statement 1.0.15 → 1.1
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/feed.rb +1 -1
- data/lib/statement/scraper.rb +16 -15
- data/lib/statement/version.rb +1 -1
- metadata +4 -4
data/lib/statement/feed.rb
CHANGED
@@ -49,12 +49,12 @@ module Statement
|
|
49
49
|
def self.from_rss(url)
|
50
50
|
doc = open_rss(url)
|
51
51
|
return unless doc
|
52
|
-
return if Mime::Type.new(doc).html?
|
53
52
|
parse_rss(doc, url)
|
54
53
|
end
|
55
54
|
|
56
55
|
def self.parse_rss(doc, url)
|
57
56
|
links = doc.xpath('//item')
|
57
|
+
return if links.empty?
|
58
58
|
results = links.map do |link|
|
59
59
|
abs_link = Utils.absolute_link(url, link.xpath('link').text)
|
60
60
|
abs_link = "http://www.burr.senate.gov/public/"+ link.xpath('link').text if url == 'http://www.burr.senate.gov/public/index.cfm?FuseAction=RSS.Feed'
|
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, :mccain, :vitter, :donnelly, :inhofe, :levin, :reid, :palazzo, :
|
32
|
+
[:capuano, :cold_fusion, :conaway, :chabot, :susandavis, :freshman_senators, :klobuchar, :lujan, :billnelson, :lautenberg, :crapo, :coburn, :boxer, :mccain, :vitter, :donnelly, :inhofe, :levin, :reid, :palazzo, :roe, :document_query, :swalwell, :fischer]
|
33
33
|
end
|
34
34
|
|
35
35
|
def self.committee_methods
|
@@ -38,8 +38,8 @@ module Statement
|
|
38
38
|
|
39
39
|
def self.member_scrapers
|
40
40
|
year = Date.today.year
|
41
|
-
results = [freshman_senators, capuano, cold_fusion(year, 0), conaway, chabot, susandavis, klobuchar, lujan, palazzo(page=1), billnelson(year=year),
|
42
|
-
document_query(page=1), document_query(page=2),
|
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), mccain(year=year),
|
43
43
|
vitter(year=year), inhofe(year=year), reid, fischer].flatten
|
44
44
|
results = results.compact
|
45
45
|
Utils.remove_generic_urls!(results)
|
@@ -519,10 +519,22 @@ module Statement
|
|
519
519
|
end
|
520
520
|
results
|
521
521
|
end
|
522
|
+
|
523
|
+
def self.roe(page=1)
|
524
|
+
results = []
|
525
|
+
domain = 'roe.house.gov'
|
526
|
+
url = "http://roe.house.gov/news/documentquery.aspx?DocumentTypeID=1532&Page=#{page}"
|
527
|
+
doc = open_html(url)
|
528
|
+
return if doc.nil?
|
529
|
+
doc.xpath("//div[@class='middlecopy']//li").each do |row|
|
530
|
+
results << { :source => url, :url => "http://roe.house.gov/news/" + row.children[1]['href'], :title => row.children[1].text.strip, :date => Date.parse(row.children[3].text.strip), :domain => domain }
|
531
|
+
end
|
532
|
+
results
|
533
|
+
end
|
522
534
|
|
523
535
|
def self.document_query(page=1)
|
524
536
|
results = []
|
525
|
-
domains = [{"
|
537
|
+
domains = [{"thornberry.house.gov" => 1776}, {"wenstrup.house.gov" => 2491}]
|
526
538
|
domains.each do |domain|
|
527
539
|
doc = open_html("http://"+domain.keys.first+"/news/documentquery.aspx?DocumentTypeID=#{domain.values.first}&Page=#{page}")
|
528
540
|
return if doc.nil?
|
@@ -533,16 +545,5 @@ module Statement
|
|
533
545
|
results.flatten
|
534
546
|
end
|
535
547
|
|
536
|
-
def self.farenthold(year=2013)
|
537
|
-
results = []
|
538
|
-
url = "http://farenthold.house.gov/index.php?flt_m=&flt_y=#{year}&option=com_content&view=article&id=1181&Itemid=100059&layout=default"
|
539
|
-
doc = open_html(url)
|
540
|
-
return if doc.nil?
|
541
|
-
doc.xpath("//div[@id='idGtReportDisplay']//li").each do |row|
|
542
|
-
results << { :source => url, :url => 'http://farenthold.house.gov'+row.children[0]['href'], :title => row.children[0].text.strip, :date => nil, :domain => "farenthold.house.gov"}
|
543
|
-
end
|
544
|
-
results
|
545
|
-
end
|
546
|
-
|
547
548
|
end
|
548
549
|
end
|
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: '1.1'
|
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: 2013-10-
|
12
|
+
date: 2013-10-29 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: 3854190129938227300
|
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: 3854190129938227300
|
227
227
|
requirements: []
|
228
228
|
rubyforge_project:
|
229
229
|
rubygems_version: 1.8.25
|