w3scraper 0.0.1

data/bin/w3scraper

@@ -0,0 +1,6 @@
+#!/usr/bin/env ruby
+#
+require 'w3scraper'
+
+app = W3scraper.new(ARGV, STDIN)
+app.run

data/lib/w3scraper.rb

@@ -0,0 +1,80 @@
+#!/usr/bin/env ruby
+
+require 'nokogiri'
+require 'date'
+require 'optparse'
+
+
+class W3scraper
+  VERSION = '0.0.1'
+
+  def initialize(arguments, stdin)
+    @arguments = arguments
+    @stdin = stdin
+    @options = {}
+    parse_options
+  end
+
+  def run
+    errors = parse_errors(w3_parser_output)
+    print_errors(errors)
+  end
+
+  protected
+
+    def w3_parser_output
+      @w3_parser_output ||= `curl -s -F output=text -F "uploaded_file=@#{@target_file};type=text/html" http://validator.w3.org/check`
+    end
+
+    def print_errors(errors)
+      errors.each do |result|
+        puts "Line #{result[:line]} Col #{result[:col]}: #{result[:msg]}"
+      end
+    end
+
+    def parse_errors(output)
+      doc = Nokogiri::HTML(output)
+      ret_val = []
+
+      doc.css("#error_loop .msg_err").each do |e|
+        next_result = {}
+
+        #extract line/column numbers
+        pos_match = e.css("em").first.to_s.match(/Line (\d*).*Column (\d*)/m)
+        next_result[:line] = pos_match[1]
+        next_result[:col] = pos_match[2]
+
+        #extract error message
+        next_result[:msg] = e.css(".msg").first.content
+
+        ret_val << next_result
+      end
+
+      ret_val
+    end
+
+
+    def parse_options
+      opts = OptionParser.new
+      opts.banner = "Usage: #{File.basename(__FILE__)} [options] [file to validate]"
+
+      opts.on('-v', '--version') do
+        puts "version: #{VERSION}"
+        exit 0
+      end
+
+      opts.on('-h', '--help') do
+        puts opts
+        exit 0
+      end
+
+      opts.parse!(@arguments)
+
+      if @arguments.size != 1
+        puts opts
+        exit 1
+      end
+
+      @target_file = @arguments.first
+    end
+end

metadata

@@ -0,0 +1,48 @@
+--- !ruby/object:Gem::Specification
+name: w3scraper
+version: !ruby/object:Gem::Version
+  version: 0.0.1
+  prerelease: 
+platform: ruby
+authors:
+- Martin Grenfell
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2012-05-02 00:00:00.000000000 Z
+dependencies: []
+description: Sends a file to http://validator.w3.org and parses/outputs any errors.
+email: martin.grenfell@gmail.com
+executables:
+- w3scraper
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/w3scraper.rb
+- bin/w3scraper
+homepage: http://rubygems.org/gems/w3scraper
+licenses: []
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 1.8.10
+signing_key: 
+specification_version: 3
+summary: Command line interface to http://validator.w3.org
+test_files: []
+has_rdoc: