Merge branch 'main' into main

Author: ex3ndr

README.md

@@ -1,6 +1,6 @@
 # Llama Coder
 
-Llama Coder is a better and self-hosted Github Copilot replacement for VS Studio Code. Llama Coder uses [Ollama](https://ollama.ai) and codellama to provide autocomplete that runs on your hardware. Works best with Mac M1/M2/M3 or with RTX 4090.
+Llama Coder is a better and self-hosted Github Copilot replacement for [VS Code](https://github.com/microsoft/vscode). Llama Coder uses [Ollama](https://ollama.ai) and codellama to provide autocomplete that runs on your hardware. Works best with Mac M1/M2/M3 or with RTX 4090.
 
 [VS Code Plugin](https://marketplace.visualstudio.com/items?itemName=ex3ndr.llama-coder)
 
@@ -14,7 +14,7 @@ Llama Coder is a better and self-hosted Github Copilot replacement for VS Studio
 
 Minimum required RAM: 16GB is a minimum, more is better since even smallest model takes 5GB of RAM.
 The best way: dedicated machine with RTX 4090. Install [Ollama](https://ollama.ai) on this machine and configure endpoint in extension settings to offload to this machine.
-Second best way: run on MacBook M1/M2/M3 with enougth RAM (more == better, but 10gb extra would be enougth).
+Second best way: run on MacBook M1/M2/M3 with enough RAM (more == better, but 10gb extra would be enough).
 For windows notebooks: it runs good with decent GPU, but dedicated machine with a good GPU is recommended. Perfect if you have a dedicated gaming PC.
 
 ## Local Installation

package.json

@@ -33,6 +33,24 @@
   "extensionKind": ["ui"],
   "main": "./out/extension.js",
   "contributes": {
+    "commands": [
+      {
+        "command": "llama.openSettings",
+        "title": "Llama Coder: Open Settings"
+      },
+      {
+        "command": "llama.pause",
+        "title": "Llama Coder: Pause"
+      },
+      {
+        "command": "llama.resume",
+        "title": "Llama Coder: Resume"
+      },
+      {
+        "command": "llama.toggle",
+        "title": "Llama Coder: Toggle"
+      }
+    ],
     "configuration": [
       {
         "title": "Llama coder",
@@ -119,7 +137,7 @@
               "codellama",
               "deepseek"
             ],
-            "default": "codellama",
+            "default": "stable-code",
             "description": "Custom model prompt format",
             "order": 5
           },

src/extension.ts

@@ -9,12 +9,12 @@ export function activate(context: vscode.ExtensionContext) {
 	info('Llama Coder is activated.');
 
 	// Create status bar
-	const openSettings = 'llama.openSettings';
-	context.subscriptions.push(vscode.commands.registerCommand(openSettings, () => {
+	context.subscriptions.push(vscode.commands.registerCommand('llama.openSettings', () => {
 		vscode.commands.executeCommand('workbench.action.openSettings', '@ext:ex3ndr.llama-coder');
 	}));
+
 	let statusBarItem = vscode.window.createStatusBarItem(vscode.StatusBarAlignment.Right, 100);
-	statusBarItem.command = openSettings;
+	statusBarItem.command = 'llama.toggle';
 	statusBarItem.text = `$(chip) Llama Coder`;
 	statusBarItem.show();
 	context.subscriptions.push(statusBarItem);
@@ -23,6 +23,17 @@ export function activate(context: vscode.ExtensionContext) {
 	const provider = new PromptProvider(statusBarItem, context);
 	let disposable = vscode.languages.registerInlineCompletionItemProvider({ pattern: '**', }, provider);
 	context.subscriptions.push(disposable);
+
+	context.subscriptions.push(vscode.commands.registerCommand('llama.pause', () => {
+		provider.paused = true;
+	}));
+	context.subscriptions.push(vscode.commands.registerCommand('llama.resume', () => {
+		provider.paused = false;
+	}));
+	context.subscriptions.push(vscode.commands.registerCommand('llama.toggle', () => {
+		provider.paused = !provider.paused;
+	}));
+
 }
 
 export function deactivate() {

src/prompts/provider.ts

@@ -9,16 +9,49 @@ import { ollamaCheckModel } from '../modules/ollamaCheckModel';
 import { ollamaDownloadModel } from '../modules/ollamaDownloadModel';
 import { config } from '../config';
 
+type Status = {
+    icon: string;
+    text: string;
+};
+
 export class PromptProvider implements vscode.InlineCompletionItemProvider {
 
     lock = new AsyncLock();
     statusbar: vscode.StatusBarItem;
     context: vscode.ExtensionContext;
+    private _paused: boolean = false;
+    private _status: Status = { icon: "chip", text: "Llama Coder" };
 
     constructor(statusbar: vscode.StatusBarItem, context: vscode.ExtensionContext) {
         this.statusbar = statusbar;
         this.context = context;
     }
+    
+    public set paused(value: boolean) {
+        this._paused = value;
+        this.update();
+    }
+
+    public get paused(): boolean {
+        return this._paused;
+    }
+
+    private update(icon?: string, text?: string): void {
+        this._status.icon = icon ? icon : this._status.icon;
+        this._status.text = text ? text : this._status.text;
+
+        let statusText = '';
+        let statusTooltip = '';
+        if (this._paused) {
+            statusText = `$(sync-ignored) ${this._status.text}`;
+            statusTooltip = `${this._status.text} (Paused)`;
+        } else {
+            statusText = `$(${this._status.icon}) ${this._status.text}`;
+            statusTooltip = `${this._status.text}`;
+        }
+        this.statusbar.text = statusText;
+        this.statusbar.tooltip = statusTooltip;
+    }
 
     async delayCompletion(delay: number, token: vscode.CancellationToken): Promise<boolean> {
         if (config.inference.delay < 0) {
@@ -37,6 +70,9 @@ export class PromptProvider implements vscode.InlineCompletionItemProvider {
         }
 
         try {
+            if (this.paused) {
+                return;
+            }
 
             // Ignore unsupported documents
             if (!isSupported(document)) {
@@ -82,7 +118,7 @@ export class PromptProvider implements vscode.InlineCompletionItemProvider {
                     let inferenceConfig = config.inference;
 
                     // Update status
-                    this.statusbar.text = `$(sync~spin) Llama Coder`;
+                    this.update('sync~spin', 'Llama Coder');
                     try {
 
                         // Check model exists
@@ -110,9 +146,9 @@ export class PromptProvider implements vscode.InlineCompletionItemProvider {
                             }
 
                             // Perform download
-                            this.statusbar.text = `$(sync~spin) Downloading`;
+                            this.update('sync~spin', 'Downloading');
                             await ollamaDownloadModel(inferenceConfig.endpoint, inferenceConfig.modelName, inferenceConfig.bearerToken);
-                            this.statusbar.text = `$(sync~spin) Llama Coder`;
+                            this.update('sync~spin', 'Llama Coder')
                         }
                         if (token.isCancellationRequested) {
                             info(`Canceled after AI completion.`);
@@ -142,7 +178,7 @@ export class PromptProvider implements vscode.InlineCompletionItemProvider {
                             value: res
                         });
                     } finally {
-                        this.statusbar.text = `$(chip) Llama Coder`;
+                        this.update('chip', 'Llama Coder');
                     }
                 } else {
                     if (cached !== null) {