Do not pass unknown encoding names to nokogiri. (#30987)
This commit is contained in:
parent
36592d10aa
commit
2ea9336b68
3 changed files with 35 additions and 1 deletions
|
@ -274,7 +274,7 @@ class LinkDetailsExtractor
|
|||
end
|
||||
|
||||
def detect_encoding_and_parse_document
|
||||
[detect_encoding, nil, @html_charset].uniq.each do |encoding|
|
||||
[detect_encoding, nil, header_encoding].uniq.each do |encoding|
|
||||
document = Nokogiri::HTML(@html, nil, encoding)
|
||||
return document if document.to_s.valid_encoding?
|
||||
end
|
||||
|
@ -286,6 +286,13 @@ class LinkDetailsExtractor
|
|||
guess&.fetch(:confidence, 0).to_i > 60 ? guess&.fetch(:encoding, nil) : nil
|
||||
end
|
||||
|
||||
def header_encoding
|
||||
Encoding.find(@html_charset).name if @html_charset
|
||||
rescue ArgumentError
|
||||
# Encoding from HTTP header is not recognized by ruby
|
||||
nil
|
||||
end
|
||||
|
||||
def detector
|
||||
@detector ||= CharlockHolmes::EncodingDetector.new.tap do |detector|
|
||||
detector.strip_tags = true
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue