# Write a function that reads the Simpsons phonebook from
# https://storage.googleapis.com/cs1111/practice/simpsons_phone_book.txt
# The function then finds and returns all phone numbers
# of people from Simpsons TV series.
# Assuming no area code included.
# (hint: use regular expression).

# For practice purpose, download the simpsons_phone_book.txt
# and write another version of the function that will
# read the local file.

import re
import urllib.request

def find_all_phone_number_from_url(url, regex):
    stream = urllib.request.urlopen(url).read().decode('UTF-8')

    # get the entire string, find all matches
    # matches = regex.finditer(stream)
    matches = regex.findall(stream)

    phone_list = []

    for match in matches:
        if match not in phone_list:
            phone_list.append(match)
            # phone_list.append(match.group())
        # print(match.group())

    return phone_list




url = "https://storage.googleapis.com/cs1111/practice/simpsons_phone_book.txt"
regex = re.compile(r"[0-9]?[0-9][0-9][0-9]-[0-9][0-9][0-9][0-9]")   # raw data, compile
print(find_all_phone_number_from_url(url, regex))



####################

# def find_all_phone_number_from_file(filename, regex):
#
#
#
#
#
#
#
#
#
# regex = re.compile(r"[0-9]?[0-9][0-9][0-9]-[0-9][0-9][0-9][0-9]")   # raw data, compile
# print(find_all_phone_number_from_file("simpsons_phone_book.txt", regex))




######################
# Challenging:
# Find all possible phone number of people from Simpsons TV series
# whose first name start with "J" and last name is "Neu"