insideCode = 32 | |
| } else { | |
| insideCode -= 65248 | |
| } | |
| if insideCode < 32 || insideCode > 126 { | |
| strLst = append(strLst, string(i)) | |
| } else { | |
| strLst = append(strLst, string(insideCode)) | |
| } | |
| } | |
| return strings.Join(strLst, "") | |
| } | |
| func oggToWav(oggFileName, wavFileName string) (err error) { | |
| // 如果文件已经存在, 则不再转换 | |
| if _, err := os.Stat(wavFileName); err == nil { | |
| return nil | |
| } | |
| f, err := os.Open(oggFileName) | |
| if err != nil { | |
| return err | |
| } | |
| defer f.Close() | |
| s, format, err := vorbis.Decode(f) | |
| if err != nil { | |
| return err | |
| } | |
| defer s.Close() | |
| sr := beep.SampleRate(22050) | |
| // 保存为 wav 文件 | |
| f, err = os.Create(wavFileName) | |
| if err != nil { | |
| return err | |
| } | |
| format.SampleRate = sr | |
| format.NumChannels = 1 | |
| if err = wav.Encode(f, s, format); err != nil { | |
| return err | |
| } | |
| return | |
| } | |
| func saveToFile(fileName string, content string) error { | |
| f, err := os.Create(fileName) | |
| if err != nil { | |
| return err | |
| } | |
| defer f.Close() | |
| _, _ = io.WriteString(f, content) | |
| return nil | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment