当用作文字正则表达式时，转义的 Ruby 正则表达式不会给出匹配项。为什么？
代码说明了一切：
teststring = "helloworld$"
string_from_user = "world$"
regexp = Regexp.escape(string_from_user) # assigns "world\\$"
p teststring =~ Regexp.new(regexp) # prints 0 => match found
p teststring =~ /regexp/ # prints nil => no match
第一个匹配在 Regexp.escape 文档。 但为什么第二个版本不匹配呢？
我很担心，因为我需要将此正则表达式传递给第三方 Ruby 代码。该字符串来自用户，所以我想转义它。然后，在某些情况下，我可能会向该用户的字符串添加额外的正则表达式符号。例如，我可能会传递 "^helloworld\\$" ，以便第三方代码匹配 "helloworld$othercontent" 等字符串。
我担心如果第三方代码使用 =~ /regexp/ 而不是 =~ Regexp.new(regexp)，我会遇到麻烦，因为会有如上面的代码所示，没有匹配项。
Code says it all:
teststring = "helloworld$"
string_from_user = "world$"
regexp = Regexp.escape(string_from_user) # assigns "world\\$"
p teststring =~ Regexp.new(regexp) # prints 0 => match found
p teststring =~ /regexp/ # prints nil => no match
That the first one matches is mentioned in the Regexp.escape docs.
But why doesn't the second version match?
I'm concerned because I need to pass this regexp to third party Ruby code. The string comes from the user, so I want to escape it. Then, in some situations, I might add additional regexp symbols to this user's string. For example, I might pass "^helloworld\\$" so that third party code would match strings like "helloworld$othercontent".
I am worried that if the third party code uses =~ /regexp/ instead of =~ Regexp.new(regexp), I will be in trouble, because there will be no match as indicated by the code above.
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（1）
因为
/regexp/是与字符串"regexp"匹配的正则表达式。也许您的意思是/#{regexp}/？编辑：通过更全面地阅读您的问题，我认为您将一个字符串传递到第三方代码中，您知道该代码将从该字符串创建正则表达式。在这种情况下，你应该是安全的。如上所述，
/regexp/不可能是他们正在做的事情，因为它只是错误。他们必须使用 Regexp.new() 或类似的东西。Because
/regexp/is a regexp matching the string"regexp". Perhaps you meant/#{regexp}/?Edit: I take it, from reading your question more fully, that you're passing a string into third party code that you know will be making a Regexp from that string. In which case, you should be safe. As noted above,
/regexp/cannot possibly be what they're doing, because it's just wrong. They must be usingRegexp.new()or something similar.