emoji_tracker_parse.py

#!/usr/bin/env python

"""
Script to parse the most frequently used emoji unicode values from http://emojitracker.com.
"""

import argparse
from bs4 import BeautifulSoup
import glob
import os

def parse_args():
    parser = argparse.ArgumentParser(
        description='filter only frequently used emoji assets')
    parser.add_argument(
        'dump', help='the raw html (from http://emojitracker.com)')
    parser.add_argument(
        'images', help='the path to the images')
    return parser.parse_args()

def main():
    args = parse_args()
    dump = args.dump
    images = args.images

    with open(dump, 'r') as f:
        soup = BeautifulSoup(f.read(), "html.parser")

    # Max # of results to keep (rankings).
    COUNT = 500

    print "### Parsing up to %s codes..." % COUNT

    # Extract all the codes, preserving order.
    # Links are in the form:
    # <a href="/details/1F52B" title="PISTOL" data-id="1F52B">
    codes = []
    rankings = soup.find("section", {"id": "rankings"})
    for link in soup.findAll('a'):
        data = link.get('data-id')
        if data:
            codes.append(data.lower())
            if len(codes) == COUNT:
                break

    print "### %s codes found..." % len(codes)

    delete_cnt = 0
    filelist = glob.glob(os.path.join(images, '*.png'))
    for f in filelist:
        filename = os.path.splitext(os.path.basename(f))[0].lower()
        if filename not in codes:
            print "### Delete: %s" % f
            os.remove(f)
            delete_cnt += 1

    print "### Deleted %s assets." % delete_cnt

if __name__ == "__main__":
    main()