#!/usr/bin/env python3

from datetime import datetime, timedelta
import json
import os.path as op
import sys
from operator import itemgetter
from urllib.parse import urlparse

from dateutil.parser import parse as parse_date
from lxml.html import document_fromstring
from requests import get
import werkzeug.contrib.atom as atom


YAHOO_JAPAN_URL = 'http://auctions.search.yahoo.co.jp/search'
YAHOO_JAPAN_PAYLOAD = {
    'p': 'おやすみプンプン',
    'disp_number': 100000,
    'n': 100
}
YAHOO_JAPAN_URL_SELECTOR = 'div#AS1m1 div#list01 tr td.a1 h3 a'
YAHOO_JAPAN_DESC_SELECTOR = 'div.untHead+div.modUsrPrv'
BUYEE_JAPAN_DATE_SELECTOR = 'ul#itemDetail_data'
BUYEE_TIMESTAMP_FORMAT = '%d %b %Y %H:%M:%S'
BUYEE_URL = 'http://buyee.jp/item/yahoo/auction/'
JSON_FILE = '~/rss/yahoo.json'
ATOM_FILE = '~/rss/yahoo.atom'


def get_selections(url, selector, params=None):
    r = get(url, params=params)
    dom = document_fromstring(r.text)
    return dom.cssselect(selector)


def get_yahoo_auctions():
    results = []
    yahoo_urls = [result.get('href') for result in
                  get_selections(YAHOO_JAPAN_URL,
                                 YAHOO_JAPAN_URL_SELECTOR,
                                 params=YAHOO_JAPAN_PAYLOAD)]

    print('{} auctions found!'.format(len(yahoo_urls)))
    for yahoo_url in yahoo_urls:
        yahoo_id = urlparse(yahoo_url).path.split('/')[-1]
        buyee_url = '{}{}{}'.format(BUYEE_URL, yahoo_id, '/lang/en')

        for item in get_selections(
                buyee_url,
                BUYEE_JAPAN_DATE_SELECTOR)[0].text_content().split('\n'):
            if 'Closing Time' in item:
                buyee_datetime = datetime.strptime(
                    item.split(')')[-1],
                    BUYEE_TIMESTAMP_FORMAT).isoformat(sep=' ')
                break

        current_datetime = datetime.now().isoformat(sep=' ')
        result = {'yahoo_id': yahoo_id, 'yahoo_url': yahoo_url,
                  'buyee_url': buyee_url, 'buyee_datetime':
                  buyee_datetime, 'current_datetime':
                  current_datetime}
        print('Auction {} processed'.format(yahoo_id))
        results.append(result)

    return(sorted(results, key=itemgetter('buyee_datetime')))


def create_yahoo_feed(feed_data):
    yahoo_feed = atom.AtomFeed(
        title='おやすみプンプン', title_type='txt', author='Wasa',
        subtitle='Search on Yahoo Auctions JP for おやすみプンプン',
        feed_url='file:///home/wasa/rss/yahoo')
    for feed in feed_data:
        title = feed['yahoo_id']
        url = feed['yahoo_url']
        description_url = feed['buyee_url']
        updated = parse_date(feed['current_datetime'])
        published = parse_date(feed['buyee_datetime'])

        content = '''Ending on {}, <a href="{}">metadata</a>'''.format(
            published, description_url)

        yahoo_feed.add(title=title, content=content,
                       content_type='html', url=url, updated=updated,
                       published=published)
    return yahoo_feed


def main():
    auctions = get_yahoo_auctions()
    new_auctions = []

    if op.exists(op.expanduser(JSON_FILE)):
        with open(op.expanduser(JSON_FILE), 'r') as json_file:
            try:
                new_auctions = json.load(json_file)
            except ValueError:
                sys.exit("Invalid JSON")
    print('{} auctions loaded'.format(len(new_auctions)))

    now = datetime.now()
    raw_feed = sorted({
        a['yahoo_id']: a for a in auctions + new_auctions
        if now - parse_date(a['buyee_datetime']) < timedelta(hours=1)
    }.values(), key=itemgetter('buyee_datetime'))

    with open(op.expanduser(JSON_FILE), 'w') as json_file:
        json_file.write(json.dumps(raw_feed, sort_keys=True, indent=2,
                                   separators=(',', ': ')))
    print('{} auctions saved'.format(len(raw_feed)))

    yahoo_feed = create_yahoo_feed(raw_feed)
    with open(op.expanduser(ATOM_FILE), 'w') as atom_file:
        atom_file.write(yahoo_feed.to_string())


if __name__ == '__main__':
    main()