Spaces:
Running
Running
File size: 16,657 Bytes
8dec77d 38efa1b 8dec77d 12a78dd 8dec77d 38efa1b 8dec77d 12a78dd 8dec77d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 |
# listen.py
from flask import Flask, Blueprint, jsonify, send_file, abort, request, send_from_directory
from flask_cors import CORS
from moviepy.editor import VideoFileClip
from google.cloud import speech
import os
print(f"GOOGLE_APPLICATION_CREDENTIALS: {os.getenv('GOOGLE_APPLICATION_CREDENTIALS')}")
import uuid
import requests
from pydub import AudioSegment
import ffmpeg
import re
import io # for streaming S3 bytes in HF/AWS mode
import json # <-- added for JSON creds parsing
# Optional (only used in AWS mode)
try:
import boto3
from botocore.exceptions import BotoCoreError, ClientError
except Exception:
boto3 = None
BotoCoreError = ClientError = Exception
# ---------- Blueprint ----------
listen_bp = Blueprint("listen", __name__)
# ---------------------- storage mode helpers ----------------------
def _is_aws_video_mode() -> bool:
"""
Switch to S3 on Hugging Face / prod. Local stays on disk.
"""
if os.getenv("USE_AWS_VIDEO", "0") == "1":
return True
if os.getenv("SPACE_ID"): # set on Hugging Face Spaces
return True
if os.getenv("ENV", "dev").lower() == "prod":
return True
return False
def _s3_clients():
if boto3 is None:
raise RuntimeError("boto3 is required in AWS video mode but not available")
region = os.getenv("AWS_DEFAULT_REGION", "eu-north-1")
s3 = boto3.client("s3", region_name=region)
return s3
def _video_s3_bucket():
bucket = os.getenv("S3_BUCKET_NAME")
if not bucket:
raise RuntimeError("S3_BUCKET_NAME is not set")
return bucket
def _video_s3_key(filename: str) -> str:
# Prefix under which listen.py stores videos in the same bucket
prefix = os.getenv("LISTEN_S3_PREFIX", "listen")
prefix = prefix.strip().strip("/")
return f"{prefix}/{filename}"
# ---------- writable working directories ----------
# Base working dir: /tmp on HF/AWS; local stays under ./static (or override via LISTEN_WORKDIR)
_BASE_WORKDIR = os.getenv(
"LISTEN_WORKDIR",
"/tmp/listen" if _is_aws_video_mode() else os.path.abspath("static")
)
VIDEO_FOLDER = os.path.join(_BASE_WORKDIR, "videos")
AUDIO_FOLDER = os.path.join(_BASE_WORKDIR, "audio")
TRANSCRIPT_FOLDER = os.path.join(_BASE_WORKDIR, "transcripts")
# Ensure directories exist (with hard fallback to /tmp if needed)
for _pname in ("videos", "audio", "transcripts"):
_p = os.path.join(_BASE_WORKDIR, _pname)
try:
os.makedirs(_p, exist_ok=True)
except Exception:
_fallback_base = "/tmp/listen"
os.makedirs(os.path.join(_fallback_base, _pname), exist_ok=True)
if _pname == "videos":
VIDEO_FOLDER = os.path.join(_fallback_base, "videos")
elif _pname == "audio":
AUDIO_FOLDER = os.path.join(_fallback_base, "audio")
else:
TRANSCRIPT_FOLDER = os.path.join(_fallback_base, "transcripts")
# ---------------- Cohere configuration (migrated to v2 Chat) ----------------
COHERE_API_KEY = os.getenv("COHERE_API_KEY", "")
COHERE_API_URL = 'https://api.cohere.com/v2/chat'
# ---------------------------------------------------------------------------
# --- Google Cloud Speech-to-Text client init (prefers HF secret JSON) ---
def _make_speech_client():
sa_json = os.getenv("GOOGLE_APPLICATION_CREDENTIALS_JSON")
if sa_json:
try:
info = json.loads(sa_json)
return speech.SpeechClient.from_service_account_info(info)
except Exception as e:
print(f"Failed to parse GOOGLE_APPLICATION_CREDENTIALS_JSON: {e}")
# fall through to default ADC
return speech.SpeechClient()
speech_client = _make_speech_client()
# -------------------------------------------------------------------------
# ------------- Cohere v2 helper (extract text from chat response) -------------
def _extract_text_v2(resp_json: dict) -> str:
"""
Cohere v2 /chat returns:
{ "message": { "content": [ { "type": "text", "text": "..." }, ... ] } }
This pulls the first text block.
"""
msg = resp_json.get("message", {})
content = msg.get("content", [])
for block in content:
if isinstance(block, dict) and block.get("type") == "text":
text = (block.get("text") or "").strip()
if text:
return text
return ""
# -----------------------------------------------------------------------------
# Convert video to audio
def convert_video_to_audio(video_path, audio_path):
try:
# Using moviepy to extract audio from video
video = VideoFileClip(video_path)
video.audio.write_audiofile(audio_path, codec='mp3')
return audio_path
except Exception as e:
print(f"Error converting video to audio: {str(e)}")
return None
# Re-encode MP3 to ensure proper format
def reencode_mp3(input_audio_path, output_audio_path):
try:
# Using pydub to convert and re-encode MP3 (ensuring correct encoding)
audio = AudioSegment.from_mp3(input_audio_path)
audio.export(output_audio_path, format="mp3", codec="libmp3lame", parameters=["-q:a", "0"])
return output_audio_path
except Exception as e:
print(f"Error re-encoding MP3: {str(e)}")
return None
# Helper function to convert audio to the proper MP3 encoding if necessary
def convert_audio_to_mp3(input_file_path, output_file_path):
"""
Converts the audio file to a valid MP3 format with proper encoding.
"""
try:
ffmpeg.input(input_file_path).output(output_file_path, acodec='libmp3lame', audio_bitrate='128k').run()
return True
except Exception as e:
print(f"Error during audio conversion: {e}")
return False
# Function to compress audio dynamically
def compress_audio(input_file_path, output_file_path, target_bitrate="128k"):
audio = AudioSegment.from_file(input_file_path)
audio.export(output_file_path, format="mp3", bitrate=target_bitrate)
return output_file_path
# ---------------------------- Routes (Blueprint) ----------------------------
@listen_bp.route('/', methods=['GET'])
def home():
return "Welcome to the Flask app! The server is running."
@listen_bp.route('/videos', methods=['GET'])
def list_videos():
"""
List available videos for users to watch.
"""
# If you maintain a VIDEOS list elsewhere, return it here.
# Returning empty list so the endpoint stays valid.
return jsonify([]), 200
@listen_bp.route('/videos/<filename>')
def serve_video(filename):
"""
Local: serve file from disk.
HF/AWS: fetch object from S3 and stream bytes (no redirect).
"""
if _is_aws_video_mode():
try:
s3 = _s3_clients()
bucket = _video_s3_bucket()
key = _video_s3_key(filename)
obj = s3.get_object(Bucket=bucket, Key=key)
data = obj["Body"].read()
return send_file(
io.BytesIO(data),
mimetype="video/mp4",
download_name=filename,
as_attachment=False
)
except (BotoCoreError, ClientError, Exception) as e:
print(f"S3 fetch failed for {filename}: {e}")
abort(404)
# Local
video_path = os.path.join(VIDEO_FOLDER, filename)
if not os.path.exists(video_path):
print(f"Video file not found: {filename}")
abort(404)
return send_file(video_path, mimetype='video/mp4')
@listen_bp.route('/upload-video', methods=['POST'])
def upload_video():
"""
Local: save to static/videos or /tmp/listen/videos (depending on mode).
HF/AWS: upload to S3 (no local original).
"""
print("Received upload request.")
if 'video' not in request.files:
print("No video file provided in the request.")
return jsonify({'error': 'No video file provided'}), 400
video = request.files['video']
if video.filename == '':
print("Empty filename detected.")
return jsonify({'error': 'No selected file'}), 400
try:
filename = f"{uuid.uuid4()}.mp4"
if _is_aws_video_mode():
try:
s3 = _s3_clients()
bucket = _video_s3_bucket()
key = _video_s3_key(filename)
s3.put_object(
Bucket=bucket,
Key=key,
Body=video.stream.read(),
ContentType="video/mp4"
)
print(f"Uploaded to S3: s3://{bucket}/{key}")
except (BotoCoreError, ClientError, Exception) as e:
print(f"S3 upload error: {e}")
return jsonify({'error': 'Failed to upload to S3'}), 500
else:
# Save locally
video_path = os.path.join(VIDEO_FOLDER, filename)
print(f"Saving video: {filename}")
video.save(video_path)
print(f"Video saved successfully at {video_path}")
return jsonify({'message': 'Video uploaded successfully!', 'filename': filename}), 200
except Exception as e:
print(f"Error saving video: {str(e)}")
return jsonify({'error': 'Failed to save video'}), 500
@listen_bp.route('/generate-questions-dynamicvideo', methods=['POST'])
def generate_questions():
try:
data = request.json
video_filename = data.get('filename')
if not video_filename:
print("Error: No filename provided in request.")
return jsonify({"error": "Filename is required"}), 400
# Resolve a local readable path for processing
video_path = os.path.join(VIDEO_FOLDER, video_filename)
if _is_aws_video_mode():
# Download object bytes to a local working file path
try:
s3 = _s3_clients()
bucket = _video_s3_bucket()
key = _video_s3_key(video_filename)
obj = s3.get_object(Bucket=bucket, Key=key)
data_bytes = obj["Body"].read()
with open(video_path, "wb") as f:
f.write(data_bytes)
except (BotoCoreError, ClientError, Exception) as e:
print(f"S3 download error for {video_filename}: {e}")
return jsonify({"error": "Video file not found"}), 404
else:
if not os.path.exists(video_path):
print(f"Error: Video file {video_filename} not found at {video_path}")
return jsonify({"error": "Video file not found"}), 404
print(f"Processing video: {video_filename}")
# Convert video to audio
audio_filename = f"{uuid.uuid4()}.mp3"
audio_path = os.path.join(AUDIO_FOLDER, audio_filename)
if not convert_video_to_audio(video_path, audio_path):
print("Error: Video to audio conversion failed.")
return jsonify({"error": "Failed to convert video to audio"}), 500
# Transcribe audio using Google Cloud Speech-to-Text
with open(audio_path, 'rb') as audio_file:
audio_content = audio_file.read()
audio = speech.RecognitionAudio(content=audio_content)
config = speech.RecognitionConfig(
encoding=speech.RecognitionConfig.AudioEncoding.MP3,
sample_rate_hertz=16000,
language_code="en-US",
)
response = speech_client.recognize(config=config, audio=audio)
transcripts = [result.alternatives[0].transcript for result in response.results]
if not transcripts:
print("Error: No transcription results found.")
return jsonify({"error": "No transcription results found"}), 500
transcription_text = " ".join(transcripts)
print(f"Transcription successful: {transcription_text[:200]}...") # Print first 200 chars
# ---------------- Cohere v2 Chat call (minimal change) ----------------
headers = {
"Authorization": f"Bearer {COHERE_API_KEY}",
"Content-Type": "application/json"
}
prompt_text = (
"Generate exactly three multiple-choice questions based on this text:\n"
f"{transcription_text}\n\n"
"Rules:\n"
"- Each question starts with a number and a period (e.g., 1.)\n"
"- Each question has exactly four options labeled A., B., C., and D.\n"
"- After the options, add a line 'Correct answer: <A|B|C|D>'\n"
"- Output plain text only."
)
cohere_payload = {
"model": "command-r-08-2024",
"messages": [
{"role": "user", "content": prompt_text}
],
"max_tokens": 300,
"temperature": 0.9
}
cohere_response = requests.post(
COHERE_API_URL,
json=cohere_payload,
headers=headers,
timeout=60
)
if cohere_response.status_code != 200:
print(f"Error: Cohere API response failed: {cohere_response.text}")
return jsonify({"error": "Failed to generate questions"}), 500
raw_text = _extract_text_v2(cohere_response.json())
if not raw_text:
print("Error: No questions text returned by Cohere Chat API.")
return jsonify({"error": "No questions generated"}), 500
# ---------------------------------------------------------------------
# Extract raw text and parse questions
structured_questions = parse_questions(raw_text)
return jsonify({"questions": structured_questions}), 200
except Exception as e:
print(f"Critical Error: {e}")
return jsonify({"error": "An error occurred while generating questions"}), 500
def parse_questions(response_text):
# Split the text into individual question blocks
question_blocks = response_text.split("\n\n")
questions = []
# Process each question block
for block in question_blocks:
print("\nProcessing Block:", block) # Debug: Log each question block
# Split the block into lines
lines = block.strip().split("\n")
print("Split Lines:", lines) # Debug: Log split lines of the block
# Ensure the block contains a question
if len(lines) < 2:
print("Skipping Invalid Block") # Debug: Log invalid blocks
continue
# Extract the question text
question_line = lines[0]
question_text = question_line.split(". ", 1)[1] if ". " in question_line else question_line
print("Question Text:", question_text) # Debug: Log extracted question text
# Extract the options and find the correct answer
options = []
correct_answer_letter = None
for line in lines[1:]:
line = line.strip()
# Handle A., B., C., D. and also a) / A) formats
if line.lower().startswith("correct answer:"):
correct_answer_letter = line.split(":")[-1].strip()
continue
match = re.match(r"^(?:[a-dA-D][\).]?\s)?(.+)$", line)
if match:
option_text = match.group(1).strip()
# We already handled "Correct answer:" above, so only options get appended
if not line.lower().startswith("correct answer:"):
options.append(option_text)
print("Extracted Options:", options) # Debug: Log extracted options
print("Correct Answer Letter:", correct_answer_letter) # Debug: Log the correct answer letter
# Map the correct answer text
correct_answer_text = ""
if correct_answer_letter:
option_index = ord(correct_answer_letter.upper()) - ord('A') # Convert 'A'→0, 'B'→1, etc.
if 0 <= option_index < len(options):
correct_answer_text = options[option_index]
print("Mapped Correct Answer Text:", correct_answer_text) # Debug: Log mapped answer
# Append the parsed question to the list
if question_text and options:
questions.append({
"question": question_text,
"options": options,
"answer": correct_answer_text # Use the full answer text
})
print("\nFinal Questions:", questions) # Debug: Log final parsed questions
return questions
# ---------- Standalone (local testing) ----------
if __name__ == '__main__':
app = Flask(__name__)
CORS(app)
app.config["COHERE_API_KEY"] = os.getenv("COHERE_API_KEY", COHERE_API_KEY)
app.register_blueprint(listen_bp, url_prefix='')
app.run(host='0.0.0.0', port=5012, debug=True)
|