strange interaction of 'uniq' and 'sort'

[chrispitude]

For the following code:

use warnings;
use strict;
use Acme::Tools;

my @vals1 = ('a', 'dup');
my @vals2 = ('z', 'dup');
my @new_values = sort uniq (@vals1, @vals2);

print "VALUES:@new_values\n";

I get the following output:

VALUES:a dup z dup

If I replace sort uniq with distinct (obviously the better coding style), then I get the expected result:

VALUES:a dup z

What would cause this? Is there some strange interaction between sort and the argument template for uniq?

[kjetillll]

I have had the same problem with sort myself. Always use ( right after sort when in doubt. It's because sort uses uniq as a comparison subroutine. As the documentation says "*If SUBNAME is specified, it gives the name of a subroutine that returns an integer less than, equal to, or greater than 0, depending on how the elements of the list are to be ordered.*" perldoc -f sort #run this I didn't change Acme::Tools, just added a comment in the doc Running this might be revealing: use strict; use warnings; sub uniq(@) { my %seen; grep !$seen{$_}++, @_ } #copied from Acme::Tools sub uniqwo { my %seen; grep !$seen{$_}++, @_ } #without signature sub uniqnew { my %seen; my @r=grep !$seen{$_}++, @_; @r } sub uniqnew2 { my @inp=@_; my %seen; my @r=grep !$seen{$_}++, @inp; @r } sub distinct { return sort keys %{{map {($_,1)} @_}} } #copied from Acme::Tools my @A1 = ('a', 'dup'); my @a2 = ('z', 'dup'); my @a12 = (@A1,@a2); my @A; sub run { eval $_[0]; die $@ if $@; printf "%-40s # @A\n",$_[0] } run '@A = sort uniq (@A1, @a2); '; run '@A = sort uniq (@A1, @a2); '; run '@A = sort uniqwo (@A1, @a2); '; run '@A = sort uniqnew (@A1, @a2); '; run '@A = sort uniqnew(@A1, @a2); '; run '@A = sort uniqnew2(@A1, @a2); '; run '@A = sort(uniq(@A1, @a2)); '; run '@A = sort(uniq(@a12)); '; run '@A = sort(uniqnew(@a12)); '; run '@A = sort uniq @a12; '; run '@A = sort uniq @a12; '; run '@A = sort @A1, @a2; '; run '@A = uniq sort @A1, @a2; '; run '@A = distinct (@A1, @a2); '; run '@A = distinct(@A1, @a2); '; run '@A = distinct @A1, @a2; '; run '@A = sort distinct(@A1, @a2); '; Result: @A = sort uniq (@A1, @a2); # dup z dup a @A = sort uniq (@A1, @a2); # dup z dup a @A = sort uniqwo (@A1, @a2); # dup z dup a @A = sort uniqnew (@A1, @a2); # dup z dup a @A = sort uniqnew(@A1, @a2); # dup z dup a @A = sort uniqnew2(@A1, @a2); # dup z dup a @A = sort(uniq(@A1, @a2)); # a dup z @A = sort(uniq(@a12)); # a dup z @A = sort(uniqnew(@a12)); # a dup z @A = sort uniq @a12; # dup z dup a @A = sort uniq @a12; # dup z dup a @A = sort @A1, @a2; # a dup dup z @A = uniq sort @A1, @a2; # a dup z @A = distinct (@A1, @a2); # a dup z @A = distinct(@A1, @a2); # a dup z @A = distinct @A1, @a2; # a dup z Use of uninitialized value in sort at (eval 17) line 1. Use of uninitialized value in sort at (eval 17) line 1. Use of uninitialized value in sort at (eval 17) line 1. Use of uninitialized value in sort at (eval 17) line 1. @A = sort distinct(@A1, @a2); # a dup z dup -- ks

…

On Tue, Jul 2, 2019 at 5:00 PM chrispitude ***@***.***> wrote: For the following code: use warnings; use strict; use Acme::Tools; my @vals1 = ('a', 'dup'); my @Vals2 = ('z', 'dup'); my @new_values = sort uniq ***@***.***, @Vals2); print ***@***.***_values\n"; I get the following output: VALUES:a dup z dup If I replace sort uniq with distinct (obviously the better coding style), then I get the expected result: VALUES:a dup z What would cause this? Is there some strange interaction between sort and the argument template for uniq? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#20?email_source=notifications&email_token=AADK7TK2OIO6XOH54GRS6HDP5NURLA5CNFSM4H44WXQKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4G45DAXQ>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADK7TKDCQ2T2OL23RW6633P5NURLANCNFSM4H44WXQA> .