snatch 1.0.2 → 1.0.3

data/README.rdoc

@@ -1,19 +1,17 @@
 = snatch
 
-Ruby wrapper around wget to recursively download a site and copy it in to a public directory, typically within Rails.
+*Snatch* a site.
 
-We use this gem to download the static files produced by a CMS used by our Marketing department and then serve them using Rack within our Rails app.
+  snatch = Snatch.new
+  snatch.run
 
-== Note on Patches/Pull Requests
- 
-* Fork the project.
-* Make your feature addition or bug fix.
-* Add tests for it. This is important so I don't break it in a
-  future version unintentionally.
-* Commit, do not mess with rakefile, version, or history.
-  (if you want to have your own version, that is fine but bump version in a commit by itself I can ignore when I pull)
-* Send me a pull request. Bonus points for topic branches.
+Usage:
+  snatch.rb [OPTIONS] SITE_URI(s)
 
-== Copyright
+  Specific Options:
+    -w, --wget                       Specify the full path to the wget binary (Must be at least version 1.12.)
+    -g, --git                        Specify the full path to the git binary.
+  Common Options:
+    -h, --help                       Show this message.
 
-Copyright (c) 2010 AlphaSights. See LICENSE for details.
+You can pass multiple URIs to snatch.rb. You need to have wget v1.12 and git installed within $PATH. snatch.rb will optionally search for wget and git if they can not be found.

data/VERSION

@@ -1 +1 @@
-1.0.2
+1.0.3

data/lib/snatch/clean.rb

@@ -20,7 +20,7 @@ class Snatch
       CSS.update(@doc, @working_directory)
       HTML.update(@doc, @working_directory)
 
-      File.open(@file_name, 'w') { |f| f.write @doc.to_html }
+      File.open(@file_name, 'w') { |f| f.write @doc.to_xhtml }
     end
   end
 end

data/lib/snatch/clean/html.rb

@@ -2,9 +2,9 @@ class Snatch
   class Clean
     class HTML
       module HrefFixMethods
-        def remove_index_html(a)
-          a['href'] = a['href'].sub(%r{index\.html?$}, '')
-        end
+        # def remove_index_html(a)
+        #   a['href'] = a['href'].sub(%r{index\.html?$}, '')
+        # end
 
         def replace_absolute(a)
           a['href'] = a['href'].sub(%r{(https?)://#{MARKETING_SITE}/}, '/')

data/snatch.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = %q{snatch}
-  s.version = "1.0.2"
+  s.version = "1.0.3"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["James Conroy-Finn"]
-  s.date = %q{2010-02-15}
+  s.date = %q{2010-02-18}
   s.description = %q{Simple site downloaded that wraps wget and converts PHP CSS files in to regular CSS files.}
   s.email = %q{james@logi.cl}
   s.extra_rdoc_files = [

data/spec/snatch/clean/html_spec.rb

@@ -33,10 +33,18 @@ describe Snatch::Clean::HTML do
   describe Snatch::Clean::HTML::HrefFixMethods do
     subject { mock.extend(Snatch::Clean::HTML::HrefFixMethods) }
 
-    it 'should remove a trailing index.html' do
-      fix_node(:remove_index_html, '<a href="/blah/index.html"></a>') do |node|
-        node.should have_href('/blah/')
-      end
+    # it 'should remove a trailing index.html' do
+    #   fix_node(:remove_index_html, '<a href="/blah/index.html"></a>') do |node|
+    #     node.should have_href('/blah/')
+    #   end
+    # end
+
+    it 'should preserve parent directories within a URL' do
+      anchor = '<a href="/folder/child/file.extension"></a>'
+      doc = Nokogiri::XML(anchor)
+      @html.doc = doc
+      @html.send(:update)
+      @html.doc.to_xhtml.strip.should == '<a href="/folder/child/file.extension"></a>'
     end
 
     it 'should replace an absolute CMS URL with a domainless absolute URL' do

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: snatch
 version: !ruby/object:Gem::Version 
-  version: 1.0.2
+  version: 1.0.3
 platform: ruby
 authors: 
 - James Conroy-Finn
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-02-15 00:00:00 +00:00
+date: 2010-02-18 00:00:00 +00:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency