ci: worflow dispatch

add cli parsing, model and .dll download prompt and so on

.github/workflows/release.yml

@@ -1,9 +1,10 @@
-name: goreleaser
+name: Build and Release
 
 on:
+  workflow_dispatch:
   push:
     tags:
-      - "*"
+      - "v[0-9]+.[0-9]+.[0-9]+"
 
 permissions:
   contents: write
@@ -29,4 +30,4 @@ jobs:
           version: ${{ env.GITHUB_REF_NAME }}
           args: release --clean
         env:
-          GITHUB_TOKEN: ${{ secrets.PUBLISHER_TOKEN }}
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.gitignore

@@ -1,5 +1,6 @@
 Whisper.dll
-ggml-medium.bin
+ggml-*
+*.exe
 
 whisper-api-server.exe
 

README.md

@@ -9,7 +9,12 @@ This API server enables audio transcription using the OpenAI Whisper models.
 - Download `.exe` from [Releases](https://github.com/xzeldon/whisper-api-server/releases/latest)
 - Just run it!
 
-# Build from source
+# Build from source (Windows)
+
+## Prerequisites
+
+- GCC Compiler Installed in your PATH (You can get it from [here](https://github.com/niXman/mingw-builds-binaries))
+- Install Go (https://go.dev/doc/install)
 
 Before build make sure that **CGO_ENABLED** env is set to **1**
 
@@ -46,7 +51,7 @@ Receive a response in JSON format:
 
 ```json
 {
-  "text": "Imagine the wildest idea that you've ever had, and you're curious about how it might scale to something that's a 100, a 1,000 times bigger. This is a place where you can get to do that."
+	"text": "Imagine the wildest idea that you've ever had, and you're curious about how it might scale to something that's a 100, a 1,000 times bigger. This is a place where you can get to do that."
 }
 ```
 
@@ -65,9 +70,8 @@ Receive a response in JSON format:
 - [x] Implement automatic `Whisper.dll` downloading from [Guthub releases](https://github.com/Const-me/Whisper/releases)
 - [x] Provide prebuilt binaries for Windows
 - [ ] Include instructions for running on Linux with Wine (likely possible).
-- [ ] Use flags to override the model path
-- [ ] Use flags to override the model type (when downloading the model)
-- [ ] Use flags to override the port
+- [x] Use flags to override the model path
+- [x] Use flags to override the port
 
 # Credits
 

go.mod

@@ -10,17 +10,20 @@ require (
 
 require (
 	github.com/golang-jwt/jwt v3.2.2+incompatible // indirect
+	github.com/inconshreveable/mousetrap v1.1.0 // indirect
 	github.com/mattn/go-runewidth v0.0.14 // indirect
 	github.com/mitchellh/colorstring v0.0.0-20190213212951-d06e56a500db // indirect
 	github.com/rivo/uniseg v0.2.0 // indirect
+	github.com/spf13/pflag v1.0.5 // indirect
 	golang.org/x/term v0.10.0 // indirect
 	golang.org/x/time v0.3.0 // indirect
 )
 
 require (
-	github.com/labstack/gommon v0.4.0
+	github.com/labstack/gommon v0.4.0 // indirect
 	github.com/mattn/go-colorable v0.1.13 // indirect
 	github.com/mattn/go-isatty v0.0.19 // indirect
+	github.com/spf13/cobra v1.8.1
 	github.com/valyala/bytebufferpool v1.0.0 // indirect
 	github.com/valyala/fasttemplate v1.2.2 // indirect
 	golang.org/x/crypto v0.11.0 // indirect

go.sum

@@ -1,8 +1,11 @@
+github.com/cpuguy83/go-md2man/v2 v2.0.4/go.mod h1:tgQtvFlXSQOSOSIRvRPT7W67SCa46tRHOmNcaadrF8o=
 github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/golang-jwt/jwt v3.2.2+incompatible h1:IfV12K8xAKAnZqdXVzCZ+TOjboZ2keLg81eXfW3O+oY=
 github.com/golang-jwt/jwt v3.2.2+incompatible/go.mod h1:8pz2t5EyA70fFQQSrl6XZXzqecmYZeUEB8OUGHkxJ+I=
+github.com/inconshreveable/mousetrap v1.1.0 h1:wN+x4NVGpMsO7ErUn/mUI3vEoE6Jt13X2s0bqwp9tc8=
+github.com/inconshreveable/mousetrap v1.1.0/go.mod h1:vpF70FUmC8bwa3OWnCshd2FqLfsEA9PFc4w1p2J65bw=
 github.com/k0kubun/go-ansi v0.0.0-20180517002512-3bf9e2903213/go.mod h1:vNUNkEQ1e29fT/6vq2aBdFsgNPmy8qMdSay1npru+Sw=
 github.com/labstack/echo/v4 v4.11.1 h1:dEpLU2FLg4UVmvCGPuk/APjlH6GDpbEPti61srUUUs4=
 github.com/labstack/echo/v4 v4.11.1/go.mod h1:YuYRTSM3CHs2ybfrL8Px48bO6BAnYIN4l8wSTMP6BDQ=
@@ -24,8 +27,13 @@ github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZb
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
 github.com/rivo/uniseg v0.2.0 h1:S1pD9weZBuJdFmowNwbpi7BJ8TNftyUImj/0WQi72jY=
 github.com/rivo/uniseg v0.2.0/go.mod h1:J6wj4VEh+S6ZtnVlnTBMWIodfgj8LQOQFoIToxlJtxc=
+github.com/russross/blackfriday/v2 v2.1.0/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
 github.com/schollz/progressbar/v3 v3.13.1 h1:o8rySDYiQ59Mwzy2FELeHY5ZARXZTVJC7iHD6PEFUiE=
 github.com/schollz/progressbar/v3 v3.13.1/go.mod h1:xvrbki8kfT1fzWzBT/UZd9L6GA+jdL7HAgq2RFnO6fQ=
+github.com/spf13/cobra v1.8.1 h1:e5/vxKd/rZsfSJMUX1agtjeTDf+qv1/JdBF8gg5k9ZM=
+github.com/spf13/cobra v1.8.1/go.mod h1:wHxEcudfqmLYa8iTfL+OuZPbBZkmvliBWKIezN3kD9Y=
+github.com/spf13/pflag v1.0.5 h1:iy+VFUOCP1a+8yFto/drg2CJ5u0yRoB7fZw3DKv/JXA=
+github.com/spf13/pflag v1.0.5/go.mod h1:McXfInJRrz4CZXVZOBLb0bTZqETkiAhM9Iw0y3An2Bg=
 github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
 github.com/stretchr/testify v1.3.0/go.mod h1:M5WIy9Dh21IEIfnGCwXGc5bZfKNJtfHm1UVUgZn+9EI=
 github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=

internal/api/handler.go

@@ -1,6 +1,7 @@
 package api
 
 import (
+	"io"
 	"net/http"
 	"strings"
 
@@ -11,7 +12,7 @@ type TranscribeResponse struct {
 	Text string `json:"text"`
 }
 
-func Transcribe(c echo.Context, whisperState *WhisperState) error {
+func TranscribeFromFile(c echo.Context, whisperState *WhisperState) error {
 	audioPath, err := saveFormFile("file", c)
 	if err != nil {
 		c.Logger().Errorf("Error reading file: %s", err)
@@ -26,6 +27,11 @@ func Transcribe(c echo.Context, whisperState *WhisperState) error {
 
 	err = whisperState.context.RunFull(whisperState.params, buffer)
 
+	if err != nil {
+		c.Logger().Errorf("Error processing audio: %s", err)
+		return err
+	}
+
 	result, err := getResult(whisperState.context)
 	if err != nil {
 		c.Logger().Error(err)
@@ -43,3 +49,59 @@ func Transcribe(c echo.Context, whisperState *WhisperState) error {
 
 	return c.JSON(http.StatusOK, response)
 }
+
+func Transcribe(c echo.Context, whisperState *WhisperState) error {
+	// Get the file header
+	fileHeader, err := c.FormFile("file")
+	if err != nil {
+		c.Logger().Errorf("Error retrieving the file: %s", err)
+		return err
+	}
+
+	// Open the file
+	file, err := fileHeader.Open()
+	if err != nil {
+		c.Logger().Errorf("Error opening the file: %s", err)
+		return err
+	}
+	defer file.Close()
+
+	// Read the file into a buffer
+	buffer, err := io.ReadAll(file)
+	if err != nil {
+		c.Logger().Errorf("Error reading the file into buffer: %s", err)
+		return err
+	}
+
+	whisperState.mutex.Lock()
+	defer whisperState.mutex.Unlock()
+
+	bufferSpecial, err := whisperState.media.LoadAudioFileData(&buffer, true)
+
+	if err != nil {
+		c.Logger().Errorf("Error loading audio file data: %s", err)
+		return err
+	}
+
+	err = whisperState.context.RunStreamed(whisperState.params, bufferSpecial)
+	if err != nil {
+		c.Logger().Errorf("Error processing audio: %s", err)
+		return err
+	}
+
+	result, err := getResult(whisperState.context)
+	if err != nil {
+		c.Logger().Error(err)
+		return err
+	}
+
+	if len(result) == 0 {
+		return c.JSON(http.StatusInternalServerError, map[string]string{"error": "Internal server error"})
+	}
+
+	response := TranscribeResponse{
+		Text: strings.TrimLeft(result, " "),
+	}
+
+	return c.JSON(http.StatusOK, response)
+}

internal/api/state.go

@@ -15,7 +15,7 @@ type WhisperState struct {
 	mutex   sync.Mutex
 }
 
-func InitializeWhisperState(modelPath string) (*WhisperState, error) {
+func InitializeWhisperState(modelPath string, lang int32) (*WhisperState, error) {
 	lib, err := whisper.New(whisper.LlDebug, whisper.LfUseStandardError, nil)
 	if err != nil {
 		return nil, err
@@ -41,6 +41,8 @@ func InitializeWhisperState(modelPath string) (*WhisperState, error) {
 		return nil, err
 	}
 
+	params.SetLanguage(lang)
+
 	fmt.Printf("Params CPU Threads : %d\n", params.CpuThreads())
 
 	return &WhisperState{

internal/resources/cliArgs.go

@@ -0,0 +1,103 @@
+package resources
+
+import (
+	_ "embed"
+	"encoding/json"
+	"fmt"
+	"os"
+	"strconv"
+	"strings"
+
+	"github.com/spf13/cobra"
+)
+
+//go:embed languageMap.json
+var languageMapData []byte // Embedded language map file as a byte slice
+
+// Arguments holds the parsed CLI arguments
+type Arguments struct {
+	Language  string
+	ModelPath string
+	Port      int
+}
+
+// ParsedArguments holds the processed arguments
+type ParsedArguments struct {
+	Language  int32
+	ModelPath string
+	Port      int
+}
+
+// LanguageMap represents the mapping of languages to their hex codes
+type LanguageMap map[string]string
+
+func processLanguageAndCode(language string) (int32, error) {
+    var languageMap LanguageMap
+    err := json.Unmarshal(languageMapData, &languageMap)
+    if err != nil {
+        return 0x6E65, fmt.Errorf("error parsing language map: %w", err)
+    }
+
+    hexCode, ok := languageMap[strings.ToLower(language)]
+    if !ok {
+        return 0x6E65, fmt.Errorf("unsupported language")
+    }
+
+    fmt.Printf("Hex Code Found: %s\n", hexCode)
+
+    languageCode, err := strconv.ParseInt(hexCode, 0, 32)
+    if err != nil {
+        return 0x6E65, fmt.Errorf("error converting hex code: %w", err)
+    }
+
+    return int32(languageCode), nil
+}
+
+func ApplyExitOnHelp(c *cobra.Command, exitCode int) {
+	helpFunc := c.HelpFunc()
+	c.SetHelpFunc(func(c *cobra.Command, s []string) {
+		helpFunc(c, s)
+		os.Exit(exitCode)
+	})
+}
+
+func ParseFlags() (*ParsedArguments, error) {
+    args := &Arguments{}
+
+    var parsedArgs *ParsedArguments
+
+    rootCmd := &cobra.Command{
+        Use:   "whisper",
+        Short: "Audio transcription using the OpenAI Whisper models",
+        RunE: func(cmd *cobra.Command, _ []string) error {
+            // Process language code with fallback
+            languageCode, err := processLanguageAndCode(args.Language)
+            if err != nil {
+                fmt.Printf("Error setting language, defaulting to English")
+                // Default to English
+                languageCode = 0x6E65
+            }
+
+            parsedArgs = &ParsedArguments{
+                Language:  languageCode,
+                ModelPath: args.ModelPath,
+                Port:      args.Port,
+            }
+            return nil
+        },
+    }
+
+    rootCmd.Flags().StringVarP(&args.Language, "language", "l", "", "Language to be processed")
+    rootCmd.Flags().StringVarP(&args.ModelPath, "modelPath", "m", "ggml-medium.bin", "Path to the model file (required)")
+    rootCmd.Flags().IntVarP(&args.Port, "port", "p", 3000, "Port to start the server on")
+
+	ApplyExitOnHelp(rootCmd, 0)
+
+    err := rootCmd.Execute()
+    if err != nil {
+        return nil, err
+    }
+
+    return parsedArgs, nil
+}
+

internal/resources/download.go

@@ -1,38 +0,0 @@
-package resources
-
-import (
-	"io"
-	"net/http"
-	"os"
-
-	"github.com/schollz/progressbar/v3"
-)
-
-func DownloadFile(url string, filepath string) error {
-	out, err := os.Create(filepath)
-	if err != nil {
-		return err
-	}
-	defer out.Close()
-
-	resp, err := http.Get(url)
-	if err != nil {
-		return err
-	}
-	defer resp.Body.Close()
-
-	fileSize := resp.ContentLength
-	bar := progressbar.DefaultBytes(
-		fileSize,
-		"Downloading",
-	)
-
-	writer := io.MultiWriter(out, bar)
-
-	_, err = io.Copy(writer, resp.Body)
-	if err != nil {
-		return err
-	}
-
-	return nil
-}

internal/resources/downloadResources.go

@@ -4,10 +4,65 @@ import (
 	"archive/zip"
 	"fmt"
 	"io"
+	"net/http"
 	"os"
 	"path/filepath"
+
+	"github.com/schollz/progressbar/v3"
 )
 
+func GetModel(modelType string) (string, error) {
+	fileURL := fmt.Sprintf("https://huggingface.co/ggerganov/whisper.cpp/resolve/main/%s", modelType)
+	filePath := modelType
+
+	isModelFileExists := IsFileExists(filePath)
+
+	if !isModelFileExists {
+		fmt.Println("Model not found.")
+		err := DownloadFile(fileURL, filePath)
+		if err != nil {
+			return "", err
+		}
+	}
+
+	absPath, err := filepath.Abs(filePath)
+	if err != nil {
+		return "", err
+	}
+
+	fmt.Printf("Model found: %s\n", absPath)
+	return filePath, nil
+}
+
+func DownloadFile(url string, filepath string) error {
+	out, err := os.Create(filepath)
+	if err != nil {
+		return err
+	}
+	defer out.Close()
+
+	resp, err := http.Get(url)
+	if err != nil {
+		return err
+	}
+	defer resp.Body.Close()
+
+	fileSize := resp.ContentLength
+	bar := progressbar.DefaultBytes(
+		fileSize,
+		"Downloading",
+	)
+
+	writer := io.MultiWriter(out, bar)
+
+	_, err = io.Copy(writer, resp.Body)
+	if err != nil {
+		return err
+	}
+
+	return nil
+}
+
 func GetWhisperDll(version string) (string, error) {
 	fileUrl := fmt.Sprintf("https://github.com/Const-me/Whisper/releases/download/%s/Library.zip", version)
 	fileToExtract := "Binary/Whisper.dll"
@@ -76,3 +131,13 @@ func extractFile(archivePath string, fileToExtract string) error {
 
 	return fmt.Errorf("File not found in the archive")
 }
+
+func IsFileExists(filename string) bool {
+	_, err := os.Stat(filename)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return false
+		}
+	}
+	return true
+}

internal/resources/languageMap.json

@@ -0,0 +1,84 @@
+{
+  "af": "0x6661",
+  "sq": "0x7173",
+  "am": "0x6D61",
+  "ar": "0x7261",
+  "hy": "0x7968",
+  "as": "0x7361",
+  "az": "0x7A61",
+  "ba": "0x6162",
+  "eu": "0x7565",
+  "be": "0x6562",
+  "bn": "0x6E62",
+  "bs": "0x7362",
+  "br": "0x7262",
+  "bg": "0x6762",
+  "ca": "0x6163",
+  "zh": "0x687A",
+  "hr": "0x7268",
+  "cs": "0x7363",
+  "da": "0x6164",
+  "nl": "0x6C6E",
+  "en": "0x6E65",
+  "et": "0x7465",
+  "fo": "0x6F66",
+  "fi": "0x6966",
+  "fr": "0x7266",
+  "gl": "0x6C67",
+  "ka": "0x616B",
+  "de": "0x7265",
+  "el": "0x6C61",
+  "gu": "0x7567",
+  "he": "0x6568",
+  "hi": "0x6968",
+  "hu": "0x7568",
+  "is": "0x7369",
+  "id": "0x6469",
+  "it": "0x7469",
+  "ja": "0x616A",
+  "kn": "0x6E6B",
+  "kk": "0x6B6B",
+  "km": "0x6D6B",
+  "ko": "0x6F6B",
+  "ky": "0x796B",
+  "lo": "0x6F6C",
+  "lv": "0x766C",
+  "lt": "0x746C",
+  "mk": "0x6B6D",
+  "ms": "0x736D",
+  "ml": "0x6C6D",
+  "mr": "0x726D",
+  "mn": "0x6E6D",
+  "ne": "0x6570",
+  "no": "0x6F6E",
+  "or": "0x726F",
+  "ps": "0x7368",
+  "fa": "0x6172",
+  "pl": "0x6C70",
+  "pt": "0x7470",
+  "pa": "0x6170",
+  "ro": "0x6F72",
+  "ru": "0x7572",
+  "sa": "0x6173",
+  "sr": "0x7273",
+  "sd": "0x6473",
+  "si": "0x6973",
+  "sk": "0x6B73",
+  "sl": "0x6C73",
+  "es": "0x6573",
+  "sw": "0x7773",
+  "sv": "0x6576",
+  "tg": "0x6769",
+  "ta": "0x6174",
+  "te": "0x6574",
+  "th": "0x6874",
+  "tr": "0x7274",
+  "uk": "0x6B75",
+  "ur": "0x7275",
+  "uz": "0x7A75",
+  "vi": "0x6976",
+  "cy": "0x7963",
+  "xh": "0x6877",
+  "yi": "0x6979",
+  "yo": "0x6F79"
+}

internal/resources/model.go

@@ -1,29 +0,0 @@
-package resources
-
-import (
-	"fmt"
-	"path/filepath"
-)
-
-func GetModel(modelType string) (string, error) {
-	fileURL := fmt.Sprintf("https://huggingface.co/ggerganov/whisper.cpp/resolve/main/%s", modelType)
-	filePath := modelType
-
-	isModelFileExists := IsFileExists(filePath)
-
-	if !isModelFileExists {
-		fmt.Println("Model not found.")
-		err := DownloadFile(fileURL, filePath)
-		if err != nil {
-			return "", err
-		}
-	}
-
-	absPath, err := filepath.Abs(filePath)
-	if err != nil {
-		return "", err
-	}
-
-	fmt.Printf("Model found: %s\n", absPath)
-	return filePath, nil
-}

internal/resources/promt.go

@@ -0,0 +1,76 @@
+package resources
+
+import (
+	"bufio"
+	"fmt"
+	"os"
+	"path/filepath"
+	"strings"
+)
+
+// PromptUser prompts the user with a question and returns true if they agree
+func PromptUser(question string) bool {
+	fmt.Printf("%s (y/n): ", question)
+	reader := bufio.NewReader(os.Stdin)
+	response, err := reader.ReadString('\n')
+	if err != nil {
+		fmt.Println("Error reading input:", err)
+		return false
+	}
+	response = strings.TrimSpace(strings.ToLower(response))
+	return response == "y" || response == "yes"
+}
+
+// HandleWhisperDll checks if Whisper.dll exists or prompts the user to download it
+func HandleWhisperDll(version string) (string, error) {
+	if IsFileExists("Whisper.dll") {
+		absPath, err := filepath.Abs("Whisper.dll")
+		if err != nil {
+			return "", err
+		}
+		fmt.Printf("Library found: %s\n", absPath)
+		return "Whisper.dll", nil
+	}
+
+	fmt.Println("Whisper DLL not found.")
+	if PromptUser("Do you want to download Whisper.dll automatically?") {
+		path, err := GetWhisperDll(version)
+		if err != nil {
+			return "", fmt.Errorf("failed to download Whisper.dll: %w", err)
+		}
+		return path, nil
+	}
+
+	fmt.Println("To use Whisper, download the DLL manually:")
+	fmt.Printf("URL: https://github.com/Const-me/Whisper/releases/download/%s/Library.zip\n", version)
+	fmt.Println("Extract 'Binary/Whisper.dll' from the archive and place it in the executable's directory.")
+	fmt.Println("You can manually specify path to .dll file using cli arguments, use --help to print available cli flags")
+	return "", fmt.Errorf("whisper.dll not found and user chose not to download")
+}
+
+// HandleDefaultModel checks if the default model exists or prompts the user to download it
+func HandleDefaultModel(modelType string) (string, error) {
+	if IsFileExists(modelType) {
+		absPath, err := filepath.Abs(modelType)
+		if err != nil {
+			return "", err
+		}
+		fmt.Printf("Model found: %s\n", absPath)
+		return modelType, nil
+	}
+
+	fmt.Println("Default model not found.")
+	if PromptUser("Do you want to download the default model (ggml-medium.bin) automatically?") {
+		path, err := GetModel(modelType)
+		if err != nil {
+			return "", fmt.Errorf("failed to download the default model: %w", err)
+		}
+		return path, nil
+	}
+
+	fmt.Println("To use Whisper, download the model manually:")
+	fmt.Println("URL: https://huggingface.co/ggerganov/whisper.cpp/tree/main")
+	fmt.Println("Place the model file in the executable's directory or specify its path using cli arguments.")
+	fmt.Println("You can manually specify path to model file using cli arguments, use --help to print available cli flags")
+	return "", fmt.Errorf("default model not found and user chose not to download")
+}

internal/resources/utils.go

@@ -1,13 +0,0 @@
-package resources
-
-import "os"
-
-func IsFileExists(filename string) bool {
-	_, err := os.Stat(filename)
-	if err != nil {
-		if os.IsNotExist(err) {
-			return false
-		}
-	}
-	return true
-}

main.go

@@ -1,41 +1,71 @@
 package main
 
 import (
+	"fmt"
+	"os"
+	"path/filepath"
+
 	"github.com/labstack/echo/v4"
 	"github.com/labstack/echo/v4/middleware"
-	"github.com/labstack/gommon/log"
 	"github.com/xzeldon/whisper-api-server/internal/api"
 	"github.com/xzeldon/whisper-api-server/internal/resources"
 )
 
+const (
+	defaultModelType      = "ggml-medium.bin"
+	defaultWhisperVersion = "1.12.0"
+)
+
+func changeWorkingDirectory(e *echo.Echo) {
+	exePath, err := os.Executable()
+	if err != nil {
+		e.Logger.Error("Error getting executable path: ", err)
+		return
+	}
+
+	exeDir := filepath.Dir(exePath)
+	if err := os.Chdir(exeDir); err != nil {
+		e.Logger.Error("Error changing working directory: ", err)
+		return
+	}
+
+	cwd, _ := os.Getwd()
+	fmt.Println("Current working directory:", cwd)
+}
+
 func main() {
 	e := echo.New()
 	e.HideBanner = true
+	changeWorkingDirectory(e)
+
+	args, err := resources.ParseFlags()
+	if err != nil {
+		e.Logger.Error("Error parsing flags: ", err)
+		return
+	}
+
+	if _, err := resources.HandleWhisperDll(defaultWhisperVersion); err != nil {
+		e.Logger.Error("Error handling Whisper.dll: ", err)
+		return
+	}
+
+	if _, err := resources.HandleDefaultModel(defaultModelType); err != nil {
+		e.Logger.Error("Error handling model file: ", err)
+		return
+	}
 
 	e.Use(middleware.CORS())
 
-	if l, ok := e.Logger.(*log.Logger); ok {
-		l.SetHeader("${time_rfc3339} ${level}")
-	}
-
-	_, err := resources.GetWhisperDll("1.12.0")
+	whisperState, err := api.InitializeWhisperState(args.ModelPath, args.Language)
 	if err != nil {
-		e.Logger.Error(err)
-	}
-
-	model, err := resources.GetModel("ggml-medium.bin")
-	if err != nil {
-		e.Logger.Error(err)
-	}
-
-	whisperState, err := api.InitializeWhisperState(model)
-	if err != nil {
-		e.Logger.Error(err)
+		e.Logger.Error("Error initializing Whisper state: ", err)
+		return
 	}
 
 	e.POST("/v1/audio/transcriptions", func(c echo.Context) error {
 		return api.Transcribe(c, whisperState)
 	})
 
-	e.Logger.Fatal(e.Start("127.0.0.1:3000"))
+	address := fmt.Sprintf("127.0.0.1:%d", args.Port)
+	e.Logger.Fatal(e.Start(address))
 }

pkg/whisper/FullParams.go

@@ -86,14 +86,14 @@ func (this *FullParams) RemoveFlags(newflag eFullParamsFlags) {
 	this.cStruct.Flags = this.cStruct.Flags ^ newflag
 }
 
-func (this *FullParams) SetLanguage(language eLanguage) {
+func (this *FullParams) SetLanguage(language int32) {
 	if this == nil {
 		return
 	} else if this.cStruct == nil {
 		return
 	}
 
-	this.cStruct.Language = language
+	this.cStruct.Language = eLanguage(language)
 }
 
 /*using pfnNewSegment = HRESULT( __cdecl* )( iContext* ctx, uint32_t n_new, void* user_data ) noexcept;*/
@@ -146,9 +146,9 @@ func (this *FullParams) TestDefaultsOK() bool {
 		return false
 	}
 
-	if this.cStruct.Language != English {
-		return false
-	}
+	// if this.cStruct.Language != English {
+	// 	return false
+	// }
 
 	// Todo ... why do these not line up as expected.. is our struct out of alignment ?
 	/*
@@ -214,6 +214,7 @@ func NewFullParams(cstruct *_FullParams) *FullParams {
 }
 
 func _newFullParams_cStruct() *_FullParams {
+
 	return &_FullParams{
 
 		strategy:       0,

pkg/whisper/context.go

@@ -207,6 +207,7 @@ func (context *IContext) FullDefaultParams(strategy eSamplingStrategy) (*FullPar
 		return nil, errors.New("FullDefaultParams did not return params")
 	}
 	ParamObj := NewFullParams(params)
+	// ParamObj.SetLanguage(Polish)
 
 	if ParamObj.TestDefaultsOK() {
 		return ParamObj, nil

pkg/whisper/winversion.go

@@ -2,7 +2,6 @@
 // Use of this source code is governed by a BSD
 // license that can be found in the LICENSE file.
 // Adapted mainly from github.com/gonutz/w32
-
 //go:build windows
 // +build windows
 
@@ -10,7 +9,7 @@ package whisper
 
 import (
 	"errors"
-	"syscall"
+	"fmt"
 	"unsafe"
 
 	"golang.org/x/sys/windows"
@@ -51,34 +50,46 @@ func (fi VS_FIXEDFILEINFO) FileVersion() uint64 {
 	return uint64(fi.FileVersionMS)<<32 | uint64(fi.FileVersionLS)
 }
 
-func GetFileVersionInfoSize(path string) uint32 {
+func GetFileVersionInfoSize(path string) (uint32, error) {
+	pathPtr, err := windows.UTF16PtrFromString(path)
+	if err != nil {
+		return 0, err
+	}
 	ret, _, _ := getFileVersionInfoSize.Call(
-		uintptr(unsafe.Pointer(syscall.StringToUTF16Ptr(path))),
+		uintptr(unsafe.Pointer(pathPtr)),
 		0,
 	)
-	return uint32(ret)
+	return uint32(ret), nil
 }
 
-func GetFileVersionInfo(path string, data []byte) bool {
+func GetFileVersionInfo(path string, data []byte) (bool, error) {
+	pathPtr, err := windows.UTF16PtrFromString(path)
+	if err != nil {
+		return false, err
+	}
 	ret, _, _ := getFileVersionInfo.Call(
-		uintptr(unsafe.Pointer(syscall.StringToUTF16Ptr(path))),
+		uintptr(unsafe.Pointer(pathPtr)),
 		0,
 		uintptr(len(data)),
 		uintptr(unsafe.Pointer(&data[0])),
 	)
-	return ret != 0
+	return ret != 0, nil
 }
 
 // VerQueryValueRoot calls VerQueryValue
 // (https://msdn.microsoft.com/en-us/library/windows/desktop/ms647464(v=vs.85).aspx)
-// with `\` (root) to retieve the VS_FIXEDFILEINFO.
+// with \ (root) to retrieve the VS_FIXEDFILEINFO.
 func VerQueryValueRoot(block []byte) (VS_FIXEDFILEINFO, error) {
 	var offset uintptr
 	var length uint
 	blockStart := unsafe.Pointer(&block[0])
+	rootPathPtr, err := windows.UTF16PtrFromString(`\`)
+	if err != nil {
+		return VS_FIXEDFILEINFO{}, err
+	}
 	ret, _, _ := verQueryValue.Call(
 		uintptr(blockStart),
-		uintptr(unsafe.Pointer(syscall.StringToUTF16Ptr(`\`))),
+		uintptr(unsafe.Pointer(rootPathPtr)),
 		uintptr(unsafe.Pointer(&offset)),
 		uintptr(unsafe.Pointer(&length)),
 	)
@@ -97,27 +108,24 @@ func VerQueryValueRoot(block []byte) (VS_FIXEDFILEINFO, error) {
 
 func GetFileVersion(path string) (WinVersion, error) {
 	var result WinVersion
-	size := GetFileVersionInfoSize(path)
-	if size <= 0 {
+	size, err := GetFileVersionInfoSize(path)
+	fmt.Println(path)
+	if err != nil || size <= 0 {
 		return result, errors.New("GetFileVersionInfoSize failed")
 	}
-
 	info := make([]byte, size)
-	ok := GetFileVersionInfo(path, info)
-	if !ok {
+	ok, err := GetFileVersionInfo(path, info)
+	if err != nil || !ok {
 		return result, errors.New("GetFileVersionInfo failed")
 	}
-
 	fixed, err := VerQueryValueRoot(info)
 	if err != nil {
 		return result, err
 	}
 	version := fixed.FileVersion()
-
 	result.Major = uint32(version & 0xFFFF000000000000 >> 48)
 	result.Minor = uint32(version & 0x0000FFFF00000000 >> 32)
 	result.Patch = uint32(version & 0x00000000FFFF0000 >> 16)
 	result.Build = uint32(version & 0x000000000000FFFF)
-
 	return result, nil
 }