The attached tiny patch will fix the problem Greg Sabino Mullane had with a shared lexical $_TD, by making it a global and just pushing a local value in the trigger function.

I don't think what we had is strictly a bug, so I don't thinbk we need top backpatch this.

It will, however, require use of perl 5.6 at a minimum, because that's when the "our" function came in. Since that was over 6 years ago, I think this is not unreasonable. If there are squawks, I have another slightly longer and slightly more old-fashioned way to do the same thing, but this is the best modern way.

I don't think a docs change is needed.

If there's no objection I will apply thin in a few days.

cheers

andrew


Index: src/pl/plperl/plperl.c
===================================================================
RCS file: /cvsroot/pgsql/src/pl/plperl/plperl.c,v
retrieving revision 1.108
diff -c -r1.108 plperl.c
*** src/pl/plperl/plperl.c	4 Apr 2006 19:35:37 -0000	1.108
--- src/pl/plperl/plperl.c	22 May 2006 20:46:37 -0000
***************
*** 765,771 ****
  	ENTER;
  	SAVETMPS;
  	PUSHMARK(SP);
! 	XPUSHs(sv_2mortal(newSVpv("my $_TD=$_[0]; shift;", 0)));
  	XPUSHs(sv_2mortal(newSVpv(s, 0)));
  	PUTBACK;
  
--- 765,771 ----
  	ENTER;
  	SAVETMPS;
  	PUSHMARK(SP);
! 	XPUSHs(sv_2mortal(newSVpv("our $_TD; local $_TD=$_[0]; shift;", 0)));
  	XPUSHs(sv_2mortal(newSVpv(s, 0)));
  	PUTBACK;
  
---------------------------(end of broadcast)---------------------------
TIP 5: don't forget to increase your free space map settings

Reply via email to