what_how.py

from argparse import ArgumentParser
import moviepy.editor as mp

class EditBase():
	def __init__(self):
		self.YOUTUBE_SIZE = (854, 480)
		self.INSTAGRAM_SIZE = (480,480)
	
	def relative_to(self, target_res, position):
	# target_res and position are both tuples
		return tuple( round(target_res[i] * position[i])
					  for i in range(2)
					)

class WhatHow(EditBase):
	def __init__(self, args):
		super().__init__()
		self.TARGET_SIZE = self.INSTAGRAM_SIZE
		self.args = args
	
	def process(self):
		sequence = []
		
		audio = mp.AudioFileClip('audio/sanctuary-guardian.ogg')
		blank_video = mp.ColorClip(color=(0,0,0), duration=audio.duration, size=self.TARGET_SIZE)
		
		if (self.args.image):
			freeze_frame = mp.ImageClip(self.args.input, duration=audio.duration)
		else:
			orig_clip = mp.VideoFileClip(self.args.input).resize(newsize=self.TARGET_SIZE)
			
			if(self.args.no_prepend == False):
				sequence.append(orig_clip)
			
			freeze_frame = orig_clip.to_ImageClip(orig_clip.duration-0.1, duration=audio.duration)
		
		frame_size = (0.8, 0.6)
		gap_size = 0.005
		
		border_white = mp.ColorClip(color=(255,255,255),
									duration=audio.duration,
									size=self.relative_to(self.TARGET_SIZE, frame_size)
									)
		border_black = mp.ColorClip(color=(0,0,0),
									duration=audio.duration,
									size=self.relative_to(self.TARGET_SIZE, tuple(size-gap_size for size in frame_size))
									)
		freeze_frame = freeze_frame.resize(self.relative_to(self.TARGET_SIZE, tuple(size-(gap_size*2) for size in frame_size)))
		
		layers = [blank_video,border_white,border_black,freeze_frame]
		
		for i in range(len(layers)):
			layers[i] = layers[i].set_position(("center",0.1+(gap_size*i*0.5)),relative=True)
		
		top_text = mp.TextClip(self.args.top, color='white', font="Times-New-Roman", fontsize=36)
		top_text = top_text.set_position(("center",frame_size[1]+0.15),relative=True)
		top_text = top_text.set_duration(audio.duration)
		
		bottom_text = mp.TextClip(self.args.bottom, color='white', font="Times-New-Roman", fontsize=18)
		bottom_text = bottom_text.set_position(("center",frame_size[1]+0.3),relative=True)
		bottom_text = bottom_text.set_duration(audio.duration)
		
		layers.append(top_text)
		layers.append(bottom_text)
		
		part = mp.CompositeVideoClip(layers, size=self.TARGET_SIZE)
		part = part.set_audio(audio)
		sequence.append(part)
		
		final = mp.concatenate_videoclips(sequence)
		final.write_videofile(self.args.output, fps=30, codec='libx264', audio_codec='aac', ffmpeg_params=['-crf',str(self.args.crf)])
	
if __name__ == '__main__':
	ap = ArgumentParser(
						description="freeze frame at the very end of the video and plays music"
						)
	ap.add_argument(
						'-i', '--input',
						help='input video', required=True
					)
	ap.add_argument(
						'-o', '--output',
						help='output video', required=True
					)
	ap.add_argument(
						'-t', '--top',
						help='top text', default="WHAT"
					)
	ap.add_argument(
						'-b', '--bottom',
						help='bottom text', default="HOW"
					)
	ap.add_argument(
						'-c', '--crf',
						help='CRF', type=int, default=36
					)
	ap.add_argument(
						'-n', '--no-prepend',
						help="don't prepend the original clip", action='store_true'
					)
	ap.add_argument(
						'-j', '--image',
						help="is an image", action='store_true'
					)
	
	args = ap.parse_args()
	
	wh = WhatHow(args)
	wh.process()