179 lines
4.6 KiB
Go
179 lines
4.6 KiB
Go
|
package main
|
||
|
|
||
|
import (
|
||
|
"context"
|
||
|
"crypto/sha256"
|
||
|
"encoding/hex"
|
||
|
"encoding/json"
|
||
|
"errors"
|
||
|
"fmt"
|
||
|
"io"
|
||
|
"log"
|
||
|
"os"
|
||
|
"path/filepath"
|
||
|
"strings"
|
||
|
|
||
|
texttospeech "cloud.google.com/go/texttospeech/apiv1"
|
||
|
"cloud.google.com/go/texttospeech/apiv1/texttospeechpb"
|
||
|
)
|
||
|
|
||
|
type Job struct {
|
||
|
Name string `json:"name"`
|
||
|
Voice string `json:"voice"`
|
||
|
Language string `json:"language"`
|
||
|
Message string `json:"message"`
|
||
|
}
|
||
|
|
||
|
func loadJobs(directory string) ([]Job, error) {
|
||
|
jobs := make([]Job, 0)
|
||
|
err := filepath.Walk(directory, func(path string, info os.FileInfo, err error) error {
|
||
|
if !info.IsDir() && strings.HasSuffix(path, ".json") {
|
||
|
f, err := os.Open(fmt.Sprintf("%s", path))
|
||
|
defer f.Close()
|
||
|
if err != nil {
|
||
|
return fmt.Errorf("failed to open file '%s': %v", path, err)
|
||
|
}
|
||
|
|
||
|
b, err := io.ReadAll(f)
|
||
|
if err != nil {
|
||
|
return fmt.Errorf("failed to read file '%s': %v", path, err)
|
||
|
}
|
||
|
|
||
|
job := Job{}
|
||
|
err = json.Unmarshal(b, &job)
|
||
|
if err != nil {
|
||
|
return fmt.Errorf("failed to unmarshal file '%s' into Job struct: %v", path, err)
|
||
|
}
|
||
|
jobs = append(jobs, job)
|
||
|
}
|
||
|
return nil
|
||
|
})
|
||
|
|
||
|
if err != nil {
|
||
|
return nil, fmt.Errorf("failed to walk current directory for json files: %v", err)
|
||
|
}
|
||
|
return jobs, nil
|
||
|
}
|
||
|
|
||
|
func main() {
|
||
|
|
||
|
if len(os.Args) > 1 {
|
||
|
if os.Args[1] == "list" {
|
||
|
err := ListVoices()
|
||
|
if err != nil {
|
||
|
log.Fatalf("failed to list voices: %v", err)
|
||
|
}
|
||
|
return
|
||
|
}
|
||
|
}
|
||
|
|
||
|
jobs, err := loadJobs("jobs")
|
||
|
if err != nil {
|
||
|
log.Fatalf("failed to load jobs: %v", err)
|
||
|
}
|
||
|
|
||
|
// Check each message if MP3 file already exists, if not then it will synthesize and save the audio to file
|
||
|
// file names are SHA256 sums of the spoken text
|
||
|
for _, job := range jobs {
|
||
|
h := sha256.New()
|
||
|
h.Write([]byte(job.Message))
|
||
|
sha := h.Sum(nil) // "sha" is uint8 type, encoded in base16
|
||
|
|
||
|
filename := fmt.Sprintf("%s-%s.mp3", job.Name, hex.EncodeToString(sha))
|
||
|
|
||
|
if _, err := os.Stat(fmt.Sprintf("%s", filename)); errors.Is(err, os.ErrNotExist) {
|
||
|
mp3, err := getTTS(job.Voice, job.Message, job.Language)
|
||
|
if err != nil {
|
||
|
log.Printf("failed to get TTS: %v", err)
|
||
|
continue
|
||
|
}
|
||
|
|
||
|
err = os.WriteFile(fmt.Sprintf("%s", filename), mp3, 0644)
|
||
|
if err != nil {
|
||
|
log.Printf("failed to write mp3 file: %v", err)
|
||
|
continue
|
||
|
}
|
||
|
log.Printf("Audio content written to file: %v\n", filename)
|
||
|
}
|
||
|
}
|
||
|
}
|
||
|
|
||
|
// Accepts a string and returns a byteslice of the message in mp3 format, and an error
|
||
|
func getTTS(voice, message, language string) ([]byte, error) {
|
||
|
|
||
|
// Instantiates a client.
|
||
|
ctx := context.Background()
|
||
|
|
||
|
client, err := texttospeech.NewClient(ctx)
|
||
|
if err != nil {
|
||
|
return nil, fmt.Errorf("failed to create TTS client: %v", err)
|
||
|
}
|
||
|
defer client.Close()
|
||
|
|
||
|
// Perform the text-to-speech request on the text input with the selected
|
||
|
// voice parameters and audio file type.
|
||
|
req := texttospeechpb.SynthesizeSpeechRequest{
|
||
|
// Set the text input to be synthesized.
|
||
|
Input: &texttospeechpb.SynthesisInput{
|
||
|
InputSource: &texttospeechpb.SynthesisInput_Text{Text: message},
|
||
|
},
|
||
|
// Build the voice request, select the language code ("en-US") and the SSML
|
||
|
// voice gender ("neutral").
|
||
|
Voice: &texttospeechpb.VoiceSelectionParams{
|
||
|
LanguageCode: language,
|
||
|
Name: voice,
|
||
|
//Name: "en-US-Journey-F",
|
||
|
//Name: "tr-TR-Wavenet-E",
|
||
|
},
|
||
|
// Select the type of audio file you want returned.
|
||
|
AudioConfig: &texttospeechpb.AudioConfig{
|
||
|
AudioEncoding: texttospeechpb.AudioEncoding_MP3,
|
||
|
EffectsProfileId: []string{"headphone-class-device"},
|
||
|
SpeakingRate: 0.8,
|
||
|
},
|
||
|
}
|
||
|
|
||
|
resp, err := client.SynthesizeSpeech(ctx, &req)
|
||
|
if err != nil {
|
||
|
return nil, fmt.Errorf("failed to synthesize speech: %v", err)
|
||
|
}
|
||
|
|
||
|
return resp.AudioContent, nil
|
||
|
}
|
||
|
|
||
|
// ListVoices lists the available text to speech voices.
|
||
|
func ListVoices() error {
|
||
|
ctx := context.Background()
|
||
|
|
||
|
client, err := texttospeech.NewClient(ctx)
|
||
|
if err != nil {
|
||
|
return err
|
||
|
}
|
||
|
defer client.Close()
|
||
|
|
||
|
// Performs the list voices request.
|
||
|
resp, err := client.ListVoices(ctx, &texttospeechpb.ListVoicesRequest{})
|
||
|
if err != nil {
|
||
|
return err
|
||
|
}
|
||
|
|
||
|
for _, voice := range resp.Voices {
|
||
|
// Display the voice's name. Example: tpc-vocoded
|
||
|
fmt.Printf("Name: %v\n", voice.Name)
|
||
|
|
||
|
// Display the supported language codes for this voice. Example: "en-US"
|
||
|
for _, languageCode := range voice.LanguageCodes {
|
||
|
fmt.Printf(" Supported language: %v\n", languageCode)
|
||
|
}
|
||
|
|
||
|
// Display the SSML Voice Gender.
|
||
|
fmt.Printf(" SSML Voice Gender: %v\n", voice.SsmlGender.String())
|
||
|
|
||
|
// Display the natural sample rate hertz for this voice. Example: 24000
|
||
|
fmt.Printf(" Natural Sample Rate Hertz: %v\n",
|
||
|
voice.NaturalSampleRateHertz)
|
||
|
}
|
||
|
|
||
|
return nil
|
||
|
}
|