elixir_feed_parser

defmodule ElixirFeedParser.Parsers.RSS2 do
  import ElixirFeedParser.Parsers.Helper
  alias ElixirFeedParser.XmlNode

  @date_format "RFC_1123"

  def can_parse?(xml) do
    xml
    |> XmlNode.find("/rss")
    |> XmlNode.attr("version") == "2.0"
  end

  def parse(xml) do
    feed = XmlNode.find(xml, "/rss/channel")

    link      = feed |> element("link")
    atom_link = feed |> element("atom:link[@rel='self']", [attr: "href"])
    url       = link || atom_link

    %{
      title:                 feed |> element("title"),
      description:           feed |> element("description"),
      url:                   url,
      hubs:                  [feed |> element("atom:link[@rel='self']", [attr: "href"])],
      "atom:link":           atom_link,
      "rss2:link":           feed |> element("link"),
      language:              feed |> element("language"),
      copyright:             feed |> element("copyright"),
      "rss2:managingEditor": feed |> element("managingEditor"),
      "rss2:webMaster":      feed |> element("webMaster"),
      # TODO: also work with pubdate or publicationDate
      updated:               feed |> element("pubDate") |> to_date_time(@date_format),
      "rss2:pubDate":        feed |> element("pubDate") |> to_date_time(@date_format),
      "rss2:lastBuildDate":  feed |> element("lastBuildDate") |> to_date_time(@date_format),
      categories:            feed |> elements("category"),
      generator:             feed |> element("generator"),
      "rss2:ttl":            feed |> element("ttl"),
      # TODO: integer
      skip_hours:            feed |> elements("skipHours/hour"),
      "rss2:skipHours":      feed |> elements("skipHours/hour"),
      skip_days:             feed |> elements("skipDays/day"),
      "rss2:skipDays":       feed |> elements("skipDays/day"),

      image:                 feed |> XmlNode.find("image") |> parse_image,
      entries:               parse_entries(feed)
    }
  end

  def parse_entry(_feed, entry) do
    author     = entry |> element("author")
    dc_creator = entry |> element("dc:creator")

    %{
      title:             entry |> element("title"),
      url:               entry |> element("link"),
      "rss2:link":       entry |> element("link"),
      description:       entry |> element("description"),
      author:            author || dc_creator,
      "rss2:dc:creator": entry |> element("dc:creator"),
      categories:        entry |> elements("category"),
      # support dc:identifier too
      id:                entry |> element("guid"),
      "rss2:guid":       entry |> element("guid"),
      comments:          entry |> element("comments"),
      # TODO: also work with pubdate or publicationDate, dc:date, dc:Date, dcterms:created
      updated:           entry |> element("pubDate") |> to_date_time(@date_format),
      "rss2:pubDate":    entry |> element("pubDate") |> to_date_time(@date_format),
      source:            entry |> element("source", [attr: "url"]),
      content:           entry |> element("content:encoded"),
      enclosure:         entry |> XmlNode.find("enclosure") |> parse_enclosure
    }
  end

  defp parse_image(nil), do: nil
  defp parse_image(image) do
    %{
      title:       image |> element("title"),
      link:        image |> element("link"),
      url:         image |> element("url"),
      width:       image |> element("width"),
      height:      image |> element("height"),
      description: image |> element("description")
    }
  end

  defp parse_entries(feed) do
    XmlNode.map_children(feed, "item", fn(e) -> parse_entry(feed, e) end)
  end

  defp parse_enclosure(nil), do: nil
  defp parse_enclosure(enclosure) do
    %{
      url:    enclosure |> XmlNode.attr("url"),
      length: enclosure |> XmlNode.attr("length"),
      type:   enclosure |> XmlNode.attr("type")
    }
  end
end