import requests
import re
import sys
from timeit import timeit


# HTML des "Netzwerk"-Unterforums
dfde = requests.get("https://debianforum.de/forum/viewforum.php?f=30").text


def parse(html):
    # (?:.*?) sind non-capturing groups mit soviel wie nötig matchenden Zeichen (non-greedy)
    threadex = re.compile(r"<li class=.row.bg"
                      r"(?:.*?)" 
                      r"(?:class=.topictitle.>)(?P<topictitle>[^<]+)(?:.*?)"
                      r"(?:username.>)(?P<username>[^<]+)(?:</a> &raquo; )"
                      r"(?:.*?)"
                      r"(?P<date>[0-9][0-9.,: ]{18})"
                      r"(?:.*?)"
                      r"(?:posts.>)(?P<posts>[0-9]+)", 
                      re.S)
    for thread in threadex.finditer(dfde):
        yield "{topictitle};{date};{username};{posts}".format(**thread.groupdict())

if __name__ == "__main__":

    if "timeit" in sys.argv:
        timing_results = timeit(lambda: parse(dfde))
        print(timing_results)
    else:
        for thread in parse(dfde):
            print(thread)