urlparse

#!/usr/bin/env python2
# -*- coding: utf-8 -*-
from __future__ import unicode_literals, print_function

import itertools as it, operator as op, functools as ft
from collections import OrderedDict
import os, sys, urlparse, pyaml


def dissect_url(url):
	url = dict(it.izip(
		[ 'scheme', 'netloc', 'path', 'params', 'query',
			'fragment', 'username', 'password', 'hostname', 'port' ],
		urlparse.urlparse(url) ))
	if url.get('query'):
		url['query'] = sorted(urlparse.parse_qsl(url['query']))
	url = OrderedDict(sorted(url.viewitems()))
	# Make "query" always a last element
	url['query'] = url.pop('query')
	return url


def main():
	import argparse
	parser = argparse.ArgumentParser(
		usage='%(prog)s [options] URL1 [URL2] [ -- <diff options> ]',
		description='Process and dump URLs in a human-readable format.')
	parser.add_argument('url', help='URL to process.')
	parser.add_argument('url2', nargs='?', help='URL to show diff against.')
	parser.add_argument('--debug', action='store_true', help='Verbose output.')

	argv, diff_optz = sys.argv[1:], list()
	try: idx = argv.index('--')
	except ValueError: pass
	else: argv, diff_optz = argv[:idx], argv[idx+1:]
	optz = parser.parse_args(argv)

	import logging
	logging.basicConfig(level='DEBUG' if optz.debug else 'WARNING')
	log = logging.getLogger()

	url = dissect_url(optz.url)

	if not optz.url2: pyaml.dump(url, sys.stdout)
	else:
		from tempfile import NamedTemporaryFile
		from subprocess import Popen

		with NamedTemporaryFile() as tmp1, NamedTemporaryFile() as tmp2:
			pyaml.dump(url, tmp1)
			url2 = dissect_url(optz.url2)
			pyaml.dump(url2, tmp2)
			tmp1.flush(), tmp2.flush()
			diff_cmd = ['colordiff', '-uw'] + diff_optz + [tmp1.name, tmp2.name]
			log.debug('Diff command: {}'.format(' '.join(diff_cmd)))
			sys.exit(Popen(diff_cmd).wait())


if __name__ == '__main__': main()