from PIL import Image, ImageDraw, ImageFont
import cv2
import numpy as np

import textwrap

def generate_lyric_frame(lyric, image_size):
    # If image_size is None, provide a default value
    if image_size is None:
        image_size = (800, 600) # Example default size, you can change this

    # Ensure image_size is a tuple
    image_size = tuple(image_size)

    image = Image.new('RGB', image_size, color='black')
    draw = ImageDraw.Draw(image)

    # Use PIL's default font
    font = ImageFont.load_default()

    # Wrap text
    max_width = image_size[0] - 40  # Keeping 20 pixels margin on each side
    wrapped_text = textwrap.fill(lyric, width=40)
    lines = wrapped_text.split('\n')

    # Calculate text width and height
    text_width = max(draw.textlength(line, font=font) for line in lines)  # Corrected line here
    text_height = sum([font.getmask(line).getbbox()[3] for line in lines if font.getmask(line).getbbox()]) # Corrected line here
    
    # Calculate starting coordinates to center the text block
    text_x = (image_size[0] - text_width) / 2
    text_y = (image_size[1] - text_height) / 2

    # Draw the wrapped text on the image
    y = text_y
    for line in lines:
        draw.text((text_x, y), line, font=font, fill='white')
        y += font.getmask(line).getbbox()[3] if font.getmask(line).getbbox() else 0  # Corrected line here

    return image

def generate_video(lyrics_with_timing, image_size=(800, 600), frame_rate=30):
    frames = []
    for start_time, end_time, lyric in lyrics_with_timing:
        # Calculate the number of frames for this lyric segment
        num_frames = int((end_time - start_time) * frame_rate)

        # Create frames with the lyric text
        for _ in range(num_frames):
            frame = generate_lyric_frame(lyric, image_size)
            frames.append(frame)

    # Combine frames into a video
    video = combine_frames_into_video(frames, frame_rate)
    return video

def combine_frames_into_video(frames, frame_rate):
    # Convert the first frame to a NumPy array to get its shape
    first_frame_np = np.array(frames[0])
    height, width, _ = first_frame_np.shape

    # Define the codec and create a video writer object
    fourcc = cv2.VideoWriter_fourcc(*'XVID')
    out = cv2.VideoWriter('output.avi', fourcc, frame_rate, (width, height))

    # Write each frame to the video file
    for frame in frames:
        # Convert PIL image to NumPy array
        frame_np = np.array(frame)

        # Convert RGB to BGR as OpenCV uses BGR format
        frame_bgr = cv2.cvtColor(frame_np, cv2.COLOR_RGB2BGR)

        # Write the BGR frame to the video
        out.write(frame_bgr)

    # Release the video writer object
    out.release()

    return 'output.avi'