#!/usr/bin/env ruby
# bk1.rb
# ISBN-search -> detail-parse
# original; isbn2bib.rb for bk1 by takahashi
# コミック等が対照となる場合、通常と微妙にフォーマットが変わる為、幾つかのデータを拾い損ねる場合有り。未対応。
# ver.0.5 2003.12.20 書誌情報HTMLの形式が変更？　それに追応。
# ver.0.6 2002.3.21-2004.01.07

# 内容情報は書評の一環としてrevidが与えられてる。別途読み直し要

#デイリー/ウィークリーランキング+rss,オススメ？　以上はまだ
#simageはコメントアウト中（一度アクセスして存在確認してから組むべきか？）

# reviewの形で「内容紹介」があった場合、
# その内容をdesc(文)にし、more.gifがあれば文に(more)を付し、review_numを減ず。
# →本来のdescとの衝突確認まだ。
# 以上ToDo

if __FILE__ == $0
  require 'cookie'
  require 'webagent'
  require 'cache'
  require 'cachebib'
end

class Review_bk1
  attr_accessor :url, :info, :review_id
  attr_accessor :title, :reviewer_title, :reviewer, :date, :point, :content
  attr_accessor :insystemKcode, :outputKcode, :messageKcode
  
  def initialize ()
    @insystemKcode = 3 #sjis #1-jis,2-euc,3-sjis
    @outputKcode   = 2 #euc
    @messagetKcode = 3 #sjis
  end
  
  def parse_review (str)
    #----- set key and data ------
    mode_tbl = {
      '<!--書評タイトル-->' => :title,
      '<!--書評者肩書-->'   => :reviewer_title,
      '<!--書評者名-->'     => :reviewer,
      '<!--書評日-->'       => :date,
      '<!--読者評価値-->'   => :point,
      '<!--書評-->'         => :content,
      '<!--読者書評なら賛同ボタン表示-->' => :none
    }
    mode = :none
    data = Hash.new()
    
    #----- read and sepalate by key(mode_tbl) ------
    str.each{|line|
      l = line.chomp.chomp("\r").kconv(@insystemKcode)
      l.strip!
      if mode_tbl[l]
        mode = mode_tbl[l]
      end
      if mode != :none
        if data[mode]
          data[mode] << line
        else
          data[mode] = line
        end
      end
    }
    
    #----- parse -----
    title           = data[:title].to_s.gsub(/<[^>]+>/,'').strip
    reviewer_title  = data[:reviewer_title].to_s.gsub(/<[^>]+>/,'').strip
    reviewer        = data[:reviewer].to_s.gsub(/<[^>]+>/,'').strip
    date            = data[:date].to_s.gsub(/<[^>]+>/,'').strip
    point           = data[:point].to_s.gsub(/<[^>]+>/,'').sub(/評価（★マーク）/s,'').sub(/　/s,'').strip
    content         = data[:content].to_s.gsub(/<[^>]+>/,'').strip
    content.gsub!(/\n/,"\t")
    @title          = title          if title          != ''
    @reviewer_title = reviewer_title if reviewer_title != ''
    @reviewer       = reviewer       if reviewer       != ''
    @date           = date           if date           != ''
    @point          = point          if point          != ''
    @content        = content        if content        != ''
  end
  
  def get_review ()
    agent = WebAgent.new()
    agent.uri = @url
    begin
      agent.get()
    rescue
      return false #timeout
    end
    body = agent.body.kconv(@insystemKcode)
    parse_review(body)
    
    @info = {}
    @info["title"]          = @title          unless @title.nil?
    @info["reviewer_title"] = @reviewer_title unless @reviewer_title.nil?
    @info["reviewer"]       = @reviewer       unless @reviewer.nil?
    @info["date"]           = @date           unless @date.nil?
    @info["point"]          = @point          unless @point.nil?
    @info["content"]        = @content        unless @content.nil?
    @info["review_id"]      = @review_id      unless @review_id.nil?
                             #@review_idはCacheBib_bk1内で与えられる
    return true
  end
end

module CacheBib_bk1_xml
#  alias orig_url_bibinfo   url_bibinfo   if defined?(url_bibinfo)
#  alias orig_parse_bibinfo parse_bibinfo if defined?(parse_bibinfo)
  
  #bk1.jp
  def url_bibinfo ( id = @id )
    id = '' if id.nil?
    url = "http://bk1.jp/#{id[0..3]}/#{id}.html"
    url = "http://breeder.bk1.jp/rd/#{id}/#{@affiliate_code}/noentry" if !@affiliate_code.nil?
    return url
 end
 
 def url_bibinfo_xml ( id = @id )
    url = "http://cgi.bk1.jp/xml.cgi?"
    url << "bibid=#{id}"
    url << "&aid=#{@affiliate_code}" if !@affiliate_code.nil?
    message("CHK; access-url; #{url}")
    return url
  end
  
  def parse_bibinfo_xml ( str ) #正規XMLパーサの組込の方がいいか？
    data = {}               #それとも組込まなくてもいい利便性を採るか？
    if str =~ %r|<Detail url="(.+?)"|
      data['detail'] = $1
    end
    str.scan(%r|<([^>]+?)>([^<]+?)</[^>]+?>|){
      data[$1] = $2
    }
#    @id        = data['BIBID'] #既存のはずなのでオミット
    @title     = data['ProductName']
    @label     = data['SeriesName']
    @author    = data['Authors']
    @publisher = data['Manufacturer']
    @isbn      = data['ISBN']
    @delivery  = data['Availability']
    @simage    = data['ImageUrlSmall']
    @image     = data['ImageUrlLarge']
    @price     = data['OurPrice']
    @pubdate   = data['DateOfIssue']
  end
  
  def get_info
    unless @isbn.nil?
      message("CHK; access; #{@url_site} -> ISBN:#{@isbn}")
    else
      message("CHK; access; #{@url_site} -> ID:#{@id}")
    end
    #----- get detail-code -----
    if !@id
      @id = @isbn.dup
    end
    #----- get book-info -----
    agent = WebAgent.new()
    agent.uri = url_bibinfo_xml
    begin
      agent.get()
    rescue
      message("WAR; access-timeout.")
      return
    end
    #body = agent.body.kconv(@insystemKcode)
    body = agent.body #UTF対策。ここだけ標準から変更
    begin
      body = Iconv::iconv("UTF-8", "EUC-JP", body)
    rescue NameError
      begin
        body = Uconv::u8toeuc(body)
      rescue
      end
    end
    body = body.kconv(@insystemKcode)
    parse_bibinfo_xml(body)
    #----- reget book-info -----
    if @id.to_s == '' || @id == '02397767' #携帯ストラップはエラーとみなす
      agent = WebAgent.new()
      agent.uri = url_isbnsearch
      begin
        agent.get()
      rescue
        message("WAR; access-timeout.")
        return
      end
      body = agent.body.kconv(@insystemKcode)
      parse_isbnsearch(body)
      agent.uri = url_bibinfo
      begin
        agent.get()
      rescue
        message("WAR; access-timeout.")
        return
      end
      body = agent.body.kconv(@insystemKcode)
      parse_bibinfo(body)
    end
    #----- get book-image -----
    if !@image.nil? && @flag_image
#      filename = @dir_image + File.basename(@image)
      isbn = @isbn
#      ext = File.extname(@image) # for ruby1.7-
      ext = '.jpg'
      if @image =~ /(\.+?)$/
        ext = $1.strip
      end
      filename = @dir_image + 'isbn' + isbn.to_s.delete('-').downcase + ext
      urlname  = @url_image + 'isbn' + isbn.to_s.delete('-').downcase + ext
      if @isbn.to_s == '' && !@id.nil?
        filename = @dir_image + 'id' + @id + ext
        urlname  = @url_image + 'id' + @id + ext
      end
      unless File.exist?(filename)
        message("CHK; download-image; #{@image} -> #{filename}")
        agent = WebAgent.new()
        agent.uri = @image
        agent.get()
        open(filename, 'wb'){ |fp|
          fp.write(agent.body)
          fp.flush
        }
        @image_local = urlname
      else
        @image_local = urlname
      end
    end
    if !@image_local.nil?
      isbn = @isbn
      ext = '.jpg'
      if @image_local =~ /(\.+?)$/
        ext = $1.strip
      end
      filename = @dir_image + 'isbn' + isbn.to_s.delete('-').downcase + ext
      if @isbn.to_s == '' && !@id.nil?
        filename = @dir_image + 'id' + @id + ext
      end
      if File.exist?(filename)
        begin
          require 'imagesize'
        rescue LoadError
          require 'lib/imagesize'
        end
        w, h = ImageSize::size(filename)
        @image_width, @image_height = w, h unless w.zero?
      end
    end
    set_info
  end
end

class CacheBib_bk1 < CacheBib
  begin
    require 'iconv'
  rescue LoadError
    begin
      require 'uconv'
    rescue LoadError
    end
  end
  attr_accessor :chiefeditor, :author_intro, :user, :okidoki, :simage
  attr_accessor :author_id, :translator_id, :chiefeditor_id
  attr_accessor :url_today_recommend
  
  def initialize ()
    super
    @shop_id = 'bk1'
    @url_site = 'http://www.bk1.co.jp/'
    @affiliate_code = 'p-ippo57167'
    @reviews = Array.new()
    @flag_review = true
  end
  
  #===== ===== parts ===== =====
  def word_delivery ()
    delivery = ''
    flag = false
    if !@title.nil? # title-check -> no-info / get-info check
      flag = true
      if @delivery.nil?
        delivery = "ない"
      else
        if @delivery =~ /24/
          delivery = "24時間"
        elsif @delivery =~ /3/
          delivery = "2〜3日"
        elsif @delivery =~ /取/
          delivery = "取寄せ"
        else
          delivery = "？"
        end
      end
    else
      message("WAR; maybe data-error, no-title.")
    end
    return delivery, flag
  end
  
  #===== ===== url ===== =====
  def url_isbnsearch ( isbn = @isbn )
    url = @url_site + "cgi-bin/srch/srch_result_book.cgi/?"
    url << "isbn=#{isbn}"
    url << "&aid=#{@affiliate_code}" if !@affiliate_code.nil?
    url << "&idx=3"
    return url
  end
  
  def url_bibinfo ( id = @id )
    url = @url_site + "cgi-bin/srch/srch_detail.cgi/?"
    url << "bibid=#{id}"
    url << "&aid=#{@affiliate_code}" if !@affiliate_code.nil?
    url << "&idx=3"
    return url
  end
  
  def url_review ( id = @id, revid = @revid )
    url = @url_site + "cgi-bin/srch/srch_rev.cgi/?"
    url << "bibid=#{id}&volno=#{@volno}&revid=#{revid}"
    url << "&aid=#{@affiliate_code}" if !@affiliate_code.nil?
    return url
  end
  
  def url_authorinfo ( author_id = @author_id )
    url = @url_site + "cgi-bin/srch/srch_result_book.cgi/?"
    url << "auid=#{author_id}"
    url << "&aid=#{@affiliate_code}" if !@affiliate_code.nil?
    url << "&srch=4"
    return url
  end
  
  def url_top
    url = @url_site + "cgi-bin/srch/srch_top.cgi/?"
    url << "aid=#{@affiliate_code}" if !@affiliate_code.nil?
    return url
  end
  
  #キーワード検索主体。そのうち別途詳細検索用を用意した方がいいか？
  def url_search ( srch = 1, kywd = nil, s1 = nil, dp = nil, pg = nil, aid = @affiliate_code, codes = {} )
#検索モードにおけるFORM値
#srch (1=キーワード検索,2=詳細検索,5=書評検索)
#キーワード検索(srch=1)
#kywd キーワード
#詳細検索(srch=2)
#(http://www.bk1.co.jp/cgi-bin/srch/srch_search.cgi)
#ol キーワード
#ti タイトル
#au 著者
#pb 出版社
#isbn ISBN
#idx 発送可能時間(0,1,2,3:24時間以内,2〜3日以内,お取り寄せ,全書籍)
#gu ジャンル・フィーリング('',...)
#"01010000"文芸・ミステリ・ＳＦ・ホラー...)
#"01040000"人文・社会・ノンフィクション
#"01080000"科学・技術・医学・建築
#"01060000"ビジネス・経済・法律・資格
#"01070000"コンピュータ・インターネット
#"01050000"教育・福祉・語学・辞書
#"01030000"暮らし・実用・旅行
#"01020000"芸能・娯楽・スポーツ
#"01020900"美術・デザイン・写真
#"01090000"児童書・絵本
#"01100000"マンガ
#"01110000"文庫・新書
#"01120000"ビデオ・ＤＶＤ・限定本ほか -->
#"02010000"ホッとひと息つける本
#"02020000"感動する本
#"02030000"元気が出る本
#"02040000"ドキドキする本
#"02050000"幸せな気分になれる本
#"02060000"頑張りたい気分になれる本
#s1 並べ替え条件(''=一致度順,...)
#"za"発送時間順
#"dd"発行日順
#"da"発行日逆順
#"ta"タイトル50音順
#"td"タイトル50音逆順
#"aa"著者名50音順
#"ad"著者名50音逆順
#"pa"出版社50音順
#"pd"出版社50音逆順
#"rd"価格（高〜）順
#"ra"価格（低〜）順
#dp 表示件数(10,25,50)
#pby 出版日(yyyy or yyyymm or yyyymmdd)
#pbrg 出版日条件(0,1,2:以前,以後,頃)
#age 対象('',...)
#"AA"幼児
#"BB"小学生
#"CC"小学生1-2
#"DD"小学生3-4
#"EE"小学生5-6
#"FF"中学生
#"GG"高校生
    if !kywd.nil?
      kywd = CGI::escape(kywd.toeuc)
    end
    url = @url_site + "cgi-bin/srch/srch_result_book.cgi/?"
    url << "aid=#{aid}"    #affiliate-code
    url << "&srch=#{srch}" if !srch.nil? #search-mode(num=1=:keyword)
    url << "&kywd=#{kywd}" if !kywd.nil? #キーワード(string)(srch=1時)
    url << "&s1=#{s1}"     if !s1.nil?   #並べ替え条件(code)(''=一致度順)
    url << "&dp=#{dp}"     if !dp.nil?   #表示件数(num=10)(10,25,50)
    url << "&pg=#{pg}"     if !pg.nil?   #page(num)
    sort, ol, ti, au, pb, isbn, idx, gu, pby, pbrg, age, st = nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil
    codes.each{ |key, val| eval("#{key} = #{val}") }
    url << "&sort=#{sort}" if !sort.nil? #sort(string)#検索結果表示時のみ
    url << "&ol=#{ol}"     if !ol.nil?   #キーワード(詳細検索=srch=2時)
    url << "&ti=#{ti}"     if !ti.nil?   #タイトル
    url << "&au=#{au}"     if !au.nil?   #著者
    url << "&pb=#{pb}"     if !pb.nil?   #出版社
    url << "&isbn=#{isbn}" if !isbn.nil? #ISBN
    url << "&idx=#{idx}"   if !idx.nil?  #発送可能時間(0,1,2,3=全書籍)
    url << "&gu=#{gu}"     if !gu.nil?   #ジャンル・フィーリング
    url << "&pby=#{pby}"   if !pby.nil?  #出版日(yyyy or yyyymm or yyyymmdd)
    url << "&pgrg=#{pbrg}" if !pbrg.nil? #出版日条件(0,1,2:以前,以後,頃)
    url << "&age=#{age}"   if !age.nil?  #対象('',...)
    url << "&st=#{st}"     if !st.nil?   #??? おそらくsortと、つまりs1と被る
    return url
  end
  
  #===== ===== parse ===== =====
  def parse_isbnsearch ( str )
    count = []
    str.scan(%r|\s*<a href="/cgi-bin/srch/srch_detail.cgi/.*\?.*&bibid=(\d+)&.*">(.*)</a>|){
      count << [$1, $2] unless $1.nil?
    } #複数みつけた場合最期に見つけたものが有効
    if count.uniq.size > 1
      message("WAR; ISBN-search hit multi-column. ISBN; #{@isbn}")
    end
    @id, @title = count.last
    return count
  end
  
  def parse_bibinfo ( str )
    #----- bibinfo-area ------
    str_pre, str_post = '', ''
    if str =~ %r|<!-- %Cut Start% -->.+<!-- %Cut End% -->|mi
      str_pre, str, str_post = $`, $&, $'
    end
    tables = str.scan(%r|<tr>.*?</tr>|mi)
    
    #----- parse -----
    if tables[0] =~ %r|<img src="(.*?)"|
      image = $1
      if image !~ /no_img/ && image =~ /bookimages/
        @image = @url_site.chomp('/') + image
#        @simage = @image.chomp('.jpg') + 'S.jpg' #存在確実かどうか不明
      end
    end
    
    titles = tables[0].scan(%r|<b>(.*?)</b>|)
    @title = titles[0].first unless titles[0].nil?
    if !titles[-1].nil? && titles[-1].to_s.strip =~ /^\((.*?)\)$/
      @label = $1.strip
      if @label =~ /^(.*?)\s(.*)$/
        @label, @series = $1, $2
      end
    end
#    data[:title] =~ Regexp.new("<b>([^<]*)</b>")
    
    authors = []
    tables[1].scan(%r|<a href="/cgi-bin/srch/srch_result_book.cgi/.*?&auid=(\d+)">(.*?)</a>|){ |code, name|
      authors << [code, name]
    }
    if authors == []
      tables[1].scan(%r|著者：(.*?)<br>|mi){ |name|
      authors << [nil, name.first.strip]
      }
    end
    @author_id = nil if authors.size > 0 #preset_infoしてる分を場合に応じて消去
    authors.each { |auid, name_str|
#      str = line.to_s.gsub(/<[^>]+>/,'').delete("\n").strip
      if name_str =~ /(著$|編$|〔著〕$)/
        name = $`
          name += '編'if name_str =~ /編$/
          if @author.nil? then @author = '' else @author << "\t" end
          @author << name
          if @author_id.nil? then @author_id = '' else @author_id << "\t" end
          @author_id << auid unless auid.nil?
      elsif name_str =~ /訳$/
        name = $`
        if @translator.nil? then @translator = '' else @translator << "\t" end
        @translator << name
        if @translator_id.nil? then @translator_id = '' else @translator_id << "\t" end
        @translator_id << auid
      elsif name_str =~ /監修$/
        name = $`
        if @chiefeditor.nil? then @chiefeditor = '' else @chiefeditor << "\t" end
        @chiefeditor << name
        if @chiefeditor_id.nil? then @chiefeditor_id = '' else @chiefeditor_id << "\t" end
        @chiefeditor_id << auid
      end
    }
    
    if tables[1] =~ %r|本体価格：(.*?)<br>|mi
      @price = $1.strip
    end
    if tables[1] =~ %r|出版：(.*?)<br>|mi
      @publisher = $1.strip
    end
    if tables[1] =~ %r|サイズ：(.*?)<br>|mi
      @pubsize = $1.strip
    end
    if tables[1] =~ %r|ISBN：(.*?)<br>|mi
      isbn = $1.strip
      @isbn = isbn if isbn != ''
    end
    if tables[1] =~ %r|発行年月：(.*?)<br>|mi
      @pubdate = $1.strip
    end
    if tables[1] =~ %r|利用対象：(.*?)<br>|mi
      @user = $1.delete("　").gsub(/\s+/,'').gsub(/<!--.*?-->/,'').strip
    end
    
    if tables[1] =~ %r|発送可能時間.*?<img src="/images/.*?alt="(.*?)"|smi
      @delivery = $1
    end #注：文字コード特定option(SJIS)が無いと駄目な場合あり
    
    if tables[1] =~ %r|OkiDoki対象|
      @okidoki = true
    end
    
    if str =~ %r|内容説明(.+?)<hr|mi
      @desc = $1.gsub(/<[^>]+>/,'').gsub(/\s+/, ' ').strip
#      @desc = desc if desc.size > 0
    end
    
    str_post.scan(%r|著者紹介(.+?)<hr|mi){ |intro|
      intro = $1.gsub(/<[^>]+>/,'').gsub(/\s+/, ' ').strip
      if @author_intro.nil? then @author_intro = '' else @author_intro << "\t" end
      @author_intro << intro
    }
    
    #-----review-----
    reviewlist = [] #タイトル、評者名、（と評者時間は入れるか？）
    str_post.scan(%r|<b>出版社コメント</b>.+?<a href="/cgi-bin/srch/srch_result_rev.cgi/.+?>(.+?)</a>|mi){ |nick|
      reviewlist << '出版社コメント'
      reviewlist << nick
    }
    str_post.scan(%r|<b>内容紹介</b>.+?<a href="/cgi-bin/srch/srch_result_rev.cgi/.+?>(.+?)</a>(.+?)<hr|mi){ |nick, desc|
      reviewlist << '内容紹介'
      reviewlist << nick
      @desc = desc.gsub(/<[^>]+>/, '').gsub(/\s+/, ' ').strip
      @desc += ' (more)' if desc =~ %r|/images/bk1/more.gif|
    }
#    str_post.scan(%r|書評</b>.+?<a href="/cgi-bin/srch/srch_result_rev.cgi/.+?>(.+?)</a>|mi){ |nick|
    str_post.scan(%r|<b>([^<]+?)</b><br>\s*<a href="/cgi-bin/srch/srch_result_rev.cgi/.+?>(.+?)</a>|mi){ |nick|
      reviewlist << nick[0]
      reviewlist << nick[1]
    }
    if reviewlist.size > 0
      @reviewlist = reviewlist.join("\t")
      @review_num = reviewlist.size / 2 #ID, name, ID, name... # (time) or /3
      @review_num -= 1 if reviewlist.include?('内容紹介')
    end
    
    if @flag_review
      str_post.scan(%r|<a href="/cgi-bin/srch/srch_rev.cgi.+?volno=(\d+).+?revid=(\d+)|mi){ |volno, revid|
        @volno, @revid = volno, revid
        review = Review_bk1.new()
        review.url = url_review
        review.review_id = @revid
        if review.get_review
          @reviews << review.info
        else
          message("WAR; access-timeout; review.")
        end
      }
    end
  end
  
  def parse_top ( str )
    str = str.kconv(@insystemKcode)
    urls = []
    str.scan(%r|"(http://ans.bk1.co.jp/cgi-bin/ans/[^\"]+)"|){ urls << $1 }
    @url_today_recommend = urls.first
  end
  
  def parse_today_recommend ( str )
    str = str.kconv(@insystemKcode)
    list = []
    str.scan(%r|<a href="http://ans.bk1.co.jp/cgi-bin/ans/sfw/.+?/(\d\d+)/.+?">(.+?)</a>(.+?)</td>|){ |code, title, author|
      title.strip!
      author.gsub!(/<[^>]+?>/, '').strip!
      if author =~ /^\//
        author = $'.strip
      end
      if author =~ /著$/s
        author = $`.strip
      end
      list << [code, title, author]
    }
    @today_recommend = list
    return @today_recommend
  end
  
  #===== ===== get/set info ===== =====
  def preset_info #===== preset local-value from @info (from cache) =====
    #----- set book-info (from cache) -----
    super
    @author_id = @info['author_id'].split("\t").first unless @info['author_id'].nil?
  end
  
  def set_info #===== set book-info(@info) from local-value =====
    super
    #----- set info add ----- #bk1独自の情報類
    @info["chiefeditor"]      = @chiefeditor
    @info["author_id"]        = @author_id
    @info["translator_id"]    = @translator_id
    @info["chiefeditor_id"]   = @chiefeditor_id
    @info["author_intro"]     = @author_intro
    @info["user"]             = @user          #利用対象
    @info["okidoki"]          = @okidoki       #OkiDoki対象
    @info["simage"]           = @simage        #縮小画像(縦90px) ???
  end
end

#class CacheBib_bk1
#  include CacheBib_bk1_xml
#end

if __FILE__ == $0
#  p CacheBib_bk1.new.get_data('4-15-010739-4')
  book = CacheBib_bk1.new
  book.flag_image = true
#  book.flag_review = false
  p book.get_data('4-15-010739-4')
end