Add Tokyo Toshokan search engine

This commit is contained in:
Kirill Isakov 2016-03-27 00:49:57 +06:00
parent 7fbc12ee4e
commit 547b8a8765
3 changed files with 218 additions and 0 deletions

View File

@ -0,0 +1,102 @@
"""
Tokyo Toshokan (A BitTorrent Library for Japanese Media)
@website https://www.tokyotosho.info/
@provide-api no
@using-api no
@results HTML
@stable no (HTML can change)
@parse url, title, publishedDate, seed, leech,
filesize, magnetlink, content
"""
import re
from cgi import escape
from urllib import urlencode
from lxml import html
from searx.engines.xpath import extract_text
from datetime import datetime
from searx.engines.nyaa import int_or_zero, get_filesize_mul
# engine dependent config
categories = ['files', 'videos', 'music']
paging = True
# search-url
base_url = 'https://www.tokyotosho.info/'
search_url = base_url + 'search.php?{query}'
# do search-request
def request(query, params):
query = urlencode({'page': params['pageno'],
'terms': query})
params['url'] = search_url.format(query=query)
return params
# get response from search-request
def response(resp):
results = []
dom = html.fromstring(resp.text)
rows = dom.xpath('//table[@class="listing"]//tr[contains(@class, "category_0")]')
# check if there are no results or page layout was changed so we cannot parse it
# currently there are two rows for each result, so total count must be even
if len(rows) == 0 or len(rows) % 2 != 0:
return []
# regular expression for parsing torrent size strings
size_re = re.compile('Size:\s*([\d.]+)(TB|GB|MB|B)', re.IGNORECASE)
# processing the results, two rows at a time
for i in xrange(0, len(rows), 2):
# parse the first row
name_row = rows[i]
links = name_row.xpath('./td[@class="desc-top"]/a')
params = {
'template': 'torrent.html',
'url': links[-1].attrib.get('href'),
'title': extract_text(links[-1])
}
# I have not yet seen any torrents without magnet links, but
# it's better to be prepared to stumble upon one some day
if len(links) == 2:
magnet = links[0].attrib.get('href')
if magnet.startswith('magnet'):
# okay, we have a valid magnet link, let's add it to the result
params['magnetlink'] = magnet
# no more info in the first row, start parsing the second one
info_row = rows[i + 1]
desc = extract_text(info_row.xpath('./td[@class="desc-bot"]')[0])
for item in desc.split('|'):
item = item.strip()
if item.startswith('Size:'):
try:
# ('1.228', 'GB')
groups = size_re.match(item).groups()
multiplier = get_filesize_mul(groups[1])
params['filesize'] = int(multiplier * float(groups[0]))
except Exception as e:
pass
elif item.startswith('Date:'):
try:
# Date: 2016-02-21 21:44 UTC
date = datetime.strptime(item, 'Date: %Y-%m-%d %H:%M UTC')
params['publishedDate'] = date
except Exception as e:
pass
elif item.startswith('Comment:'):
params['content'] = item
stats = info_row.xpath('./td[@class="stats"]/span')
# has the layout not changed yet?
if len(stats) == 3:
params['seed'] = int_or_zero(extract_text(stats[0]))
params['leech'] = int_or_zero(extract_text(stats[1]))
results.append(params)
return results

View File

@ -271,6 +271,12 @@ engines:
shortcut : sw shortcut : sw
disabled : True disabled : True
- name : tokyotoshokan
engine : tokyotoshokan
shortcut : tt
timeout : 6.0
disabled : True
- name : torrentz - name : torrentz
engine : torrentz engine : torrentz
timeout : 5.0 timeout : 5.0

View File

@ -0,0 +1,110 @@
import mock
from collections import defaultdict
from searx.engines import tokyotoshokan
from searx.testing import SearxTestCase
from datetime import datetime
class TestTokyotoshokanEngine(SearxTestCase):
def test_request(self):
query = 'test_query'
dic = defaultdict(dict)
dic['pageno'] = 1
params = tokyotoshokan.request(query, dic)
self.assertTrue('url' in params)
self.assertTrue(query in params['url'])
self.assertTrue('tokyotosho.info' in params['url'])
def test_response(self):
resp = mock.Mock(text='<html></html>')
self.assertEqual(tokyotoshokan.response(resp), [])
html = """
<table class="listing">
<tbody>
<tr class="shade category_0">
<td rowspan="2">
<a href="/?cat=7"><span class="sprite_cat-raw"></span></a>
</td>
<td class="desc-top">
<a href="magnet:?xt=urn:btih:4c19eb46b5113685fbd2288ed2531b0b">
<span class="sprite_magnet"></span>
</a>
<a rel="nofollow" type="application/x-bittorrent" href="http://www.nyaa.se/f">
Koyomimonogatari
</a>
</td>
<td class="web"><a rel="nofollow" href="details.php?id=975700">Details</a></td>
</tr>
<tr class="shade category_0">
<td class="desc-bot">
Authorized: <span class="auth_ok">Yes</span>
Submitter: <a href="?username=Ohys">Ohys</a> |
Size: 10.5MB |
Date: 2016-03-26 16:41 UTC |
Comment: sample comment
</td>
<td style="color: #BBB; font-family: monospace" class="stats" align="right">
S: <span style="color: red">53</span>
L: <span style="color: red">18</span>
C: <span style="color: red">0</span>
ID: 975700
</td>
</tr>
<tr class="category_0">
<td rowspan="2">
<a href="/?cat=7"><span class="sprite_cat-raw"></span></a>
</td>
<td class="desc-top">
<a rel="nofollow" type="application/x-bittorrent" href="http://google.com/q">
Owarimonogatari
</a>
</td>
<td class="web"><a rel="nofollow" href="details.php?id=975700">Details</a></td>
</tr>
<tr class="category_0">
<td class="desc-bot">
Submitter: <a href="?username=Ohys">Ohys</a> |
Size: 932.84EB |
Date: QWERTY-03-26 16:41 UTC
</td>
<td style="color: #BBB; font-family: monospace" class="stats" align="right">
S: <span style="color: red">0</span>
</td>
</tr>
</tbody>
</table>
"""
resp = mock.Mock(text=html)
results = tokyotoshokan.response(resp)
self.assertEqual(type(results), list)
self.assertEqual(len(results), 2)
# testing the first result, which has correct format
# and should have all information fields filled
r = results[0]
self.assertEqual(r['url'], 'http://www.nyaa.se/f')
self.assertEqual(r['title'], 'Koyomimonogatari')
self.assertEqual(r['magnetlink'], 'magnet:?xt=urn:btih:4c19eb46b5113685fbd2288ed2531b0b')
self.assertEqual(r['filesize'], int(1024 * 1024 * 10.5))
self.assertEqual(r['publishedDate'], datetime(2016, 03, 26, 16, 41))
self.assertEqual(r['content'], 'Comment: sample comment')
self.assertEqual(r['seed'], 53)
self.assertEqual(r['leech'], 18)
# testing the second result, which does not include magnet link,
# seed & leech info, and has incorrect size & creation date
r = results[1]
self.assertEqual(r['url'], 'http://google.com/q')
self.assertEqual(r['title'], 'Owarimonogatari')
self.assertFalse('magnetlink' in r)
self.assertFalse('filesize' in r)
self.assertFalse('content' in r)
self.assertFalse('publishedDate' in r)
self.assertFalse('seed' in r)
self.assertFalse('leech' in r)