Detects long sentences. See https://twitter.com/doodlewhale/status/833825449289580545

simple_sentence.py

# author N Ernst
# thanks to https://stackoverflow.com/questions/4576077/python-split-text-on-sentences?rq=1

import nltk.data
import argparse

parser = argparse.ArgumentParser()
parser.add_argument("length", help="recommended sentence length [default 5]",  type=int, default=0)
parser.add_argument("filename", help="what text file to parse")
args = parser.parse_args()

#nltk.download() punkt
tokenizer = nltk.data.load('tokenizers/punkt/english.pickle')
with open(args.filename) as fp:
    data = fp.read()
    sentences = (tokenizer.tokenize(data))
    gt = list(filter(lambda x: len(x.split(' ')) > args.length, sentences))

ratio = len(gt)/len(sentences)
print ("Sentences: {}, longer than {}: {}, ratio: {:.2f}".format(len(sentences), args.length, len(gt), ratio))