valda-video_scraper 1.0.1 → 1.0.2
Sign up to get free protection for your applications and to get access to all the features.
- data/README +8 -7
- data/Rakefile +1 -0
- data/lib/www/video_scraper.rb +8 -5
- data/lib/www/video_scraper/base.rb +3 -0
- data/lib/www/video_scraper/tube8.rb +28 -0
- data/test/www/video_scraper/test_tube8.rb +24 -0
- metadata +13 -2
data/README
CHANGED
@@ -11,18 +11,19 @@ Web scraping library for video sharing sites.
|
|
11
11
|
|
12
12
|
Supported sites
|
13
13
|
|
14
|
-
* YouTube
|
15
|
-
* NICO NICO DOUGA
|
16
14
|
* AmebaVision
|
17
15
|
* Dailymotion
|
18
|
-
* Veoh
|
19
|
-
* YourFileHost
|
20
|
-
* RedTube
|
21
|
-
* Pornhub
|
22
|
-
* Ura Agesage
|
23
16
|
* MoroTube
|
17
|
+
* NICO NICO DOUGA
|
18
|
+
* Pornhub
|
24
19
|
* Pornotube
|
20
|
+
* RedTube
|
21
|
+
* Tube8
|
22
|
+
* Ura Agesage
|
23
|
+
* Veoh
|
25
24
|
* YouPorn
|
25
|
+
* YouTube
|
26
|
+
* YourFileHost
|
26
27
|
|
27
28
|
== SYNOPSIS:
|
28
29
|
|
data/Rakefile
CHANGED
data/lib/www/video_scraper.rb
CHANGED
@@ -2,24 +2,23 @@
|
|
2
2
|
|
3
3
|
require 'rubygems'
|
4
4
|
require 'open-uri'
|
5
|
+
require 'hpricot'
|
5
6
|
require 'mechanize'
|
6
7
|
require 'kconv'
|
7
8
|
require 'json'
|
8
9
|
require 'uri'
|
9
10
|
begin
|
10
|
-
require 'cgialt'
|
11
|
+
require 'cgialt' unless defined? CGI
|
11
12
|
rescue LoadError
|
12
13
|
require 'cgi'
|
13
14
|
end
|
14
15
|
|
15
16
|
module WWW
|
16
17
|
module VideoScraper
|
17
|
-
|
18
|
-
class FileNotFound < RuntimeError; end
|
19
|
-
VERSION = '1.0.1'
|
18
|
+
VERSION = '1.0.2'
|
20
19
|
|
21
20
|
MODULES_NAME = %w(age_sage ameba_vision dailymotion moro_tube
|
22
|
-
nico_video pornhub pornotube red_tube veoh
|
21
|
+
nico_video pornhub pornotube red_tube tube8 veoh
|
23
22
|
you_porn you_tube your_file_host)
|
24
23
|
|
25
24
|
@@modules = MODULES_NAME.map do |name|
|
@@ -34,6 +33,10 @@ module WWW
|
|
34
33
|
}
|
35
34
|
|
36
35
|
class << self
|
36
|
+
def modules
|
37
|
+
@@nodules
|
38
|
+
end
|
39
|
+
|
37
40
|
def options
|
38
41
|
@@options
|
39
42
|
end
|
@@ -0,0 +1,28 @@
|
|
1
|
+
# -*- mode:ruby; coding:utf-8 -*-
|
2
|
+
|
3
|
+
require File.expand_path(File.dirname(__FILE__) + '/base')
|
4
|
+
|
5
|
+
module WWW
|
6
|
+
module VideoScraper
|
7
|
+
class Tube8 < Base
|
8
|
+
url_regex %r!\Ahttp://www\.tube8\.com/.*/(\d+)(?:/|$)!
|
9
|
+
|
10
|
+
def initialize(url, opt = nil)
|
11
|
+
super
|
12
|
+
do_query
|
13
|
+
end
|
14
|
+
|
15
|
+
private
|
16
|
+
def do_query
|
17
|
+
html = http_get(@page_url)
|
18
|
+
doc = Hpricot(html.toutf8)
|
19
|
+
raise FileNotFound unless flashvars = doc.at('//object //param[@name="FlashVars"]')
|
20
|
+
flashvars = CGI.parse(flashvars.attributes['value'])
|
21
|
+
@video_url = flashvars['videoUrl'][0]
|
22
|
+
uri = URI.parse(@page_url)
|
23
|
+
@thumb_url = URI.join("#{uri.scheme}://#{uri.host}", flashvars['imageUrl'][0]).to_s
|
24
|
+
@title = doc.at('//h1[@class="text"]').inner_html rescue nil
|
25
|
+
end
|
26
|
+
end
|
27
|
+
end
|
28
|
+
end
|
@@ -0,0 +1,24 @@
|
|
1
|
+
# -*- mode:ruby; coding:utf-8 -*-
|
2
|
+
|
3
|
+
require File.dirname(__FILE__) + '/../../test_helper'
|
4
|
+
|
5
|
+
class TestTube8 < Test::Unit::TestCase
|
6
|
+
def setup
|
7
|
+
@cache_root = '/tmp/test_video_scraper_cache'
|
8
|
+
WWW::VideoScraper.configure do |conf|
|
9
|
+
conf[:cache] = FileCache.new('TestVideoScraper', @cache_root, 60*60*24)
|
10
|
+
end
|
11
|
+
end
|
12
|
+
|
13
|
+
def teardown
|
14
|
+
# FileUtils.remove_entry_secure(@cache_root, true)
|
15
|
+
end
|
16
|
+
|
17
|
+
def test_scrape
|
18
|
+
vs = WWW::VideoScraper.scrape('http://www.tube8.com/anal/alexis-amore-pov/56983/')
|
19
|
+
assert_equal 'http://www.tube8.com/anal/alexis-amore-pov/56983/', vs.page_url
|
20
|
+
assert_match %r|http://medianl\d+\.tube8\.com/flv/[[:alnum:]]{32}/\d{8}/\d{4}/\d{2}/[[:alnum:]]+/[[:alnum:]]+\.flv|, vs.video_url
|
21
|
+
assert_equal 'http://www.tube8.com/vs/83/56983.jpg', vs.thumb_url
|
22
|
+
assert_equal 'Alexis Amore POV', vs.title
|
23
|
+
end
|
24
|
+
end
|
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: valda-video_scraper
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.0.
|
4
|
+
version: 1.0.2
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- YAMAGUCHI Seiji
|
@@ -9,7 +9,7 @@ autorequire:
|
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
11
|
|
12
|
-
date: 2009-01-
|
12
|
+
date: 2009-01-18 00:00:00 -08:00
|
13
13
|
default_executable:
|
14
14
|
dependencies:
|
15
15
|
- !ruby/object:Gem::Dependency
|
@@ -21,6 +21,15 @@ dependencies:
|
|
21
21
|
- !ruby/object:Gem::Version
|
22
22
|
version: 0.8.4
|
23
23
|
version:
|
24
|
+
- !ruby/object:Gem::Dependency
|
25
|
+
name: hpricot
|
26
|
+
version_requirement:
|
27
|
+
version_requirements: !ruby/object:Gem::Requirement
|
28
|
+
requirements:
|
29
|
+
- - ">="
|
30
|
+
- !ruby/object:Gem::Version
|
31
|
+
version: 0.6.164
|
32
|
+
version:
|
24
33
|
description: Web scraping library for video sharing sites.
|
25
34
|
email: valda@underscore.jp
|
26
35
|
executables: []
|
@@ -41,6 +50,7 @@ files:
|
|
41
50
|
- test/www/video_scraper/test_dailymotion.rb
|
42
51
|
- test/www/video_scraper/test_age_sage.rb
|
43
52
|
- test/www/video_scraper/test_pornotube.rb
|
53
|
+
- test/www/video_scraper/test_tube8.rb
|
44
54
|
- test/www/video_scraper/test_your_file_host.rb
|
45
55
|
- test/www/video_scraper/test_moro_tube.rb
|
46
56
|
- test/www/video_scraper/test_veoh.rb
|
@@ -65,6 +75,7 @@ files:
|
|
65
75
|
- lib/www/video_scraper/red_tube.rb
|
66
76
|
- lib/www/video_scraper/base.rb
|
67
77
|
- lib/www/video_scraper/your_file_host.rb
|
78
|
+
- lib/www/video_scraper/tube8.rb
|
68
79
|
- lib/www/video_scraper.rb
|
69
80
|
has_rdoc: true
|
70
81
|
homepage: http://github.com/valda/video_scraper
|