colread 0.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 20aa61c77b99f49fd87290fe5061307ebd517d83
+  data.tar.gz: 103f1857d93809b17a6f2d850033555c3061bc41
+SHA512:
+  metadata.gz: ea3687991d2e945d896e8069ad7a85df8e7e0f68bd50662139f7ea073bcfc7c28387b88c3121e7a98f54a9b857d35b2653af6690d1947aa7b305ae37fd8721b4
+  data.tar.gz: b070e16493c92b5b90f24c4013561fe85aaa442c7974e883ae3bd29ed13facf835f00a5b5cb9d83531803c627466faddcb5bc6b495c416b7e84d02611bed0095

data/bin/colread

@@ -0,0 +1,28 @@
+#!/usr/bin/env ruby
+
+require 'colread'
+require 'rubygems'
+require 'optparse'
+#require File.dirname(__FILE__) + '/../lib/readability'
+
+options = { :debug => false, :images => false }
+options_parser = OptionParser.new do |opts|
+  opts.banner = "Usage: #{File.basename($0)} [options] URL FileName"
+
+  opts.on("-d", "--debug", "Show debug output") do |v|
+    options[:debug] = v
+  end
+
+  opts.on_tail("-h", "--help", "Show this message") do
+    puts opts
+    exit
+  end
+end
+options_parser.parse!
+
+if ARGV.length != 2
+  STDERR.puts options_parser
+  exit 1
+end
+
+ColRead::Text.new(url: ARGV.first, output: ARGV.last).start

data/lib/colread.rb

@@ -0,0 +1,49 @@
+require 'nokogiri'
+require 'open-uri'
+require 'readability'
+
+module Nokogiri
+  module XML
+    class Element
+      def indent
+        self['href'].gsub /\d+/, 'x' rescue nil
+      end
+    end
+  end
+end
+
+module ColRead
+  module Core
+    def chapters url
+      doc = Nokogiri::HTML(open(url))
+      doc.css('a').group_by{|a| a.indent}.sort_by{|a| a.last.count}.last.last
+    end
+
+    def contents chapters
+      chapters.each do |a|
+        href=a['href'].start_with?('http') ? a['href'] : @root+a['href']
+        source = open(href).read.encode('utf-8')
+        content=Nokogiri::HTML(Readability::Document.new(source).content).text
+        yield [a.text, content]
+      end
+    end
+  end
+  
+  class Text
+    include ColRead::Core
+    def initialize(options={url: '',output: ''})
+      @url=options[:url]
+      @output=options[:output]
+      @root=@url.split(/(?<!\/)\/(?!\/)/).first
+    end
+    def start
+      f=open(@output,'w')
+      contents(chapters(@url)) do |c|
+        puts c.first
+        f.puts c.first
+        f.puts c.last
+      end
+      f.close
+    end
+  end
+end

metadata

@@ -0,0 +1,45 @@
+--- !ruby/object:Gem::Specification
+name: colread
+version: !ruby/object:Gem::Version
+  version: 0.0.0
+platform: ruby
+authors:
+- yunjie chen
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2013-06-03 00:00:00.000000000 Z
+dependencies: []
+description: download a list of text on a page
+email: kikyous@163.com
+executables:
+- colread
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/colread.rb
+- bin/colread
+homepage: http://rubygems.org/gems/colread
+licenses: []
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.0.3
+signing_key: 
+specification_version: 4
+summary: colread!
+test_files: []