term_extraction 0.1.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (16) hide show
  1. data/.gitignore +4 -0
  2. data/LICENSE +22 -0
  3. data/README +31 -0
  4. data/Rakefile +55 -0
  5. data/VERSION.yml +4 -0
  6. data/lib/term_extraction/yahoo.rb +54 -0
  7. data/lib/term_extraction/zemanta.rb +52 -0
  8. data/lib/term_extraction.rb +14 -0
  9. data/term_extraction.gemspec +59 -0
  10. data/test/fixtures/yahoo.xml +3 -0
  11. data/test/fixtures/yahoo2.xml +3 -0
  12. data/test/fixtures/zemanta.xml +104 -0
  13. data/test/fixtures/zemanta2.xml +104 -0
  14. data/test/term_extraction_test.rb +62 -0
  15. data/test/test_helper.rb +22 -0
  16. metadata +80 -0
data/.gitignore ADDED
@@ -0,0 +1,4 @@
1
+ *.sw?
2
+ .DS_Store
3
+ coverage
4
+ pkg
data/LICENSE ADDED
@@ -0,0 +1,22 @@
1
+ (The MIT License)
2
+
3
+ Copyright (c) 2009 Stateless Systems (http://statelesssystems.com)
4
+
5
+ Permission is hereby granted, free of charge, to any person obtaining
6
+ a copy of this software and associated documentation files (the
7
+ 'Software'), to deal in the Software without restriction, including
8
+ without limitation the rights to use, copy, modify, merge, publish,
9
+ distribute, sublicense, and/or sell copies of the Software, and to
10
+ permit persons to whom the Software is furnished to do so, subject to
11
+ the following conditions:
12
+
13
+ The above copyright notice and this permission notice shall be
14
+ included in all copies or substantial portions of the Software.
15
+
16
+ THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND,
17
+ EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
18
+ MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
19
+ IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
20
+ CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
21
+ TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
22
+ SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
data/README ADDED
@@ -0,0 +1,31 @@
1
+ = term_extraction
2
+
3
+ == DESCRIPTION:
4
+
5
+ Provides access to term extraction APIs such as Yahoo! Term Extraction API and
6
+ Zemanta.
7
+
8
+ == SYNOPSIS:
9
+
10
+ # Query Yahoo! for terms
11
+ yahoo = TermExtraction::Yahoo.new(:api_key => 'myApiKey', :context => 'xbox 360 gears of war')
12
+ yahoo.terms # => ["gears of war", "xbox 360", "gears", "xbox"]
13
+
14
+ # Query Zemanta for terms
15
+ zemanta = TermExtraction::Zemanta.new(:api_key => 'myApiKey', :context => 'apple imac')
16
+ zemanta.terms # => ["Apple", "IMac", "Rumor", "Hardware", "Nvidia", "Macintosh", "Desktops", "AllInOne"]
17
+
18
+ == INSTALL:
19
+
20
+ * Via git:
21
+
22
+ git clone git://github.com/alexrabarts/term_extraction.git
23
+
24
+ * Via gem:
25
+
26
+ gem install alexrabarts-term_extraction -s http://gems.github.com
27
+
28
+ COPYRIGHT
29
+ =========
30
+
31
+ Copyright (c) 2009 Stateless Systems (http://statelesssystems.com). See LICENSE for details.
data/Rakefile ADDED
@@ -0,0 +1,55 @@
1
+ require 'rake'
2
+
3
+ begin
4
+ require 'jeweler'
5
+ Jeweler::Tasks.new do |s|
6
+ s.name = "term_extraction"
7
+ s.summary = %Q{
8
+ Provides access to term extraction APIs such as Yahoo! Term Extraction API and
9
+ Zemanta.
10
+ }
11
+ s.email = "alexrabarts@gmail.com"
12
+ s.homepage = "http://github.com/alexrabarts/term_extraction"
13
+ s.description = "Term extraction library"
14
+ s.authors = ["alex"]
15
+ s.add_dependency 'nokogiri', ['>=1.0.7']
16
+ end
17
+ rescue LoadError
18
+ puts "Jeweler not available. Install it with: sudo gem install technicalpickles-jeweler -s http://gems.github.com"
19
+ end
20
+
21
+ require 'rake/rdoctask'
22
+ Rake::RDocTask.new do |rdoc|
23
+ rdoc.rdoc_dir = 'rdoc'
24
+ rdoc.title = 'term_extraction'
25
+ rdoc.options << '--line-numbers' << '--inline-source'
26
+ rdoc.rdoc_files.include('README*')
27
+ rdoc.rdoc_files.include('lib/**/*.rb')
28
+ end
29
+
30
+ require 'rake/testtask'
31
+ Rake::TestTask.new(:test) do |t|
32
+ t.libs << 'lib' << 'test'
33
+ t.pattern = 'test/**/*_test.rb'
34
+ t.verbose = false
35
+ end
36
+
37
+ begin
38
+ require 'rcov/rcovtask'
39
+ Rcov::RcovTask.new do |t|
40
+ t.libs << 'test'
41
+ t.test_files = FileList['test/**/*_test.rb']
42
+ t.verbose = true
43
+ end
44
+ rescue LoadError
45
+ puts "RCov is not available. In order to run rcov, you must: sudo gem install spicycode-rcov"
46
+ end
47
+
48
+ begin
49
+ require 'cucumber/rake/task'
50
+ Cucumber::Rake::Task.new(:features)
51
+ rescue LoadError
52
+ puts "Cucumber is not available. In order to run features, you must: sudo gem install cucumber"
53
+ end
54
+
55
+ task :default => :test
data/VERSION.yml ADDED
@@ -0,0 +1,4 @@
1
+ ---
2
+ :minor: 1
3
+ :patch: 4
4
+ :major: 0
data/lib/term_extraction/yahoo.rb ADDED
@@ -0,0 +1,54 @@
1
+ require 'nokogiri'
2
+ require 'addressable/uri'
3
+ require 'open-uri'
4
+
5
+ class TermExtraction
6
+ class Yahoo < TermExtraction
7
+ def terms
8
+ terms = []
9
+ data = Nokogiri::XML.parse(remote_xml)
10
+
11
+ unless data.nil?
12
+ data.search('//s:Result', ns).each do |n|
13
+ terms << n.text
14
+ end
15
+ end
16
+
17
+ terms
18
+ end
19
+
20
+ def uri
21
+ api_uri = Addressable::URI.parse(gateway)
22
+ api_uri.query_values = {
23
+ 'appid' => @api_key,
24
+ 'output' => 'xml',
25
+ 'context' => @context
26
+ }
27
+ api_uri
28
+ end
29
+
30
+ class << self
31
+ def canonical_name
32
+ 'yahoo'
33
+ end
34
+ end
35
+
36
+ private
37
+ def ns
38
+ {'s' => 'urn:yahoo:cate'}
39
+ end
40
+
41
+ def gateway
42
+ 'http://search.yahooapis.com/ContentAnalysisService/V1/termExtraction'
43
+ end
44
+
45
+ def remote_xml
46
+ begin
47
+ open(uri).read
48
+ rescue => e
49
+ $stderr.puts "Couldn't fetch from API: #{e.message}" if $VERBOSE
50
+ nil
51
+ end
52
+ end
53
+ end
54
+ end
data/lib/term_extraction/zemanta.rb ADDED
@@ -0,0 +1,52 @@
1
+ require 'net/http'
2
+ require 'nokogiri'
3
+ require 'addressable/uri'
4
+
5
+ class TermExtraction
6
+ class Zemanta < TermExtraction
7
+ def terms
8
+ terms = []
9
+ data = Nokogiri::XML.parse(remote_xml)
10
+
11
+ data.search('//name').each do |n|
12
+ terms << n.text
13
+ end
14
+
15
+ terms
16
+ end
17
+
18
+ def uri
19
+ Addressable::URI.parse(gateway)
20
+ end
21
+
22
+ def post_params
23
+ {
24
+ 'method' =>'zemanta.suggest',
25
+ 'api_key' => @api_key,
26
+ 'return_images' => 0,
27
+ 'text' => @context,
28
+ 'format' => 'xml'
29
+ }
30
+ end
31
+
32
+ class << self
33
+ def canonical_name
34
+ 'zemanta'
35
+ end
36
+ end
37
+
38
+ private
39
+ def gateway
40
+ 'http://api.zemanta.com/services/rest/0.0/'
41
+ end
42
+
43
+ def remote_xml
44
+ begin
45
+ Net::HTTP.post_form(uri, post_params).body
46
+ rescue => e
47
+ $stderr.puts "Couldn't fetch from API: #{e.message}" if $VERBOSE
48
+ nil
49
+ end
50
+ end
51
+ end
52
+ end
data/lib/term_extraction.rb ADDED
@@ -0,0 +1,14 @@
1
+ class TermExtraction
2
+ attr_accessor :context, :api_key
3
+
4
+ def initialize(options={})
5
+ @context = options[:context]
6
+ @api_key = options[:api_key]
7
+ end
8
+
9
+ def canonical_name
10
+ self.class.canonical_name
11
+ end
12
+ end
13
+
14
+ %w{yahoo zemanta}.each{|t| require "term_extraction/#{t}"}
data/term_extraction.gemspec ADDED
@@ -0,0 +1,59 @@
1
+ # Generated by jeweler
2
+ # DO NOT EDIT THIS FILE DIRECTLY
3
+ # Instead, edit Jeweler::Tasks in Rakefile, and run the gemspec command
4
+ # -*- encoding: utf-8 -*-
5
+
6
+ Gem::Specification.new do |s|
7
+ s.name = %q{term_extraction}
8
+ s.version = "0.1.4"
9
+
10
+ s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
11
+ s.authors = ["alex"]
12
+ s.date = %q{2009-12-20}
13
+ s.description = %q{Term extraction library}
14
+ s.email = %q{alexrabarts@gmail.com}
15
+ s.extra_rdoc_files = [
16
+ "LICENSE",
17
+ "README"
18
+ ]
19
+ s.files = [
20
+ ".gitignore",
21
+ "LICENSE",
22
+ "README",
23
+ "Rakefile",
24
+ "VERSION.yml",
25
+ "lib/term_extraction.rb",
26
+ "lib/term_extraction/yahoo.rb",
27
+ "lib/term_extraction/zemanta.rb",
28
+ "term_extraction.gemspec",
29
+ "test/fixtures/yahoo.xml",
30
+ "test/fixtures/yahoo2.xml",
31
+ "test/fixtures/zemanta.xml",
32
+ "test/fixtures/zemanta2.xml",
33
+ "test/term_extraction_test.rb",
34
+ "test/test_helper.rb"
35
+ ]
36
+ s.homepage = %q{http://github.com/alexrabarts/term_extraction}
37
+ s.rdoc_options = ["--charset=UTF-8"]
38
+ s.require_paths = ["lib"]
39
+ s.rubygems_version = %q{1.3.5}
40
+ s.summary = %q{Provides access to term extraction APIs such as Yahoo! Term Extraction API and Zemanta.}
41
+ s.test_files = [
42
+ "test/term_extraction_test.rb",
43
+ "test/test_helper.rb"
44
+ ]
45
+
46
+ if s.respond_to? :specification_version then
47
+ current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
48
+ s.specification_version = 3
49
+
50
+ if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then
51
+ s.add_runtime_dependency(%q<nokogiri>, [">= 1.0.7"])
52
+ else
53
+ s.add_dependency(%q<nokogiri>, [">= 1.0.7"])
54
+ end
55
+ else
56
+ s.add_dependency(%q<nokogiri>, [">= 1.0.7"])
57
+ end
58
+ end
59
+
data/test/fixtures/yahoo.xml ADDED
@@ -0,0 +1,3 @@
1
+ <?xml version="1.0"?>
2
+ <ResultSet xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="urn:yahoo:cate" xsi:schemaLocation="urn:yahoo:cate http://api.search.yahoo.com/ContentAnalysisService/V1/TermExtractionResponse.xsd"><Result>gears of war</Result><Result>gears</Result></ResultSet>
3
+ <!-- ws04.search.scd.yahoo.com uncompressed/chunked Thu Feb 26 21:04:16 PST 2009 -->
data/test/fixtures/yahoo2.xml ADDED
@@ -0,0 +1,3 @@
1
+ <?xml version="1.0"?>
2
+ <ResultSet xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="urn:yahoo:cate" xsi:schemaLocation="urn:yahoo:cate http://api.search.yahoo.com/ContentAnalysisService/V1/TermExtractionResponse.xsd"><Result>fears of war</Result><Result>fears</Result></ResultSet>
3
+ <!-- ws04.search.scd.yahoo.com uncompressed/chunked Thu Feb 26 21:04:16 PST 2009 -->
data/test/fixtures/zemanta.xml ADDED
@@ -0,0 +1,104 @@
1
+ <?xml version="1.0"?>
2
+ <rsp>
3
+ <status>ok</status>
4
+ <articles>
5
+ <article>
6
+ <url>http://www.crunchgear.com/2009/02/24/nvidia-based-imacs-coming-soon/</url>
7
+ <confidence>0.033153</confidence>
8
+ <published_datetime>2009-02-24T18:00:45Z</published_datetime>
9
+ <zemified>0</zemified>
10
+ <title>NVIDIA-based iMacs coming soon?</title>
11
+ </article><article>
12
+ <url>http://www.tuaw.com/2009/02/24/rumor-new-imacs-around-the-bend/</url>
13
+ <confidence>0.028595</confidence>
14
+ <published_datetime>2009-02-24T22:00:00Z</published_datetime>
15
+ <zemified>0</zemified>
16
+ <title>Rumor: New iMacs around the bend</title>
17
+ </article><article>
18
+ <url>http://www.engadget.com/2009/02/24/a-few-new-rumors-point-to-two-new-nvidia-packing-imacs/</url>
19
+ <confidence>0.027595</confidence>
20
+ <published_datetime>2009-02-24T13:03:00Z</published_datetime>
21
+ <zemified>0</zemified>
22
+ <title>A few new rumors point to two new NVIDIA-packing iMacs</title>
23
+ </article><article>
24
+ <url>http://www.techmeme.com/090126/p31</url>
25
+ <confidence>0.019056</confidence>
26
+ <published_datetime>2009-01-26T17:00:22Z</published_datetime>
27
+ <zemified>0</zemified>
28
+ <title>Chip complex delaying Apple's new iMac line, says analyst (Zach Spear/AppleInsider)</title>
29
+ </article><article>
30
+ <url>http://i.gizmodo.com/5143476/apple-warns-resellers-of-reduced-imac-availability-new-models-on-the-way</url>
31
+ <confidence>0.018611</confidence>
32
+ <published_datetime>2009-01-31T16:30:00Z</published_datetime>
33
+ <zemified>0</zemified>
34
+ <title>Apple Warns Resellers of Reduced iMac Availability: New Models On the Way? [Apple]</title>
35
+ </article><article>
36
+ <url>http://cultofmac.com/analyst-new-imacs-delayed-for-chips-snow-leopard/7549</url>
37
+ <confidence>0.01685</confidence>
38
+ <published_datetime>2009-01-26T16:43:44Z</published_datetime>
39
+ <zemified>0</zemified>
40
+ <title>Analyst: New iMacs Delayed For Chips, Snow Leopard</title>
41
+ </article><article>
42
+ <url>http://www.ubergizmo.com/15/archives/2008/12/new_imac_allinone_pc_confirmed.html</url>
43
+ <confidence>0.016311</confidence>
44
+ <published_datetime>2008-12-24T01:06:56Z</published_datetime>
45
+ <zemified>0</zemified>
46
+ <title>New iMac All-In-One PC Confirmed</title>
47
+ </article><article>
48
+ <url>http://www.labnol.org/gadgets/exchange-windows-pc-for-apple-mac/5775/</url>
49
+ <confidence>0.016238</confidence>
50
+ <published_datetime>2008-12-03T08:17:35Z</published_datetime>
51
+ <zemified>0</zemified>
52
+ <title>Exchange Your Windows PC or TV for an Apple iMac</title>
53
+ </article><article>
54
+ <url>http://cultofmac.com/class-action-lawsuit-over-imac-display-problems/6562</url>
55
+ <confidence>0.015608</confidence>
56
+ <published_datetime>2009-01-02T16:17:06Z</published_datetime>
57
+ <zemified>0</zemified>
58
+ <title>Class-Action Lawsuit Over iMac Display Problems</title>
59
+ </article><article>
60
+ <url>http://www.crunchgear.com/2008/12/29/new-imacs-to-include-hotter-components-new-cooling/</url>
61
+ <confidence>0.015093</confidence>
62
+ <published_datetime>2008-12-29T23:50:04Z</published_datetime>
63
+ <zemified>0</zemified>
64
+ <title>New iMacs to include hotter components, new cooling?</title>
65
+ </article>
66
+ </articles><markup>
67
+ <text>apple imac</text>
68
+ </markup><signature>&lt;div class="zemanta-pixie"&gt;&lt;a class="zemanta-pixie-a" href="http://reblog.zemanta.com/zemified/3695b55a-d68e-4328-bf86-c3f01be8b2da/" title="Zemified by Zemanta"&gt;&lt;img class="zemanta-pixie-img" src="http://img.zemanta.com/reblog_e.png?x-id=3695b55a-d68e-4328-bf86-c3f01be8b2da" alt="Reblog this post [with Zemanta]" /&gt;&lt;/a&gt;&lt;/div&gt;</signature>
69
+ <keywords>
70
+ <keyword>
71
+ <confidence>0.404748</confidence>
72
+ <scheme>general</scheme>
73
+ <name>Apple</name>
74
+ </keyword><keyword>
75
+ <confidence>0.277753</confidence>
76
+ <scheme>general</scheme>
77
+ <name>IMac</name>
78
+ </keyword><keyword>
79
+ <confidence>0.094392</confidence>
80
+ <scheme>general</scheme>
81
+ <name>Rumor</name>
82
+ </keyword><keyword>
83
+ <confidence>0.076399</confidence>
84
+ <scheme>general</scheme>
85
+ <name>Hardware</name>
86
+ </keyword><keyword>
87
+ <confidence>0.072186</confidence>
88
+ <scheme>general</scheme>
89
+ <name>Nvidia</name>
90
+ </keyword><keyword>
91
+ <confidence>0.069183</confidence>
92
+ <scheme>general</scheme>
93
+ <name>Macintosh</name>
94
+ </keyword><keyword>
95
+ <confidence>0.050487</confidence>
96
+ <scheme>general</scheme>
97
+ <name>Desktops</name>
98
+ </keyword><keyword>
99
+ <confidence>0.039205</confidence>
100
+ <scheme>general</scheme>
101
+ <name>AllInOne</name>
102
+ </keyword>
103
+ </keywords><rid>3695b55a-d68e-4328-bf86-c3f01be8b2da</rid>
104
+ </rsp>
data/test/fixtures/zemanta2.xml ADDED
@@ -0,0 +1,104 @@
1
+ <?xml version="1.0"?>
2
+ <rsp>
3
+ <status>ok</status>
4
+ <articles>
5
+ <article>
6
+ <url>http://www.crunchgear.com/2009/02/24/nvidia-based-imacs-coming-soon/</url>
7
+ <confidence>0.033153</confidence>
8
+ <published_datetime>2009-02-24T18:00:45Z</published_datetime>
9
+ <zemified>0</zemified>
10
+ <title>NVIDIA-based iMacs coming soon?</title>
11
+ </article><article>
12
+ <url>http://www.tuaw.com/2009/02/24/rumor-new-imacs-around-the-bend/</url>
13
+ <confidence>0.028595</confidence>
14
+ <published_datetime>2009-02-24T22:00:00Z</published_datetime>
15
+ <zemified>0</zemified>
16
+ <title>Rumor: New iMacs around the bend</title>
17
+ </article><article>
18
+ <url>http://www.engadget.com/2009/02/24/a-few-new-rumors-point-to-two-new-nvidia-packing-imacs/</url>
19
+ <confidence>0.027595</confidence>
20
+ <published_datetime>2009-02-24T13:03:00Z</published_datetime>
21
+ <zemified>0</zemified>
22
+ <title>A few new rumors point to two new NVIDIA-packing iMacs</title>
23
+ </article><article>
24
+ <url>http://www.techmeme.com/090126/p31</url>
25
+ <confidence>0.019056</confidence>
26
+ <published_datetime>2009-01-26T17:00:22Z</published_datetime>
27
+ <zemified>0</zemified>
28
+ <title>Chip complex delaying Apple's new iMac line, says analyst (Zach Spear/AppleInsider)</title>
29
+ </article><article>
30
+ <url>http://i.gizmodo.com/5143476/apple-warns-resellers-of-reduced-imac-availability-new-models-on-the-way</url>
31
+ <confidence>0.018611</confidence>
32
+ <published_datetime>2009-01-31T16:30:00Z</published_datetime>
33
+ <zemified>0</zemified>
34
+ <title>Apple Warns Resellers of Reduced iMac Availability: New Models On the Way? [Apple]</title>
35
+ </article><article>
36
+ <url>http://cultofmac.com/analyst-new-imacs-delayed-for-chips-snow-leopard/7549</url>
37
+ <confidence>0.01685</confidence>
38
+ <published_datetime>2009-01-26T16:43:44Z</published_datetime>
39
+ <zemified>0</zemified>
40
+ <title>Analyst: New iMacs Delayed For Chips, Snow Leopard</title>
41
+ </article><article>
42
+ <url>http://www.ubergizmo.com/15/archives/2008/12/new_imac_allinone_pc_confirmed.html</url>
43
+ <confidence>0.016311</confidence>
44
+ <published_datetime>2008-12-24T01:06:56Z</published_datetime>
45
+ <zemified>0</zemified>
46
+ <title>New iMac All-In-One PC Confirmed</title>
47
+ </article><article>
48
+ <url>http://www.labnol.org/gadgets/exchange-windows-pc-for-apple-mac/5775/</url>
49
+ <confidence>0.016238</confidence>
50
+ <published_datetime>2008-12-03T08:17:35Z</published_datetime>
51
+ <zemified>0</zemified>
52
+ <title>Exchange Your Windows PC or TV for an Apple iMac</title>
53
+ </article><article>
54
+ <url>http://cultofmac.com/class-action-lawsuit-over-imac-display-problems/6562</url>
55
+ <confidence>0.015608</confidence>
56
+ <published_datetime>2009-01-02T16:17:06Z</published_datetime>
57
+ <zemified>0</zemified>
58
+ <title>Class-Action Lawsuit Over iMac Display Problems</title>
59
+ </article><article>
60
+ <url>http://www.crunchgear.com/2008/12/29/new-imacs-to-include-hotter-components-new-cooling/</url>
61
+ <confidence>0.015093</confidence>
62
+ <published_datetime>2008-12-29T23:50:04Z</published_datetime>
63
+ <zemified>0</zemified>
64
+ <title>New iMacs to include hotter components, new cooling?</title>
65
+ </article>
66
+ </articles><markup>
67
+ <text>apple imac</text>
68
+ </markup><signature>&lt;div class="zemanta-pixie"&gt;&lt;a class="zemanta-pixie-a" href="http://reblog.zemanta.com/zemified/3695b55a-d68e-4328-bf86-c3f01be8b2da/" title="Zemified by Zemanta"&gt;&lt;img class="zemanta-pixie-img" src="http://img.zemanta.com/reblog_e.png?x-id=3695b55a-d68e-4328-bf86-c3f01be8b2da" alt="Reblog this post [with Zemanta]" /&gt;&lt;/a&gt;&lt;/div&gt;</signature>
69
+ <keywords>
70
+ <keyword>
71
+ <confidence>0.404748</confidence>
72
+ <scheme>general</scheme>
73
+ <name>Zapple</name>
74
+ </keyword><keyword>
75
+ <confidence>0.277753</confidence>
76
+ <scheme>general</scheme>
77
+ <name>AMac</name>
78
+ </keyword><keyword>
79
+ <confidence>0.094392</confidence>
80
+ <scheme>general</scheme>
81
+ <name>Tumor</name>
82
+ </keyword><keyword>
83
+ <confidence>0.076399</confidence>
84
+ <scheme>general</scheme>
85
+ <name>Shardware</name>
86
+ </keyword><keyword>
87
+ <confidence>0.072186</confidence>
88
+ <scheme>general</scheme>
89
+ <name>Mvidia</name>
90
+ </keyword><keyword>
91
+ <confidence>0.069183</confidence>
92
+ <scheme>general</scheme>
93
+ <name>Hackintosh</name>
94
+ </keyword><keyword>
95
+ <confidence>0.050487</confidence>
96
+ <scheme>general</scheme>
97
+ <name>Zesktops</name>
98
+ </keyword><keyword>
99
+ <confidence>0.039205</confidence>
100
+ <scheme>general</scheme>
101
+ <name>AllInNone</name>
102
+ </keyword>
103
+ </keywords><rid>3695b55a-d68e-4328-bf86-c3f01be8b2da</rid>
104
+ </rsp>
data/test/term_extraction_test.rb ADDED
@@ -0,0 +1,62 @@
1
+ require File.dirname(__FILE__) + '/test_helper'
2
+
3
+ class TermExtractionTest < Test::Unit::TestCase
4
+ should 'return correct terms from Yahoo!' do
5
+ yahoo = TermExtraction::Yahoo.new
6
+ fake_uri(:get, yahoo.uri, 'yahoo.xml')
7
+ assert_equal yahoo.terms, correct_yahoo_terms
8
+ end
9
+
10
+ should 'return correct terms from Zemanta' do
11
+ zemanta = TermExtraction::Zemanta.new
12
+ fake_uri(:post, zemanta.uri, 'zemanta.xml')
13
+ assert_equal zemanta.terms, correct_zemanta_terms
14
+ end
15
+
16
+ should 'be able to set the context after initialization' do
17
+ yahoo = TermExtraction::Yahoo.new
18
+ context = 'foo'
19
+ yahoo.context = context
20
+ assert_equal yahoo.context, context
21
+ end
22
+
23
+ should 'be able to set the api key after initialization' do
24
+ zemanta = TermExtraction::Zemanta.new
25
+ context = 'bar'
26
+ zemanta.context = context
27
+ assert_equal zemanta.context, context
28
+ end
29
+
30
+ should 'return different response on subsequent calls when different data is returned from Yahoo!' do
31
+ yahoo = TermExtraction::Yahoo.new
32
+ fake_uri(:get, yahoo.uri, 'yahoo.xml')
33
+ original_terms = yahoo.terms
34
+ fake_uri(:get, yahoo.uri, 'yahoo2.xml')
35
+ assert_not_equal original_terms, yahoo.terms
36
+ end
37
+
38
+ should 'return different response on subsequent calls when different data is returned from Zemanta' do
39
+ zemanta = TermExtraction::Zemanta.new
40
+ fake_uri(:post, zemanta.uri, 'zemanta.xml')
41
+ original_terms = zemanta.terms
42
+ fake_uri(:post, zemanta.uri, 'zemanta2.xml')
43
+ assert_not_equal original_terms, zemanta.terms
44
+ end
45
+
46
+ context 'Yahoo!' do
47
+ should 'be able to handle a context with "%" in it' do
48
+ yahoo = TermExtraction::Yahoo.new(:context => '%')
49
+ fake_uri(:get, yahoo.uri, 'yahoo.xml')
50
+ assert_nothing_thrown{ yahoo.terms }
51
+ end
52
+ end
53
+
54
+ private
55
+ def correct_yahoo_terms
56
+ ['gears of war', 'gears']
57
+ end
58
+
59
+ def correct_zemanta_terms
60
+ ['Apple', 'IMac', 'Rumor', 'Hardware', 'Nvidia', 'Macintosh', 'Desktops', 'AllInOne']
61
+ end
62
+ end
data/test/test_helper.rb ADDED
@@ -0,0 +1,22 @@
1
+ require 'rubygems'
2
+ require 'test/unit'
3
+ require 'shoulda'
4
+ require 'mocha'
5
+ require 'fake_web'
6
+
7
+ $LOAD_PATH.unshift(File.dirname(__FILE__))
8
+ require 'term_extraction'
9
+
10
+ class Test::Unit::TestCase
11
+ def fake_uri(method, uri, fixture)
12
+ FakeWeb.clean_registry
13
+ FakeWeb.allow_net_connect = false
14
+ response = File.open(File.join(File.dirname(__FILE__), 'fixtures', fixture)).read
15
+ FakeWeb.register_uri(method, uri.to_s, :string => response)
16
+ end
17
+
18
+ def teardown
19
+ FakeWeb.allow_net_connect = true
20
+ FakeWeb.clean_registry
21
+ end
22
+ end
metadata ADDED
@@ -0,0 +1,80 @@
1
+ --- !ruby/object:Gem::Specification
2
+ name: term_extraction
3
+ version: !ruby/object:Gem::Version
4
+ version: 0.1.4
5
+ platform: ruby
6
+ authors:
7
+ - alex
8
+ autorequire:
9
+ bindir: bin
10
+ cert_chain: []
11
+
12
+ date: 2009-12-20 00:00:00 -07:00
13
+ default_executable:
14
+ dependencies:
15
+ - !ruby/object:Gem::Dependency
16
+ name: nokogiri
17
+ type: :runtime
18
+ version_requirement:
19
+ version_requirements: !ruby/object:Gem::Requirement
20
+ requirements:
21
+ - - ">="
22
+ - !ruby/object:Gem::Version
23
+ version: 1.0.7
24
+ version:
25
+ description: Term extraction library
26
+ email: alexrabarts@gmail.com
27
+ executables: []
28
+
29
+ extensions: []
30
+
31
+ extra_rdoc_files:
32
+ - LICENSE
33
+ - README
34
+ files:
35
+ - .gitignore
36
+ - LICENSE
37
+ - README
38
+ - Rakefile
39
+ - VERSION.yml
40
+ - lib/term_extraction.rb
41
+ - lib/term_extraction/yahoo.rb
42
+ - lib/term_extraction/zemanta.rb
43
+ - term_extraction.gemspec
44
+ - test/fixtures/yahoo.xml
45
+ - test/fixtures/yahoo2.xml
46
+ - test/fixtures/zemanta.xml
47
+ - test/fixtures/zemanta2.xml
48
+ - test/term_extraction_test.rb
49
+ - test/test_helper.rb
50
+ has_rdoc: true
51
+ homepage: http://github.com/alexrabarts/term_extraction
52
+ licenses: []
53
+
54
+ post_install_message:
55
+ rdoc_options:
56
+ - --charset=UTF-8
57
+ require_paths:
58
+ - lib
59
+ required_ruby_version: !ruby/object:Gem::Requirement
60
+ requirements:
61
+ - - ">="
62
+ - !ruby/object:Gem::Version
63
+ version: "0"
64
+ version:
65
+ required_rubygems_version: !ruby/object:Gem::Requirement
66
+ requirements:
67
+ - - ">="
68
+ - !ruby/object:Gem::Version
69
+ version: "0"
70
+ version:
71
+ requirements: []
72
+
73
+ rubyforge_project:
74
+ rubygems_version: 1.3.5
75
+ signing_key:
76
+ specification_version: 3
77
+ summary: Provides access to term extraction APIs such as Yahoo! Term Extraction API and Zemanta.
78
+ test_files:
79
+ - test/term_extraction_test.rb
80
+ - test/test_helper.rb