email_crawler 0.0.6 → 0.0.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: a260b02f463c94ff01de5957eb1ec7ea95e8e150
-  data.tar.gz: aac88c2198ed95902045c7ee2fc874bf6d5b65f4
+  metadata.gz: c66a31ba29e57a09952acc7e1816093ba8aaf613
+  data.tar.gz: 094c99616e62b5c92d892faa02d44e322b44c2b2
 SHA512:
-  metadata.gz: 00f1003d1e385527d0bcceae8fdddd042e06b074f4f1d3447b1f79506fcc173aa8d7ed750a5698738a5c59a28c0e8306e4f9ea4859bc318c5697f36d6ce0b3c1
-  data.tar.gz: c53d9f2e6cede921ec98c1ce42986d628e4394124514cca1aaf0b0a2fc174d966e20abed152368545714b4e11da724e12b294f491b10c48b8d1fa5dd1d930e21
+  metadata.gz: 387b8ef9db7125a79ed935801b07b5cfb3175895c43fdbfcc922135b150ce6535fa3508b565245905568f5ac0241d9fa57c2f50452046185984bbd96d7c40e15
+  data.tar.gz: 9e51f77812283e4d02ecb8130d19030d17e19716be3cad0521f6ff2f7dfe6d271134fa5fd9c4f52e422b194fba94675292b2a15bac411943e0beb7cb6abb4560

data/lib/email_crawler/page_links.rb

@@ -49,6 +49,8 @@ module EmailCrawler
             @logger.error "Giving up grabbing link for '#{@url}' after #{retries} retries"
             break
           end
+        rescue URI::InvalidComponentError => err
+          @logger.warn err.inspect
         else
           retries = 0
         end

data/lib/email_crawler/version.rb

@@ -1,3 +1,3 @@
 module EmailCrawler
-  VERSION = "0.0.6"
+  VERSION = "0.0.7"
 end

data/spec/lib/email_crawler/scraper_spec.rb

@@ -4,10 +4,12 @@ require File.expand_path("lib/email_crawler")
 
 module EmailCrawler
   describe Scraper do
-    subject { Scraper.new("google.de") }
+    let(:max_results) { 10 }
+
+    subject { Scraper.new("google.de", max_results) }
 
     it "returns the top 10 URLs for a given search term/expression" do
-      subject.top_ten_urls_for("berlin tours").length.must_equal 10
+      subject.search_result_urls_for("berlin tours").length.must_equal max_results
     end
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: email_crawler
 version: !ruby/object:Gem::Version
-  version: 0.0.6
+  version: 0.0.7
 platform: ruby
 authors:
 - Cristian Rasch
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-03-08 00:00:00.000000000 Z
+date: 2014-03-19 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mechanize