Chapter 18: Regular expressions

declare namespace functx = "http://www.functx.com";
declare function functx:get-matches-and-non-matches
($string as xs:string?, $regex as xs:string) as element()* {
  let $iomf := functx:index-of-match-first($string, $regex)
  return
   if (empty($iomf))
   then <non-match>{$string}</non-match>
   else if ($iomf > 1)
   then (<non-match>{substring($string,1,$iomf - 1)}</non-match>,
         functx:get-matches-and-non-matches(
           substring($string,$iomf),$regex))
   else
   let $length :=
        string-length($string) -
        string-length(functx:replace-first($string, $regex,''))
    return (<match>{substring($string,1,$length)}</match>,
            if (string-length($string) > $length)
            then functx:get-matches-and-non-matches(
                  substring($string,$length + 1),$regex)
            else ())
 } ;
      
declare function functx:index-of-match-first
($arg as xs:string?, $pattern as xs:string) as xs:integer? {
  if (matches($arg,$pattern))
  then string-length(tokenize($arg, $pattern)[1]) + 1
  else ()
 } ;

declare function functx:replace-first
($arg as xs:string?, $pattern as xs:string,
$replacement as xs:string ) as xs:string {
  replace($arg, concat('(^.*?)', $pattern),
           concat('$1',$replacement))
 } ;