def main()

in split_data.py [0:0]

• 29 lines of code
• 2 McCabe index (conditional complexity)

def main(args):
	#load parsed wiktionary data
	q_path = os.path.join(args.raw_dir, 'quotations.txt')
	quotes = load_quotations(q_path)
	s_path = os.path.join(args.raw_dir, 'senses.txt')
	senses = load_senses(s_path)
	ex_path = os.path.join(args.raw_dir, 'examples.txt')
	examples = load_examples(ex_path)

	#split quotes into train/dev/test
	data, monosemous_data = filter_monosemous_data(quotes, senses)
	train, fs_dev, zs_dev, fs_test, zs_test = split_data(data, senses)

	#save train data
	train_path = os.path.join(args.save_dir, 'train.txt')
	save_examples(train_path, train)

	#create and save train extended 
	#(adds examples as extra train data)
	random.shuffle(examples)
	#filter monosymous senses from examples
	ext, monosemous_ext = filter_monosemous_data(examples, senses)
	#filter senses in zero-shot splits from examples
	zero_shot_examples = zs_dev+zs_test
	ext = filter_senses(ext, zero_shot_examples) 
	ext = train+ext
	ext_path = os.path.join(args.save_dir, 'train.ext.txt')
	save_examples(ext_path, ext)

	#save dev data
	fs_dev_path = os.path.join(args.save_dir, 'dev.few-shot.txt')
	save_examples(fs_dev_path, fs_dev)
	zs_dev_path = os.path.join(args.save_dir, 'dev.zero-shot.txt')
	save_examples(zs_dev_path, zs_dev)

	#save test data
	fs_test_path = os.path.join(args.save_dir, 'test.few-shot.txt')
	save_examples(fs_test_path, fs_test)
	zs_test_path = os.path.join(args.save_dir, 'test.zero-shot.txt')
	save_examples(zs_test_path, zs_test)

	#save monosemous examples as extra data
	mono_path = os.path.join(args.save_dir, 'monosemous.txt')
	mono_examples = monosemous_data+monosemous_ext
	save_examples(mono_path, mono_examples)