はてなブログをtweetから探すスクリプト

twitter_hatena.rb

require "rubygems"
require "bundler/setup"
require "twitter"
require 'net/http'
require 'uri'
require "json"
require 'rss'
require "sqlite3"
require "time"

$subscribes = 10
$sqlitePath = "xxx";

$twClient = Twitter::REST::Client.new do |config|
    config.consumer_key        = "xxxx"
    config.consumer_secret     = "xxxx"
    config.access_token        = "xxxx"
    config.access_token_secret = "xxxx"
end

def expand_url(url)
    begin
      response = Net::HTTP.get_response(URI.parse(url))
    rescue
      return url
    end
    case response
    when Net::HTTPRedirection
      expand_url(response['location'])
    else
      url
    end
end

def get_hatena_blog_info(url)
    uri = URI.parse("http://blog.hatena.ne.jp/api/init?blog=" + url)
    req = Net::HTTP::Get.new(uri.request_uri)
    req['X-Requested-With'] = 'XMLHttpRequest'

    res = Net::HTTP.start(uri.host, uri.port) do |http|
        http.request(req)
    end

    return JSON.parse(res.body)
end

$db = SQLite3::Database.new($sqlitePath);
result = $db.execute("select count(*) from sqlite_master where type='table' and name='blog';");

if result.first.first == 0
	$db.execute("create table blog (id integer primary key , title text , url text unique );");
end


blog_result = []
$twClient.search("#はてなブログ", result_type: "recent").take(100).each{ | tweet |
    
    if m = tweet.text.match( %r{(https://t.co/[a-zA-Z0-9…]+)})
        if m.length >= 2
            url = m[1]
            if url.empty? == false && url.include?("…") == false
                url =  expand_url(url)
                blogInfo = get_hatena_blog_info(url)
                if blogInfo["subscribes"].to_i >= $subscribes
                    blog_name = JSON.parse( %Q{["#{blogInfo["blog_name"]}"]} )[0]

                    result = $db.execute( "select count(*) from blog where url = ?;" , blogInfo["blog_url"] );
                    if result.first.first == 0
                        blog_result << {
                            "blog_name" => blog_name ,
                            "entry_url" => url ,
                            "blog_url" => blogInfo["blog_url"] ,
                            "subscribes" => blogInfo["subscribes"] ,
                            "twurl" => tweet.uri.to_s
                        }

                        sql = "insert into blog (title,url) values (?,?);"
			            $db.execute( sql , blog_name , blogInfo["blog_url"] );
                    end
                end
            end
        end
    end
}

$db.close

exit(0) if blog_result.empty?

feed = RSS::Maker.make("2.0") do |feed|
    feed.channel.title = "hatena bloger"
    feed.channel.link = "https://nasust.hatenablog.com"
    feed.channel.about = "http://xxxx/feed.xml"
    feed.channel.description = "twitter hatena bloger"
    feed.channel.updated = Time.now

    blog_result.each do | blog |
        feed.items.new_item do |item|
            item.title = blog["blog_name"]
            item.link = blog["blog_url"]
            item.description = "subscribes:" + blog["subscribes"] + " " +  blog["twurl"] 
        end
    end
end


File.open("xxx/feed.xml", 'w') { |f| f.puts(feed) }