From ec73ac968be0c362bae5751fbb4d23bf7a45d9c1 Mon Sep 17 00:00:00 2001 From: keerthikarankv <100576900+keerthikarankv@users.noreply.github.com> Date: Tue, 17 Mar 2026 18:31:21 +0530 Subject: [PATCH 1/3] UG(1015625): AI Services UG for Smart Rich Text Editor component. --- blazor-toc.html | 8 + .../azure-openai-service.md | 349 +++++++++++ .../custom-inference-backend.md | 544 ++++++++++++++++++ blazor/smart-rich-text-editor/ollama.md | 386 +++++++++++++ .../smart-rich-text-editor/openai-service.md | 236 ++++++++ 5 files changed, 1523 insertions(+) create mode 100644 blazor/smart-rich-text-editor/azure-openai-service.md create mode 100644 blazor/smart-rich-text-editor/custom-inference-backend.md create mode 100644 blazor/smart-rich-text-editor/ollama.md create mode 100644 blazor/smart-rich-text-editor/openai-service.md diff --git a/blazor-toc.html b/blazor-toc.html index 79da90bded..3962a4a3a4 100644 --- a/blazor-toc.html +++ b/blazor-toc.html @@ -208,6 +208,14 @@
  • Blazor WASM App
    • +
  • AI Services + +
  • AI AssistView diff --git a/blazor/smart-rich-text-editor/azure-openai-service.md b/blazor/smart-rich-text-editor/azure-openai-service.md new file mode 100644 index 0000000000..09feaaa142 --- /dev/null +++ b/blazor/smart-rich-text-editor/azure-openai-service.md @@ -0,0 +1,349 @@ +--- +layout: post +title: Azure OpenAI Configuration for Syncfusion Smart Rich Text Editor +description: Configure Azure OpenAI for Syncfusion Blazor Smart Rich Text Editor: authentication, client setup, DI registration, and usage examples. +platform: Blazor +control: Smart Rich Text Editor +documentation: ug +--- + +# Azure OpenAI Configuration + +The Syncfusion® Blazor Smart Rich Text Editor supports Azure OpenAI for enterprise-grade AI capabilities with enhanced security and compliance features. + +## Prerequisites + +* Active Azure subscription +* Azure OpenAI Service resource deployed +* Deployed model (e.g., gpt-4, gpt-35-turbo) +* Azure credentials with necessary permissions + +## Deploy Azure OpenAI Service + +### Step 1: Create Azure OpenAI Resource + +1. Sign in to [Azure Portal](https://portal.azure.com/) +2. Click **Create a resource** +3. Search for **Azure OpenAI** +4. Click **Create** +5. Fill in the resource details: + - **Subscription**: Select your subscription + - **Resource group**: Create or select existing + - **Region**: Choose appropriate region + - **Name**: Give your resource a unique name + - **Pricing tier**: Select S0 or higher + +### Step 2: Deploy a Model + +1. Go to **Azure AI Studio** (https://oai.azure.com/) +2. Select your Azure OpenAI resource +3. Navigate to **Deployments** +4. Click **Create new deployment** +5. Configure: + - **Deployment name**: e.g., `gpt-35-turbo-deployment` + - **Model name**: Select model (e.g., `gpt-35-turbo`, `gpt-4`) + - **Model version**: Choose version + - **Deployment type**: Standard + +### Step 3: Obtain Credentials + +From your Azure OpenAI resource in Azure Portal, copy: +- **Endpoint**: `https://.openai.azure.com/` +- **Key**: Found under **Keys and Endpoint** +- **Deployment name**: Created in Step 2 + +## Installation + +Install required NuGet packages: + +{% tabs %} +{% highlight c# tabtitle="Package Manager" %} + +Install-Package Syncfusion.Blazor.SmartRichTextEditor +Install-Package Syncfusion.Blazor.Themes +Install-Package Microsoft.Extensions.AI +Install-Package Microsoft.Extensions.AI.OpenAI +Install-Package Azure.AI.OpenAI + +{% endhighlight %} +{% endtabs %} + +## Configuration + +### Step 1: Setup in Program.cs + +Add the following configuration to your **Program.cs** file: + +```csharp +using Syncfusion.Blazor; +using Syncfusion.Blazor.AI; +using Azure.AI.OpenAI; +using Microsoft.Extensions.AI; +using System.ClientModel; + +var builder = WebApplication.CreateBuilder(args); + +// Add services to the container +builder.Services.AddRazorPages(); +builder.Services.AddServerSideBlazor(); + +// Register Syncfusion Blazor Service +builder.Services.AddSyncfusionBlazor(); + +// Configure Azure OpenAI +string azureOpenAIKey = "your-azure-key"; +string azureOpenAIEndpoint = "https://your-resource-name.openai.azure.com/"; +string azureOpenAIDeployment = "your-deployment-name"; + +AzureOpenAIClient azureOpenAIClient = new AzureOpenAIClient( + new Uri(azureOpenAIEndpoint), + new ApiKeyCredential(azureOpenAIKey) +); + +IChatClient azureOpenAIChatClient = azureOpenAIClient + .GetChatClient(azureOpenAIDeployment) + .AsIChatClient(); + +builder.Services.AddChatClient(azureOpenAIChatClient); + +// Register Smart Rich Text Editor Components with Azure OpenAI +builder.Services.AddSingleton(); + +var app = builder.Build(); + +// ... rest of your application setup +``` + +### Step 2: Add Imports to _Imports.razor + +Update **~/_Imports.razor**: + +```razor +@using Syncfusion.Blazor +@using Syncfusion.Blazor.SmartRichTextEditor +``` + +### Step 3: Use Azure OpenAI with Smart Rich Text Editor Component + +```razor +@page "/editor" +@using Syncfusion.Blazor.SmartRichTextEditor + + + +
    + Tips: +
      +
    • Select text and click AI Commands for quick improvements
      • +
    • Press Alt+Enter to open AI Query dialog
      • +
    • Use AI to fix grammar, adjust tone, or rephrase content
      • +
    +

    Welcome to the Smart Rich Text Editor — try selecting a sentence to see AI suggestions.

    +
    +     +``` + +## Using Configuration Files + +### appsettings.json + +Store Azure credentials in configuration: + +```json +{ + "AzureOpenAI": { + "Key": "your-azure-key", + "Endpoint": "https://your-resource-name.openai.azure.com/", + "DeploymentName": "your-deployment-name" + } +} +``` + +### Reading from Configuration + +```csharp +string azureOpenAIKey = builder.Configuration["AzureOpenAI:Key"]; +string azureOpenAIEndpoint = builder.Configuration["AzureOpenAI:Endpoint"]; +string azureOpenAIDeployment = builder.Configuration["AzureOpenAI:DeploymentName"]; + +AzureOpenAIClient azureOpenAIClient = new AzureOpenAIClient( + new Uri(azureOpenAIEndpoint), + new ApiKeyCredential(azureOpenAIKey) +); + +IChatClient azureOpenAIChatClient = azureOpenAIClient + .GetChatClient(azureOpenAIDeployment) + .AsIChatClient(); + +builder.Services.AddChatClient(azureOpenAIChatClient); +``` + +## Using User Secrets + +For development environment: + +```bash +dotnet user-secrets init +dotnet user-secrets set "AzureOpenAI:Key" "your-azure-key" +dotnet user-secrets set "AzureOpenAI:Endpoint" "https://your-resource-name.openai.azure.com/" +dotnet user-secrets set "AzureOpenAI:DeploymentName" "your-deployment-name" +``` + +## Advanced Configuration + +### Model Parameters + +Customize Azure OpenAI behavior: + +```csharp +builder.Services.AddSingleton(); + .ConfigureAzureOpenAI(options => + { + options.MaxTokens = 500; + options.Temperature = 0.7f; + options.TopP = 0.9f; + options.FrequencyPenalty = 0.5f; + options.PresencePenalty = 0.5f; + }); +``` + +### Multiple Deployments + +For production with multiple model deployments: + +```csharp +// Create clients for different deployments +var gpt35Client = azureOpenAIClient + .GetChatClient("gpt-35-turbo-deployment") + .AsIChatClient(); + +var gpt4Client = azureOpenAIClient + .GetChatClient("gpt-4-deployment") + .AsIChatClient(); + +// Register based on use case +builder.Services.AddKeyedSingleton("gpt-35-turbo", gpt35Client); +builder.Services.AddKeyedSingleton("gpt-4", gpt4Client); +``` + +## Supported Models + +Azure OpenAI supports various models: + +| Model | Deployment Name | Use Case | +|-------|-----------------|----------| +| GPT-4 | `gpt-4` | Complex reasoning, high quality | +| GPT-4 Turbo | `gpt-4-turbo` | Latest capabilities, balanced cost | +| GPT-3.5 Turbo | `gpt-35-turbo` | Fast responses, cost-effective | + +## Security Features + +### Benefits of Azure OpenAI + +1. **Enterprise Security** + - Virtual Network support + - Private endpoints + - Azure security standards + +2. **Compliance** + - HIPAA compliance + - SOC 2 compliance + - Data residency options + +3. **Access Control** + - Azure AD integration + - Role-based access control + - Managed identities + +4. **Monitoring** + - Azure Monitor integration + - Detailed logging + - Usage analytics + +### Managed Identity (Recommended) + +For enhanced security using Managed Identity: + +```csharp +// Enable Managed Identity in Azure +// In Program.cs +using Azure.Identity; + +var credential = new DefaultAzureCredential(); +string azureOpenAIEndpoint = "https://your-resource-name.openai.azure.com/"; +string azureOpenAIDeployment = "your-deployment-name"; + +AzureOpenAIClient azureOpenAIClient = new AzureOpenAIClient( + new Uri(azureOpenAIEndpoint), + credential +); + +IChatClient azureOpenAIChatClient = azureOpenAIClient + .GetChatClient(azureOpenAIDeployment) + .AsIChatClient(); + +builder.Services.AddChatClient(azureOpenAIChatClient); +``` + +## Troubleshooting + +### Common Issues + +**Error: ResourceNotFound (404)** +- Verify endpoint URL is correct +- Check resource name matches your Azure resource +- Ensure resource exists in specified region + +**Error: InvalidAuthenticationTokenTenant (401)** +- Verify API key is correct +- Check key hasn't expired +- Ensure using the correct region's key + +**Error: Model not found (404)** +- Verify deployment name matches your Azure deployment +- Check deployment is active and ready +- Ensure model is properly deployed + +**Timeout Issues** +- Check Azure OpenAI resource capacity +- Verify network connectivity +- Consider timeout configuration + +## Monitoring and Analytics + +### Azure Monitor Integration + +Monitor your Smart Rich Text Editor usage: + +1. Go to your Azure OpenAI resource +2. Click **Monitoring** > **Metrics** +3. View metrics: + - Requests per minute + - Token usage + - Response latency + - Error rates + +### Application Insights + +Add Application Insights for detailed tracing: + +```csharp +builder.Services.AddApplicationInsightsTelemetry(); +``` + +## Cost Optimization + +### Tips for Azure OpenAI + +1. **Right-size deployments**: Use appropriate TPM (tokens per minute) +2. **Monitor usage**: Check metrics regularly +3. **Use appropriate models**: GPT-3.5-turbo for most cases +4. **Implement caching**: Reduce repeated requests +5. **Batch operations**: Process multiple requests efficiently + +## See also + +* [Getting Started with Smart Rich Text Editor](getting-started.md) +* [OpenAI Configuration](openai-service.md) +* [AI Features and Customization](ai-features.md) +* [Azure OpenAI Documentation](https://learn.microsoft.com/en-us/azure/ai-services/openai/) \ No newline at end of file diff --git a/blazor/smart-rich-text-editor/custom-inference-backend.md b/blazor/smart-rich-text-editor/custom-inference-backend.md new file mode 100644 index 0000000000..4a053f657f --- /dev/null +++ b/blazor/smart-rich-text-editor/custom-inference-backend.md @@ -0,0 +1,544 @@ +--- +layout: post +title: Custom Inference Backend for Syncfusion Smart Rich Text Editor +description: Learn how to configure custom AI inference back-ends for the Syncfusion Blazor Smart Rich Text Editor component. +platform: Blazor +control: Smart Rich Text Editor +documentation: ug +--- + +# Custom Inference Backend + +The Syncfusion® Blazor Smart Rich Text Editor supports custom inference back-ends, allowing you to integrate with proprietary AI services or internal AI infrastructure. + +## Overview + +Custom back-ends enable: + +- Integration with internal AI services +- Compliance with corporate data policies +- Custom model fine-tuning +- Specialized domain models +- Proprietary inference engines + +## Prerequisites + +* Existing AI/ML inference service +* API endpoint for inference +* Authentication credentials (if required) +* Understanding of your backend's API format + +## Creating a Custom Backend + +### Step 1: Implement IChatClient Interface + +Create a custom class implementing `IChatClient`: + +```csharp +using Microsoft.Extensions.AI; +using System.Collections.Generic; +using System.Threading.Tasks; + +namespace YourNamespace.AI +{ + public class CustomAIBackend : IChatClient + { + private readonly string _endpoint; + private readonly string _apiKey; + private readonly HttpClient _httpClient; + + public CustomAIBackend(string endpoint, string apiKey) + { + _endpoint = endpoint; + _apiKey = apiKey; + _httpClient = new HttpClient(); + } + + public async ValueTask CompleteAsync( + IList chatMessages, + ChatOptions? options = null, + CancellationToken cancellationToken = default) + { + // Format request for your custom backend + var requestBody = new + { + messages = ConvertChatMessages(chatMessages), + max_tokens = options?.MaxCompletionTokens ?? 500, + temperature = options?.Temperature ?? 0.7f + }; + + // Call your custom endpoint + var request = new HttpRequestMessage(HttpMethod.Post, _endpoint) + { + Content = new StringContent( + System.Text.Json.JsonSerializer.Serialize(requestBody), + System.Text.Encoding.UTF8, + "application/json" + ) + }; + + // Add authentication if needed + if (!string.IsNullOrEmpty(_apiKey)) + { + request.Headers.Add("Authorization", $"Bearer {_apiKey}"); + } + + var response = await _httpClient.SendAsync(request, cancellationToken); + response.EnsureSuccessStatusCode(); + + var responseContent = await response.Content.ReadAsStringAsync(); + var result = System.Text.Json.JsonSerializer.Deserialize(responseContent); + + return new ChatCompletion(new ChatMessage( + ChatRole.Assistant, + result?.choices?[0]?.message?.content ?? "No response" + )); + } + + private object ConvertChatMessages(IList messages) + { + return messages.Select(m => new + { + role = m.Role.ToString().ToLower(), + content = m.Content?[0]?.Text ?? string.Empty + }).ToList(); + } + + public void Dispose() + { + _httpClient?.Dispose(); + } + } + + public class CustomBackendResponse + { + public Choice[]? choices { get; set; } + } + + public class Choice + { + public Message? message { get; set; } + } + + public class Message + { + public string? content { get; set; } + } +} +``` + +### Step 2: Register in Program.cs + +Register your custom backend: + +```csharp +using Syncfusion.Blazor; +using Syncfusion.Blazor.AI; +using YourNamespace.AI; + +var builder = WebApplication.CreateBuilder(args); + +// Add services +builder.Services.AddRazorPages(); +builder.Services.AddServerSideBlazor(); +builder.Services.AddSyncfusionBlazor(); + +// Register custom backend +string customEndpoint = "https://your-ai-service.com/api/inference"; +string customApiKey = "your-api-key"; + +var customBackend = new CustomAIBackend(customEndpoint, customApiKey); +builder.Services.AddChatClient(customBackend); + +// Register Smart Rich Text Editor Components +builder.Services.AddSingleton(); + +var app = builder.Build(); + +// ... rest of setup +``` + +### Step 3: Use in Blazor Component + +```razor +@page "/editor" +@using Syncfusion.Blazor.SmartRichTextEditor + + + +
    +

    Welcome to the AI-assisted editor

    +

    Try selecting a sentence and use AI Commands to improve tone or clarity.

    +
      +
    • Select text and request suggestions
      • +
    • Press Alt+Enter for AI Query
      • +
    +
    +     +``` + +## Advanced Implementation + +### With Configuration Support + +```csharp +public class CustomAIBackend : IChatClient +{ + private readonly IConfiguration _configuration; + private readonly string _endpoint; + private readonly string _apiKey; + private readonly HttpClient _httpClient; + + public CustomAIBackend(IConfiguration configuration) + { + _configuration = configuration; + _endpoint = _configuration["CustomAI:Endpoint"] + ?? throw new InvalidOperationException("CustomAI:Endpoint not configured"); + _apiKey = _configuration["CustomAI:ApiKey"] + ?? throw new InvalidOperationException("CustomAI:ApiKey not configured"); + _httpClient = new HttpClient(); + } + + public async ValueTask CompleteAsync( + IList chatMessages, + ChatOptions? options = null, + CancellationToken cancellationToken = default) + { + // Implementation... + throw new NotImplementedException(); + } + + public void Dispose() + { + _httpClient?.Dispose(); + } +} +``` + +### Configuration in appsettings.json + +```json +{ + "CustomAI": { + "Endpoint": "https://your-ai-service.com/api/inference", + "ApiKey": "your-api-key", + "MaxRetries": 3, + "TimeoutSeconds": 30 + } +} +``` + +### Registration with Configuration + +```csharp +var customBackend = new CustomAIBackend(builder.Configuration); +builder.Services.AddSingleton(customBackend); +``` + +## Streaming Responses + +For better UX with custom back-ends: + +```csharp +public class StreamingCustomAIBackend : IChatClient +{ + private readonly string _endpoint; + private readonly HttpClient _httpClient; + + public StreamingCustomAIBackend(string endpoint) + { + _endpoint = endpoint; + _httpClient = new HttpClient(); + } + + public async ValueTask CompleteAsync( + IList chatMessages, + ChatOptions? options = null, + CancellationToken cancellationToken = default) + { + var request = new HttpRequestMessage(HttpMethod.Post, _endpoint); + request.Content = new StringContent( + System.Text.Json.JsonSerializer.Serialize(new { messages = chatMessages }), + System.Text.Encoding.UTF8, + "application/json" + ); + + // Enable streaming + request.Headers.Add("Accept", "text/event-stream"); + + var response = await _httpClient.SendAsync( + request, + HttpCompletionOption.ResponseHeadersRead, + cancellationToken + ); + + using var stream = await response.Content.ReadAsStreamAsync(cancellationToken); + using var reader = new System.IO.StreamReader(stream); + + var fullContent = new System.Text.StringBuilder(); + + while (!reader.EndOfStream) + { + var line = await reader.ReadLineAsync(); + if (line?.StartsWith("data: ") == true) + { + var data = line.Substring("data: ".Length); + var chunk = System.Text.Json.JsonSerializer.Deserialize(data); + if (chunk?.content != null) + { + fullContent.Append(chunk.content); + } + } + } + + return new ChatCompletion(new ChatMessage( + ChatRole.Assistant, + fullContent.ToString() + )); + } + + public void Dispose() + { + _httpClient?.Dispose(); + } + + private class StreamChunk + { + public string? content { get; set; } + } +} +``` + +## Error Handling + +Implement robust error handling: + +```csharp +public class RobustCustomAIBackend : IChatClient +{ + private readonly string _endpoint; + private readonly HttpClient _httpClient; + private readonly ILogger _logger; + private readonly int _maxRetries = 3; + + public RobustCustomAIBackend(string endpoint, ILogger logger) + { + _endpoint = endpoint; + _logger = logger; + _httpClient = new HttpClient(); + } + + public async ValueTask CompleteAsync( + IList chatMessages, + ChatOptions? options = null, + CancellationToken cancellationToken = default) + { + for (int attempt = 1; attempt <= _maxRetries; attempt++) + { + try + { + return await AttemptCompletion(chatMessages, options, cancellationToken); + } + catch (HttpRequestException ex) when (attempt < _maxRetries) + { + _logger.LogWarning($"Attempt {attempt} failed: {ex.Message}. Retrying..."); + await Task.Delay(TimeSpan.FromSeconds(attempt * 2), cancellationToken); + } + catch (Exception ex) + { + _logger.LogError($"Unrecoverable error: {ex.Message}"); + throw; + } + } + + throw new InvalidOperationException("Failed to get response after max retries"); + } + + private async ValueTask AttemptCompletion( + IList chatMessages, + ChatOptions? options, + CancellationToken cancellationToken) + { + using var cts = CancellationTokenSource.CreateLinkedTokenSource(cancellationToken); + cts.CancelAfter(TimeSpan.FromSeconds(30)); // 30 second timeout + + var request = new HttpRequestMessage(HttpMethod.Post, _endpoint); + // ... request setup ... + + var response = await _httpClient.SendAsync(request, cts.Token); + response.EnsureSuccessStatusCode(); + + var content = await response.Content.ReadAsStringAsync(); + return ParseResponse(content); + } + + private ChatCompletion ParseResponse(string content) + { + try + { + var result = System.Text.Json.JsonSerializer.Deserialize(content); + return new ChatCompletion(new ChatMessage( + ChatRole.Assistant, + result?.output ?? "No response" + )); + } + catch (System.Text.Json.JsonException ex) + { + _logger.LogError($"Failed to parse response: {ex.Message}"); + throw; + } + } + + public void Dispose() + { + _httpClient?.Dispose(); + } + + private class CustomResponse + { + public string? output { get; set; } + } +} +``` + +## Example: Internal AI Service Integration + +### Scenario: Company-Internal AI API + +```csharp +public class CorporateAIBackend : IChatClient +{ + private readonly string _internalEndpoint; + private readonly string _bearerToken; + private readonly HttpClient _httpClient; + private readonly ILogger _logger; + + public CorporateAIBackend(string endpoint, string token, ILogger logger) + { + _internalEndpoint = endpoint; + _bearerToken = token; + _logger = logger; + _httpClient = new HttpClient(); + _httpClient.Timeout = TimeSpan.FromSeconds(60); + } + + public async ValueTask CompleteAsync( + IList chatMessages, + ChatOptions? options = null, + CancellationToken cancellationToken = default) + { + try + { + var requestBody = new + { + messages = chatMessages.Select(m => new + { + role = m.Role == ChatRole.User ? "user" : "assistant", + content = m.Content?[0]?.Text ?? string.Empty + }), + parameters = new + { + temperature = options?.Temperature ?? 0.7f, + max_tokens = options?.MaxCompletionTokens ?? 1000 + } + }; + + var request = new HttpRequestMessage(HttpMethod.Post, _internalEndpoint); + request.Headers.Authorization = new System.Net.Http.Headers.AuthenticationHeaderValue("Bearer", _bearerToken); + request.Content = new StringContent( + System.Text.Json.JsonSerializer.Serialize(requestBody), + System.Text.Encoding.UTF8, + "application/json" + ); + + var response = await _httpClient.SendAsync(request, cancellationToken); + + if (!response.IsSuccessStatusCode) + { + _logger.LogError($"API Error: {response.StatusCode}"); + throw new InvalidOperationException($"AI service returned: {response.StatusCode}"); + } + + var content = await response.Content.ReadAsStringAsync(cancellationToken); + var result = System.Text.Json.JsonSerializer.Deserialize(content); + + return new ChatCompletion(new ChatMessage( + ChatRole.Assistant, + result?.result?.text ?? "Unable to generate response" + )); + } + catch (Exception ex) + { + _logger.LogError($"Error calling corporate AI: {ex.Message}"); + throw; + } + } + + public void Dispose() + { + _httpClient?.Dispose(); + } + + private class CorporateResponse + { + public ResultData? result { get; set; } + } + + private class ResultData + { + public string? text { get; set; } + } +} +``` + +## Testing Your Custom Backend + +### Unit Test Example + +```csharp +[TestClass] +public class CustomAIBackendTests +{ + [TestMethod] + public async Task CompleteAsync_WithValidMessages_ReturnsResponse() + { + // Arrange + var backend = new CustomAIBackend("https://test-endpoint.com", "test-key"); + var messages = new List + { + new ChatMessage(ChatRole.User, "Hello") + }; + + // Act + var result = await backend.CompleteAsync(messages); + + // Assert + Assert.IsNotNull(result); + Assert.IsNotNull(result.Content); + } + + [TestMethod] + public async Task CompleteAsync_WithInvalidEndpoint_ThrowsException() + { + // Arrange + var backend = new CustomAIBackend("https://invalid-endpoint.com", "test-key"); + var messages = new List + { + new ChatMessage(ChatRole.User, "Test") + }; + + // Act & Assert + await Assert.ThrowsExceptionAsync( + () => backend.CompleteAsync(messages).AsTask() + ); + } +} +``` + +## See also + +* [Getting Started with Smart Rich Text Editor](getting-started.md) +* [OpenAI Configuration](openai-service.md) +* [Azure OpenAI Configuration](azure-openai-service.md) +* [Ollama Configuration](ollama.md) +* [AI Features and Customization](ai-features.md) \ No newline at end of file diff --git a/blazor/smart-rich-text-editor/ollama.md b/blazor/smart-rich-text-editor/ollama.md new file mode 100644 index 0000000000..4950d9c84b --- /dev/null +++ b/blazor/smart-rich-text-editor/ollama.md @@ -0,0 +1,386 @@ +--- +layout: post +title: Ollama Configuration for Blazor Smart Rich Text Editor | Syncfusion +description: Step-by-step guide to configure Ollama for the Syncfusion Blazor Smart Rich Text Editor, covering installation, client setup, and usage examples for local AI. +platform: Blazor +control: Smart Rich Text Editor +documentation: ug +--- + +# Ollama Configuration + +The Syncfusion® Blazor Smart Rich Text Editor supports Ollama for running open-source models locally. This is ideal for privacy-conscious applications and development environments. + +## What is Ollama? + +Ollama is a lightweight, open-source framework for running large language models locally on your machine. It provides: + +- **Privacy**: All processing happens locally +- **No API costs**: No cloud service fees +- **Offline capability**: Works without internet connection +- **Model variety**: Access to many open-source models +- **Easy setup**: Simple installation and management + +## Prerequisites + +* Windows 10/11, macOS, or Linux +* At least 8GB RAM (16GB+ recommended) +* GPU support optional but recommended for performance +* Docker (optional, for containerized deployment) + +## Installation + +### Step 1: Download and Install Ollama + +Visit [Ollama's Official Website](https://ollama.com) and download the installer for your operating system. + +#### Windows + +1. Download the Windows installer +2. Run the installer and follow the setup wizard +3. Accept default installation path or customize +4. Complete installation + +#### macOS + +1. Download the macOS installer +2. Open the `.dmg file` +3. Drag Ollama to Applications folder +4. Launch from Applications + +#### Linux + +Use the installation script: + +```bash +curl https://ollama.ai/install.sh | sh +``` + +### Step 2: Verify Installation + +Open a terminal/command prompt and verify: + +```bash +ollama --version +``` + +### Step 3: Start Ollama Service + +#### Windows + +Ollama starts automatically. Access at `http://localhost:11434` + +#### macOS/Linux + +Start the Ollama service: + +```bash +ollama serve +``` + +## Installing Models + +### Available Models + +Browse available models at [Ollama Library](https://ollama.com/library) + +Popular models for text generation: + +- **Mistral**: `mistral` - Fast, efficient +- **Llama 2**: `llama2` - General purpose +- **Neural Chat**: `neural-chat` - Conversational +- **Orca Mini**: `orca-mini` - Lightweight +- **Dolphin**: `dolphin-mixtral` - Advanced + +### Install a Model + +```bash +ollama pull mistral +``` + +This downloads and prepares the model (may take several minutes). + +### List Installed Models + +```bash +ollama list +``` + +### Run a Model Directly + +Test the model: + +```bash +ollama run mistral +``` + +Type your prompt and press Enter. Type `/bye` to exit. + +## Configuration in Blazor + +### Installation + +Install required NuGet packages: + +{% tabs %} +{% highlight c# tabtitle="Package Manager" %} + +Install-Package Syncfusion.Blazor.SmartRichTextEditor +Install-Package Syncfusion.Blazor.Themes +Install-Package Microsoft.Extensions.AI +Install-Package OllamaSharp + +{% endhighlight %} +{% endtabs %} + +### Setup in Program.cs + +```csharp +using Syncfusion.Blazor; +using Syncfusion.Blazor.AI; +using Microsoft.Extensions.AI; +using OllamaSharp; + +var builder = WebApplication.CreateBuilder(args); + +// Add services +builder.Services.AddRazorPages(); +builder.Services.AddServerSideBlazor(); +builder.Services.AddSyncfusionBlazor(); + +// Configure Ollama +string ollamaEndpoint = "http://localhost:11434"; +string modelName = "mistral"; // or any other installed model + +// Create Ollama client +IOllamaApiClient ollamaClient = new OllamaApiClient(ollamaEndpoint, modelName); + +// Convert to IChatClient +IChatClient chatClient = ollamaClient; + +builder.Services.AddChatClient(chatClient); + +// Register Smart Rich Text Editor Components with Azure OpenAI +builder.Services.AddSingleton(); + +var app = builder.Build(); + +// ... rest of setup +``` + +### Add to _Imports.razor + +```razor +@using Syncfusion.Blazor +@using Syncfusion.Blazor.SmartRichTextEditor +``` + +### Use Ollama AI with Smart Rich Text Editor Component + +```razor +@page "/editor" +@using Syncfusion.Blazor.SmartRichTextEditor + + + +
    +

    Ollama + Smart Rich Text Editor

    +

    Use AI Commands to improve selected text or change tone.

    +
      +
    • Select text and request suggestions
      • +
    • Press Alt+Enter for AI Query
      • +
    +
    +     +``` + +## Configuration Options + +### Custom Endpoint + +If Ollama runs on different host/port: + +```csharp +string ollamaEndpoint = "http://192.168.1.100:11434"; +IOllamaApiClient ollamaClient = new OllamaApiClient(ollamaEndpoint, "mistral"); +``` + +### Model Parameters + +Configure model behavior: + +```csharp +// Create client with parameters +IOllamaApiClient ollamaClient = new OllamaApiClient("http://localhost:11434", "mistral"); + +// Set parameters (varies by OllamaSharp version) +var request = new GenerateRequest +{ + Model = "mistral", + Prompt = "Your prompt", + Temperature = 0.7f, + TopK = 40, + TopP = 0.9f, +}; +``` + +### Dynamic Model Selection + +```csharp +// Read model name from configuration +string modelName = builder.Configuration["Ollama:ModelName"] ?? "mistral"; +string endpoint = builder.Configuration["Ollama:Endpoint"] ?? "http://localhost:11434"; + +IOllamaApiClient ollamaClient = new OllamaApiClient(endpoint, modelName); +``` + +## Docker Deployment + +### Docker Compose Setup + +```yaml +version: '3.8' +services: + ollama: + image: ollama/ollama:latest + ports: + - "11434:11434" + volumes: + - ollama_data:/root/.ollama + environment: + - OLLAMA_HOST=0.0.0.0:11434 + +volumes: + ollama_data: +``` + +### Start with Docker + +```bash +docker-compose up -d +``` + +### Pull Models in Container + +```bash +docker exec -it ollama pull mistral +``` + +## Advanced Configuration + +### Multiple Models + +Run different models for different purposes: + +```csharp +// For general editing +var generalClient = new OllamaApiClient("http://localhost:11434", "mistral"); + +// For code-related tasks +var codeClient = new OllamaApiClient("http://localhost:11434", "neural-chat"); + +// Conditionally use based on context +IChatClient chatClient = useForCode ? codeClient : generalClient; +``` + +## Performance Optimization + +### Tips for Better Performance + +1. **Use GPU Acceleration** + - NVIDIA GPUs: Install CUDA toolkit + - AMD GPUs: Install ROCm + - Intel GPUs: Install Intel oneAPI + +2. **Increase Memory Allocation** + ```bash + # Linux/macOS + export OLLAMA_NUM_PARALLEL=1 + ollama serve + ``` + +3. **Choose Efficient Models** + - Mistral is fastest + - `Orca-mini` for lightweight deployment + - `Dolphin-mixtral` for best quality + +4. **Implement Caching** + - Cache common prompts + - Reuse model responses + +5. **Optimize Request Size** + - Keep prompts concise + - Limit response length with max tokens + +## Troubleshooting + +### Connection Issues + +**Error: Unable to connect to Ollama** + +1. Verify Ollama is running: + ```bash + curl http://localhost:11434/api/tags + ``` + +2. Check endpoint configuration in Program.cs + +3. If running on different machine, update endpoint URL + +### Model Issues + +**Model not found** +```bash +# List available models +ollama list + +# Pull the model +ollama pull mistral +``` + +**Out of memory errors** +- Use smaller models (`orca-mini`, `mistral`) +- Close other applications +- Restart Ollama service + +### Performance Issues + +**Slow response times** +- Check CPU/GPU usage +- Enable GPU acceleration if available +- Use faster models +- Increase system RAM + +## Model Recommendations + +| Use Case | Model | Pros | Cons | +|----------|-------|------|------| +| General editing | Mistral | Fast, good quality | Less context | +| Content creation | Llama 2 | Balanced | Larger model | +| Code assistance | Neural-chat | Good reasoning | Slower | +| Lightweight | Orca-mini | Very fast | Limited capability | +| Best quality | Dolphin-mixtral | Excellent | Resource heavy | + +## Security Considerations + +### Local Processing Benefits + +- All data stays on your machine +- No external API calls +- Compliance with data regulations +- Full control over data retention + +### Best Practices + +1. Restrict network access to Ollama +2. Use firewall rules if networked +3. Keep Ollama updated +4. Monitor resource usage + +## See also + +* [Getting Started with Smart Rich Text Editor](getting-started.md) +* [OpenAI Configuration](openai-service.md) +* [Azure OpenAI Configuration](azure-openai-service.md) +* [AI Features and Customization](ai-features.md) +* [Ollama Official Documentation](https://ollama.ai/docs) \ No newline at end of file diff --git a/blazor/smart-rich-text-editor/openai-service.md b/blazor/smart-rich-text-editor/openai-service.md new file mode 100644 index 0000000000..aeb776a22c --- /dev/null +++ b/blazor/smart-rich-text-editor/openai-service.md @@ -0,0 +1,236 @@ +--- +layout: post +title: OpenAI Configuration for Blazor Smart Rich Text Editor | Syncfusion +description: Configure OpenAI for Syncfusion Blazor Smart Rich Text Editor: API keys, client setup, DI registration, usage examples, and best practices. +platform: Blazor +control: Smart Rich Text Editor +documentation: ug +--- + +# OpenAI Configuration + +The Syncfusion® Blazor Smart Rich Text Editor supports OpenAI's GPT models for intelligent content assistance. + +## Prerequisites + +* Active OpenAI account with API access +* Valid API key from OpenAI +* Internet connectivity to access OpenAI services + +## Get Your OpenAI API Key + +1. Visit [OpenAI Platform](https://platform.openai.com/) +2. Sign in with your OpenAI account (create one if needed) +3. Navigate to **API Keys** section +4. Click **Create new secret key** +5. Copy and securely store your API key + +## Supported Models + +OpenAI offers various models for different use cases: + +- `gpt-4-turbo`: Most capable, latest model +- `gpt-4`: Advanced reasoning and complex tasks +- `gpt-3.5-turbo`: Fast and cost-effective +- `gpt-3.5-turbo-16k`: Extended context window + +> **Note**: Model availability and names may change. Refer to [OpenAI documentation](https://developers.openai.com/api/docs/models) for current options. + +## Installation + +Install the required NuGet packages: + +{% tabs %} +{% highlight c# tabtitle="Package Manager" %} + +Install-Package Syncfusion.Blazor.SmartRichTextEditor +Install-Package Syncfusion.Blazor.Themes +Install-Package Microsoft.Extensions.AI +Install-Package Microsoft.Extensions.AI.OpenAI + +{% endhighlight %} +{% endtabs %} + +## Configuration + +### Step 1: Setup in Program.cs + +Add the following configuration to your **Program.cs** file: + +```csharp +using Syncfusion.Blazor; +using Syncfusion.Blazor.AI; +using Microsoft.Extensions.AI; +using OpenAI; + +var builder = WebApplication.CreateBuilder(args); + +// Add services to the container +builder.Services.AddRazorPages(); +builder.Services.AddServerSideBlazor(); + +// Register Syncfusion Blazor Service +builder.Services.AddSyncfusionBlazor(); + +// Configure OpenAI +string openAIApiKey = "your-api-key-here"; +string openAIModel = "gpt-3.5-turbo"; // or gpt-4, gpt-4-turbo, etc. + +OpenAIClient openAIClient = new OpenAIClient(openAIApiKey); +IChatClient openAIChatClient = openAIClient.GetChatClient(openAIModel).AsIChatClient(); +builder.Services.AddChatClient(openAIChatClient); + +// Register Smart Rich Text Editor Components with OpenAI +builder.Services.AddSingleton(); + + +var app = builder.Build(); + +// ... rest of your application setup +``` + +### Step 2: Add Imports to _Imports.razor + +Update **~/_Imports.razor** to include necessary namespaces: + +```razor +@using Syncfusion.Blazor +@using Syncfusion.Blazor.SmartRichTextEditor +``` + +### Step 3: Use OpenAI with Smart Rich Text Editor Component + +Add the Smart Rich Text Editor to your Blazor page: + +```razor +@page "/editor" +@using Syncfusion.Blazor.SmartRichTextEditor + + + +
    + Tips: +
      +
    • Select text and click "AI Commands" for quick improvements
      • +
    • Press Alt+Enter to open AI Query dialog
      • +
    • Use AI to fix grammar, adjust tone, or rephrase content
      • +
    +
    +     +``` + +## Using Environment Variables + +For security, store your API key in environment variables: + +### Windows + +```powershell +$env:OPENAI_API_KEY = "your-api-key" +``` + +### Linux/macOS + +```bash +export OPENAI_API_KEY="your-api-key" +``` + +### Reading from Environment in Program.cs + +```csharp +string openAIApiKey = Environment.GetEnvironmentVariable("OPENAI_API_KEY") + ?? throw new InvalidOperationException("OpenAI API key not found in environment variables"); +``` + +## Using User Secrets + +For development, use User Secrets: + +```bash +dotnet user-secrets init +dotnet user-secrets set "OpenAI:ApiKey" "your-api-key" +``` + +Access in Program.cs: + +```csharp +var openAIApiKey = builder.Configuration["OpenAI:ApiKey"]; +``` + +## Advanced Configuration + +### Model Parameters + +Customize OpenAI model behavior: + +```csharp +builder.Services.AddSingleton(); + .ConfigureOpenAI(options => + { + options.MaxTokens = 500; // Limit response length + options.Temperature = 0.7f; // Creativity level (0-2) + options.TopP = 0.9f; // Diversity control + options.FrequencyPenalty = 0.5f; // Repeat penalty + options.PresencePenalty = 0.5f; // New topic penalty + }); +``` + +### Parameter Explanation + +| Parameter | Range | Default | Description | +|-----------|-------|---------|-------------| +| MaxTokens | 1-4096 | 500 | Maximum length of AI response | +| Temperature | 0-2 | 0.7 | Higher = more creative, Lower = more focused | +| TopP | 0-1 | 0.9 | Nucleus sampling - controls diversity | +| FrequencyPenalty | -2-2 | 0 | Reduces likelihood of repetition | +| PresencePenalty | -2-2 | 0 | Encourages new topics | + +## Troubleshooting + +### Common Issues + +**Authentication Error: 401 Unauthorized** +- Verify API key is correct +- Check API key is still valid and hasn't expired +- Ensure API key has necessary permissions + +**Rate Limit Error: 429 Too Many Requests** +- Reduce request frequency +- Implement request throttling +- Check usage limits on OpenAI account + +**Model Not Found Error** +- Verify model name is correct (check [model list](https://developers.openai.com/api/docs/models)) +- Ensure model is available in your region/account + +**No Response from AI** +- Check internet connectivity +- Verify API key permissions +- Review OpenAI service status + +## Cost Optimization + +### Tips for Reducing Costs + +1. **Use Efficient Models**: `gpt-3.5-turbo` is significantly cheaper than `gpt-4` +2. **Limit Token Count**: Set reasonable `MaxTokens` values +3. **Implement Caching**: Cache common requests +4. **Monitor Usage**: Track API calls and costs +5. **Batch Operations**: Process multiple requests efficiently + +## Security Best Practices + +1. **Never hardcode API keys** - Use environment variables or secure configuration +2. **Use User Secrets** for development - Not for production +3. **Implement Rate Limiting** - Prevent abuse of your application +4. **Validate User Input** - Before sending to OpenAI +5. **Review Content** - Always review AI-generated content before publishing +6. **Set Spending Limits** - Configure usage limits on OpenAI account + +## See also + +* [Getting Started with Smart Rich Text Editor](getting-started.md) +* [AI Features and Customization](ai-features.md) +* [Azure OpenAI Configuration](azure-openai-service.md) +* [Ollama Configuration](ollama.md) +* [OpenAI Documentation](https://developers.openai.com/api/docs) \ No newline at end of file From bb2ced25a5f72b05d14925e57ab30e96240159e1 Mon Sep 17 00:00:00 2001 From: keerthikarankv <100576900+keerthikarankv@users.noreply.github.com> Date: Wed, 18 Mar 2026 13:57:52 +0530 Subject: [PATCH 2/3] UG(1015625): AI Services UG for Smart Rich Text Editor component. --- blazor/smart-rich-text-editor/getting-started.md | 2 -- 1 file changed, 2 deletions(-) diff --git a/blazor/smart-rich-text-editor/getting-started.md b/blazor/smart-rich-text-editor/getting-started.md index 7a3ba8e642..c694a10050 100644 --- a/blazor/smart-rich-text-editor/getting-started.md +++ b/blazor/smart-rich-text-editor/getting-started.md @@ -108,8 +108,6 @@ Open **~/_Imports.razor** file and import the `Syncfusion.Blazor` and `Syncfusio Now, register the Syncfusion® Blazor Service in the **~/Program.cs** file of your Blazor Server App. -Now, register the Syncfusion® Blazor Service in the **~/Program.cs** file of your Blazor Server App. - {% tabs %} {% highlight C# tabtitle="Blazor Server App" hl_lines="4 11" %} From d604b3ab0a508bc96277706b8ea760b432dd7e7e Mon Sep 17 00:00:00 2001 From: keerthikarankv <100576900+keerthikarankv@users.noreply.github.com> Date: Wed, 18 Mar 2026 18:36:59 +0530 Subject: [PATCH 3/3] UG(1015625): AI Services UG for Smart Rich Text Editor component. --- blazor-toc.html | 2 +- blazor/smart-rich-text-editor/getting-started.md | 9 ++++++++- 2 files changed, 9 insertions(+), 2 deletions(-) diff --git a/blazor-toc.html b/blazor-toc.html index 5208149fef..70e49f00ee 100644 --- a/blazor-toc.html +++ b/blazor-toc.html @@ -205,7 +205,7 @@
  • Getting Started
  • AI Services diff --git a/blazor/smart-rich-text-editor/getting-started.md b/blazor/smart-rich-text-editor/getting-started.md index c694a10050..25f0fca679 100644 --- a/blazor/smart-rich-text-editor/getting-started.md +++ b/blazor/smart-rich-text-editor/getting-started.md @@ -310,7 +310,14 @@ N> Check out the [Blazor Themes](https://blazor.syncfusion.com/documentation/app ## Add Syncfusion® Blazor Smart Rich Text Editor component -Add the Syncfusion® **Blazor Smart Rich Text Editor** component in the **~/Pages/Index.razor** file. +Add the Syncfusion® **Blazor Smart Rich Text Editor** component in the **~/Pages/Index.razor** file. If an interactivity location as `per page/component`, define a render mode at the top of the `Index.razor` page. + +N> If an Interactivity Location is set to `Global` and the **Render Mode** is set to `Server`, the render mode is configured in the `App.razor` file by default. + +``` +@* desired render mode define here *@ +@rendermode InteractiveServer +``` {% tabs %} {% highlight razor tabtitle="~/Index.razor" %}