XPath to return only elements containing the text, and not its parents

Do you want to find elements that contain "match", or that equal "match"?

This will find elements that have text nodes that equal 'match' (matches none of the elements because of leading and trailing whitespace in random2):

//*[text()='match']

This will find all elements that have text nodes that equal "match", after removing leading and trailing whitespace(matches random2):

//*[normalize-space(text())='match']

This will find all elements that contain 'match' in the text node value (matches random2 and random3):

//*[contains(text(),'match')]

This XPATH 2.0 solution uses the matches() function and a regex pattern that looks for text nodes that contain 'match' and begin at the start of the string(i.e. ^) or a word boundary (i.e. \W) and terminated by the end of the string (i.e. $) or a word boundary. The third parameter i evaluates the regex pattern case-insensitive. (matches random2)

//*[matches(text(),'(^|\W)match($|\W)','i')]