How to use Heredocs with XPath

Question

I struggle with the following problem:

Have an XPath like so (as you can see quite long):

'//h2[contains(text(),"Kontakt")]/../following-sibling::div[contains(@class,"body") and child::dl[contains(@class,"horizontal")]]'

Now, to make it more readable i tried to split it into multiple lines using Heredoc:

xpath = <<-XPath.gsub(/\s+/,'')
  //h2[contains(text(),"Kontakt")]/..
  /following-sibling::div[contains(@class,"body") and
  child::dl[contains(@class,"horizontal")]]
XPath

The catch, however, is that the .gsub(/\s+/,'') will remove all whitespaces. I have tried different combinations with just removing \n or .strip, but I can't seem to find a solution that outputs the xpath as it is shown above.

Any thougts?

Answer 1

You attempt to handle the same cases (1st→2nd string and 2nd→3rd string) in different manner: concatenate strings in first case and insert space in the latter. There is an AI required to understand what did you mean :)

I would suggest you not to break the string at space position. In such a case /\s*\n\s*|\A\s+|\s+\Z/m regexp (remove carriage returns surrounded with spaces, beginning and trailing spaces) would work:

xpath = <<-XPath.gsub(/\s*\n\s*|\A\s+|\s+\Z/m,'')
  //h2[contains(text(),"Kontakt")]/..
  /following-sibling::div[contains(@class,"body") and child
  ::dl[contains(@class,"horizontal")]]
XPath

# "//h2[contains(text(),\"Kontakt\")]/../following-sibling::div[contains(@class,\"body\") and child::dl[contains(@class,\"horizontal\")]]"