Maintain html_safe? on sliced HTML safe strings - #33808

Thanks for the pull request, and welcome! The Rails team is excited to review your changes, and you should hear from @pixeltrix (or someone else) soon.

If any changes to this PR are deemed necessary, please add them as extra commits. This ensures that the reviewer can see what has changed since they last reviewed the code. Due to the way GitHub handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed. Large or tricky changes may require several passes of review and changes.

This repository is being automatically checked for code quality issues using Code Climate. You can see results for this analysis in the PR status below. Newly introduced issues should be fixed before a Pull Request is considered ready to review.

Please see the contribution instructions for more information.

Thanks! Can you add a CHANGELOG entry?

The railties failures seem unrelated, so I'll merge this.

I thought that I already did 🤦‍♂️. Thanks

I'm coming way late to the party but how is this not extremely dangerous?

"<wholetag>".slice[-1..-1] would then give you a closing tag that's marked as html_safe to inject. is this not a problem because it would have to be developer error to use that piece anyways?

@SampsonCrowley I think the assumption here is that the string being operated on is marked "html safe" when it has been html-sanitized (for example, by ActionView::Helpers::SanitizeHelper#sanitize) or escaped (for example, by ERB::Util.html_escape).

In that case, any comments or dangerous tags (like <script>) will have been removed or escaped during sanitization, and so while a slice may not be well-formed, it's not malicious. If you are able to generate a string that, once it's been sanitized and marked safe, would have a malicious substring in it, please open an issue so we can discuss?

As a starting point, here's a script to generate two different "safe" HTML strings:

#! /usr/bin/env ruby

require "action_view"
require "active_support"

class Sanitizer
  include ActionView::Helpers::SanitizeHelper
end

class Escaper
  include ERB::Util

  def escape(string)
    html_escape(string)
  end
end

original = %{<div><script>alert("xss")</script></div>}
# => "<div><script>alert(\"xss\")</script></div>"

safe = Sanitizer.new.sanitize(original)
# => "<div>alert(\"xss\")</div>"
safe.html_safe? # => true

safe = Escaper.new.escape(original)
# => "&lt;div&gt;&lt;script&gt;alert(&quot;xss&quot;)&lt;/script&gt;&lt;/div&gt;"
safe.html_safe? # => true

@flavorjones sorry my original reply was missing the part marking the string as html_safe

Taking the original original comment as an example:

In the behavior before this PR: str = "<script>".html_safe would not be escaped, and it shouldn't be because it is being explicitly marked as an actual html "safe" string, i.e. any unsafe characters have already been properly formatted

But if you were to then call slice on it, str[-1..-1], the previous behavior was that the string was then decidedly not assumed safe for direct consumption because any html may no longer be properly formatted or escaped. In the previous call you would now have just a closing carrat ">", if you know as a developer that the string is still safe you could mark it as such, but otherwise it was previously assumed to no longer be safe for direct consumption

Now

str[-1..-1] will give you a closing carrat that is still marked html "safe", even though it's not the original structure and may be improperly formatted or contain partial html.

So I don't understand how it's is safe to assume by default that a string that has been sliced from an html_safe string is still html safe just because it was html safe in its original full string

Or maybe I'm just misunderstanding the use case