wogiflow 1.0.0

package/scripts/flow-voice-input.js

@@ -0,0 +1,638 @@
+#!/usr/bin/env node
+
+/**
+ * Wogi Flow - Voice Input
+ *
+ * Voice-to-transcript support with multiple provider options:
+ * - Local: Whisper.cpp (no API key required)
+ * - Cloud: OpenAI Whisper API
+ * - Cloud: Groq (free tier available)
+ *
+ * Usage:
+ *   flow voice-input                    # Record and transcribe
+ *   flow voice-input --duration 30      # Record for 30 seconds
+ *   flow voice-input --provider openai  # Use specific provider
+ *   flow voice-input --to-story         # Create story from transcript
+ *   flow voice-input setup              # Interactive setup
+ */
+
+const fs = require('fs');
+const path = require('path');
+const { spawn, execSync } = require('child_process');
+const { HttpClient } = require('./flow-http-client');
+const readline = require('readline');
+const { getConfig, getProjectRoot, colors: c } = require('./flow-utils');
+
+const PROJECT_ROOT = getProjectRoot();
+const WORKFLOW_DIR = path.join(PROJECT_ROOT, '.workflow');
+const CONFIG_PATH = path.join(WORKFLOW_DIR, 'config.json');
+
+// ============================================================
+// Voice Provider Types
+// ============================================================
+
+const VOICE_PROVIDERS = {
+  LOCAL: 'local',
+  OPENAI: 'openai',
+  GROQ: 'groq'
+};
+
+const PROVIDER_INFO = {
+  local: {
+    name: 'Local (Whisper.cpp)',
+    description: 'Run transcription locally - no API key required',
+    requiresKey: false,
+    binaryName: 'whisper'
+  },
+  openai: {
+    name: 'OpenAI Whisper',
+    description: 'Cloud transcription via OpenAI API',
+    requiresKey: true,
+    endpoint: 'https://api.openai.com/v1/audio/transcriptions',
+    model: 'whisper-1'
+  },
+  groq: {
+    name: 'Groq',
+    description: 'Fast cloud transcription - free tier available',
+    requiresKey: true,
+    endpoint: 'https://api.groq.com/openai/v1/audio/transcriptions',
+    model: 'whisper-large-v3'
+  }
+};
+
+// ============================================================
+// Configuration
+// ============================================================
+
+/**
+ * Get voice configuration from config.json
+ */
+function getVoiceConfig() {
+  const config = getConfig();
+  return config.voice || {
+    enabled: false,
+    provider: null,
+    openaiApiKey: null,
+    groqApiKey: null,
+    localModelPath: null,
+    defaultDuration: 30,
+    sampleRate: 16000,
+    channels: 1
+  };
+}
+
+/**
+ * Check if voice input is enabled
+ */
+function isVoiceEnabled() {
+  const config = getVoiceConfig();
+  return config.enabled === true && config.provider !== null;
+}
+
+/**
+ * Get the active provider
+ */
+function getActiveProvider() {
+  const config = getVoiceConfig();
+  return config.provider;
+}
+
+// ============================================================
+// Audio Recording
+// ============================================================
+
+/**
+ * Check if recording dependencies are available
+ */
+function checkRecordingDependencies() {
+  const issues = [];
+
+  // Check for sox (required for most recorders)
+  try {
+    execSync('which sox', { stdio: 'pipe' });
+  } catch {
+    issues.push('sox not found - install with: brew install sox');
+  }
+
+  // Check for rec (part of sox)
+  try {
+    execSync('which rec', { stdio: 'pipe' });
+  } catch {
+    issues.push('rec not found - install with: brew install sox');
+  }
+
+  return issues;
+}
+
+/**
+ * Record audio from microphone
+ * Returns path to recorded WAV file
+ */
+async function recordAudio(durationSeconds = 30, options = {}) {
+  const {
+    sampleRate = 16000,
+    channels = 1,
+    showProgress = true
+  } = options;
+
+  const tempFile = path.join('/tmp', `wogi-voice-${Date.now()}.wav`);
+
+  return new Promise((resolve, reject) => {
+    if (showProgress) {
+      console.log(`${c.cyan}Recording for ${durationSeconds} seconds...${c.reset}`);
+      console.log(`${c.dim}Press Ctrl+C to stop early${c.reset}\n`);
+    }
+
+    // Use sox's rec command for cross-platform recording
+    const rec = spawn('rec', [
+      '-r', String(sampleRate),
+      '-c', String(channels),
+      '-b', '16',
+      tempFile,
+      'trim', '0', String(durationSeconds)
+    ], {
+      stdio: showProgress ? ['inherit', 'inherit', 'inherit'] : 'pipe'
+    });
+
+    rec.on('close', (code) => {
+      if (code === 0 && fs.existsSync(tempFile)) {
+        resolve(tempFile);
+      } else {
+        reject(new Error(`Recording failed with code ${code}`));
+      }
+    });
+
+    rec.on('error', (err) => {
+      reject(new Error(`Recording error: ${err.message}`));
+    });
+  });
+}
+
+// ============================================================
+// Transcription Providers
+// ============================================================
+
+/**
+ * Transcribe with local Whisper.cpp
+ */
+async function transcribeLocal(audioPath, options = {}) {
+  const config = getVoiceConfig();
+  const modelPath = config.localModelPath || 'base.en';
+
+  try {
+    // Check if whisper is available
+    execSync('which whisper', { stdio: 'pipe' });
+  } catch {
+    throw new Error(
+      'Local Whisper not found. Install with:\n' +
+      '  brew install openai-whisper\n' +
+      'Or download whisper.cpp from: https://github.com/ggerganov/whisper.cpp'
+    );
+  }
+
+  return new Promise((resolve, reject) => {
+    const whisper = spawn('whisper', [
+      audioPath,
+      '--model', modelPath,
+      '--output_format', 'txt',
+      '--output_dir', '/tmp'
+    ]);
+
+    let stderr = '';
+    whisper.stderr.on('data', (data) => {
+      stderr += data.toString();
+    });
+
+    whisper.on('close', (code) => {
+      if (code === 0) {
+        // Read the output file
+        const outputPath = audioPath.replace('.wav', '.txt');
+        if (fs.existsSync(outputPath)) {
+          const text = fs.readFileSync(outputPath, 'utf-8').trim();
+          fs.unlinkSync(outputPath); // Cleanup
+          resolve({ text, provider: 'local', model: modelPath });
+        } else {
+          reject(new Error('Whisper output file not found'));
+        }
+      } else {
+        reject(new Error(`Whisper failed: ${stderr}`));
+      }
+    });
+  });
+}
+
+/**
+ * Transcribe with OpenAI Whisper API using shared HttpClient
+ */
+async function transcribeOpenAI(audioPath, options = {}) {
+  const config = getVoiceConfig();
+  const apiKey = config.openaiApiKey || process.env.OPENAI_API_KEY;
+
+  if (!apiKey) {
+    throw new Error(
+      'OpenAI API key not found. Set it in config.json:\n' +
+      '  "voice": { "openaiApiKey": "sk-..." }\n' +
+      'Or set OPENAI_API_KEY environment variable'
+    );
+  }
+
+  const audioData = fs.readFileSync(audioPath);
+  const client = new HttpClient('https://api.openai.com', {
+    headers: { 'Authorization': `Bearer ${apiKey}` },
+    timeout: 60000
+  });
+
+  const response = await client.postMultipart('/v1/audio/transcriptions', [
+    { name: 'file', value: audioData, filename: 'audio.wav', contentType: 'audio/wav' },
+    { name: 'model', value: 'whisper-1' }
+  ]);
+
+  if (response.data?.error) {
+    throw new Error(response.data.error.message);
+  }
+
+  return { text: response.data.text, provider: 'openai', model: 'whisper-1' };
+}
+
+/**
+ * Transcribe with Groq API using shared HttpClient
+ */
+async function transcribeGroq(audioPath, options = {}) {
+  const config = getVoiceConfig();
+  const apiKey = config.groqApiKey || process.env.GROQ_API_KEY;
+
+  if (!apiKey) {
+    throw new Error(
+      'Groq API key not found. Set it in config.json:\n' +
+      '  "voice": { "groqApiKey": "gsk_..." }\n' +
+      'Or set GROQ_API_KEY environment variable\n\n' +
+      'Get a free key at: https://console.groq.com'
+    );
+  }
+
+  const audioData = fs.readFileSync(audioPath);
+  const client = new HttpClient('https://api.groq.com', {
+    headers: { 'Authorization': `Bearer ${apiKey}` },
+    timeout: 60000
+  });
+
+  const response = await client.postMultipart('/openai/v1/audio/transcriptions', [
+    { name: 'file', value: audioData, filename: 'audio.wav', contentType: 'audio/wav' },
+    { name: 'model', value: 'whisper-large-v3' }
+  ]);
+
+  if (response.data?.error) {
+    throw new Error(response.data.error.message);
+  }
+
+  return { text: response.data.text, provider: 'groq', model: 'whisper-large-v3' };
+}
+
+/**
+ * Main transcription function - routes to appropriate provider
+ */
+async function transcribe(audioPath, providerOverride = null) {
+  const config = getVoiceConfig();
+  const provider = providerOverride || config.provider || 'openai';
+
+  switch (provider) {
+    case 'local':
+      return transcribeLocal(audioPath);
+    case 'openai':
+      return transcribeOpenAI(audioPath);
+    case 'groq':
+      return transcribeGroq(audioPath);
+    default:
+      throw new Error(`Unknown provider: ${provider}`);
+  }
+}
+
+// ============================================================
+// Interactive Setup
+// ============================================================
+
+/**
+ * Interactive setup wizard for voice input
+ */
+async function runSetup() {
+  const rl = readline.createInterface({
+    input: process.stdin,
+    output: process.stdout
+  });
+
+  const question = (q) => new Promise(resolve => rl.question(q, resolve));
+
+  console.log(`\n${c.cyan}=== Wogi Flow Voice Input Setup ===${c.reset}\n`);
+
+  // Check recording dependencies
+  const issues = checkRecordingDependencies();
+  if (issues.length > 0) {
+    console.log(`${c.yellow}Recording dependencies missing:${c.reset}`);
+    issues.forEach(i => console.log(`  - ${i}`));
+    console.log('');
+  }
+
+  // Enable voice input
+  const enable = await question(`Enable voice input? (y/n): `);
+  if (enable.toLowerCase() !== 'y') {
+    console.log(`${c.dim}Voice input disabled.${c.reset}`);
+    rl.close();
+    return;
+  }
+
+  // Choose provider
+  console.log(`\n${c.cyan}Choose transcription provider:${c.reset}`);
+  console.log('  1. Local (Whisper.cpp) - No API key, works offline');
+  console.log('  2. OpenAI - Best accuracy, requires API key');
+  console.log('  3. Groq - Fast, free tier available');
+
+  const providerChoice = await question(`\nSelect provider (1-3): `);
+  let provider, apiKey = null;
+
+  switch (providerChoice) {
+    case '1':
+      provider = 'local';
+      console.log(`\n${c.green}Local provider selected.${c.reset}`);
+      console.log(`${c.dim}Ensure whisper is installed: brew install openai-whisper${c.reset}`);
+      break;
+    case '2':
+      provider = 'openai';
+      apiKey = await question(`\nEnter OpenAI API key (sk-...): `);
+      if (!apiKey.startsWith('sk-')) {
+        console.log(`${c.yellow}Warning: Key doesn't look like an OpenAI key${c.reset}`);
+      }
+      break;
+    case '3':
+      provider = 'groq';
+      apiKey = await question(`\nEnter Groq API key (gsk_...): `);
+      if (!apiKey.startsWith('gsk_')) {
+        console.log(`${c.yellow}Warning: Key doesn't look like a Groq key${c.reset}`);
+      }
+      break;
+    default:
+      console.log(`${c.red}Invalid choice${c.reset}`);
+      rl.close();
+      return;
+  }
+
+  // Save configuration
+  const config = getConfig();
+  config.voice = {
+    enabled: true,
+    provider,
+    openaiApiKey: provider === 'openai' ? apiKey : (config.voice?.openaiApiKey || null),
+    groqApiKey: provider === 'groq' ? apiKey : (config.voice?.groqApiKey || null),
+    localModelPath: config.voice?.localModelPath || 'base.en',
+    defaultDuration: 30,
+    sampleRate: 16000,
+    channels: 1
+  };
+
+  fs.writeFileSync(CONFIG_PATH, JSON.stringify(config, null, 2));
+
+  console.log(`\n${c.green}Voice input configured!${c.reset}`);
+  console.log(`Provider: ${PROVIDER_INFO[provider].name}`);
+  console.log(`\nTest with: ${c.cyan}./scripts/flow voice-input${c.reset}`);
+
+  rl.close();
+}
+
+// ============================================================
+// Main CLI Handler
+// ============================================================
+
+async function main() {
+  const args = process.argv.slice(2);
+  const command = args[0];
+
+  // Parse options
+  const options = {
+    duration: 30,
+    provider: null,
+    toStory: false,
+    output: null
+  };
+
+  for (let i = 0; i < args.length; i++) {
+    switch (args[i]) {
+      case '--duration':
+      case '-d':
+        options.duration = parseInt(args[++i]) || 30;
+        break;
+      case '--provider':
+      case '-p':
+        options.provider = args[++i];
+        break;
+      case '--to-story':
+        options.toStory = true;
+        break;
+      case '--output':
+      case '-o':
+        options.output = args[++i];
+        break;
+    }
+  }
+
+  // Handle commands
+  switch (command) {
+    case 'setup':
+      await runSetup();
+      break;
+
+    case 'status':
+      showStatus();
+      break;
+
+    case 'test':
+      await testVoiceInput();
+      break;
+
+    case undefined:
+    case 'record':
+      await recordAndTranscribe(options);
+      break;
+
+    default:
+      showHelp();
+  }
+}
+
+/**
+ * Show voice input status
+ */
+function showStatus() {
+  const config = getVoiceConfig();
+
+  console.log(`\n${c.cyan}Voice Input Status${c.reset}\n`);
+  console.log(`Enabled: ${config.enabled ? c.green + 'Yes' : c.red + 'No'}${c.reset}`);
+
+  if (config.enabled) {
+    const providerInfo = PROVIDER_INFO[config.provider] || {};
+    console.log(`Provider: ${providerInfo.name || config.provider}`);
+    console.log(`Default duration: ${config.defaultDuration}s`);
+
+    if (config.provider === 'openai') {
+      console.log(`API Key: ${config.openaiApiKey ? c.green + 'Configured' : c.red + 'Missing'}${c.reset}`);
+    } else if (config.provider === 'groq') {
+      console.log(`API Key: ${config.groqApiKey ? c.green + 'Configured' : c.red + 'Missing'}${c.reset}`);
+    }
+  }
+
+  // Check dependencies
+  const issues = checkRecordingDependencies();
+  if (issues.length > 0) {
+    console.log(`\n${c.yellow}Missing dependencies:${c.reset}`);
+    issues.forEach(i => console.log(`  - ${i}`));
+  } else {
+    console.log(`\n${c.green}Recording dependencies OK${c.reset}`);
+  }
+}
+
+/**
+ * Test voice input with a short recording
+ */
+async function testVoiceInput() {
+  const config = getVoiceConfig();
+
+  if (!config.enabled) {
+    console.log(`${c.yellow}Voice input not enabled. Run: ./scripts/flow voice-input setup${c.reset}`);
+    return;
+  }
+
+  console.log(`${c.cyan}Testing voice input...${c.reset}`);
+  console.log(`Provider: ${PROVIDER_INFO[config.provider]?.name || config.provider}\n`);
+
+  try {
+    console.log('Recording 5 seconds of audio...\n');
+    const audioPath = await recordAudio(5);
+
+    console.log(`\n${c.cyan}Transcribing...${c.reset}`);
+    const result = await transcribe(audioPath);
+
+    console.log(`\n${c.green}Success!${c.reset}`);
+    console.log(`Transcript: "${result.text}"`);
+    console.log(`Provider: ${result.provider}, Model: ${result.model}`);
+
+    // Cleanup
+    if (fs.existsSync(audioPath)) {
+      fs.unlinkSync(audioPath);
+    }
+  } catch (error) {
+    console.error(`${c.red}Test failed: ${error.message}${c.reset}`);
+  }
+}
+
+/**
+ * Main record and transcribe flow
+ */
+async function recordAndTranscribe(options) {
+  const config = getVoiceConfig();
+
+  if (!config.enabled) {
+    console.log(`${c.yellow}Voice input not enabled.${c.reset}`);
+    console.log(`Run: ${c.cyan}./scripts/flow voice-input setup${c.reset}`);
+    return;
+  }
+
+  try {
+    // Record
+    const duration = options.duration || config.defaultDuration || 30;
+    const audioPath = await recordAudio(duration);
+
+    // Transcribe
+    console.log(`\n${c.cyan}Transcribing...${c.reset}`);
+    const result = await transcribe(audioPath, options.provider);
+
+    console.log(`\n${c.green}Transcript:${c.reset}`);
+    console.log(result.text);
+
+    // Output to file if requested
+    if (options.output) {
+      fs.writeFileSync(options.output, result.text);
+      console.log(`\n${c.dim}Saved to: ${options.output}${c.reset}`);
+    }
+
+    // Create story if requested
+    if (options.toStory) {
+      console.log(`\n${c.cyan}Creating story from transcript...${c.reset}`);
+      // This would integrate with flow-story.js
+      console.log(`${c.dim}[Integration with story creation would go here]${c.reset}`);
+    }
+
+    // Cleanup
+    if (fs.existsSync(audioPath)) {
+      fs.unlinkSync(audioPath);
+    }
+
+    // Return result for programmatic use
+    return result;
+  } catch (error) {
+    console.error(`${c.red}Error: ${error.message}${c.reset}`);
+    process.exit(1);
+  }
+}
+
+/**
+ * Show help
+ */
+function showHelp() {
+  console.log(`
+${c.cyan}Wogi Flow - Voice Input${c.reset}
+
+${c.bold}Usage:${c.reset}
+  flow voice-input [command] [options]
+
+${c.bold}Commands:${c.reset}
+  setup              Interactive setup wizard
+  status             Show voice input configuration
+  test               Test with a 5-second recording
+  record             Record and transcribe (default)
+
+${c.bold}Options:${c.reset}
+  --duration, -d     Recording duration in seconds (default: 30)
+  --provider, -p     Override provider (local, openai, groq)
+  --to-story         Create a story from the transcript
+  --output, -o       Save transcript to file
+
+${c.bold}Examples:${c.reset}
+  flow voice-input setup              # Configure voice input
+  flow voice-input                    # Record and transcribe
+  flow voice-input -d 60              # Record for 60 seconds
+  flow voice-input -p groq            # Use Groq provider
+  flow voice-input --to-story         # Create story from voice
+
+${c.bold}Providers:${c.reset}
+  local   - Whisper.cpp (no API key, works offline)
+  openai  - OpenAI Whisper API (best accuracy)
+  groq    - Groq API (fast, free tier available)
+`);
+}
+
+// ============================================================
+// Exports
+// ============================================================
+
+module.exports = {
+  getVoiceConfig,
+  isVoiceEnabled,
+  getActiveProvider,
+  recordAudio,
+  transcribe,
+  transcribeLocal,
+  transcribeOpenAI,
+  transcribeGroq,
+  checkRecordingDependencies,
+  VOICE_PROVIDERS,
+  PROVIDER_INFO
+};
+
+// ============================================================
+// Run CLI
+// ============================================================
+
+if (require.main === module) {
+  main().catch(err => {
+    console.error(`${c.red}Error: ${err.message}${c.reset}`);
+    process.exit(1);
+  });
+}