使用 Nokogiri 查找文本链接时如何避免“无效字节序列”

Question

我正在使用带有 Ruby 4.2 的 Rails 5 并扫描我用 Nokogiri 解析的文档，以不区分大小写的方式查找带有文本的链接：

a_elt = doc ? doc.xpath('//a').detect { |node| /link[[:space:]]+text/i === node.text } : nil 

在content 中获取我网页的 HTML 后，我使用以下方法将其解析为 Nokogiri 文档：

doc = Nokogiri::HTML(content) 

问题是，我得到了

ArgumentError invalid byte sequence in UTF-8

在某些网页上使用上述正则表达式时。

2.4.0 :002 > doc.encoding
 => "UTF-8" 
2.4.0 :003 > doc.xpath('//a').detect { |node| /individual[[:space:]]+results/i === node.text }
ArgumentError: invalid byte sequence in UTF-8
    from (irb):3:in `==='
    from (irb):3:in `block in irb_binding'
    from /Users/davea/.rvm/gems/ruby-2.4.0@global/gems/nokogiri-1.7.0/lib/nokogiri/xml/node_set.rb:187:in `block in each'
    from /Users/davea/.rvm/gems/ruby-2.4.0@global/gems/nokogiri-1.7.0/lib/nokogiri/xml/node_set.rb:186:in `upto'
    from /Users/davea/.rvm/gems/ruby-2.4.0@global/gems/nokogiri-1.7.0/lib/nokogiri/xml/node_set.rb:186:in `each'
    from (irb):3:in `detect'
    from (irb):3
    from /Users/davea/.rvm/gems/ruby-2.4.0@global/gems/railties-5.0.1/lib/rails/commands/console.rb:65:in `start'
    from /Users/davea/.rvm/gems/ruby-2.4.0@global/gems/railties-5.0.1/lib/rails/commands/console_helper.rb:9:in `start'
    from /Users/davea/.rvm/gems/ruby-2.4.0@global/gems/railties-5.0.1/lib/rails/commands/commands_tasks.rb:78:in `console'
    from /Users/davea/.rvm/gems/ruby-2.4.0@global/gems/railties-5.0.1/lib/rails/commands/commands_tasks.rb:49:in `run_command!'
    from /Users/davea/.rvm/gems/ruby-2.4.0@global/gems/railties-5.0.1/lib/rails/commands.rb:18:in `<top (required)>'
    from bin/rails:4:in `require'
    from bin/rails:4:in `<main>' 

有没有办法我可以重写上面的内容来自动解释编码或奇怪的字符而不是翻转？

Answer 1

您的问题之前可能已经得到解答。你试过“Is there any way to clean a file of "invalid byte sequence in UTF-8" errors in Ruby?”的方法吗？

特别是在detect块之前，尝试去除除换行之外的无效字节和控制字符：

doc.scrub!("")
doc.gsub!(/[[:cntrl:]&&[^\n\r]]/,"")

请记住，scrub! 是 Ruby 2.1+ method。