refactor: remove unused whispher transcribe logic

This commit is contained in:
Haileyesus
2026-04-10 16:24:33 +03:00
parent a8dab0edcf
commit 91954daedd
23 changed files with 12 additions and 855 deletions

View File

@@ -1984,155 +1984,6 @@ function handleShellConnection(ws) {
console.error('[ERROR] Shell WebSocket error:', error);
});
}
// Audio transcription endpoint
app.post('/api/transcribe', authenticateToken, async (req, res) => {
try {
const multer = (await import('multer')).default;
const upload = multer({ storage: multer.memoryStorage() });
// Handle multipart form data
upload.single('audio')(req, res, async (err) => {
if (err) {
return res.status(400).json({ error: 'Failed to process audio file' });
}
if (!req.file) {
return res.status(400).json({ error: 'No audio file provided' });
}
const apiKey = process.env.OPENAI_API_KEY;
if (!apiKey) {
return res.status(500).json({ error: 'OpenAI API key not configured. Please set OPENAI_API_KEY in server environment.' });
}
try {
// Create form data for OpenAI
const FormData = (await import('form-data')).default;
const formData = new FormData();
formData.append('file', req.file.buffer, {
filename: req.file.originalname,
contentType: req.file.mimetype
});
formData.append('model', 'whisper-1');
formData.append('response_format', 'json');
formData.append('language', 'en');
// Make request to OpenAI
const response = await fetch('https://api.openai.com/v1/audio/transcriptions', {
method: 'POST',
headers: {
'Authorization': `Bearer ${apiKey}`,
...formData.getHeaders()
},
body: formData
});
if (!response.ok) {
const errorData = await response.json().catch(() => ({}));
throw new Error(errorData.error?.message || `Whisper API error: ${response.status}`);
}
const data = await response.json();
let transcribedText = data.text || '';
// Check if enhancement mode is enabled
const mode = req.body.mode || 'default';
// If no transcribed text, return empty
if (!transcribedText) {
return res.json({ text: '' });
}
// If default mode, return transcribed text without enhancement
if (mode === 'default') {
return res.json({ text: transcribedText });
}
// Handle different enhancement modes
try {
const OpenAI = (await import('openai')).default;
const openai = new OpenAI({ apiKey });
let prompt, systemMessage, temperature = 0.7, maxTokens = 800;
switch (mode) {
case 'prompt':
systemMessage = 'You are an expert prompt engineer who creates clear, detailed, and effective prompts.';
prompt = `You are an expert prompt engineer. Transform the following rough instruction into a clear, detailed, and context-aware AI prompt.
Your enhanced prompt should:
1. Be specific and unambiguous
2. Include relevant context and constraints
3. Specify the desired output format
4. Use clear, actionable language
5. Include examples where helpful
6. Consider edge cases and potential ambiguities
Transform this rough instruction into a well-crafted prompt:
"${transcribedText}"
Enhanced prompt:`;
break;
case 'vibe':
case 'instructions':
case 'architect':
systemMessage = 'You are a helpful assistant that formats ideas into clear, actionable instructions for AI agents.';
temperature = 0.5; // Lower temperature for more controlled output
prompt = `Transform the following idea into clear, well-structured instructions that an AI agent can easily understand and execute.
IMPORTANT RULES:
- Format as clear, step-by-step instructions
- Add reasonable implementation details based on common patterns
- Only include details directly related to what was asked
- Do NOT add features or functionality not mentioned
- Keep the original intent and scope intact
- Use clear, actionable language an agent can follow
Transform this idea into agent-friendly instructions:
"${transcribedText}"
Agent instructions:`;
break;
default:
// No enhancement needed
break;
}
// Only make GPT call if we have a prompt
if (prompt) {
const completion = await openai.chat.completions.create({
model: 'gpt-4o-mini',
messages: [
{ role: 'system', content: systemMessage },
{ role: 'user', content: prompt }
],
temperature: temperature,
max_tokens: maxTokens
});
transcribedText = completion.choices[0].message.content || transcribedText;
}
} catch (gptError) {
console.error('GPT processing error:', gptError);
// Fall back to original transcription if GPT fails
}
res.json({ text: transcribedText });
} catch (error) {
console.error('Transcription error:', error);
res.status(500).json({ error: error.message });
}
});
} catch (error) {
console.error('Endpoint error:', error);
res.status(500).json({ error: 'Internal server error' });
}
});
// Image upload endpoint
app.post('/api/projects/:projectName/upload-images', authenticateToken, async (req, res) => {
try {