semantic-crawler 0.0.2
Sign up to get free protection for your applications and to get access to all the features.
- data/MIT-LICENSE +20 -0
- data/README.rdoc +70 -0
- data/Rakefile +38 -0
- data/lib/semantic_crawler.rb +13 -0
- data/lib/semantic_crawler/factbook/country.rb +124 -0
- data/lib/semantic_crawler/freebase/country.rb +33 -0
- data/lib/semantic_crawler/freebase/crawler.rb +21 -0
- data/lib/semantic_crawler/freebase/entity.rb +3 -0
- data/lib/semantic_crawler/version.rb +3 -0
- data/lib/tasks/semantic_crawler_tasks.rake +4 -0
- data/test/dummy/README.rdoc +261 -0
- data/test/dummy/Rakefile +7 -0
- data/test/dummy/app/assets/javascripts/application.js +15 -0
- data/test/dummy/app/assets/stylesheets/application.css +13 -0
- data/test/dummy/app/controllers/application_controller.rb +3 -0
- data/test/dummy/app/helpers/application_helper.rb +2 -0
- data/test/dummy/app/views/layouts/application.html.erb +14 -0
- data/test/dummy/config.ru +4 -0
- data/test/dummy/config/application.rb +56 -0
- data/test/dummy/config/boot.rb +10 -0
- data/test/dummy/config/database.yml +25 -0
- data/test/dummy/config/environment.rb +5 -0
- data/test/dummy/config/environments/development.rb +37 -0
- data/test/dummy/config/environments/production.rb +67 -0
- data/test/dummy/config/environments/test.rb +37 -0
- data/test/dummy/config/initializers/backtrace_silencers.rb +7 -0
- data/test/dummy/config/initializers/inflections.rb +15 -0
- data/test/dummy/config/initializers/mime_types.rb +5 -0
- data/test/dummy/config/initializers/secret_token.rb +7 -0
- data/test/dummy/config/initializers/session_store.rb +8 -0
- data/test/dummy/config/initializers/wrap_parameters.rb +14 -0
- data/test/dummy/config/locales/en.yml +5 -0
- data/test/dummy/config/routes.rb +58 -0
- data/test/dummy/db/development.sqlite3 +0 -0
- data/test/dummy/db/test.sqlite3 +0 -0
- data/test/dummy/log/development.log +55 -0
- data/test/dummy/log/test.log +2888 -0
- data/test/dummy/public/404.html +26 -0
- data/test/dummy/public/422.html +26 -0
- data/test/dummy/public/500.html +25 -0
- data/test/dummy/public/favicon.ico +0 -0
- data/test/dummy/script/rails +6 -0
- data/test/factbook_test.rb +86 -0
- data/test/semantic_crawler_test.rb +27 -0
- data/test/test_helper.rb +10 -0
- metadata +230 -0
@@ -0,0 +1,26 @@
|
|
1
|
+
<!DOCTYPE html>
|
2
|
+
<html>
|
3
|
+
<head>
|
4
|
+
<title>The page you were looking for doesn't exist (404)</title>
|
5
|
+
<style type="text/css">
|
6
|
+
body { background-color: #fff; color: #666; text-align: center; font-family: arial, sans-serif; }
|
7
|
+
div.dialog {
|
8
|
+
width: 25em;
|
9
|
+
padding: 0 4em;
|
10
|
+
margin: 4em auto 0 auto;
|
11
|
+
border: 1px solid #ccc;
|
12
|
+
border-right-color: #999;
|
13
|
+
border-bottom-color: #999;
|
14
|
+
}
|
15
|
+
h1 { font-size: 100%; color: #f00; line-height: 1.5em; }
|
16
|
+
</style>
|
17
|
+
</head>
|
18
|
+
|
19
|
+
<body>
|
20
|
+
<!-- This file lives in public/404.html -->
|
21
|
+
<div class="dialog">
|
22
|
+
<h1>The page you were looking for doesn't exist.</h1>
|
23
|
+
<p>You may have mistyped the address or the page may have moved.</p>
|
24
|
+
</div>
|
25
|
+
</body>
|
26
|
+
</html>
|
@@ -0,0 +1,26 @@
|
|
1
|
+
<!DOCTYPE html>
|
2
|
+
<html>
|
3
|
+
<head>
|
4
|
+
<title>The change you wanted was rejected (422)</title>
|
5
|
+
<style type="text/css">
|
6
|
+
body { background-color: #fff; color: #666; text-align: center; font-family: arial, sans-serif; }
|
7
|
+
div.dialog {
|
8
|
+
width: 25em;
|
9
|
+
padding: 0 4em;
|
10
|
+
margin: 4em auto 0 auto;
|
11
|
+
border: 1px solid #ccc;
|
12
|
+
border-right-color: #999;
|
13
|
+
border-bottom-color: #999;
|
14
|
+
}
|
15
|
+
h1 { font-size: 100%; color: #f00; line-height: 1.5em; }
|
16
|
+
</style>
|
17
|
+
</head>
|
18
|
+
|
19
|
+
<body>
|
20
|
+
<!-- This file lives in public/422.html -->
|
21
|
+
<div class="dialog">
|
22
|
+
<h1>The change you wanted was rejected.</h1>
|
23
|
+
<p>Maybe you tried to change something you didn't have access to.</p>
|
24
|
+
</div>
|
25
|
+
</body>
|
26
|
+
</html>
|
@@ -0,0 +1,25 @@
|
|
1
|
+
<!DOCTYPE html>
|
2
|
+
<html>
|
3
|
+
<head>
|
4
|
+
<title>We're sorry, but something went wrong (500)</title>
|
5
|
+
<style type="text/css">
|
6
|
+
body { background-color: #fff; color: #666; text-align: center; font-family: arial, sans-serif; }
|
7
|
+
div.dialog {
|
8
|
+
width: 25em;
|
9
|
+
padding: 0 4em;
|
10
|
+
margin: 4em auto 0 auto;
|
11
|
+
border: 1px solid #ccc;
|
12
|
+
border-right-color: #999;
|
13
|
+
border-bottom-color: #999;
|
14
|
+
}
|
15
|
+
h1 { font-size: 100%; color: #f00; line-height: 1.5em; }
|
16
|
+
</style>
|
17
|
+
</head>
|
18
|
+
|
19
|
+
<body>
|
20
|
+
<!-- This file lives in public/500.html -->
|
21
|
+
<div class="dialog">
|
22
|
+
<h1>We're sorry, but something went wrong.</h1>
|
23
|
+
</div>
|
24
|
+
</body>
|
25
|
+
</html>
|
File without changes
|
@@ -0,0 +1,6 @@
|
|
1
|
+
#!/usr/bin/env ruby18
|
2
|
+
# This command will automatically be run when you run "rails" with Rails 3 gems installed from the root of your application.
|
3
|
+
|
4
|
+
APP_PATH = File.expand_path('../../config/application', __FILE__)
|
5
|
+
require File.expand_path('../../config/boot', __FILE__)
|
6
|
+
require 'rails/commands'
|
@@ -0,0 +1,86 @@
|
|
1
|
+
require 'test_helper'
|
2
|
+
require 'pp'
|
3
|
+
|
4
|
+
class FactbookTest < ActiveSupport::TestCase
|
5
|
+
test "truth" do
|
6
|
+
assert_kind_of Module, SemanticCrawler
|
7
|
+
end
|
8
|
+
|
9
|
+
test "init austria country information" do
|
10
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
11
|
+
assert austria.country_name.eql?("austria")
|
12
|
+
assert austria.url.eql?("http://www4.wiwiss.fu-berlin.de/factbook/data/austria")
|
13
|
+
end
|
14
|
+
|
15
|
+
test "get austria name" do
|
16
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
17
|
+
value = austria.name
|
18
|
+
assert value.to_s.eql?("Austria")
|
19
|
+
end
|
20
|
+
|
21
|
+
test "get austria background information" do
|
22
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
23
|
+
value = austria.background
|
24
|
+
assert !value.nil?
|
25
|
+
end
|
26
|
+
|
27
|
+
test "get austria population information" do
|
28
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
29
|
+
value = austria.population_total
|
30
|
+
assert !value.nil?
|
31
|
+
end
|
32
|
+
|
33
|
+
test "get austria legalsystem property (not wrapped)" do
|
34
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
35
|
+
value = austria.get_factbook_property("legalsystem")
|
36
|
+
assert !value.nil?
|
37
|
+
end
|
38
|
+
|
39
|
+
test "get austria latitude" do
|
40
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
41
|
+
value = austria.latitude
|
42
|
+
assert value.to_s.eql?("47.3333333333333")
|
43
|
+
end
|
44
|
+
|
45
|
+
test "get austria longitude" do
|
46
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
47
|
+
value = austria.longitude
|
48
|
+
assert value.to_s.eql?("13.3333333333333")
|
49
|
+
end
|
50
|
+
|
51
|
+
test "get austria landboundary" do
|
52
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
53
|
+
value = austria.landboundary
|
54
|
+
assert value.count == 7
|
55
|
+
end
|
56
|
+
|
57
|
+
test "get austria terrain" do
|
58
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
59
|
+
value = austria.terrain
|
60
|
+
assert !value.nil?
|
61
|
+
end
|
62
|
+
|
63
|
+
test "get austria total airports" do
|
64
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
65
|
+
value = austria.airports_total
|
66
|
+
assert value.to_s.eql?("55")
|
67
|
+
end
|
68
|
+
|
69
|
+
test "get austria heliports" do
|
70
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
71
|
+
value = austria.heliports
|
72
|
+
assert value.to_s.eql?("1")
|
73
|
+
end
|
74
|
+
|
75
|
+
test "get austria climate" do
|
76
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
77
|
+
value = austria.climate
|
78
|
+
assert !value.nil?
|
79
|
+
end
|
80
|
+
|
81
|
+
test "get austria location description" do
|
82
|
+
austria = SemanticCrawler::Factbook::Country.new("Austria")
|
83
|
+
value = austria.location
|
84
|
+
assert !value.nil?
|
85
|
+
end
|
86
|
+
end
|
@@ -0,0 +1,27 @@
|
|
1
|
+
require 'test_helper'
|
2
|
+
require 'pp'
|
3
|
+
|
4
|
+
class SemanticCrawlerTest < ActiveSupport::TestCase
|
5
|
+
test "truth" do
|
6
|
+
assert_kind_of Module, SemanticCrawler
|
7
|
+
end
|
8
|
+
|
9
|
+
test "init austria country information" do
|
10
|
+
austria = SemanticCrawler::Freebase::Crawler.new
|
11
|
+
countryObj = austria.getCountry("Austria")
|
12
|
+
countryNameNormalized = countryObj.input_name
|
13
|
+
assert countryNameNormalized.eql?("austria")
|
14
|
+
end
|
15
|
+
|
16
|
+
test "get austria country information" do
|
17
|
+
austria = SemanticCrawler::Freebase::Crawler.new
|
18
|
+
countryObj = austria.getCountry("Austria")
|
19
|
+
#pp countryObj.get["properties"]["/location/statistical_region/population"]
|
20
|
+
end
|
21
|
+
|
22
|
+
test "get empty crawler object" do
|
23
|
+
empty = SemanticCrawler::Freebase::Crawler.new
|
24
|
+
assert !empty.nil?
|
25
|
+
end
|
26
|
+
|
27
|
+
end
|
data/test/test_helper.rb
ADDED
@@ -0,0 +1,10 @@
|
|
1
|
+
# Configure Rails Environment
|
2
|
+
ENV["RAILS_ENV"] = "test"
|
3
|
+
|
4
|
+
require File.expand_path("../dummy/config/environment.rb", __FILE__)
|
5
|
+
require "rails/test_help"
|
6
|
+
|
7
|
+
Rails.backtrace_cleaner.remove_silencers!
|
8
|
+
|
9
|
+
# Load support files
|
10
|
+
Dir["#{File.dirname(__FILE__)}/support/**/*.rb"].each { |f| require f }
|
metadata
ADDED
@@ -0,0 +1,230 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: semantic-crawler
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
hash: 27
|
5
|
+
prerelease:
|
6
|
+
segments:
|
7
|
+
- 0
|
8
|
+
- 0
|
9
|
+
- 2
|
10
|
+
version: 0.0.2
|
11
|
+
platform: ruby
|
12
|
+
authors:
|
13
|
+
- Alex Oberhauser
|
14
|
+
autorequire:
|
15
|
+
bindir: bin
|
16
|
+
cert_chain: []
|
17
|
+
|
18
|
+
date: 2012-03-25 00:00:00 Z
|
19
|
+
dependencies:
|
20
|
+
- !ruby/object:Gem::Dependency
|
21
|
+
name: yard
|
22
|
+
prerelease: false
|
23
|
+
requirement: &id001 !ruby/object:Gem::Requirement
|
24
|
+
none: false
|
25
|
+
requirements:
|
26
|
+
- - ">="
|
27
|
+
- !ruby/object:Gem::Version
|
28
|
+
hash: 3
|
29
|
+
segments:
|
30
|
+
- 0
|
31
|
+
version: "0"
|
32
|
+
type: :runtime
|
33
|
+
version_requirements: *id001
|
34
|
+
- !ruby/object:Gem::Dependency
|
35
|
+
name: httparty
|
36
|
+
prerelease: false
|
37
|
+
requirement: &id002 !ruby/object:Gem::Requirement
|
38
|
+
none: false
|
39
|
+
requirements:
|
40
|
+
- - ">="
|
41
|
+
- !ruby/object:Gem::Version
|
42
|
+
hash: 3
|
43
|
+
segments:
|
44
|
+
- 0
|
45
|
+
version: "0"
|
46
|
+
type: :runtime
|
47
|
+
version_requirements: *id002
|
48
|
+
- !ruby/object:Gem::Dependency
|
49
|
+
name: json
|
50
|
+
prerelease: false
|
51
|
+
requirement: &id003 !ruby/object:Gem::Requirement
|
52
|
+
none: false
|
53
|
+
requirements:
|
54
|
+
- - ">="
|
55
|
+
- !ruby/object:Gem::Version
|
56
|
+
hash: 3
|
57
|
+
segments:
|
58
|
+
- 0
|
59
|
+
version: "0"
|
60
|
+
type: :runtime
|
61
|
+
version_requirements: *id003
|
62
|
+
- !ruby/object:Gem::Dependency
|
63
|
+
name: nokogiri
|
64
|
+
prerelease: false
|
65
|
+
requirement: &id004 !ruby/object:Gem::Requirement
|
66
|
+
none: false
|
67
|
+
requirements:
|
68
|
+
- - ">="
|
69
|
+
- !ruby/object:Gem::Version
|
70
|
+
hash: 3
|
71
|
+
segments:
|
72
|
+
- 0
|
73
|
+
version: "0"
|
74
|
+
type: :runtime
|
75
|
+
version_requirements: *id004
|
76
|
+
- !ruby/object:Gem::Dependency
|
77
|
+
name: rails
|
78
|
+
prerelease: false
|
79
|
+
requirement: &id005 !ruby/object:Gem::Requirement
|
80
|
+
none: false
|
81
|
+
requirements:
|
82
|
+
- - ~>
|
83
|
+
- !ruby/object:Gem::Version
|
84
|
+
hash: 11
|
85
|
+
segments:
|
86
|
+
- 3
|
87
|
+
- 2
|
88
|
+
- 2
|
89
|
+
version: 3.2.2
|
90
|
+
type: :development
|
91
|
+
version_requirements: *id005
|
92
|
+
- !ruby/object:Gem::Dependency
|
93
|
+
name: sqlite3
|
94
|
+
prerelease: false
|
95
|
+
requirement: &id006 !ruby/object:Gem::Requirement
|
96
|
+
none: false
|
97
|
+
requirements:
|
98
|
+
- - ">="
|
99
|
+
- !ruby/object:Gem::Version
|
100
|
+
hash: 3
|
101
|
+
segments:
|
102
|
+
- 0
|
103
|
+
version: "0"
|
104
|
+
type: :development
|
105
|
+
version_requirements: *id006
|
106
|
+
description: SemanticCrawler is a ruby library that encapsulates data gathering from different sources.
|
107
|
+
email:
|
108
|
+
- alex.oberhauser@sigimera.org
|
109
|
+
executables: []
|
110
|
+
|
111
|
+
extensions: []
|
112
|
+
|
113
|
+
extra_rdoc_files: []
|
114
|
+
|
115
|
+
files:
|
116
|
+
- lib/semantic_crawler.rb
|
117
|
+
- lib/tasks/semantic_crawler_tasks.rake
|
118
|
+
- lib/semantic_crawler/freebase/country.rb
|
119
|
+
- lib/semantic_crawler/freebase/crawler.rb
|
120
|
+
- lib/semantic_crawler/freebase/entity.rb
|
121
|
+
- lib/semantic_crawler/factbook/country.rb
|
122
|
+
- lib/semantic_crawler/version.rb
|
123
|
+
- MIT-LICENSE
|
124
|
+
- Rakefile
|
125
|
+
- README.rdoc
|
126
|
+
- test/test_helper.rb
|
127
|
+
- test/factbook_test.rb
|
128
|
+
- test/semantic_crawler_test.rb
|
129
|
+
- test/dummy/log/test.log
|
130
|
+
- test/dummy/log/development.log
|
131
|
+
- test/dummy/public/422.html
|
132
|
+
- test/dummy/public/favicon.ico
|
133
|
+
- test/dummy/public/500.html
|
134
|
+
- test/dummy/public/404.html
|
135
|
+
- test/dummy/config.ru
|
136
|
+
- test/dummy/config/locales/en.yml
|
137
|
+
- test/dummy/config/environments/production.rb
|
138
|
+
- test/dummy/config/environments/test.rb
|
139
|
+
- test/dummy/config/environments/development.rb
|
140
|
+
- test/dummy/config/boot.rb
|
141
|
+
- test/dummy/config/application.rb
|
142
|
+
- test/dummy/config/database.yml
|
143
|
+
- test/dummy/config/environment.rb
|
144
|
+
- test/dummy/config/routes.rb
|
145
|
+
- test/dummy/config/initializers/secret_token.rb
|
146
|
+
- test/dummy/config/initializers/inflections.rb
|
147
|
+
- test/dummy/config/initializers/backtrace_silencers.rb
|
148
|
+
- test/dummy/config/initializers/wrap_parameters.rb
|
149
|
+
- test/dummy/config/initializers/mime_types.rb
|
150
|
+
- test/dummy/config/initializers/session_store.rb
|
151
|
+
- test/dummy/script/rails
|
152
|
+
- test/dummy/README.rdoc
|
153
|
+
- test/dummy/db/development.sqlite3
|
154
|
+
- test/dummy/db/test.sqlite3
|
155
|
+
- test/dummy/Rakefile
|
156
|
+
- test/dummy/app/helpers/application_helper.rb
|
157
|
+
- test/dummy/app/assets/stylesheets/application.css
|
158
|
+
- test/dummy/app/assets/javascripts/application.js
|
159
|
+
- test/dummy/app/controllers/application_controller.rb
|
160
|
+
- test/dummy/app/views/layouts/application.html.erb
|
161
|
+
homepage: https://github.com/obale/semantic_crawler
|
162
|
+
licenses: []
|
163
|
+
|
164
|
+
post_install_message:
|
165
|
+
rdoc_options: []
|
166
|
+
|
167
|
+
require_paths:
|
168
|
+
- lib
|
169
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
170
|
+
none: false
|
171
|
+
requirements:
|
172
|
+
- - ">="
|
173
|
+
- !ruby/object:Gem::Version
|
174
|
+
hash: 3
|
175
|
+
segments:
|
176
|
+
- 0
|
177
|
+
version: "0"
|
178
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
179
|
+
none: false
|
180
|
+
requirements:
|
181
|
+
- - ">="
|
182
|
+
- !ruby/object:Gem::Version
|
183
|
+
hash: 3
|
184
|
+
segments:
|
185
|
+
- 0
|
186
|
+
version: "0"
|
187
|
+
requirements: []
|
188
|
+
|
189
|
+
rubyforge_project:
|
190
|
+
rubygems_version: 1.8.21
|
191
|
+
signing_key:
|
192
|
+
specification_version: 3
|
193
|
+
summary: SemanticCrawler is a ruby library that encapsulates data gathering from different sources.
|
194
|
+
test_files:
|
195
|
+
- test/test_helper.rb
|
196
|
+
- test/factbook_test.rb
|
197
|
+
- test/semantic_crawler_test.rb
|
198
|
+
- test/dummy/log/test.log
|
199
|
+
- test/dummy/log/development.log
|
200
|
+
- test/dummy/public/422.html
|
201
|
+
- test/dummy/public/favicon.ico
|
202
|
+
- test/dummy/public/500.html
|
203
|
+
- test/dummy/public/404.html
|
204
|
+
- test/dummy/config.ru
|
205
|
+
- test/dummy/config/locales/en.yml
|
206
|
+
- test/dummy/config/environments/production.rb
|
207
|
+
- test/dummy/config/environments/test.rb
|
208
|
+
- test/dummy/config/environments/development.rb
|
209
|
+
- test/dummy/config/boot.rb
|
210
|
+
- test/dummy/config/application.rb
|
211
|
+
- test/dummy/config/database.yml
|
212
|
+
- test/dummy/config/environment.rb
|
213
|
+
- test/dummy/config/routes.rb
|
214
|
+
- test/dummy/config/initializers/secret_token.rb
|
215
|
+
- test/dummy/config/initializers/inflections.rb
|
216
|
+
- test/dummy/config/initializers/backtrace_silencers.rb
|
217
|
+
- test/dummy/config/initializers/wrap_parameters.rb
|
218
|
+
- test/dummy/config/initializers/mime_types.rb
|
219
|
+
- test/dummy/config/initializers/session_store.rb
|
220
|
+
- test/dummy/script/rails
|
221
|
+
- test/dummy/README.rdoc
|
222
|
+
- test/dummy/db/development.sqlite3
|
223
|
+
- test/dummy/db/test.sqlite3
|
224
|
+
- test/dummy/Rakefile
|
225
|
+
- test/dummy/app/helpers/application_helper.rb
|
226
|
+
- test/dummy/app/assets/stylesheets/application.css
|
227
|
+
- test/dummy/app/assets/javascripts/application.js
|
228
|
+
- test/dummy/app/controllers/application_controller.rb
|
229
|
+
- test/dummy/app/views/layouts/application.html.erb
|
230
|
+
has_rdoc:
|