Add tests for lvalue subs returning vec() or substr() and called twice in one expression

[t-a-k]

I found that Perl prior to v5.20 (at least v5.16.3 bundled in CentOS Linux 7 and locally-built v5.18.4) fail to return correct value from subroutine returning vec() and substr() as lvalue but called multiple times in rvalue context of one expression:

$ perl -wle 'my $x = "\002"; sub myvec1 :lvalue { vec($x, $_[0], 1) }' \
    -e 'print myvec1(0); print myvec1(1); print myvec1(0) + myvec1(1)'
0
1
2
$

This apparently had been fixed in the commit 169504d (in 2013), but testcases for these symptoms doesn't seem to exist so I'd like to add ones to prevent accidental regression.

[jkeenan]

I extracted the proposed changes to a program outside my checkout, then ran that program against perl-5.18.4 and perl-5.20.0, respectively.

$ perlbrew use perl-5.18.4
$ prove -v gh-22386.t 
gh-22386.t .. 
1..2
not ok 1 - wrap_vec
not ok 2 - wrap_substr

#   Failed test 'wrap_vec'
#   at gh-22386.t line 16.
#          got: 10
#     expected: 8

#   Failed test 'wrap_substr'
#   at gh-22386.t line 26.
#          got: 'kk'
#     expected: 'ok'
# Looks like you failed 2 tests of 2.
Dubious, test returned 2 (wstat 512, 0x200)
Failed 2/2 subtests 

Test Summary Report
-------------------
gh-22386.t (Wstat: 512 Tests: 2 Failed: 2)
  Failed tests:  1-2
  Non-zero exit status: 2
Files=1, Tests=2,  0 wallclock secs ( 0.01 usr +  0.00 sys =  0.01 CPU)
Result: FAIL

$ perlbrew use perl-5.20.0
$ prove -v gh-22386.t 
gh-22386.t .. 
1..2
ok 1 - wrap_vec
ok 2 - wrap_substr
ok
All tests successful.
Files=1, Tests=2,  0 wallclock secs ( 0.01 usr +  0.00 sys =  0.01 CPU)
Result: PASS

So I think that once the grammar in the inline comment is corrected, this will be good to go into blead.

[tonycoz]

Could use grammer to grammar in the second commit message (which a squash would likely fix)

[t-a-k]

Could use grammer to grammar in the second commit message (which a squash would likely fix)

Sorry for the typo. I've force-pushed the second commit.