Made a proper python package out of this.

pixelterm.py

@@ -1,12 +1,10 @@
 #!/usr/bin/env python
 
-import os, sys, argparse, os.path, json
 #NOTE: This script uses pygments for RGB->X256 conversion since pygments is
 #readily available. If you do not like pygments (e.g. because it is large),
 #you could patch in something like https://github.com/magarcia/python-x256
 #(but don't forget to send me a pull request ;)
 from pygments.formatters import terminal256
-from PIL import Image, PngImagePlugin
 
 formatter = terminal256.Terminal256Formatter()
 reset_sequence = terminal256.EscapeSequence(fg=formatter._closest_color(0,0,0), bg=formatter._closest_color(0,0,0)).reset_string()
@@ -76,26 +74,3 @@ def termify_pixels(img):
 		out = (out.rstrip() if bg == (0,0,0,0) else out) + '\n'
 	return out[:-1] + reset_sequence + '\n'
 
-if __name__ == '__main__':
-	parser = argparse.ArgumentParser(description='Render pixel images on 256-color ANSI terminals')
-	parser.add_argument('image', type=str, nargs='*')
-	parser.add_argument('-d', '--output-dir', type=str, help='Output directory (if not given, output to stdout)')
-	args = parser.parse_args()
-	for f in args.image:
-		img = Image.open(f).convert("RGBA")
-		if args.output_dir:
-			print(f)
-			foo, _, _ = f.rpartition('.png')
-			output = os.path.join(args.output_dir, os.path.basename(foo)+'.pony')
-			metadata = json.loads(img.info.get('pixelterm-metadata'))
-			comment = metadata['_comment']
-			del metadata['_comment']
-			metadataarea = '$$$\n' +\
-				'\n'.join([ '\n'.join([ k.upper() + ': ' + v for v in metadata[k] ]) for k in sorted(metadata.keys()) ]) +\
-				'\n' + comment +\
-				'\n$$$\n'
-			with open(output, 'w') as of:
-				of.write(metadataarea)
-				of.write(termify_pixels(img))
-		else:
-			print(termify_pixels(img))