Module: Feedium

Defined in:
lib/feedium.rb,
lib/feedium/version.rb,
lib/feedium/exceptions.rb

Defined Under Namespace

Classes: Error, Request, RequestError

Constant Summary collapse

CONTENT_TYPES =
%w(
  application/x.atom+xml
  application/atom+xml
  application/xml
  text/xml
  application/rss+xml
  application/rdf+xml
  application/json
).freeze
VERSION =
'0.1.15'

Class Method Summary collapse

Class Method Details

.feed?(url) ⇒ Boolean

Returns:

  • (Boolean)


71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
# File 'lib/feedium.rb', line 71

def self.feed?(url)
  @request = Request.new(url, @base_uri)
  @request.send
  meta_type = @request.io.meta['content-type']

  content_type = @request.io.content_type.downcase
  content_type = meta_type.gsub(/;.*$/, '') if content_type == 'application/octet-stream' && !meta_type.nil?

  is_feed = CONTENT_TYPES.include?(content_type) ? true : false

  if !is_feed
    begin
      @body = @request.io.read
      is_feed = !self.parse(url, @body).class.name.index('Feedjira::Parser').nil?
      rescue
    end
  end

  is_feed
end

.find(url) ⇒ Object



23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
# File 'lib/feedium.rb', line 23

def self.find(url)
  begin
    uri   = URI.parse(url.downcase.index('http') ? url : "http://#{url}")
    query = uri.query.nil? ? '' : "?#{uri.query}"
    url   = "#{uri.scheme}://#{uri.host}#{uri.path}#{query}"
  rescue URI::InvalidURIError => e
    raise Feedium::RequestError.new(e.message)
  end

  return @request.url if self.feed?(url)

  @body = @body || @request.io.read

  doc = Nokogiri::HTML(@body)

  if doc.at('base') && doc.at('base')['href']
    @base_uri = doc.at('base')['href']
  else
    @base_uri = "#{@request.uri.scheme}://#{@request.uri.host}"
  end

  doc.xpath("//link[@rel='alternate'][@href][@type]").each do |l|
    if CONTENT_TYPES.include?(l['type'].downcase.strip)
      return @request.url if self.feed?(l['href'])
    end
  end

  doc.xpath("//a[@href]").each do |a|
    found = %w(feed rss atom).detect {|k| !a['href'].index(k).nil? }

    begin
      return @request.url if found && self.feed?(a['href'])
    rescue Feedium::RequestError
      next
    end
  end

  nil
end

.find!(url) ⇒ Object



63
64
65
66
67
68
69
# File 'lib/feedium.rb', line 63

def self.find!(url)
  feed = self.find(url)

  raise Feedium::RequestError.new('Feed not found') if feed.nil?

  feed
end

.parse(url, content = nil) ⇒ Object



92
93
94
95
96
97
98
99
100
# File 'lib/feedium.rb', line 92

def self.parse(url, content = nil)
  if content.nil?
    request = Request.new(url)
    request.send
    content = request.io.read
  end

  Feedjira::Feed.parse(content)
end