99 lines
3.8 KiB
Python
99 lines
3.8 KiB
Python
#!/usr/bin/env python3
|
|
from flask import Flask, request, Response
|
|
import requests
|
|
import base64
|
|
import uuid
|
|
import time
|
|
|
|
app = Flask(__name__)
|
|
GROQ_BASE = "https://api.groq.com"
|
|
|
|
CORS = {
|
|
"Access-Control-Allow-Origin": "*",
|
|
"Access-Control-Allow-Methods": "GET, POST, OPTIONS",
|
|
"Access-Control-Allow-Headers": "Authorization, Content-Type",
|
|
}
|
|
|
|
# Чанковая сборка аудио (DPI bypass — каждый POST <10KB)
|
|
_sessions = {} # session_id -> {chunks, mime, total, token, expires}
|
|
|
|
@app.route("/<path:path>", methods=["OPTIONS"])
|
|
def options(path):
|
|
return Response(status=204, headers=CORS)
|
|
|
|
@app.route("/v1/transcribe", methods=["POST", "OPTIONS"])
|
|
def transcribe_chunked():
|
|
if request.method == "OPTIONS":
|
|
return Response(status=204, headers=CORS)
|
|
|
|
data = request.get_json(force=True) or {}
|
|
idx = data.get("idx", 0)
|
|
total = data.get("total", 1)
|
|
chunk = data.get("chunk", "")
|
|
mime = data.get("mime", "audio/webm")
|
|
token = data.get("token", "")
|
|
sid = data.get("sid", "")
|
|
|
|
# Очистка старых сессий
|
|
now = time.time()
|
|
for k in list(_sessions.keys()):
|
|
if _sessions[k]["expires"] < now:
|
|
del _sessions[k]
|
|
|
|
if not sid:
|
|
sid = uuid.uuid4().hex[:12]
|
|
_sessions[sid] = {"chunks": {}, "mime": mime, "total": total, "token": token, "expires": now + 120}
|
|
else:
|
|
s = _sessions.get(sid)
|
|
if not s:
|
|
return Response('{"error":"session not found"}', status=404, headers=CORS, content_type="application/json")
|
|
|
|
_sessions[sid]["chunks"][idx] = chunk
|
|
_sessions[sid]["expires"] = now + 120
|
|
|
|
if len(_sessions[sid]["chunks"]) >= total:
|
|
s = _sessions.pop(sid)
|
|
audio_b64 = "".join(s["chunks"][i] for i in sorted(s["chunks"]))
|
|
audio_data = base64.b64decode(audio_b64)
|
|
|
|
ext = "webm"
|
|
if "ogg" in s["mime"]: ext = "ogg"
|
|
elif "mp4" in s["mime"] or "aac" in s["mime"]: ext = "mp4"
|
|
elif "wav" in s["mime"]: ext = "wav"
|
|
|
|
boundary = "----ChunkedBoundary" + uuid.uuid4().hex[:16]
|
|
payload = b"".join([
|
|
f'--{boundary}\r\nContent-Disposition: form-data; name="file"; filename="audio.{ext}"\r\nContent-Type: {s["mime"]}\r\n\r\n'.encode(),
|
|
audio_data,
|
|
f'\r\n--{boundary}\r\nContent-Disposition: form-data; name="model"\r\n\r\nwhisper-large-v3'.encode(),
|
|
f'\r\n--{boundary}\r\nContent-Disposition: form-data; name="language"\r\n\r\nit'.encode(),
|
|
f'\r\n--{boundary}--\r\n'.encode(),
|
|
])
|
|
|
|
r = requests.post(f"{GROQ_BASE}/openai/v1/audio/transcriptions",
|
|
headers={"Authorization": f"Bearer {s['token']}", "Content-Type": f"multipart/form-data; boundary={boundary}"},
|
|
data=payload, timeout=60)
|
|
result = r.json()
|
|
out = dict(CORS)
|
|
out["Content-Type"] = "application/json"
|
|
return Response(r.text, status=r.status_code, headers=out)
|
|
else:
|
|
return Response(f'{{"ok":true,"sid":"{sid}"}}', status=200, headers=CORS, content_type="application/json")
|
|
|
|
@app.route("/<path:subpath>", methods=["GET", "POST", "PUT", "DELETE", "PATCH"])
|
|
def proxy(subpath):
|
|
url = f"{GROQ_BASE}/{subpath}"
|
|
hdrs = {}
|
|
if request.headers.get("Authorization"):
|
|
hdrs["Authorization"] = request.headers["Authorization"]
|
|
if request.headers.get("Content-Type"):
|
|
hdrs["Content-Type"] = request.headers["Content-Type"]
|
|
r = requests.request(request.method, url, headers=hdrs,
|
|
data=request.get_data(), timeout=60, stream=True)
|
|
out = dict(CORS)
|
|
out["Content-Type"] = r.headers.get("Content-Type", "application/json")
|
|
return Response(r.iter_content(8192), status=r.status_code, headers=out)
|
|
|
|
if __name__ == "__main__":
|
|
app.run(host="127.0.0.1", port=8765)
|