xref: /onnv-gate/usr/src/cmd/perl/5.8.4/distrib/t/op/repeat.t (revision 0:68f95e015346)
1#!./perl
2
3BEGIN {
4    chdir 't' if -d 't';
5    @INC = '../lib';
6}
7
8require './test.pl';
9plan(tests => 33);
10
11# compile time
12
13is('-' x 5, '-----',    'compile time x');
14is('-' x 1, '-',        '  x 1');
15is('-' x 0, '',         '  x 0');
16is('-' x -1, '',        '  x -1');
17is('-' x undef, '',     '  x undef');
18
19is('ab' x 3, 'ababab',  '  more than one char');
20
21# run time
22
23$a = '-';
24is($a x 5, '-----',     'run time x');
25is($a x 1, '-',         '  x 1');
26is($a x 0, '',          '  x 0');
27is($a x -3, '',         '  x -3');
28is($a x undef, '',      '  x undef');
29
30$a = 'ab';
31is($a x 3, 'ababab',    '  more than one char');
32$a = 'ab';
33is($a x 0, '',          '  more than one char');
34$a = 'ab';
35is($a x -12, '',        '  more than one char');
36
37$a = 'xyz';
38$a x= 2;
39is($a, 'xyzxyz',        'x=2');
40$a x= 1;
41is($a, 'xyzxyz',        'x=1');
42$a x= 0;
43is($a, '',              'x=0');
44
45@x = (1,2,3);
46
47is(join('', @x x 4),        '3333',                 '@x x Y');
48is(join('', (@x) x 4),      '123123123123',         '(@x) x Y');
49is(join('', (@x,()) x 4),   '123123123123',         '(@x,()) x Y');
50is(join('', (@x,1) x 4),    '1231123112311231',     '(@x,1) x Y');
51is(join(':', () x 4),       '',                     '() x Y');
52is(join(':', (9) x 4),      '9:9:9:9',              '(X) x Y');
53is(join(':', (9,9) x 4),    '9:9:9:9:9:9:9:9',      '(X,X) x Y');
54is(join('', (split(//,"123")) x 2), '123123',       'split and x');
55
56is(join('', @x x -12),      '',                     '@x x -12');
57is(join('', (@x) x -14),    '',                     '(@x) x -14');
58
59
60# This test is actually testing for Digital C compiler optimizer bug,
61# present in Dec C versions 5.* and 6.0 (used in Digital UNIX and VMS),
62# found in December 1998.  The bug was reported to Digital^WCompaq as
63#     DECC 2745 (21-Dec-1998)
64# GEM_BUGS 7619 (23-Dec-1998)
65# As of April 1999 the bug has been fixed in Tru64 UNIX 5.0 and is planned
66# to be fixed also in 4.0G.
67#
68# The bug was as follows: broken code was produced for util.c:repeatcpy()
69# (a utility function for the 'x' operator) in the case *all* these
70# four conditions held:
71#
72# (1) len == 1
73# (2) "from" had the 8th bit on in its single character
74# (3) count > 7 (the 'x' count > 16)
75# (4) the highest optimization level was used in compilation
76#     (which is the default when compiling Perl)
77#
78# The bug looked like this (. being the eight-bit character and ? being \xff):
79#
80# 16 ................
81# 17 .........???????.
82# 18 .........???????..
83# 19 .........???????...
84# 20 .........???????....
85# 21 .........???????.....
86# 22 .........???????......
87# 23 .........???????.......
88# 24 .........???????.???????
89# 25 .........???????.???????.
90#
91# The bug was triggered in the "if (len == 1)" branch.  The fix
92# was to introduce a new temporary variable.  In diff -u format:
93#
94#     register char *frombase = from;
95#
96#     if (len == 1) {
97#-       todo = *from;
98#+       register char c = *from;
99#        while (count-- > 0)
100#-           *to++ = todo;
101#+           *to++ = c;
102#        return;
103#     }
104#
105# The bug could also be (obscurely) avoided by changing "from" to
106# be an unsigned char pointer.
107#
108# This obscure bug was not found by the then test suite but instead
109# by Mark.Martinec@nsc.ijs.si while trying to install Digest-MD5-2.00.
110#
111# jhi@iki.fi
112#
113is("\xdd" x 24, "\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd\xdd", 'Dec C bug');
114
115
116# When we use a list repeat in a scalar context, it behaves like
117# a scalar repeat. Make sure that works properly, and doesn't leave
118# extraneous values on the stack.
119#  -- robin@kitsite.com
120
121my ($x, $y) = scalar ((1,2)x2);
122is($x, "22",    'list repeat in scalar context');
123is($y, undef,   '  no extra values on stack');
124
125# Make sure the stack doesn't get truncated too much - the left
126# operand of the eq binop needs to remain!
127is(77, scalar ((1,7)x2),    'stack truncation');
128
129
130# perlbug 20011113.110 works in 5.6.1, broken in 5.7.2
131{
132    my $x= [("foo") x 2];
133    is( join('', @$x), 'foofoo', 'list repeat in anon array ref broken [ID 20011113.110]' );
134}
135
136# [ID 20010809.028] x operator not copying elements in 'for' list?
137{
138    local $TODO = "x operator not copying elements in 'for' list? [ID 20010809.028]";
139    my $x = 'abcd';
140    my $y = '';
141    for (($x =~ /./g) x 2) {
142	$y .= chop;
143    }
144    is($y, 'abcdabcd');
145}
146
147