Wrap punctuation APIs to C#. (#945)

.github/scripts/test-dot-net.sh

@@ -2,7 +2,10 @@
 
 cd dotnet-examples/
 
-cd speaker-identification
+cd offline-punctuation
+./run.sh
+
+cd ../speaker-identification
 ./run.sh
 
 cd ../streaming-hlg-decoding/

.github/workflows/test-dot-net.yaml

@@ -196,6 +196,7 @@ jobs:
           cp -v scripts/dotnet/examples/spoken-language-identification.csproj dotnet-examples/spoken-language-identification/
           cp -v scripts/dotnet/examples/streaming-hlg-decoding.csproj dotnet-examples/streaming-hlg-decoding
           cp -v scripts/dotnet/examples/speaker-identification.csproj dotnet-examples/speaker-identification
+          cp -v scripts/dotnet/examples/offline-punctuation.csproj dotnet-examples/offline-punctuation
 
           ls -lh /tmp
 

dotnet-examples/offline-punctuation/Program.cs

@@ -0,0 +1,45 @@
+// Copyright (c)  2024  Xiaomi Corporation
+//
+// This file shows how to add punctuations to text.
+//
+// 1. Download a model from
+// https://github.com/k2-fsa/sherpa-onnx/releases/tag/punctuation-models
+//
+// wget https://github.com/k2-fsa/sherpa-onnx/releases/download/punctuation-models/sherpa-onnx-punct-ct-transformer-zh-en-vocab272727-2024-04-12.tar.bz2
+//
+// 3. Now run it
+//
+// dotnet run
+
+using SherpaOnnx;
+using System.Collections.Generic;
+using System;
+
+class OfflinePunctuationDemo
+{
+
+  static void Main(string[] args)
+  {
+    var config = new OfflinePunctuationConfig();
+    config.Model.CtTransformer = "./sherpa-onnx-punct-ct-transformer-zh-en-vocab272727-2024-04-12/model.onnx";
+    config.Model.Debug = 1;
+    config.Model.NumThreads = 1;
+    var punct = new OfflinePunctuation(config);
+
+    string[] textList = new string[] {
+        "这是一个测试你好吗How are you我很好thank you are you ok谢谢你",
+        "我们都是木头人不会说话不会动",
+        "The African blogosphere is rapidly expanding bringing more voices online in the form of commentaries opinions analyses rants and poetry",
+    };
+
+    Console.WriteLine("---------");
+    foreach (string text in textList)
+    {
+      string textWithPunct = punct.AddPunct(text);
+      Console.WriteLine("Input text: {0}", text);
+      Console.WriteLine("Output text: {0}", textWithPunct);
+      Console.WriteLine("---------");
+    }
+  }
+}
+

dotnet-examples/offline-punctuation/offline-punctuation.csproj

@@ -0,0 +1,15 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+  <PropertyGroup>
+    <OutputType>Exe</OutputType>
+    <TargetFramework>net6.0</TargetFramework>
+    <RootNamespace>offline_punctuation</RootNamespace>
+    <ImplicitUsings>enable</ImplicitUsings>
+    <Nullable>enable</Nullable>
+  </PropertyGroup>
+
+  <ItemGroup>
+    <PackageReference Include="org.k2fsa.sherpa.onnx" Version="*" />
+  </ItemGroup>
+
+</Project>

dotnet-examples/offline-punctuation/run.sh

@@ -0,0 +1,11 @@
+#!/usr/bin/env bash
+
+set -ex
+
+if [ ! -e ./sherpa-onnx-punct-ct-transformer-zh-en-vocab272727-2024-04-12/model.onnx ]; then
+  curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/punctuation-models/sherpa-onnx-punct-ct-transformer-zh-en-vocab272727-2024-04-12.tar.bz2
+  tar xvf sherpa-onnx-punct-ct-transformer-zh-en-vocab272727-2024-04-12.tar.bz2
+  rm sherpa-onnx-punct-ct-transformer-zh-en-vocab272727-2024-04-12.tar.bz2
+fi
+
+dotnet run

dotnet-examples/sherpa-onnx.sln

@@ -19,6 +19,8 @@ Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "streaming-hlg-decoding", "s
 EndProject
 Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "speaker-identification", "speaker-identification\speaker-identification.csproj", "{2B1B140E-A92F-426B-B0DF-5D916B67304F}"
 EndProject
+Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "offline-punctuation", "offline-punctuation\offline-punctuation.csproj", "{42D85582-BB63-4259-A4EA-837D66AC078B}"
+EndProject
 Global
 	GlobalSection(SolutionConfigurationPlatforms) = preSolution
 		Debug|Any CPU = Debug|Any CPU
@@ -60,5 +62,9 @@ Global
 		{2B1B140E-A92F-426B-B0DF-5D916B67304F}.Debug|Any CPU.Build.0 = Debug|Any CPU
 		{2B1B140E-A92F-426B-B0DF-5D916B67304F}.Release|Any CPU.ActiveCfg = Release|Any CPU
 		{2B1B140E-A92F-426B-B0DF-5D916B67304F}.Release|Any CPU.Build.0 = Release|Any CPU
+		{42D85582-BB63-4259-A4EA-837D66AC078B}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
+		{42D85582-BB63-4259-A4EA-837D66AC078B}.Debug|Any CPU.Build.0 = Debug|Any CPU
+		{42D85582-BB63-4259-A4EA-837D66AC078B}.Release|Any CPU.ActiveCfg = Release|Any CPU
+		{42D85582-BB63-4259-A4EA-837D66AC078B}.Release|Any CPU.Build.0 = Release|Any CPU
 	EndGlobalSection
 EndGlobal

scripts/dotnet/OfflineLMConfig.cs

@@ -8,7 +8,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     [StructLayout(LayoutKind.Sequential)]
     public struct OfflineLMConfig
     {
@@ -22,5 +21,4 @@ namespace SherpaOnnx
 
         public float Scale;
     }
-
 }

scripts/dotnet/OfflineModelConfig.cs

@@ -8,7 +8,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     [StructLayout(LayoutKind.Sequential)]
     public struct OfflineModelConfig
     {
@@ -44,6 +43,4 @@ namespace SherpaOnnx
         [MarshalAs(UnmanagedType.LPStr)]
         public string ModelType;
     }
-
-
 }

scripts/dotnet/OfflineNemoEncDecCtcModelConfig.cs

@@ -8,7 +8,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     [StructLayout(LayoutKind.Sequential)]
     public struct OfflineNemoEncDecCtcModelConfig
     {

scripts/dotnet/OfflineParaformerModelConfig.cs

@@ -18,5 +18,4 @@ namespace SherpaOnnx
         [MarshalAs(UnmanagedType.LPStr)]
         public string Model;
     }
-
 }

scripts/dotnet/OfflinePunctuation.cs

@@ -0,0 +1,88 @@
+/// Copyright (c)  2024  Xiaomi Corporation (authors: Fangjun Kuang)
+
+using System.Linq;
+using System.Collections.Generic;
+using System.Runtime.InteropServices;
+using System.Text;
+using System;
+
+namespace SherpaOnnx
+{
+    public class OfflinePunctuation : IDisposable
+    {
+        public OfflinePunctuation(OfflinePunctuationConfig config)
+        {
+            IntPtr h = SherpaOnnxCreateOfflinePunctuation(ref config);
+            _handle = new HandleRef(this, h);
+        }
+
+        public String AddPunct(String text)
+        {
+            IntPtr p = SherpaOfflinePunctuationAddPunct(_handle.Handle, text);
+
+            string s = "";
+            int length = 0;
+
+            unsafe
+            {
+                byte* b = (byte*)p;
+                if (b != null)
+                {
+                    while (*b != 0)
+                    {
+                        ++b;
+                        length += 1;
+                    }
+                }
+            }
+
+            if (length > 0)
+            {
+                byte[] stringBuffer = new byte[length];
+                Marshal.Copy(p, stringBuffer, 0, length);
+                s = Encoding.UTF8.GetString(stringBuffer);
+            }
+
+            SherpaOfflinePunctuationFreeText(p);
+
+            return s;
+        }
+
+        public void Dispose()
+        {
+            Cleanup();
+            // Prevent the object from being placed on the
+            // finalization queue
+            System.GC.SuppressFinalize(this);
+        }
+
+        ~OfflinePunctuation()
+        {
+            Cleanup();
+        }
+
+        private void Cleanup()
+        {
+            SherpaOnnxDestroyOfflinePunctuation(_handle.Handle);
+
+            // Don't permit the handle to be used again.
+            _handle = new HandleRef(this, IntPtr.Zero);
+        }
+
+        private HandleRef _handle;
+
+
+        [DllImport(Dll.Filename)]
+        private static extern IntPtr SherpaOnnxCreateOfflinePunctuation(ref OfflinePunctuationConfig config);
+
+        [DllImport(Dll.Filename)]
+        private static extern void SherpaOnnxDestroyOfflinePunctuation(IntPtr handle);
+
+        [DllImport(Dll.Filename)]
+        private static extern IntPtr SherpaOfflinePunctuationAddPunct(IntPtr handle, [MarshalAs(UnmanagedType.LPStr)] string text);
+
+        [DllImport(Dll.Filename)]
+        private static extern void SherpaOfflinePunctuationFreeText(IntPtr p);
+    }
+}
+

scripts/dotnet/OfflinePunctuationConfig.cs

@@ -0,0 +1,21 @@
+/// Copyright (c)  2024  Xiaomi Corporation (authors: Fangjun Kuang)
+
+using System.Linq;
+using System.Collections.Generic;
+using System.Runtime.InteropServices;
+using System.Text;
+using System;
+
+namespace SherpaOnnx
+{
+    [StructLayout(LayoutKind.Sequential)]
+    public struct OfflinePunctuationConfig
+    {
+        public OfflinePunctuationConfig()
+        {
+            Model = new OfflinePunctuationModelConfig();
+        }
+        public OfflinePunctuationModelConfig Model;
+    }
+}
+

scripts/dotnet/OfflinePunctuationModelConfig.cs

@@ -0,0 +1,32 @@
+/// Copyright (c)  2024  Xiaomi Corporation (authors: Fangjun Kuang)
+
+using System.Linq;
+using System.Collections.Generic;
+using System.Runtime.InteropServices;
+using System.Text;
+using System;
+
+namespace SherpaOnnx
+{
+    [StructLayout(LayoutKind.Sequential)]
+    public struct OfflinePunctuationModelConfig
+    {
+        public OfflinePunctuationModelConfig()
+        {
+            CtTransformer = "";
+            NumThreads = 1;
+            Debug = 0;
+            Provider = "cpu";
+        }
+
+        [MarshalAs(UnmanagedType.LPStr)]
+        public string CtTransformer;
+
+        public int NumThreads;
+
+        public int Debug;
+
+        [MarshalAs(UnmanagedType.LPStr)]
+        public string Provider;
+    }
+}

scripts/dotnet/OfflineRecognizer.cs

@@ -72,5 +72,4 @@ namespace SherpaOnnx
         [DllImport(Dll.Filename, EntryPoint = "DecodeMultipleOfflineStreams")]
         private static extern void Decode(IntPtr handle, IntPtr[] streams, int n);
     }
-
 }

scripts/dotnet/OfflineRecognizerConfig.cs

@@ -8,7 +8,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     [StructLayout(LayoutKind.Sequential)]
     public struct OfflineRecognizerConfig
     {
@@ -38,6 +37,4 @@ namespace SherpaOnnx
 
         public float HotwordsScore;
     }
-
-
 }

scripts/dotnet/OfflineRecognizerResult.cs

@@ -8,7 +8,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     public class OfflineRecognizerResult
     {
         public OfflineRecognizerResult(IntPtr handle)
@@ -44,6 +43,4 @@ namespace SherpaOnnx
         private String _text;
         public String Text => _text;
     }
-
-
 }

scripts/dotnet/OfflineStream.cs

@@ -8,7 +8,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     public class OfflineStream : IDisposable
     {
         public OfflineStream(IntPtr p)
@@ -68,5 +67,4 @@ namespace SherpaOnnx
         [DllImport(Dll.Filename, EntryPoint = "DestroyOfflineRecognizerResult")]
         private static extern void DestroyResult(IntPtr handle);
     }
-
 }

scripts/dotnet/OfflineTdnnModelConfig.cs

@@ -18,5 +18,4 @@ namespace SherpaOnnx
         [MarshalAs(UnmanagedType.LPStr)]
         public string Model;
     }
-
 }

scripts/dotnet/OfflineTransducerModelConfig.cs

@@ -26,5 +26,4 @@ namespace SherpaOnnx
         [MarshalAs(UnmanagedType.LPStr)]
         public string Joiner;
     }
-
 }

scripts/dotnet/OfflineTtsConfig.cs

@@ -28,5 +28,4 @@ namespace SherpaOnnx
         [MarshalAs(UnmanagedType.LPStr)]
         public string RuleFars;
     }
-
 }

scripts/dotnet/OfflineTtsModelConfig.cs

@@ -8,7 +8,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     [StructLayout(LayoutKind.Sequential)]
     public struct OfflineTtsModelConfig
     {

scripts/dotnet/OfflineWhisperModelConfig.cs

@@ -33,5 +33,4 @@ namespace SherpaOnnx
 
         public int TailPaddings;
     }
-
 }

scripts/dotnet/OnlineCtcFstDecoderConfig.cs

@@ -24,5 +24,4 @@ namespace SherpaOnnx
 
         public int MaxActive;
     }
-
 }

scripts/dotnet/OnlineModelConfig.cs

@@ -10,7 +10,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     [StructLayout(LayoutKind.Sequential)]
     public struct OnlineModelConfig
     {
@@ -45,5 +44,4 @@ namespace SherpaOnnx
         [MarshalAs(UnmanagedType.LPStr)]
         public string ModelType;
     }
-
 }

scripts/dotnet/OnlineParaformerModelConfig.cs

@@ -10,7 +10,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     [StructLayout(LayoutKind.Sequential)]
     public struct OnlineParaformerModelConfig
     {
@@ -26,5 +25,4 @@ namespace SherpaOnnx
         [MarshalAs(UnmanagedType.LPStr)]
         public string Decoder;
     }
-
 }

scripts/dotnet/OnlineRecognizerConfig.cs

@@ -10,7 +10,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     [StructLayout(LayoutKind.Sequential)]
     public struct OnlineRecognizerConfig
     {
@@ -66,5 +65,4 @@ namespace SherpaOnnx
 
         public OnlineCtcFstDecoderConfig CtcFstDecoderConfig;
     }
-
 }

scripts/dotnet/OnlineRecognizerResult.cs

@@ -10,7 +10,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     public class OnlineRecognizerResult
     {
         public OnlineRecognizerResult(IntPtr handle)

scripts/dotnet/OnlineStream.cs

@@ -10,7 +10,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     public class OnlineStream : IDisposable
     {
         public OnlineStream(IntPtr p)
@@ -61,5 +60,4 @@ namespace SherpaOnnx
         [DllImport(Dll.Filename)]
         private static extern void InputFinished(IntPtr handle);
     }
-
 }

scripts/dotnet/OnlineTransducerModelConfig.cs

@@ -10,7 +10,6 @@ using System;
 
 namespace SherpaOnnx
 {
-
     [StructLayout(LayoutKind.Sequential)]
     public struct OnlineTransducerModelConfig
     {
@@ -30,5 +29,4 @@ namespace SherpaOnnx
         [MarshalAs(UnmanagedType.LPStr)]
         public string Joiner;
     }
-
 }

scripts/dotnet/examples/offline-punctuation.csproj

@@ -0,0 +1,19 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+  <PropertyGroup>
+    <OutputType>Exe</OutputType>
+    <TargetFramework>net6.0</TargetFramework>
+    <RootNamespace>offline_punctuation</RootNamespace>
+    <ImplicitUsings>enable</ImplicitUsings>
+    <Nullable>enable</Nullable>
+  </PropertyGroup>
+
+  <PropertyGroup>
+    <RestoreSources>/tmp/packages;$(RestoreSources);https://api.nuget.org/v3/index.json</RestoreSources>
+  </PropertyGroup>
+
+  <ItemGroup>
+    <PackageReference Include="org.k2fsa.sherpa.onnx" Version="*" />
+  </ItemGroup>
+
+</Project>