#!/usr/bin/env ruby
#
# Author: IKARASHI Yoshinori <yoosee at gmail dot com >
# Created: 2005-09-17 21:45:01 JST
# Modified: 2008-04-07 23:18:56
# Licence:  GPLv2
# $Id: mixi2gmail.rb 38 2008-04-07 23:22:01Z yoosee $
#
=begin
= mixi2gmail description

mixi2gmail feeds mixi diary to some e-mail account, such as
gmail.

= configuration
this program requires configuration file at $HOME/.mixi/mixirc .
It should be including parameters as follows:

username = foo@example.com      # mixi account e-mail address
password = xxxxxxxx             # mixi login password
mailaddress = foo@gmail.com     # e-mail address mail to send
mailfrom  = mixi@init.org       # dummy email From address
smtpserver = localhost          # mail server
excludelist = 1111, 34567       # skip list
watchlist = 98765, 43210, 2222  # watch list who are not your friend

= website

http://trac.yoosee.net/mixi2gmail/

=end
begin
  require 'rubygems'
rescue LoadError
end
require 'nkf'
require 'mechanize'
require 'net/smtp'
require 'time'
require 'optparse'
require 'logger'

CID_SUFFIX = 'init.org' # just used for multipart content-id part

module Mixi

  class Config
    attr_reader :config
    def initialize(rcfile)
      @rcfile = rcfile
      @config = Hash.new
      config_file_parser
    end

    def config_file_parser(rcfile=@rcfile)
      return nil unless File.exist? rcfile
      File.open(rcfile) do |file|
        file.each do |line|
          key, value = line.chomp.split(/\s*=\s*/)
          if /,/ =~ value
            values = value.split(/\s*,\s*/)
            value = values
          end
          @config[key] = value
        end
      end
    end

    def command_line_parser
      opt = OptionParser.new
    end
  end


  class Diary
    attr_accessor :author, :datetime, :title, :imagepages, :images,
      :body, :url, :comments
    def initialize
      @comments = Array.new
      @imagepages = Array.new
    end

    def mkdatetime date, time
      if(/(\d{4})\D+(\d{2})\D+(\d{2})/) =~ date
        year, month, date = $1, $2, $3
      end
      if(/(\d{2}):(\d{2})/) =~ time
        hour, minute = $1, $2
      end

      Time.local(year,month,date,hour,minute)
    end

    def parse doc # doc is Hpricot::Doc object
      author, datetime, title, body = nil, nil, nil, nil
      @author = nil

      author = doc.search("//div[@class='diaryTitleFriend clearfix']/h2").
        inner_html.gsub("さんの日記",'').chomp

      dbox  = doc.search("//div[@class='viewDiaryBox']")

      ### journal title
      dhead = dbox.search("/div[@class='listDiaryTitle']/dl")
      title = dhead.search("dt").inner_html.gsub(/<img.+>/, '').chomp
      datetime = dhead.search("dd").inner_html
      date = datetime.gsub(/\d\d:\d\d/, '')
      time = datetime.gsub(/\d+年\d+月\d+日/, '')

      dbody = dbox.search("//div[@class='txtconfirmArea']")

      ### images
      @imagepages = Array.new
      @imagesrcs  = Array.new
      # pick up images from diary (standard image upload)
      dbody.search("//div[@class='diaryPhoto']//td") do |i|
        @imagepages.push i.at("a")['onclick'].gsub(/MM_openBrWindow\('/, '').gsub(/'.+/, '')
        a = i.at("a")
        @imagesrcs.push a.at("img")['src'].gsub(/.+\//, "").gsub(/s\.jpg/, '.jpg')
      end

      # pick up images from body (from mixi album)
      dbody.search("//div[@id='diary_body']/a").each do |a|
        if onclick = a['onclick']
          @imagepages.push onclick.gsub(/MM_openBrWindow\('/,'').gsub(/'.+/,'')
          # rewrite image uri to large image filename
          imgfile = a.at("img")['src'].gsub(/.+\//, "").gsub(/s\.jpg/, '.jpg')
          contentid = imgfile.gsub(/_/, ".") + "@" + CID_SUFFIX
          a.at("img")['src'] = "cid:#{contentid}"
        end
      end

      ### journal body
      body = dbody.search("//div[@id='diary_body']").inner_html
             
      unless title.empty?
        enc = ENV['LANG'] =~ /euc-jp/i ? 'e' : 'w'
        puts NKF.nkf("-#{enc}", "  + #{author}: #{title} (#{date} #{time})")
      else
        puts "  + failed to parse diary of #{author}"
#        puts "=================================================="
#        puts doc.inner_html
#        puts "=================================================="
        return false
      end

      @author = author
      @title  = title
      @body   = body
      @datetime = mkdatetime(date, time)

      #### comments
      doc.search("//div[@class='diaryCommentbox']").each do |c|
        author   = c.search("span[@class='commentTitleName']/a").inner_html
        url      = c.at("span[@class='commentTitleName']/a")['href']
        datetime = c.search("span[@class='commentTitleDate']").inner_html
        body     = c.search("dd").inner_html
        @comments.push(DiaryComment.new(author, mkdatetime(date,time), url, body))
      end

      # pickup last comment (different class)
      doc.search("//div[@class='diaryCommentboxLast']").each do |c|
        author   = c.search("span[@class='commentTitleName']/a").inner_html
        url      = c.at("span[@class='commentTitleName']/a")['href']
        datetime = c.search("span[@class='commentTitleDate']").inner_html
        body     = c.search("dd").inner_html
        @comments.push(DiaryComment.new(author, mkdatetime(date,time), url, body))
      end

      return true
    end

    def html
      output = '<html><body>'
      output << "<h1><a href=\"http://mixi.jp/#{@url}\">#{@title}</a></h1> by #{@author} at #{@datetime.strftime('%Y-%m-%d %H:%M')}<br /><br />"
      @imagesrcs.each do |src|
        contentid = src.gsub(/_/, ".") + "@" + CID_SUFFIX
        output << "<img src=\"cid:#{contentid}\" alt=\"[IMAGE]\" /> "
      end
      output << "<p>#{@body}</p><hr />"
      @comments.each do |comment|
        output << "<h3>#{comment.author} at #{comment.datetime.strftime('%Y-%m-%d %H:%M')}</h3>#{comment.body}<hr />"
      end
      output << '</body></html>'
      output
    end
  end

  class DiaryComment
    attr_accessor :author, :datetime, :url, :body
    def initialize(author, datetime, url, body)
      @author, @datetime, @url, @body = author, datetime, url, body
    end
  end

  class DiaryFetcher
    attr_accessor :username, :password, :lastidfile, :logfile, :debug,
                  :excludelist, :wait
    attr_reader :lastid
    def initialize(username, password)
      @username, @password = username, password
      @logfile = ENV['HOME'] + '/.mixi/' + 'mixi-access.log'
      @lastidfile = ENV['HOME'] + '/.mixi/' + 'lastid.log'
      @excludelist = Array.new
      @wait  = 3
      login
      @lastid = File.open(@lastidfile).read.to_i rescue 0
      @maxid  = @lastid
      @debug = true
    end

    def login
      @agent = WWW::Mechanize.new {|a| a.log = Logger.new(@logfile) }
      @agent.user_agent_alias = 'Windows Mozilla'
      if ENV['HTTP_PROXY'] then
        host, port = ENV['HTTP_PROXY'].split( /:/, 2 )
        port = (port || 8080).to_i
        @agent.set_proxy(host, port)
      end
      puts "access to http://mixi.jp as #{@agent.user_agent}"
      page = @agent.get('http://mixi.jp/')
#      page = @agent.get('mixi.html')
      form = page.forms[0]
#      form.fields.find {|f| p f}
      form.fields.find {|f| f.name == 'email'}.value = @username
      form.fields.find {|f| f.name == 'password'}.value = @password
      form.fields.find {|f| f.name == 'next_url'}.value = '/home.pl'
      page = @agent.submit(form, form.buttons.first)
      puts "login to mixi as #{@username}." if @debug
      if /url=([^"])"/ =~ page.body
        link = 'http://mixi.jp' + $1.to_s
        @agent.get(link)
      end
      sleep @wait if @wait > 0
    end

    def fetch_user_diary ownerid
      listurl = "http://mixi.jp/list_diary.pl?id=#{ownerid}"
      page = @agent.get(listurl)
      diary = Array.new
      page.links.each do |link|
        if /続きを読む/ =~ link.node.inner_html &&
            /view_diary.pl\?id=(\d+)&owner_id=(\d+)/ =~ link.href
          diaryid = $1.to_i
          @maxid = diaryid if @maxid < diaryid
          if diaryid <= @lastid
#            puts "pre-fetched diary: #{link.href}" if @debug
            next
          end
          sleep @wait if @wait > 0
          mixidiary = Diary.new
          mixidiary.url = link.href
          mixidiary.parse(@agent.get(link.href))
          diary.push mixidiary
        end
      end
      return diary
    end

    def fetch_friend_diary
      listurl = 'http://mixi.jp/new_friend_diary.pl'
      page = @agent.get(listurl)

      list = Array.new
      page.body.gsub(/view_diary.pl\?id=(\d+)&owner_id=(\d+)/) do |href|
        diaryid, ownerid = $1.to_i, $2.to_i
        @maxid = diaryid if @maxid < diaryid
        if @lastid.to_i >= diaryid
#          puts "pre-fetched diary: #{href}" if @debug
          next
        end
        if @excludelist.include? ownerid.to_s # to_s is temporaly...
          puts "skip #{$1} because of exclude list" if @debug
          next
        end
        list.push ownerid
      end

      diary = Array.new
      list.uniq.reverse.each do |ownerid|
        sleep @wait if @wait > 0
        puts "fetching diary: id=#{ownerid}" if @debug
        userdiary = fetch_user_diary(ownerid)
				diary << userdiary unless userdiary.empty?
      end
      return diary.flatten
    end

    def fetch_image imagepage_uri
      return if imagepage_uri.empty?
#      puts "       + try to get #{imagepage_uri}"
      image, src = nil, ''
      page = @agent.get("http://mixi.jp/" + imagepage_uri)
      page.body.gsub(/<img src="(http:\/\/[^"]+\.jpg)"[^>]*>/i) {
        src = $1
        puts "    -> fetch image: #{src}"
        image = @agent.get(src).body
      }
      return image, src
    end

    def close
      File.open(@lastidfile, 'w') do |file|
        file.write(@maxid)
      end
    end
  end
end

def send_mail(text, subject, datetime, 
              mail_from, recipients, smtpserver,
              images)
  header = ''
  imagebody = ''
    boundary = '__mixi2gmail_123456789ABCDEFGHIJ__'
    header << "Content-Type: multipart/mixed; boundary=\"#{boundary}\""
    content = ''
    tmp = text
    text = "--#{boundary}\n"
    text << "Content-Type: text/html;\n\n"
    text << tmp
    text << "\n\n"

  unless images.empty?
    images.each do |src, image|
      filename = src.gsub(/.+\//, "")
#      filename = Time.now.to_i.to_s + '_' + rand(1000).to_i.to_s + '.jpg'
      contentid = filename.gsub(/_/, ".") + "@" + CID_SUFFIX
      content = "--#{boundary}\n"
#      content << "Content-Type: application/octed-stream;\n"
      content << "Content-Type: image/jpeg;\n"
      content << " name=\"#{filename}\"\n"
      content << "Content-ID: <#{contentid}>\n"
      content << "Content-Transfer-Encoding: base64\n"
      content << "Content-Disposition: inline;\n"
      content << " filename=\"#{filename}\"\n\n"
      content << [image].pack('m')
      content << "\n\n"
      imagebody << content
    end
    imagebody << "--#{boundary}\n"
  end
  text << imagebody

  message = "From: #{mail_from}
To: #{recipients}
Subject: #{NKF::nkf('-M', subject)}
Date: #{datetime.rfc2822}
MIME-Version: 1.0
X-ML-Name: mixi
Content-Type: text/plain; charset=\"iso-2022-jp\" 
Content-Transfer-Encoding: 7bit
#{header}

#{NKF.nkf('-j',text)}
"
   Net::SMTP.start(smtpserver, 25) do |smtp|
     smtp.send_message(message, mail_from, recipients)
   end

end

rcfile = "#{ENV['HOME']}/.mixi/mixirc"
config = Mixi::Config.new(rcfile)
unless config
	puts "configuration file #{rcfile} not found."
	exit
end

agent = Mixi::DiaryFetcher.new(config.config['username'], 
                               config.config['password'])
agent.debug = true
agent.excludelist = config.config['excludelist'] || []
agent.fetch_friend_diary.reverse.each do |diary|
	begin
    images = Hash.new
    diary.imagepages.each do |imagepage|
      image, src = agent.fetch_image(imagepage)
      images[src] = image
      sleep 2
    end
    send_mail(diary.html, diary.author+'さんの日記', diary.datetime,
            config.config['mailfrom'], config.config['mailaddress'],
            config.config['smtpserver'],
            images)
	rescue
		puts "Error: #{diary.url if diary.url} => #{$!}:#{$@}"
		p diary
	end
end

(config.config['watchlist'] || []).each do |ownerid|
  agent.fetch_user_diary(ownerid).each do |diary|
	  begin
      images = Hash.new
      diary.imagepages.each do |imagepage|
        image, src = agent.fetch_image(imagepage)
        images[src] = image
        sleep 2
      end
      send_mail(diary.html, diary.author+'さんの日記', diary.datetime,
                config.config['mailfrom'], config.config['mailaddress'],
                config.config['smtpserver'],
                images)
	  rescue
	  	puts "Error: #{diary.url if diary.url} => #{$!}:#{$@}"
	  end
  end
end
agent.close