statement 0.7.5 → 0.8
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.
- data/lib/statement/version.rb +1 -1
- data/lib/statement.rb +172 -0
- metadata +13 -44
data/lib/statement/version.rb
CHANGED
data/lib/statement.rb
CHANGED
|
@@ -77,6 +77,178 @@ module Statement
|
|
|
77
77
|
].flatten
|
|
78
78
|
end
|
|
79
79
|
|
|
80
|
+
def self.committee_scrapers
|
|
81
|
+
year = Date.today.year
|
|
82
|
+
[senate_approps_majority, senate_approps_minority, senate_banking(year), senate_hsag_majority(year), senate_hsag_minority(year),
|
|
83
|
+
senate_indian, senate_aging, senate_smallbiz_minority, senate_intel(113, 2013, 2014), house_energy_minority, house_homeland_security_minority,
|
|
84
|
+
house_judiciary_majority, house_rules_majority, house_ways_means_majority].flatten
|
|
85
|
+
end
|
|
86
|
+
|
|
87
|
+
## special cases for committees without RSS feeds
|
|
88
|
+
|
|
89
|
+
def self.senate_approps_majority
|
|
90
|
+
results = []
|
|
91
|
+
url = "http://www.appropriations.senate.gov/news.cfm"
|
|
92
|
+
doc = open_html(url)
|
|
93
|
+
return if doc.nil?
|
|
94
|
+
doc.xpath("//div[@class='newsDateUnderlined']").each do |date|
|
|
95
|
+
date.next.next.children.reject{|c| c.text.strip.empty?}.each do |row|
|
|
96
|
+
results << { :source => url, :url => url + row.children[0]['href'], :title => row.text, :date => Date.parse(date.text), :domain => "http://www.appropriations.senate.gov/", :party => 'majority' }
|
|
97
|
+
end
|
|
98
|
+
end
|
|
99
|
+
results
|
|
100
|
+
end
|
|
101
|
+
|
|
102
|
+
def self.senate_approps_minority
|
|
103
|
+
results = []
|
|
104
|
+
url = "http://www.appropriations.senate.gov/republican.cfm"
|
|
105
|
+
doc = open_html(url)
|
|
106
|
+
return if doc.nil?
|
|
107
|
+
doc.xpath("//div[@class='newsDateUnderlined']").each do |date|
|
|
108
|
+
date.next.next.children.reject{|c| c.text.strip.empty?}.each do |row|
|
|
109
|
+
results << { :source => url, :url => url + row.children[0]['href'], :title => row.text, :date => Date.parse(date.text), :domain => "http://www.appropriations.senate.gov/", :party => 'minority' }
|
|
110
|
+
end
|
|
111
|
+
end
|
|
112
|
+
results
|
|
113
|
+
end
|
|
114
|
+
|
|
115
|
+
def self.senate_banking(year)
|
|
116
|
+
results = []
|
|
117
|
+
url = "http://www.banking.senate.gov/public/index.cfm?FuseAction=Newsroom.PressReleases&ContentRecordType_id=b94acc28-404a-4fc6-b143-a9e15bf92da4&Region_id=&Issue_id=&MonthDisplay=0&YearDisplay=#{year}"
|
|
118
|
+
doc = open_html(url)
|
|
119
|
+
return if doc.nil?
|
|
120
|
+
doc.xpath("//tr").each do |row|
|
|
121
|
+
results << { :source => url, :url => "http://www.banking.senate.gov/public/" + row.children[2].children[1]['href'], :title => row.children[2].text.strip, :date => Date.parse(row.children[0].text.strip+", #{year}"), :domain => "http://www.banking.senate.gov/", :party => 'majority' }
|
|
122
|
+
end
|
|
123
|
+
results
|
|
124
|
+
end
|
|
125
|
+
|
|
126
|
+
def self.senate_hsag_majority(year)
|
|
127
|
+
results = []
|
|
128
|
+
url = "http://www.hsgac.senate.gov/media/majority-media?year=#{year}"
|
|
129
|
+
doc = open_html(url)
|
|
130
|
+
return if doc.nil?
|
|
131
|
+
doc.xpath("//tr").each do |row|
|
|
132
|
+
next if row.text.strip.size < 30
|
|
133
|
+
results << { :source => url, :url => row.children[2].children[0]['href'].strip, :title => row.children[2].children[0].text, :date => Date.parse(row.children[0].text), :domain => "http://www.hsgac.senate.gov/", :party => 'majority' }
|
|
134
|
+
end
|
|
135
|
+
results
|
|
136
|
+
end
|
|
137
|
+
|
|
138
|
+
def self.senate_hsag_minority(year)
|
|
139
|
+
results = []
|
|
140
|
+
url = "http://www.hsgac.senate.gov/media/minority-media?year=#{year}"
|
|
141
|
+
doc = open_html(url)
|
|
142
|
+
return if doc.nil?
|
|
143
|
+
doc.xpath("//tr").each do |row|
|
|
144
|
+
next if row.text.strip.size < 30
|
|
145
|
+
results << { :source => url, :url => row.children[2].children[0]['href'].strip, :title => row.children[2].children[0].text, :date => Date.parse(row.children[0].text), :domain => "http://www.hsgac.senate.gov/", :party => 'minority' }
|
|
146
|
+
end
|
|
147
|
+
results
|
|
148
|
+
end
|
|
149
|
+
|
|
150
|
+
def self.senate_indian
|
|
151
|
+
results = []
|
|
152
|
+
url = "http://www.indian.senate.gov/news/index.cfm"
|
|
153
|
+
doc = open_html(url)
|
|
154
|
+
return if doc.nil?
|
|
155
|
+
doc.xpath("//h3").each do |row|
|
|
156
|
+
results << { :source => url, :url => "http://www.indian.senate.gov"+row.children[0]['href'], :title => row.children[0].text, :date => Date.parse(row.previous.previous.text), :domain => "http://www.indian.senate.gov/", :party => 'majority' }
|
|
157
|
+
end
|
|
158
|
+
results
|
|
159
|
+
end
|
|
160
|
+
|
|
161
|
+
def self.senate_aging
|
|
162
|
+
results = []
|
|
163
|
+
url = "http://www.aging.senate.gov/pressroom.cfm?maxrows=100&startrow=1&&type=1"
|
|
164
|
+
doc = open_html(url)
|
|
165
|
+
return if doc.nil?
|
|
166
|
+
doc.xpath("//tr")[6..104].each do |row|
|
|
167
|
+
results << { :source => url, :url => "http://www.aging.senate.gov/"+row.children[2].children[0]['href'], :title => row.children[2].text.strip, :date => Date.parse(row.children[0].text), :domain => "http://www.aging.senate.gov/" }
|
|
168
|
+
end
|
|
169
|
+
results
|
|
170
|
+
end
|
|
171
|
+
|
|
172
|
+
def self.senate_smallbiz_minority
|
|
173
|
+
results = []
|
|
174
|
+
url = "http://www.sbc.senate.gov/public/index.cfm?p=RepublicanPressRoom"
|
|
175
|
+
doc = open_html(url)
|
|
176
|
+
return if doc.nil?
|
|
177
|
+
doc.xpath("//ul[@class='recordList']").each do |row|
|
|
178
|
+
results << { :source => url, :url => row.children[0].children[2].children[0]['href'], :title => row.children[0].children[2].children[0].text, :date => Date.parse(row.children[0].children[0].text), :domain => "http://www.sbc.senate.gov/", :party => 'minority' }
|
|
179
|
+
end
|
|
180
|
+
results
|
|
181
|
+
end
|
|
182
|
+
|
|
183
|
+
def self.senate_intel(congress, start_year, end_year)
|
|
184
|
+
results = []
|
|
185
|
+
url = "http://www.intelligence.senate.gov/press/releases.cfm?congress=#{congress}&y1=#{start_year}&y2=#{end_year}"
|
|
186
|
+
doc = open_html(url)
|
|
187
|
+
return if doc.nil?
|
|
188
|
+
doc.xpath("//tr[@valign='top']")[7..-1].each do |row|
|
|
189
|
+
results << { :source => url, :url => "http://www.intelligence.senate.gov/press/"+row.children[2].children[0]['href'], :title => row.children[2].children[0].text.strip, :date => Date.parse(row.children[0].text), :domain => "http://www.intelligence.senate.gov/" }
|
|
190
|
+
end
|
|
191
|
+
results
|
|
192
|
+
end
|
|
193
|
+
|
|
194
|
+
def self.house_energy_minority
|
|
195
|
+
results = []
|
|
196
|
+
url = "http://democrats.energycommerce.house.gov/index.php?q=news-releases"
|
|
197
|
+
doc = open_html(url)
|
|
198
|
+
return if doc.nil?
|
|
199
|
+
doc.xpath("//div[@class='views-field-title']").each do |row|
|
|
200
|
+
results << { :source => url, :url => "http://democrats.energycommerce.house.gov"+row.children[1].children[0]['href'], :title => row.children[1].children[0].text, :date => Date.parse(row.next.next.text.strip), :domain => "http://energycommerce.house.gov/", :party => 'minority' }
|
|
201
|
+
end
|
|
202
|
+
results
|
|
203
|
+
end
|
|
204
|
+
|
|
205
|
+
def self.house_homeland_security_minority
|
|
206
|
+
results = []
|
|
207
|
+
url = "http://chsdemocrats.house.gov/press/index.asp?subsection=1"
|
|
208
|
+
doc = open_html(url)
|
|
209
|
+
return if doc.nil?
|
|
210
|
+
doc.xpath("//li[@class='article']").each do |row|
|
|
211
|
+
results << { :source => url, :url => "http://chsdemocrats.house.gov"+row.children[1]['href'], :title => row.children[1].text.strip, :date => Date.parse(row.children[3].text), :domain => "http://chsdemocrats.house.gov/", :party => 'minority' }
|
|
212
|
+
end
|
|
213
|
+
results
|
|
214
|
+
end
|
|
215
|
+
|
|
216
|
+
def self.house_judiciary_majority
|
|
217
|
+
results = []
|
|
218
|
+
url = "http://judiciary.house.gov/news/press2013.html"
|
|
219
|
+
doc = open_html(url)
|
|
220
|
+
return if doc.nil?
|
|
221
|
+
doc.xpath("//p")[3..60].each do |row|
|
|
222
|
+
next if row.text.size < 30
|
|
223
|
+
results << { :source => url, :url => row.children[5]['href'], :title => row.children[0].text, :date => Date.parse(row.children[1].text.strip), :domain => "http://judiciary.house.gov/", :party => 'majority' }
|
|
224
|
+
end
|
|
225
|
+
results
|
|
226
|
+
end
|
|
227
|
+
|
|
228
|
+
def self.house_rules_majority
|
|
229
|
+
results = []
|
|
230
|
+
url = "http://www.rules.house.gov/News/Default.aspx"
|
|
231
|
+
doc = open_html(url)
|
|
232
|
+
return if doc.nil?
|
|
233
|
+
doc.xpath("//tr")[1..-2].each do |row|
|
|
234
|
+
next if row.text.strip.size < 30
|
|
235
|
+
results << { :source => url, :url => "http://www.rules.house.gov/News/"+row.children[0].children[1].children[0]['href'], :title => row.children[0].children[1].children[0].text, :date => Date.parse(row.children[2].children[1].text.strip), :domain => "http://www.rules.house.gov/", :party => 'majority' }
|
|
236
|
+
end
|
|
237
|
+
results
|
|
238
|
+
end
|
|
239
|
+
|
|
240
|
+
def self.house_ways_means_majority
|
|
241
|
+
results = []
|
|
242
|
+
url = "http://waysandmeans.house.gov/news/documentquery.aspx?DocumentTypeID=1496"
|
|
243
|
+
doc = open_html(url)
|
|
244
|
+
return if doc.nil?
|
|
245
|
+
doc.xpath("//ul[@class='UnorderedNewsList']").children.each do |row|
|
|
246
|
+
next if row.text.strip.size < 10
|
|
247
|
+
results << { :source => url, :url => "http://waysandmeans.house.gov"+row.children[1].children[1]['href'], :title => row.children[1].children[1].text, :date => Date.parse(row.children[3].children[0].text.strip), :domain => "http://waysandmeans.house.gov/", :party => 'majority' }
|
|
248
|
+
end
|
|
249
|
+
results
|
|
250
|
+
end
|
|
251
|
+
|
|
80
252
|
## special cases for members without RSS feeds
|
|
81
253
|
|
|
82
254
|
def self.capuano
|
metadata
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: statement
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.
|
|
4
|
+
version: '0.8'
|
|
5
5
|
prerelease:
|
|
6
6
|
platform: ruby
|
|
7
7
|
authors:
|
|
@@ -9,11 +9,11 @@ authors:
|
|
|
9
9
|
autorequire:
|
|
10
10
|
bindir: bin
|
|
11
11
|
cert_chain: []
|
|
12
|
-
date: 2013-05-
|
|
12
|
+
date: 2013-05-16 00:00:00.000000000 Z
|
|
13
13
|
dependencies:
|
|
14
14
|
- !ruby/object:Gem::Dependency
|
|
15
15
|
name: bundler
|
|
16
|
-
requirement: !ruby/object:Gem::Requirement
|
|
16
|
+
requirement: &2152291960 !ruby/object:Gem::Requirement
|
|
17
17
|
none: false
|
|
18
18
|
requirements:
|
|
19
19
|
- - ~>
|
|
@@ -21,15 +21,10 @@ dependencies:
|
|
|
21
21
|
version: '1.3'
|
|
22
22
|
type: :development
|
|
23
23
|
prerelease: false
|
|
24
|
-
version_requirements:
|
|
25
|
-
none: false
|
|
26
|
-
requirements:
|
|
27
|
-
- - ~>
|
|
28
|
-
- !ruby/object:Gem::Version
|
|
29
|
-
version: '1.3'
|
|
24
|
+
version_requirements: *2152291960
|
|
30
25
|
- !ruby/object:Gem::Dependency
|
|
31
26
|
name: rake
|
|
32
|
-
requirement: !ruby/object:Gem::Requirement
|
|
27
|
+
requirement: &2152239160 !ruby/object:Gem::Requirement
|
|
33
28
|
none: false
|
|
34
29
|
requirements:
|
|
35
30
|
- - ! '>='
|
|
@@ -37,15 +32,10 @@ dependencies:
|
|
|
37
32
|
version: '0'
|
|
38
33
|
type: :development
|
|
39
34
|
prerelease: false
|
|
40
|
-
version_requirements:
|
|
41
|
-
none: false
|
|
42
|
-
requirements:
|
|
43
|
-
- - ! '>='
|
|
44
|
-
- !ruby/object:Gem::Version
|
|
45
|
-
version: '0'
|
|
35
|
+
version_requirements: *2152239160
|
|
46
36
|
- !ruby/object:Gem::Dependency
|
|
47
37
|
name: webmock
|
|
48
|
-
requirement: !ruby/object:Gem::Requirement
|
|
38
|
+
requirement: &2152121140 !ruby/object:Gem::Requirement
|
|
49
39
|
none: false
|
|
50
40
|
requirements:
|
|
51
41
|
- - ! '>='
|
|
@@ -53,15 +43,10 @@ dependencies:
|
|
|
53
43
|
version: '0'
|
|
54
44
|
type: :development
|
|
55
45
|
prerelease: false
|
|
56
|
-
version_requirements:
|
|
57
|
-
none: false
|
|
58
|
-
requirements:
|
|
59
|
-
- - ! '>='
|
|
60
|
-
- !ruby/object:Gem::Version
|
|
61
|
-
version: '0'
|
|
46
|
+
version_requirements: *2152121140
|
|
62
47
|
- !ruby/object:Gem::Dependency
|
|
63
48
|
name: american_date
|
|
64
|
-
requirement: !ruby/object:Gem::Requirement
|
|
49
|
+
requirement: &2152084380 !ruby/object:Gem::Requirement
|
|
65
50
|
none: false
|
|
66
51
|
requirements:
|
|
67
52
|
- - ! '>='
|
|
@@ -69,15 +54,10 @@ dependencies:
|
|
|
69
54
|
version: '0'
|
|
70
55
|
type: :runtime
|
|
71
56
|
prerelease: false
|
|
72
|
-
version_requirements:
|
|
73
|
-
none: false
|
|
74
|
-
requirements:
|
|
75
|
-
- - ! '>='
|
|
76
|
-
- !ruby/object:Gem::Version
|
|
77
|
-
version: '0'
|
|
57
|
+
version_requirements: *2152084380
|
|
78
58
|
- !ruby/object:Gem::Dependency
|
|
79
59
|
name: nokogiri
|
|
80
|
-
requirement: !ruby/object:Gem::Requirement
|
|
60
|
+
requirement: &2151878460 !ruby/object:Gem::Requirement
|
|
81
61
|
none: false
|
|
82
62
|
requirements:
|
|
83
63
|
- - ! '>='
|
|
@@ -85,12 +65,7 @@ dependencies:
|
|
|
85
65
|
version: '0'
|
|
86
66
|
type: :runtime
|
|
87
67
|
prerelease: false
|
|
88
|
-
version_requirements:
|
|
89
|
-
none: false
|
|
90
|
-
requirements:
|
|
91
|
-
- - ! '>='
|
|
92
|
-
- !ruby/object:Gem::Version
|
|
93
|
-
version: '0'
|
|
68
|
+
version_requirements: *2151878460
|
|
94
69
|
description: Crawls congressional websites for press releases.
|
|
95
70
|
email:
|
|
96
71
|
- dwillis@gmail.com
|
|
@@ -127,21 +102,15 @@ required_ruby_version: !ruby/object:Gem::Requirement
|
|
|
127
102
|
- - ! '>='
|
|
128
103
|
- !ruby/object:Gem::Version
|
|
129
104
|
version: '0'
|
|
130
|
-
segments:
|
|
131
|
-
- 0
|
|
132
|
-
hash: -672742253743525613
|
|
133
105
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
134
106
|
none: false
|
|
135
107
|
requirements:
|
|
136
108
|
- - ! '>='
|
|
137
109
|
- !ruby/object:Gem::Version
|
|
138
110
|
version: '0'
|
|
139
|
-
segments:
|
|
140
|
-
- 0
|
|
141
|
-
hash: -672742253743525613
|
|
142
111
|
requirements: []
|
|
143
112
|
rubyforge_project:
|
|
144
|
-
rubygems_version: 1.8.
|
|
113
|
+
rubygems_version: 1.8.17
|
|
145
114
|
signing_key:
|
|
146
115
|
specification_version: 3
|
|
147
116
|
summary: Given a url, Statement returns links to press releases and official statements.
|