Module: Onebox::Helpers

Defined in:
lib/onebox/helpers.rb

Defined Under Namespace

Classes: DownloadTooLarge

Class Method Summary collapse

Class Method Details

.audio_placeholder_html ⇒ Object 



223
224
225
 
# File 'lib/onebox/helpers.rb', line 223

def self.audio_placeholder_html
  "<div class='onebox-placeholder-container'><span class='placeholder-icon audio'></span></div>"
end

.blank?(value) ⇒ Boolean

Returns:

  • (Boolean) 


149
150
151
152
153
154
155
156
157
 
# File 'lib/onebox/helpers.rb', line 149

def self.blank?(value)
  if value.nil?
    true
  elsif String === value
    value.empty? || !(/[[:^space:]]/ === value)
  else
    value.respond_to?(:empty?) ? !!value.empty? : !value
  end
end

.clean(html) ⇒ Object 



23
24
25
 
# File 'lib/onebox/helpers.rb', line 23

def self.clean(html)
  html.gsub(/<[^>]+>/, ' ').gsub(/\n/, '')
end

.click_to_scroll_div(width = 690, height = 400) ⇒ Object 



145
146
147
 
# File 'lib/onebox/helpers.rb', line 145

def self.click_to_scroll_div(width = 690, height = 400)
  "<div style=\"background:transparent;position:relative;width:#{width}px;height:#{height}px;top:#{height}px;margin-top:-#{height}px;\" onClick=\"style.pointerEvents='none'\"></div>"
end

.fetch_content_length(location) ⇒ Object 



106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
 
# File 'lib/onebox/helpers.rb', line 106

def self.fetch_content_length(location)
  uri = URI(location)

  Net::HTTP.start(uri.host, uri.port, use_ssl: uri.is_a?(URI::HTTPS)) do |http|
    http.open_timeout = Onebox.options.connect_timeout
    http.read_timeout = Onebox.options.timeout
    if uri.is_a?(URI::HTTPS)
      http.use_ssl = true
      http.verify_mode = OpenSSL::SSL::VERIFY_NONE
    end

    http.request_head([uri.path, uri.query].join("?")) do |response|
      code = response.code.to_i
      unless code === 200 || Onebox::Helpers.blank?(response.header['content-length'])
        return nil
      end
      return response.header['content-length']
    end
  end
end

.fetch_html_doc(url, headers = nil) ⇒ Object 



27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
 
# File 'lib/onebox/helpers.rb', line 27

def self.fetch_html_doc(url, headers = nil)
  response = (fetch_response(url, nil, nil, headers) rescue nil)
  doc = Nokogiri::HTML(response)
  uri = URI(url)

  ignore_canonical_tag = doc.at('meta[property="og:ignore_canonical"]')
  should_ignore_canonical = IGNORE_CANONICAL_DOMAINS.map { |hostname| uri.hostname.match?(hostname) }.any?

  unless (ignore_canonical_tag && ignore_canonical_tag['content'].to_s == 'true') || should_ignore_canonical
    # prefer canonical link
    canonical_link = doc.at('//link[@rel="canonical"]/@href')
    if canonical_link && "#{URI(canonical_link).host}#{URI(canonical_link).path}" != "#{uri.host}#{uri.path}"
      response = (fetch_response(canonical_link, nil, nil, headers) rescue nil)
      doc = Nokogiri::HTML(response) if response
    end
  end

  doc
end

.fetch_response(location, limit = nil, domain = nil, headers = nil) ⇒ Object

Raises:

  • (Net::HTTPError) 


47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
 
# File 'lib/onebox/helpers.rb', line 47

def self.fetch_response(location, limit = nil, domain = nil, headers = nil)

  limit ||= 5
  limit = Onebox.options.redirect_limit if limit > Onebox.options.redirect_limit

  raise Net::HTTPError.new('HTTP redirect too deep', location) if limit == 0

  uri = URI(location)
  uri = URI("#{domain}#{location}") if !uri.host

  result = StringIO.new
  Net::HTTP.start(uri.host, uri.port, use_ssl: uri.is_a?(URI::HTTPS)) do |http|
    http.open_timeout = Onebox.options.connect_timeout
    http.read_timeout = Onebox.options.timeout
    if uri.is_a?(URI::HTTPS)
      http.use_ssl = true
      http.verify_mode = OpenSSL::SSL::VERIFY_NONE
    end

    headers ||= {}

    if Onebox.options.user_agent && !headers['User-Agent']
      headers['User-Agent'] = Onebox.options.user_agent
    end

    request = Net::HTTP::Get.new(uri.request_uri, headers)
    start_time = Time.now

    size_bytes = Onebox.options.max_download_kb * 1024
    http.request(request) do |response|

      if cookie = response.get_fields('set-cookie')
        header = { 'Cookie' => cookie.join }
      end

      header = nil unless header.is_a? Hash

      code = response.code.to_i
      unless code === 200
        response.error! unless [301, 302].include?(code)
        return fetch_response(
          response['location'],
          limit - 1,
          "#{uri.scheme}://#{uri.host}",
          header
        )
      end

      response.read_body do |chunk|
        result.write(chunk)
        raise DownloadTooLarge.new if result.size > size_bytes
        raise Timeout::Error.new if (Time.now - start_time) > Onebox.options.timeout
      end

      return result.string
    end
  end
end

.generic_placeholder_html ⇒ Object 



231
232
233
 
# File 'lib/onebox/helpers.rb', line 231

def self.generic_placeholder_html
  "<div class='onebox-placeholder-container'><span class='placeholder-icon generic'></span></div>"
end

.get(meta, attr) ⇒ Object 



163
164
165
 
# File 'lib/onebox/helpers.rb', line 163

def self.get(meta, attr)
  (meta && !blank?(meta[attr])) ? sanitize(meta[attr]) : nil
end

.get_absolute_image_url(src, url) ⇒ Object 



183
184
185
186
187
188
189
190
191
192
193
194
195
196
 
# File 'lib/onebox/helpers.rb', line 183

def self.get_absolute_image_url(src, url)
  if src && !!(src =~ /^\/\//)
    uri = URI(url)
    src = "#{uri.scheme}:#{src}"
  elsif src && src.match(/^https?:\/\//i).nil?
    uri = URI(url)
    src = if !src.start_with?("/") && uri.path.present?
      "#{uri.scheme}://#{uri.host.sub(/\/$/, '')}#{uri.path.sub(/\/$/, '')}/#{src.sub(/^\//, '')}"
    else
      "#{uri.scheme}://#{uri.host.sub(/\/$/, '')}/#{src.sub(/^\//, '')}"
    end
  end
  src
end

.map_placeholder_html ⇒ Object 



227
228
229
 
# File 'lib/onebox/helpers.rb', line 227

def self.map_placeholder_html
  "<div class='onebox-placeholder-container'><span class='placeholder-icon map'></span></div>"
end

.normalize_url_for_output(url) ⇒ Object 



172
173
174
175
176
177
178
179
180
181
 
# File 'lib/onebox/helpers.rb', line 172

def self.normalize_url_for_output(url)
  return "" unless url
  url = url.dup
  # expect properly encoded url, remove any unsafe chars
  url.gsub!(' ', '%20')
  url.gsub!("'", "&apos;")
  url.gsub!('"', "&quot;")
  url.gsub!(/[^\w\-`.~:\/?#\[\]@!$&'\(\)*+,;=%\p{M}’]/, "")
  url
end

.pretty_filesize(size) ⇒ Object 



127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
 
# File 'lib/onebox/helpers.rb', line 127

def self.pretty_filesize(size)
  conv = [ 'B', 'KB', 'MB', 'GB', 'TB', 'PB', 'EB' ]
  scale = 1024

  ndx = 1
  if (size < 2 * (scale**ndx)) then
    return "#{(size)} #{conv[ndx - 1]}"
  end
  size = size.to_f
  [2, 3, 4, 5, 6, 7].each do |i|
    if (size < 2 * (scale**i)) then
      return "#{'%.2f' % (size / (scale**(i - 1)))} #{conv[i - 1]}"
    end
  end
  ndx = 7
  "#{'%.2f' % (size / (scale**(ndx - 1)))} #{conv[ndx - 1]}"
end

.sanitize(value, length = 50) ⇒ Object 



167
168
169
170
 
# File 'lib/onebox/helpers.rb', line 167

def self.sanitize(value, length = 50)
  return nil if blank?(value)
  Sanitize.fragment(value).strip
end

.symbolize_keys(hash) ⇒ Object 



12
13
14
15
16
17
18
19
20
21
 
# File 'lib/onebox/helpers.rb', line 12

def self.symbolize_keys(hash)
  return {} if hash.nil?

  hash.inject({}) do |result, (key, value)|
    new_key = key.is_a?(String) ? key.to_sym : key
    new_value = value.is_a?(Hash) ? symbolize_keys(value) : value
    result[new_key] = new_value
    result
  end
end

.truncate(string, length = 50) ⇒ Object 



159
160
161
 
# File 'lib/onebox/helpers.rb', line 159

def self.truncate(string, length = 50)
  string.size > length ? string[0...(string.rindex(" ", length) || length)] + "..." : string
end

.uri_encode(url) ⇒ Object

Percent-encodes a URI string per RFC3986 - tools.ietf.org/html/rfc3986



199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
 
# File 'lib/onebox/helpers.rb', line 199

def self.uri_encode(url)
  return "" unless url

  uri = Addressable::URI.parse(url)

  encoded_uri = Addressable::URI.new(
    scheme: Addressable::URI.encode_component(uri.scheme, Addressable::URI::CharacterClasses::SCHEME),
    authority: Addressable::URI.encode_component(uri.authority, Addressable::URI::CharacterClasses::AUTHORITY),
    path: Addressable::URI.encode_component(uri.path, Addressable::URI::CharacterClasses::PATH + "\\%"),
    query: Addressable::URI.encode_component(uri.query, "a-zA-Z0-9\\-\\.\\_\\~\\$\\&\\*\\,\\=\\:\\@\\?\\%"),
    fragment: Addressable::URI.encode_component(uri.fragment, "a-zA-Z0-9\\-\\.\\_\\~\\!\\$\\&\\'\\(\\)\\*\\+\\,\\;\\=\\:\\/\\?\\%")
  )

  encoded_uri.to_s
end

.uri_unencode(url) ⇒ Object 



215
216
217
 
# File 'lib/onebox/helpers.rb', line 215

def self.uri_unencode(url)
  Addressable::URI.unencode(url)
end

.video_placeholder_html ⇒ Object 



219
220
221
 
# File 'lib/onebox/helpers.rb', line 219

def self.video_placeholder_html
  "<div class='onebox-placeholder-container'><span class='placeholder-icon video'></span></div>"
end