statement 2.0.3 → 2.0.4
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.
- checksums.yaml +4 -4
- data/lib/statement/scraper.rb +71 -11
- data/lib/statement/version.rb +1 -1
- data/spec/bill_nelson_press-coldfusion.html +593 -0
- data/spec/bill_nelson_press.html +453 -572
- data/spec/rand_paul_press.html +5492 -0
- data/spec/statement_spec.rb +31 -3
- metadata +6 -2
data/spec/statement_spec.rb
CHANGED
@@ -40,13 +40,30 @@ describe Statement do
|
|
40
40
|
@results.last[:url].must_equal "http://www.gop.gov/republicans/other/relative_url_test.html"
|
41
41
|
end
|
42
42
|
|
43
|
-
it "scrapes a senate cold fusion page" do
|
43
|
+
it "scrapes a senate cold fusion page (old bill nelson)" do
|
44
|
+
skip "This test no longer relevant as Bill Nelson is off of cold fusion"
|
44
45
|
@url = "http://www.billnelson.senate.gov/news/media.cfm?year=2013"
|
45
|
-
WebMock.stub_request(:
|
46
|
+
WebMock.stub_request(:get, "http://www.billnelson.senate.gov/newsroom/press-releases?page=2013").
|
47
|
+
with(:headers => {'Accept'=>'*/*', 'Accept-Encoding'=>'gzip;q=1.0,deflate;q=0.6,identity;q=0.3', 'User-Agent'=>'Ruby'}).
|
48
|
+
to_return(:status => 200, :body => File.new(File.join(File.dirname(__FILE__), 'bill_nelson_press-coldfusion.html')), :headers => {})
|
49
|
+
|
50
|
+
|
46
51
|
@results = Scraper.billnelson(year=2013)
|
47
|
-
@results.last[:url].must_equal "http://www.billnelson.senate.gov/
|
52
|
+
@results.last[:url].must_equal "http://www.billnelson.senate.gov/newsroom/press-releases/lawmakers-call-on-feds-to-investigate-use-of-supercookies"
|
53
|
+
end
|
54
|
+
|
55
|
+
|
56
|
+
it "scrapes a bill nelson page" do
|
57
|
+
@url = "http://www.billnelson.senate.gov/newsroom/press-releases?page=1"
|
58
|
+
WebMock.stub_request(:get, "http://www.billnelson.senate.gov/newsroom/press-releases?page=1").
|
59
|
+
to_return(:status => 200,
|
60
|
+
:body => File.new(File.join(File.dirname(__FILE__), 'bill_nelson_press.html')), :headers => {})
|
61
|
+
|
62
|
+
@results = Scraper.billnelson(page = 1)
|
63
|
+
@results.last[:url].must_equal "http://www.billnelson.senate.gov/newsroom/press-releases/lawmakers-call-on-feds-to-investigate-use-of-supercookies"
|
48
64
|
end
|
49
65
|
|
66
|
+
|
50
67
|
it "scrapes vitter pages for 2013" do
|
51
68
|
@vitter = "http://www.vitter.senate.gov/newsroom/press?year=2013"
|
52
69
|
WebMock.stub_request(:any, @vitter).to_return(:body => File.new(File.join(File.dirname(__FILE__), 'vitter_press.html')), :status => 200)
|
@@ -113,4 +130,15 @@ describe Statement do
|
|
113
130
|
@results.length.must_equal 10
|
114
131
|
@results.first.must_equal expected_result
|
115
132
|
end
|
133
|
+
|
134
|
+
it "scrapes Rand Paul's first page of press releases" do
|
135
|
+
@rand_paul = "http://www.paul.senate.gov/news/press?PageNum_rs=1"
|
136
|
+
WebMock.stub_request(:any, @rand_paul).to_return(:body => File.new(File.join(File.dirname(__FILE__), 'rand_paul_press.html')), :status => 200)
|
137
|
+
@results = Scraper.rand_paul(page = 1)
|
138
|
+
@results.length.must_equal 20
|
139
|
+
@results[0][:url].must_equal 'http://www.paul.senate.gov/news/press/sens-rand-paul-and-mark-warner-introduce-the-bonuses-for-cost-cutters-act-of-2015'
|
140
|
+
@results[0][:title].must_equal 'Sens. Rand Paul & Mark Warner Introduce the Bonuses for Cost-Cutters Act of 2015'
|
141
|
+
end
|
142
|
+
|
143
|
+
|
116
144
|
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: statement
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 2.0.
|
4
|
+
version: 2.0.4
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Derek Willis
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2015-
|
11
|
+
date: 2015-05-29 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: bundler
|
@@ -170,6 +170,7 @@ files:
|
|
170
170
|
- lib/statement/utils.rb
|
171
171
|
- lib/statement/version.rb
|
172
172
|
- scraper_guide.md
|
173
|
+
- spec/bill_nelson_press-coldfusion.html
|
173
174
|
- spec/bill_nelson_press.html
|
174
175
|
- spec/butterfield_press.html
|
175
176
|
- spec/cowan_press.html
|
@@ -178,6 +179,7 @@ files:
|
|
178
179
|
- spec/ed_perlmutter_press.html
|
179
180
|
- spec/house_gop_releases.html
|
180
181
|
- spec/keating_press.html
|
182
|
+
- spec/rand_paul_press.html
|
181
183
|
- spec/richard_burr.xml
|
182
184
|
- spec/ruiz_rss.xml
|
183
185
|
- spec/statement_spec.rb
|
@@ -208,6 +210,7 @@ signing_key:
|
|
208
210
|
specification_version: 4
|
209
211
|
summary: Given a url, Statement returns links to press releases and official statements.
|
210
212
|
test_files:
|
213
|
+
- spec/bill_nelson_press-coldfusion.html
|
211
214
|
- spec/bill_nelson_press.html
|
212
215
|
- spec/butterfield_press.html
|
213
216
|
- spec/cowan_press.html
|
@@ -216,6 +219,7 @@ test_files:
|
|
216
219
|
- spec/ed_perlmutter_press.html
|
217
220
|
- spec/house_gop_releases.html
|
218
221
|
- spec/keating_press.html
|
222
|
+
- spec/rand_paul_press.html
|
219
223
|
- spec/richard_burr.xml
|
220
224
|
- spec/ruiz_rss.xml
|
221
225
|
- spec/statement_spec.rb
|