Class: Kudzu::Agent::Util::CharsetDetector

Inherits:
Object
  • Object
show all
Defined in:
lib/kudzu/agent/util/charset_detector.rb

Constant Summary collapse

CORRECTION =
{
  'utf_8' => 'utf-8',
  'shift-jis' => 'shift_jis',
  'x-sjis' => 'shift_jis',
  'euc_jp' => 'euc-jp'
}

Class Method Summary collapse

Class Method Details

.detect(response) ⇒ Object


13
14
15
16
17
18
19
20
21
22
23
24
# File 'lib/kudzu/agent/util/charset_detector.rb', line 13

def detect(response)
  if response.html?
    from_html(response.body) || from_text(response.body)
  elsif response.xml?
    from_xml(response.body) || from_text(response.body)
  elsif response.text?
    from_text(response.body)
  end
rescue => e
  Kudzu.log :warn, "failed to detect charset: #{response.url}", error: e
  nil
end