# -*- coding: utf-8 -*-

# = Config example
#
# - module: Web::GetEntries
#   config: 
#     url: http://exmaple.com/
#     extract_xpath:
#       title: {xpath}
#       link: {xpath}
#       description: {xpath}
#       date: {xpath}
#
# extract_xpath は必須ではなく、
# 1. まず url だけを指定して取得する
# 2. それで取得できない場合、または取得できたものでは不都合な場合は
#    extract_xpath をユーザが定義する。
#
# = 備考
# title, date などが取得できない・不正な場合は
# nil を返し、呼び出し元で処理する。

require "pp"
require 'yapra/plugin/mechanize_base'

module Yapra::Plugin::Feed
  class Extract < Yapra::Plugin::MechanizeBase

    def initialize
      super

      @xconfig_for_hatom = {
        "capture"     => "//body",
        "split"       => xpath_attr_contains("class", "hentry"),
        "title"       => xpath_attr_contains("class", "entry-title"),
        "link"        => xpath_attr_contains("rel",   "bookmark"),
        "description" => xpath_attr_contains("class", "entry-content"),
        "date"        => xpath_attr_contains("class", "published")
      }
    end
    
    # Returns XPath that explains {attr} contains {value}.
    def xpath_attr_contains(attr, value)
      ".//*[contains(concat(' ',normalize-space(@#{attr}),' '), ' #{value} ')]"
    end

    def get_title(element, xpath)
      if element.xpath(xpath).size > 0
        element.xpath(xpath)[0].text.strip
      else
        nil
      end
    end

    def get_link(element, xpath, url)
      href = element.xpath( xpath )[0].attr("href")
      unless /^http/ =~ href
        url_base = url.split("/")[0..2].join("/")
        href = url_base + href
      end
      href
    end

    def get_description(element, xpath)
      if xpath
        element.xpath(xpath)[0]
      else
        element
      end
    end
    
    def get_date(element, xpath)
      if xpath && element.xpath(xpath).size > 0
        date_str = element.xpath(xpath)[0].text
        begin
          Time.parse( Date.parse(date_str).to_s )
        rescue ArgumentError
          nil
        end
      else
        nil
      end
    end
    
    def get_entries(page, config)
      xconfig = config["extract_xpath"] ? config["extract_xpath"] : @xconfig_for_hatom
      capture = xconfig['capture']

      root = capture ? page.root.xpath(capture) : page.root
      root.xpath(xconfig['split'])
    end

    def run(data)
      url = config["url"]
      page = agent.get(url)

      xconfig = if config["extract_xpath"]
                  config["extract_xpath"]
                else
                  @xconfig_for_hatom
                end

      get_entries(page, config).map{|entry|
        item = RSS::RDF::Item.new
        item.title       = get_title(entry, xconfig["title"])
        item.link        = get_link(entry, xconfig["link"], url)
        item.description = get_description(entry, xconfig["description"])
        item.date        = get_date(entry, xconfig["date"])
        item
      }
    end
  end
end