tumblr-2-epub

index.coffee

tumblr = require 'tumblr'
cheerio = require 'cheerio'

Tumblr2Peepub = (tumblrConfig) ->
  that = this
  this.tumblrConfig = tumblrConfig

  allPosts   = []
  totalPosts = false
  bookDoc    = {}

  # fetch all posts recursively
  fetchPosts = (cb, offset) ->
    setTimeout ->
      if !totalPosts || totalPosts > allPosts.length
        that.blog.posts { offset : offset || 0 }, (err, res) ->
          return cb err if err

          bookDoc = 
            title       : res.blog.title
            url         : 'http://' + that.blogUrl
            subtitle    : res.blog.description
            description : res.blog.description

          totalPosts = res.total_posts
          allPosts = allPosts.concat res.posts
          fetchPosts cb, allPosts.length
      else
        cb null, allPosts
    , 500

  this.fetch = (tumblrPrefix, cb) ->
    this.blogUrl = tumblrPrefix + '.tumblr.com'
    this.blog = new tumblr.Blog this.blogUrl, this.tumblrConfig

    allPosts   = []
    totalPosts = false
    bookDoc    = {}

    fetchPosts (err, posts) ->
      return cb err if err

      photos = posts.filter((p) -> p.photos).map (p) ->
        p.photos[0].original_size.url

      # first image as cover, there's a conflict with first image inside the book
      bookDoc.cover = photos[(photos.length-1)];

      # description
      bookDoc.description = "Tumblr archived using http://streambooks.thepeoplesebook.net/"

      bookDoc.pages = posts
      .map (p) ->
        # possibly use a function for those
        # post title
        if p.title
          ptitle = "<h1>" + p.title + "</h1>"
        else
          ptitle = ""
        # post source
        if p.source_url
          psource = "<p>Source: <a href=\"" + p.source_url + "\">" + p.source_title + "</a></p>"
        else
          psource = ""
        # post body
        if p.body
          pbody = p.body
        else
          pbody = ""
        # post quote
        if p.text
          ptext = "<h1><blockquote>“" + p.text + "”</blockquote></h1>"
        else 
          ptext = ""
        if p.source
          pquotesrc = "<p>— " + p.source + "</p>"
        else 
          pquotesrc = ""
        # post link
        if p.url
          if p.title
            ptitle = "<h1><a href=\"" + p.url + "\">" + p.title + "</a></h1>"
          else
            ptitle = "<h1><a href=\"" + p.url + "\">" + p.url + "</a></h1>"
        if p.description
          pdesc = p.description
        else 
          pdesc = ""
        # post video
        if p.player
          # sometimes "embed_code" is false
          if p.player[0]["embed_code"]
            # get embed code
            embed_code = p.player[0]["embed_code"]
            # get url from embed_code
            $ = cheerio.load(embed_code)
            video_url = $('iframe').attr('src')
            # remove // for vimeo, etc.
            if video_url
              if video_url.indexOf('//') == 0
                video_url = video_url.replace('//','http://')
              # get domain name
              video_domain = video_url.replace('http://','').replace('https://','').split(/[/?#]/)[0]
              if video_domain == "" 
                video_domain = video_url
              pvideo = "<h1><p>Video: <a href=\"" + video_url + "\">" + video_domain + "</a></p></h1>"
          else
            pvideo = ""
        else
          pvideo = ""
        # post photo
        if p.photos
          pphotos = (p.photos.map((photo) -> "<p><img src=\"" + photo.original_size.url + "\" /></p>").join(''))
        else
          pphotos = ""
        if p.caption
          pcap = p.caption 
        else 
          pcap = ""
        # create titles (use date)
        title = p.date.replace(" GMT", "");
        # create the page body
        body = ptitle + ptext + pphotos + pvideo + pcap + pbody + pquotesrc + pdesc + psource
        # remove iframes from body
        exp = /<iframe.+<\/iframe>/g
        regex = new RegExp(exp)
        body = body.replace(regex, '')
        return {
          # create title
          title : title
          body : body
          toc : true
        }
      .filter (p) ->
        return p.body?
      cb null, bookDoc

  return this
  

  
module.exports = Tumblr2Peepub