#!/usr/bin/env python

import os
import random
import sre
import sys

poststart = sre.compile(r'^\[\[meta author="(?:[^\\"]|\\.)*"\]\]')
dateline  = sre.compile(r'^\[\[meta date="([0-9]+)/([0-9]+)/([0-9]+) +([0-9]+):([0-9]+):([0-9]+) +([AP]M)"\]\]')

breakre = sre.compile(r'< *[bB][rR] */>')
emptydivre = sre.compile(r'< *div(?:[^"]|"(?:[^\\"]|\\.)*")*> *< */div *>')
commentre = sre.compile(r'^Comment by .* at')

def dumppost(lines):
    # Find the date
    fn_stem = None
    for l in lines:
        m = dateline.match(l)
        if m:
            hour = int(m.group(4))
            if m.group(7).upper() == 'PM':
                hour = hour + 12
            fn_stem = '%02d-%02d-%02d--%02d-%02d-%02d' \
                %(int(m.group(3)),
                  int(m.group(1)),
                  int(m.group(2)),
                  hour,
                  int(m.group(5)),
                  int(m.group(6)))
    fn = '%s.mdwn'%fn_stem
    # NB: not race-condition-safe!
    while os.path.exists(fn):
        fn = '%s-%x.mdwn'%(fn_stem, random.randint(0, 1E30))

    file(fn, 'w').writelines(lines)

def find_and_dump_entries(lines):
    currpost = []
    for l in lines:
        if poststart.match(l):
            if len(currpost) > 0:
                dumppost(currpost)
                currpost = []

        # Drop the bogus empty divs blogspot inserts and
        # interpret <br/> to line breaks.
        remove_divs = emptydivre.sub('', l)
        if commentre.match(remove_divs):
            remove_divs = '### ' + remove_divs
        for real_line in breakre.split(remove_divs):
            currpost.append(real_line.rstrip() + '\n')

    if len(currpost) > 0:
        dumppost(currpost)
        currpost = []





in_file = file(sys.argv[1])
find_and_dump_entries(in_file)