logo

searx

My custom branche(s) on searx, a meta-search engine git clone https://anongit.hacktivis.me/git/searx.git/

tokyotoshokan.py (3405B)

    

  1. """
  2.  Tokyo Toshokan (A BitTorrent Library for Japanese Media)
  4.  @website      https://www.tokyotosho.info/
  5.  @provide-api  no
  6.  @using-api    no
  7.  @results      HTML
  8.  @stable       no (HTML can change)
  9.  @parse        url, title, publishedDate, seed, leech,
  10.                filesize, magnetlink, content
  11. """
  13. import re
  14. from lxml import html
  15. from searx.engines.xpath import extract_text
  16. from datetime import datetime
  17. from searx.url_utils import urlencode
  18. from searx.utils import get_torrent_size, int_or_zero
  20. # engine dependent config
  21. categories = ['files', 'videos', 'music']
  22. paging = True
  24. # search-url
  25. base_url = 'https://www.tokyotosho.info/'
  26. search_url = base_url + 'search.php?{query}'
  29. # do search-request
  30. def request(query, params):
  31.     query = urlencode({'page': params['pageno'], 'terms': query})
  32.     params['url'] = search_url.format(query=query)
  33.     return params
  36. # get response from search-request
  37. def response(resp):
  38.     results = []
  40.     dom = html.fromstring(resp.text)
  41.     rows = dom.xpath('//table[@class="listing"]//tr[contains(@class, "category_0")]')
  43.     # check if there are no results or page layout was changed so we cannot parse it
  44.     # currently there are two rows for each result, so total count must be even
  45.     if len(rows) == 0 or len(rows) % 2 != 0:
  46.         return []
  48.     # regular expression for parsing torrent size strings
  49.     size_re = re.compile(r'Size:\s*([\d.]+)(TB|GB|MB|B)', re.IGNORECASE)
  51.     # processing the results, two rows at a time
  52.     for i in range(0, len(rows), 2):
  53.         # parse the first row
  54.         name_row = rows[i]
  56.         links = name_row.xpath('./td[@class="desc-top"]/a')
  57.         params = {
  58.             'template': 'torrent.html',
  59.             'url': links[-1].attrib.get('href'),
  60.             'title': extract_text(links[-1])
  61.         }
  62.         # I have not yet seen any torrents without magnet links, but
  63.         # it's better to be prepared to stumble upon one some day
  64.         if len(links) == 2:
  65.             magnet = links[0].attrib.get('href')
  66.             if magnet.startswith('magnet'):
  67.                 # okay, we have a valid magnet link, let's add it to the result
  68.                 params['magnetlink'] = magnet
  70.         # no more info in the first row, start parsing the second one
  71.         info_row = rows[i + 1]
  72.         desc = extract_text(info_row.xpath('./td[@class="desc-bot"]')[0])
  73.         for item in desc.split('|'):
  74.             item = item.strip()
  75.             if item.startswith('Size:'):
  76.                 try:
  77.                     # ('1.228', 'GB')
  78.                     groups = size_re.match(item).groups()
  79.                     params['filesize'] = get_torrent_size(groups[0], groups[1])
  80.                 except:
  81.                     pass
  82.             elif item.startswith('Date:'):
  83.                 try:
  84.                     # Date: 2016-02-21 21:44 UTC
  85.                     date = datetime.strptime(item, 'Date: %Y-%m-%d %H:%M UTC')
  86.                     params['publishedDate'] = date
  87.                 except:
  88.                     pass
  89.             elif item.startswith('Comment:'):
  90.                 params['content'] = item
  91.         stats = info_row.xpath('./td[@class="stats"]/span')
  92.         # has the layout not changed yet?
  93.         if len(stats) == 3:
  94.             params['seed'] = int_or_zero(extract_text(stats[0]))
  95.             params['leech'] = int_or_zero(extract_text(stats[1]))
  97.         results.append(params)
  99.     return results