statement 1.4 → 1.5
Sign up to get free protection for your applications and to get access to all the features.
- data/lib/statement/scraper.rb +4 -28
- data/lib/statement/version.rb +1 -1
- metadata +4 -4
data/lib/statement/scraper.rb
CHANGED
@@ -29,7 +29,7 @@ module Statement
|
|
29
29
|
end
|
30
30
|
|
31
31
|
def self.member_methods
|
32
|
-
[:capuano, :cold_fusion, :conaway, :chabot, :susandavis, :freshman_senators, :klobuchar, :lujan, :billnelson, :lautenberg, :crapo, :coburn, :boxer, :
|
32
|
+
[:capuano, :cold_fusion, :conaway, :chabot, :susandavis, :freshman_senators, :klobuchar, :lujan, :billnelson, :lautenberg, :crapo, :coburn, :boxer, :vitter, :donnelly, :inhofe, :reid, :palazzo, :roe, :document_query, :swalwell, :fischer]
|
33
33
|
end
|
34
34
|
|
35
35
|
def self.committee_methods
|
@@ -39,7 +39,7 @@ module Statement
|
|
39
39
|
def self.member_scrapers
|
40
40
|
year = Date.today.year
|
41
41
|
results = [freshman_senators, capuano, cold_fusion(year, 0), conaway, chabot, susandavis, klobuchar, lujan, palazzo(page=1), roe(page=1), billnelson(year=year),
|
42
|
-
document_query(page=1), document_query(page=2), swalwell(page=1), donnelly(year=year), crapo, coburn, boxer(start=1),
|
42
|
+
document_query(page=1), document_query(page=2), swalwell(page=1), donnelly(year=year), crapo, coburn, boxer(start=1),
|
43
43
|
vitter(year=year), inhofe(year=year), reid, fischer].flatten
|
44
44
|
results = results.compact
|
45
45
|
Utils.remove_generic_urls!(results)
|
@@ -48,7 +48,7 @@ module Statement
|
|
48
48
|
def self.backfill_from_scrapers
|
49
49
|
results = [cold_fusion(2012, 0), cold_fusion(2011, 0), cold_fusion(2010, 0), billnelson(year=2012), document_query(page=3),
|
50
50
|
document_query(page=4), coburn(year=2012), coburn(year=2011), coburn(year=2010), boxer(start=11), boxer(start=21),
|
51
|
-
boxer(start=31), boxer(start=41),
|
51
|
+
boxer(start=31), boxer(start=41), vitter(year=2012), vitter(year=2011), swalwell(page=2), swalwell(page=3)
|
52
52
|
].flatten
|
53
53
|
Utils.remove_generic_urls!(results)
|
54
54
|
end
|
@@ -300,7 +300,7 @@ module Statement
|
|
300
300
|
url = base_url + "documentquery.aspx?DocumentTypeID=2508&Year=#{year}"
|
301
301
|
doc = open_html(url)
|
302
302
|
return if doc.nil?
|
303
|
-
doc.xpath("//li")[
|
303
|
+
doc.xpath("//li")[40..48].each do |row|
|
304
304
|
results << { :source => url, :url => base_url + row.children[1]['href'], :title => row.children[1].children.text.strip, :date => Date.parse(row.children[3].text.strip), :domain => "chabot.house.gov" }
|
305
305
|
end
|
306
306
|
results
|
@@ -433,18 +433,6 @@ module Statement
|
|
433
433
|
results
|
434
434
|
end
|
435
435
|
|
436
|
-
def self.mccain(year=Date.today.year)
|
437
|
-
results = []
|
438
|
-
url = "http://www.mccain.senate.gov/public/index.cfm?FuseAction=PressOffice.PressReleases&ContentRecordType_id=75e7e4a0-6088-44b6-8061-089d80513dc4&Region_id=&Issue_id=&MonthDisplay=0&YearDisplay=#{year}"
|
439
|
-
domain = 'www.mccain.senate.gov'
|
440
|
-
doc = open_html(url)
|
441
|
-
return if doc.nil?
|
442
|
-
doc.xpath("//li")[7..-1].each do |row|
|
443
|
-
results << { :source => url, :url => "http://"+domain+'/public/'+row.children[3].children[1].children[4].children[0]['href'], :title => row.children[3].children[1].children[4].text, :date => Date.strptime(row.children[3].children[1].children[0].text, "%m/%d/%y"), :domain => domain}
|
444
|
-
end
|
445
|
-
results
|
446
|
-
end
|
447
|
-
|
448
436
|
def self.vitter(year=Date.today.year)
|
449
437
|
results = []
|
450
438
|
url = "http://www.vitter.senate.gov/newsroom/"
|
@@ -484,18 +472,6 @@ module Statement
|
|
484
472
|
results
|
485
473
|
end
|
486
474
|
|
487
|
-
def self.levin(page=1)
|
488
|
-
results = []
|
489
|
-
url = "http://www.levin.senate.gov/newsroom/index.cfm?PageNum_rs=#{page}§ion=press"
|
490
|
-
domain = "www.levin.senate.gov"
|
491
|
-
doc = open_html(url)
|
492
|
-
return if doc.nil?
|
493
|
-
doc.xpath('//tr').each do |row|
|
494
|
-
results << { :source => url, :url => row.children[2].children[0]['href'].gsub(/\s+/, ""), :title => row.children[2].children[0].text, :date => Date.parse(row.children[0].text), :domain => domain}
|
495
|
-
end
|
496
|
-
results
|
497
|
-
end
|
498
|
-
|
499
475
|
def self.reid
|
500
476
|
results = []
|
501
477
|
url = "http://www.reid.senate.gov/newsroom/press_releases.cfm"
|
data/lib/statement/version.rb
CHANGED
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: '1.
|
4
|
+
version: '1.5'
|
5
5
|
prerelease:
|
6
6
|
platform: ruby
|
7
7
|
authors:
|
@@ -9,7 +9,7 @@ authors:
|
|
9
9
|
autorequire:
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
|
-
date: 2013-11-
|
12
|
+
date: 2013-11-23 00:00:00.000000000 Z
|
13
13
|
dependencies:
|
14
14
|
- !ruby/object:Gem::Dependency
|
15
15
|
name: bundler
|
@@ -214,7 +214,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
|
|
214
214
|
version: '0'
|
215
215
|
segments:
|
216
216
|
- 0
|
217
|
-
hash:
|
217
|
+
hash: -2681532689604515141
|
218
218
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
219
219
|
none: false
|
220
220
|
requirements:
|
@@ -223,7 +223,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
223
223
|
version: '0'
|
224
224
|
segments:
|
225
225
|
- 0
|
226
|
-
hash:
|
226
|
+
hash: -2681532689604515141
|
227
227
|
requirements: []
|
228
228
|
rubyforge_project:
|
229
229
|
rubygems_version: 1.8.25
|