video_processor.py

import cv2
import os
import uuid
import numpy as np
from utils import extract_frames, save_video

TEMP_DIR = "temp_videos"
os.makedirs(TEMP_DIR, exist_ok=True)

def process_live_video():
    # Simulated live stream (replace with real buffer system if needed)
    live_source = "sample_videos/lbw_sample.mp4"
    
    cap = cv2.VideoCapture(live_source)
    fps = cap.get(cv2.CAP_PROP_FPS)
    total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
    duration = total_frames / fps

    # Trim last 10 seconds
    trim_start = max(0, duration - 10)
    start_frame = int(trim_start * fps)

    cap.set(cv2.CAP_PROP_POS_FRAMES, start_frame)
    frames = []
    while True:
        ret, frame = cap.read()
        if not ret:
            break
        frames.append(frame)
    cap.release()

    video_id = str(uuid.uuid4())
    trimmed_path = os.path.join(TEMP_DIR, f"{video_id}_trimmed.mp4")
    save_video(frames, trimmed_path, fps)

    # Extract & Analyze
    decision = "Analyzing..."  # Placeholder, will use lbw_detector later
    return trimmed_path, decision


def process_uploaded_video(video_file):
    cap = cv2.VideoCapture(video_file)
    fps = cap.get(cv2.CAP_PROP_FPS)

    frames = []
    while True:
        ret, frame = cap.read()
        if not ret:
            break
        frames.append(frame)
    cap.release()

    video_id = str(uuid.uuid4())
    saved_path = os.path.join(TEMP_DIR, f"{video_id}_upload.mp4")
    save_video(frames, saved_path, fps)

    # Extract & Analyze
    decision = "Analyzing..."  # Placeholder for lbw_detector output
    return saved_path, decision