Front page | perl.perl5.porters |
Postings from October 2016
* Dave Mitchell <davem@iabyn.com> [2016-10-19 13:12]: > > This says “lhs”. It does not say “rhs” if you change the LHS of the > > assignment to (undef,undef) – which would be what it should say if > > such an assignment to undef were returning an alias to the RHS > > element. > > No, you have a typo: my($r2,$r2) rather than $r1. With that changed, > it prints 'rhs' when changing ($l1,$l2) to (undef,undef). Ah. My bad. > I can see 4 ways this could be handled. > > 1. (...,undef,...) = (...) autovififies a new undefined scalar value > then assigns the RHS value to it, returning the new scalar; > > 2. it skips the assignment for that slot, and returns the RHS element; > > 3. it skips the assignment for that slot, and returns &PL_sv_undef > (which will croak if later modified); > > 4. it skips the assignment for that slot, and returns a fresh undef > scalar. > > (2) is the current behaviour. I'd like to change it to (3). Do you have a specific reason other than that it seems that it “should” be that way? I am extremely leery of this particular option as it would cause code to crash that has run fine so far. I don’t know how much code would be affected, but it feels potentially significant. Do we get any benefits out of the breakage other than conceptual purity? > (1) and (4) seem inefficient, having to create a new mortal scalar. Sure. I thought that’s what’s going on already but since it isn’t, no need to go in an inefficient direction. > I think (2) is definitely semantically incorrect: the essence of list > assignment is that it copies values: you shouldn't end up with aliases > to RHS elements. I’m not convinced. Perl has lots of little exceptions to such axiomatic rules when it serves the conciseness of code. The very fact that you can assign to `undef` is already a violation of your principle. I wouldn’t dismiss the fact that it returns the RHS element as a mere accident with no deliberation behind it. The inconsistency you originally asked about regarding the return value of list assignment in list context seems much more clearly like a bug, since it’s contradictory both itself and the documentation. Regards, -- Aristotle Pagaltzis // <http://plasmasturm.org/>Thread Previous | Thread Next