Potential Typo in `REXML::Functions::normalize_space()` Implementation

[flatland001]

Hello Ruby/REXML developers,

The implementation can be found here:

https://github.com/ruby/rexml/blob/7e4049f6a68c99c4efec2df117057ee080680c9f/lib/rexml/functions.rb#L265C1-L273C8

The current code is as follows:

    # UNTESTED
    def Functions::normalize_space( string=nil )
      string = string(@@context[:node]) if string.nil?
      if string.kind_of? Array
        string.collect{|x| string.to_s.strip.gsub(/\s+/um, ' ') if string}
      else
        string.to_s.strip.gsub(/\s+/um, ' ')
      end
    end

It appears there may be a typo in the use of x versus string within the collect block. I believe the following adjustment might be more in line with the intended functionality:

string.collect{|x| x.to_s.strip.gsub(/\s+/um, ' ') if x}

I might be misunderstanding the intention here, so if I've overlooked something, please do let me know.
Thank you for your time.

[kou]

Good catch!
Could you open a PR for this?
(It's better that it has a test for this case.)