scraprr 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: c7ede4f1bfe1eebbca1915f4c515537ada2c71b1
-  data.tar.gz: 22e5a2d1cce4c50d0ed8d11d41eb2aaa7529ebef
+  metadata.gz: 94ea1d7615ca2011a82052a968bea0f557e1c210
+  data.tar.gz: 762eea3721533afe842d013dd3000781faddd4f2
 SHA512:
-  metadata.gz: a24d82be7501c66d1192a3900c837c27605559328ab0e2e486664fd10bb34a920a1cc028f1e24f9a1692af6d595849b614cb8d9ff5ecc8a0d4f5ea592574cc43
-  data.tar.gz: 1c1a572ad884718f93509aa1d5348681bdeb428a00a70b7acb43c5719bb6e54b6449ea1ab62342c76ef28692ecfd8edb6bdbded218c7fc323a2ec624bb63e9b5
+  metadata.gz: 149a1660d39220c2c7e484c31d5b4ef2613ce9432c6e1d2e60e69afbdd062c33d60ac7a6e87a71f381ef7b6c95adedf8df502206294b6b0cc38b54b2f86326af
+  data.tar.gz: b98578eb8c1dff1106ed6f2824aaee07965eefbc7b1eb007c19a5a4b8c8041537ad78431f8c6d4fa48d3d30f85de28a2ccece7005318ae107860f20ad9167b16

data/example/taphaus.rb

@@ -0,0 +1,19 @@
+#!/usr/bin/env ruby
+require 'pp'
+require 'open-uri'
+require 'watir'
+require 'watir-webdriver'
+require 'scraprr'
+
+
+scraper = Scraprr::Scraper.new('div[skinpart=itemsContainer] > div').
+  attribute(:name, 'div[skinpart=title]', :required => true).
+  attribute(:abv, 'div[skinpart=description]', :regexp => /([0-9.]+)%/)
+
+browser = Watir::Browser.new
+browser.goto 'http://www.thetaphaus.co.nz/#!now-on-tap/c14a4'
+sleep 10
+document = Nokogiri::HTML(browser.html)
+browser.close
+
+pp scraper.extract(document)

data/example/wikipedia.rb

@@ -0,0 +1,13 @@
+#!/usr/bin/env ruby
+require 'pp'
+require 'open-uri'
+require 'scraprr'
+
+scraper = Scraprr::Scraper.new('table tr').
+  attribute(:brewery, './td[1]', :required => true).
+  attribute(:url, './td[1]/a', :attr => 'href', :required => true).
+  attribute(:year, './td[3]').
+  attribute(:production, './td[4]')
+
+document = Nokogiri::HTML(open('http://en.wikipedia.org/wiki/Trappist_beer'))
+pp scraper.extract(document)

data/lib/scraprr/strip_filter.rb

@@ -5,7 +5,8 @@ module Scraprr
     end
 
     def run(value)
-      @chain.run(value).strip
+      value = @chain.run(value)
+      value.strip if value
     end
   end
 end

data/lib/scraprr/version.rb

@@ -1,3 +1,3 @@
 module Scraprr
-  VERSION = "0.0.1"
+  VERSION = "0.0.2"
 end

data/test/scraprr/strip_filter_test.rb

@@ -13,5 +13,13 @@ describe Scraprr::StripFilter do
       @chain = Scraprr::StripFilter.new(@chain)
       @chain.run(value).must_equal('test')
     end
+
+    it "returns nil if value was nil" do
+      value = nil
+      @chain.expect(:run, value, [value])
+
+      @chain = Scraprr::StripFilter.new(@chain)
+      @chain.run(value).must_equal(nil)
+    end
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: scraprr
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - James Fargher
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-05-03 00:00:00.000000000 Z
+date: 2013-12-12 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -78,6 +78,8 @@ files:
 - LICENSE.txt
 - README.md
 - Rakefile
+- example/taphaus.rb
+- example/wikipedia.rb
 - lib/scraprr.rb
 - lib/scraprr/attribute_scraper.rb
 - lib/scraprr/exceptions.rb
@@ -118,7 +120,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.0.0
+rubygems_version: 2.0.3
 signing_key: 
 specification_version: 4
 summary: Declarative HTML/XML scraper
@@ -132,4 +134,3 @@ test_files:
 - test/scraprr/strip_filter_test.rb
 - test/scraprr/value_extractor_test.rb
 - test/test_helper.rb
-has_rdoc: