<html> <head> <base href="https://bugs.documentfoundation.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_UNCONFIRMED " title="UNCONFIRMED - REGEX() function with an integer as 'Occurence' returns the complete 'Text' if the respective match doesn't exist. It should return #N/A." href="https://bugs.documentfoundation.org/show_bug.cgi?id=126764#c1">Comment # 1</a> on <a class="bz_bug_link bz_status_UNCONFIRMED " title="UNCONFIRMED - REGEX() function with an integer as 'Occurence' returns the complete 'Text' if the respective match doesn't exist. It should return #N/A." href="https://bugs.documentfoundation.org/show_bug.cgi?id=126764">bug 126764</a> from <a class="email" href="mailto:erack@redhat.com" title="Eike Rathke <erack@redhat.com>"> Eike Rathke</a> <pre>There is no 3rd occurrence of the expression in text, hence nothing is replaced. This essentially is similar to =REGEX("a";"b";"") and is standard replacement behaviour in tools like sed and awk, if there is no match then nothing is replaced. Specifically gawk's gensub() that knows the occurrence parameter as well behaves the same: gawk -e '{print gensub(/b/,"",1)}' Try with different input of "a" and "b". If you need to check if there actually is a 3rd occurrence to be replaced then do an explicit check: =IF(ISNA(REGEX("12-34-56";"[^-]*-";;3));NA();REGEX("12-34-56";"[^-]*-";"";3)) Unconditionally returning #N/A if there is no match in the replace case IMHO is not an option, otherwise the usual logic of replacements with regular expressions would be negated and all replacements would have to use IFNA(). What we maybe could do is add an optional parameter to the function whether #N/A shall be returned in replace cases or not. That would have to be hidden in the Function Wizard and input hints as we have UI translations freeze for 6.3, but in general would be possible even for 6.3. Earlier releases reading such additional parameter if present of course would fail then.</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the assignee for the bug.</li> </ul> </body> </html>