spider_rails 4.0.0 → 4.0.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/Rakefile +2 -13
- data/lib/spider_rails/common.rb +2 -0
- data/lib/spider_rails/ji_ying.rb +37 -19
- data/lib/spider_rails/version.rb +1 -1
- metadata +1 -1
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 45eb0723b5f87fe4d47fe672b0731b2b023a6c9c
|
4
|
+
data.tar.gz: 7c08ae9a4ba6499bcda5b4c01038a0695e378d39
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 61551419495a9dd7cd11cdb81a81e8a79fc5d0dd8e840ac7dbc3eb292979911e5f626e933c1a9f704cba42cd14933a3e5f756aab4f1c40ddb9cbeb96dc435f61
|
7
|
+
data.tar.gz: aa82ef9627e5661cdb32e4face134711249599dcde520238982469aa089cbbfb6d0fcecf13a3bf7f02556f905244b20e34b9f96836eed55dda31f15304e0b2b8
|
data/Rakefile
CHANGED
@@ -14,19 +14,8 @@ RDoc::Task.new(:rdoc) do |rdoc|
|
|
14
14
|
rdoc.rdoc_files.include('lib/**/*.rb')
|
15
15
|
end
|
16
16
|
|
17
|
-
|
18
|
-
|
19
|
-
|
20
17
|
Bundler::GemHelper.install_tasks
|
21
18
|
|
22
|
-
require '
|
23
|
-
|
24
|
-
Rake::TestTask.new(:test) do |t|
|
25
|
-
t.libs << 'lib'
|
26
|
-
t.libs << 'test'
|
27
|
-
t.pattern = 'test/**/*_test.rb'
|
28
|
-
t.verbose = false
|
29
|
-
end
|
30
|
-
|
19
|
+
require 'rspec/core/rake_task'
|
31
20
|
|
32
|
-
|
21
|
+
RSpec::Core::RakeTask.new(:spec)
|
data/lib/spider_rails/common.rb
CHANGED
@@ -43,6 +43,7 @@ module Spider
|
|
43
43
|
class << self
|
44
44
|
def open_browser url
|
45
45
|
@browser = Watir::Browser.new :chrome, switches: %w( --user-data-dir=/home/zxr/.config/google-chrome)
|
46
|
+
#@browser = Watir::Browser.new :chrome
|
46
47
|
@browser.goto url
|
47
48
|
@browser
|
48
49
|
end
|
@@ -58,3 +59,4 @@ module Spider
|
|
58
59
|
end
|
59
60
|
|
60
61
|
|
62
|
+
|
data/lib/spider_rails/ji_ying.rb
CHANGED
@@ -1,11 +1,19 @@
|
|
1
1
|
module Spider
|
2
2
|
class JiYing < Common
|
3
|
-
attr_accessor :ani, :anis
|
3
|
+
attr_accessor :ani, :anis, :page
|
4
4
|
|
5
5
|
def initialize page
|
6
6
|
@ani = Hash.new
|
7
7
|
@anis = Array.new
|
8
8
|
@page = page
|
9
|
+
@base_url = page.url.gsub!(/\b\d(.html)?/, '')
|
10
|
+
|
11
|
+
if @base_url.include?('search')
|
12
|
+
@mode = 'search'
|
13
|
+
else
|
14
|
+
@mode = 'normal'
|
15
|
+
end
|
16
|
+
|
9
17
|
end
|
10
18
|
|
11
19
|
def full_site
|
@@ -20,40 +28,50 @@ module Spider
|
|
20
28
|
end
|
21
29
|
|
22
30
|
def final_page
|
23
|
-
@
|
31
|
+
if @mode == 'search'
|
32
|
+
fp = @page.element(css: '.title h2 a').text[/\(.+\)/].gsub!(/\(|\)/, '').to_i/100 + 1
|
33
|
+
else
|
34
|
+
fp = @page.element(css: 'a.pager-last').text.to_i
|
35
|
+
end
|
36
|
+
fp
|
24
37
|
end
|
25
38
|
|
26
39
|
def full_page page_num
|
27
|
-
@
|
28
|
-
|
29
|
-
|
30
|
-
|
40
|
+
case @mode
|
41
|
+
when 'search'
|
42
|
+
@page.goto "#{@base_url}#{page_num}"
|
43
|
+
html = Nokogiri::HTML.parse @page.html
|
44
|
+
html.css('.ltext').each do |td|
|
45
|
+
sing le(td)
|
46
|
+
end
|
47
|
+
when 'normal'
|
48
|
+
@page.goto "#{@base_url}#{page_num}.html"
|
31
49
|
end
|
32
50
|
end
|
33
51
|
|
34
|
-
def single element
|
52
|
+
def single element, &block
|
35
53
|
get(element)
|
36
|
-
@anis
|
54
|
+
@anis
|
37
55
|
end
|
38
56
|
|
39
57
|
def get(element)
|
40
58
|
element.css('a.quick-down').each do |a|
|
41
|
-
@ani[:torrent] = a['href']
|
42
|
-
end
|
43
|
-
element.css('a.quick-down+a').each do |a|
|
44
|
-
@ani[:title] = a.content
|
59
|
+
@ani[:torrent] = 'http://bt.ktxp.com' + a['href']
|
45
60
|
end
|
61
|
+
@ani[:title],
|
62
|
+
@ani[:size],
|
63
|
+
@ani[:finish] = get_content(element, 'a.quick-down+a', 'td.ltext+td', 'td.ltext+td+td+td+td')
|
46
64
|
@ani
|
47
65
|
end
|
48
66
|
|
49
|
-
def
|
50
|
-
|
51
|
-
|
52
|
-
|
53
|
-
|
67
|
+
def get_content(element, *selectors)
|
68
|
+
tds = Array.new
|
69
|
+
selectors.each do |selector|
|
70
|
+
element.css(selector).each do |td|
|
71
|
+
tds << td.text
|
72
|
+
end
|
54
73
|
end
|
55
|
-
|
56
|
-
@ani if @ani
|
74
|
+
tds
|
57
75
|
end
|
58
76
|
end
|
59
77
|
end
|
data/lib/spider_rails/version.rb
CHANGED