horsefield 0.1.1 → 0.2.0
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/.gitignore +0 -5
- data/.rspec +2 -0
- data/LICENSE.txt +1 -1
- data/README.md +22 -4
- data/horsefield.gemspec +5 -7
- data/lib/horsefield.rb +2 -5
- data/lib/horsefield/node.rb +12 -0
- data/lib/horsefield/node_set.rb +51 -0
- data/lib/horsefield/scraper.rb +30 -10
- data/lib/horsefield/version.rb +1 -1
- data/spec/fixtures/monster.html +2311 -0
- data/spec/fixtures/vcr_cassettes/facebook/johnny_qiu1.yml +7105 -0
- data/spec/horsefield_spec.rb +4 -0
- data/spec/scraper_spec.rb +21 -0
- data/spec/spec_helper.rb +11 -11
- metadata +33 -65
- data/Guardfile +0 -24
- data/lib/horsefield/diggable.rb +0 -34
- data/lib/horsefield/nokogiri.rb +0 -12
- data/spec/fixtures/linkedin_biggs.html +0 -652
- data/spec/fixtures/vcr_cassettes/linkedin_lunarmobiscuit.yml +0 -880
- data/spec/lib/horsefield/scraper_spec.rb +0 -42
|
@@ -1,42 +0,0 @@
|
|
|
1
|
-
require 'spec_helper'
|
|
2
|
-
require 'horsefield/scraper'
|
|
3
|
-
|
|
4
|
-
describe Horsefield::Scraper do
|
|
5
|
-
let(:scraper) { Horsefield::Scraper.new('http://www.linkedin.com/in/lunarmobiscuit') }
|
|
6
|
-
|
|
7
|
-
it 'should accept HTML or a URL as argument' do
|
|
8
|
-
VCR.use_cassette('linkedin_lunarmobiscuit') do
|
|
9
|
-
scraper.html.should match('<!DOCTYPE html>')
|
|
10
|
-
end
|
|
11
|
-
end
|
|
12
|
-
|
|
13
|
-
describe '#scrape' do
|
|
14
|
-
it 'should take a block with attributes to scrape' do
|
|
15
|
-
data = scraper.scrape do
|
|
16
|
-
scope '.profile-header' do
|
|
17
|
-
one :name, '//span[@class="full-name"]'
|
|
18
|
-
end
|
|
19
|
-
|
|
20
|
-
many :skills, '#skills-list li'
|
|
21
|
-
|
|
22
|
-
many :experiences, '#profile-experience .position' do
|
|
23
|
-
one :headline, 'span.title'
|
|
24
|
-
|
|
25
|
-
one :start_year, '.period .dtstart' do
|
|
26
|
-
attr('title').split('-').first.to_i
|
|
27
|
-
end
|
|
28
|
-
end
|
|
29
|
-
|
|
30
|
-
one :company do
|
|
31
|
-
one :name, '.postitle h4'
|
|
32
|
-
end
|
|
33
|
-
end
|
|
34
|
-
|
|
35
|
-
p data[:company]
|
|
36
|
-
|
|
37
|
-
data[:name].should == "Michael 'Luni' Libes"
|
|
38
|
-
data[:experiences].first[:headline].should == 'Founder and Managing Director'
|
|
39
|
-
data[:skills].should == ["Entrepreneurship", "Business Planning", "Fundraising", "Team Building", "Start-ups", "Venture Capital", "Social Entrepreneurship", "Strategic Planning", "Strategic Partnerships", "Strategy", "Business Development", "Management Consulting", "Marketing Strategy", "Sustainability", "Executive Management", "Marketing", "New Business Development", "Competitive Analysis", "Go-to-market Strategy", "Thought Leadership", "Corporate Development", "Software Development", "Business Strategy", "Mobile Devices", "Wireless", "Management", "Cloud Computing", "Nonprofits", "Business Modeling", "SaaS", "Mobile Applications", "Product Marketing", "Program Management", "Analytics", "E-commerce", "Leadership", "Enterprise Software", "Consulting", "Strategic Consulting"]
|
|
40
|
-
end
|
|
41
|
-
end
|
|
42
|
-
end
|