Chat request refactoring for OpenAI-compatible providers (#722)

app/MindWork AI Studio/Provider/SelfHosted/ChatRequest.cs

@@ -1,20 +0,0 @@
-using System.Text.Json.Serialization;
-
-namespace AIStudio.Provider.SelfHosted;
-
-/// <summary>
-/// The chat request model.
-/// </summary>
-/// <param name="Model">Which model to use for chat completion.</param>
-/// <param name="Messages">The chat messages.</param>
-/// <param name="Stream">Whether to stream the chat completion.</param>
-public readonly record struct ChatRequest(
-    string Model,
-    IList<IMessageBase> Messages,
-    bool Stream
-)
-{
-    // Attention: The "required" modifier is not supported for [JsonExtensionData].
-    [JsonExtensionData]
-    public IDictionary<string, object> AdditionalApiParameters { get; init; } = new Dictionary<string, object>();
-}

app/MindWork AI Studio/Provider/SelfHosted/ProviderSelfHosted.cs

@@ -1,7 +1,5 @@
 using System.Net.Http.Headers;
 using System.Runtime.CompilerServices;
-using System.Text;
-using System.Text.Json;
 
 using AIStudio.Chat;
 using AIStudio.Provider.OpenAI;
@@ -25,58 +23,39 @@ public sealed class ProviderSelfHosted(Host host, string hostname) : BaseProvide
     /// <inheritdoc />
     public override async IAsyncEnumerable<ContentStreamChunk> StreamChatCompletion(Provider.Model chatModel, ChatThread chatThread, SettingsManager settingsManager, [EnumeratorCancellation] CancellationToken token = default)
     {
-        // Get the API key:
-        var requestedSecret = await RUST_SERVICE.GetAPIKey(this, SecretStoreType.LLM_PROVIDER, isTrying: true);
-        
-        // Prepare the system prompt:
-        var systemPrompt = new TextMessage
-        {
-            Role = "system",
-            Content = chatThread.PrepareSystemPrompt(settingsManager),
-        };
-        
-        // Parse the API parameters:
-        var apiParameters = this.ParseAdditionalApiParameters();
+        await foreach (var content in this.StreamOpenAICompatibleChatCompletion<ChatCompletionAPIRequest, ChatCompletionDeltaStreamLine, ChatCompletionAnnotationStreamLine>(
+                           "self-hosted provider",
+                           chatModel,
+                           chatThread,
+                           settingsManager,
+                           async (systemPrompt, apiParameters) =>
+                           {
+                               // Build the list of messages. The image format depends on the host:
+                               // - Ollama uses the direct image URL format: { "type": "image_url", "image_url": "data:..." }
+                               // - LM Studio, vLLM, and llama.cpp use the nested image URL format: { "type": "image_url", "image_url": { "url": "data:..." } }
+                               var messages = host switch
+                               {
+                                   Host.OLLAMA => await chatThread.Blocks.BuildMessagesUsingDirectImageUrlAsync(this.Provider, chatModel),
+                                   _ => await chatThread.Blocks.BuildMessagesUsingNestedImageUrlAsync(this.Provider, chatModel),
+                               };
 
-        // Build the list of messages. The image format depends on the host:
-        // - Ollama uses the direct image URL format: { "type": "image_url", "image_url": "data:..." }
-        // - LM Studio, vLLM, and llama.cpp use the nested image URL format: { "type": "image_url", "image_url": { "url": "data:..." } }
-        var messages = host switch
-        {
-            Host.OLLAMA => await chatThread.Blocks.BuildMessagesUsingDirectImageUrlAsync(this.Provider, chatModel),
-            _ => await chatThread.Blocks.BuildMessagesUsingNestedImageUrlAsync(this.Provider, chatModel),
-        };
-        
-        // Prepare the OpenAI HTTP chat request:
-        var providerChatRequest = JsonSerializer.Serialize(new ChatRequest
-        {
-            Model = chatModel.Id,
-            
-            // Build the messages:
-            // - First of all the system prompt
-            // - Then none-empty user and AI messages
-            Messages = [systemPrompt, ..messages],
-            
-            // Right now, we only support streaming completions:
-            Stream = true,
-            AdditionalApiParameters = apiParameters
-        }, JSON_SERIALIZER_OPTIONS);
+                               return new ChatCompletionAPIRequest
+                               {
+                                   Model = chatModel.Id,
 
-        async Task<HttpRequestMessage> RequestBuilder()
-        {
-            // Build the HTTP post request:
-            var request = new HttpRequestMessage(HttpMethod.Post, host.ChatURL());
+                                   // Build the messages:
+                                   // - First of all the system prompt
+                                   // - Then none-empty user and AI messages
+                                   Messages = [systemPrompt, ..messages],
 
-            // Set the authorization header:
-            if (requestedSecret.Success)
-                request.Headers.Authorization = new AuthenticationHeaderValue("Bearer", await requestedSecret.Secret.Decrypt(ENCRYPTION));
-
-            // Set the content:
-            request.Content = new StringContent(providerChatRequest, Encoding.UTF8, "application/json");
-            return request;
-        }
-        
-        await foreach (var content in this.StreamChatCompletionInternal<ChatCompletionDeltaStreamLine, ChatCompletionAnnotationStreamLine>("self-hosted provider", RequestBuilder, token))
+                                   // Right now, we only support streaming completions:
+                                   Stream = true,
+                                   AdditionalApiParameters = apiParameters
+                               };
+                           },
+                           isTryingSecret: true,
+                           requestPath: host.ChatURL(),
+                           token: token))
             yield return content;
     }
 
@@ -211,4 +190,4 @@ public sealed class ProviderSelfHosted(Host host, string hostname) : BaseProvide
                            filterPhrases.All( filter => model.Id.Contains(filter, StringComparison.InvariantCulture)))
             .Select(n => new Provider.Model(n.Id, null));
     }
-}
+}