13 my $Is_EBCDIC = (ord('i') == 0x89 & ord('J') == 0xd1);
15 $_ = "abcdefghijklmnopqrstuvwxyz";
19 is($_, "ABCDEFGHIJKLMNOPQRSTUVWXYZ", 'uc');
23 is($_, "abcdefghijklmnopqrstuvwxyz", 'lc');
26 is($_, "aBCDEFGHIJKLMNOPQRSTUVWXYz", 'partial uc');
29 # In EBCDIC 'I' is \xc9 and 'J' is \0xd1, 'i' is \x89 and 'j' is \x91.
30 # Yes, discontinuities. Regardless, the \xca in the below should stay
31 # untouched (and not became \x8a).
38 is($_, "i\xcaj", 'EBCDIC discontinuity');
44 (my $y = 12) =~ tr/1/3/;
45 ($f = 1.5) =~ tr/1/3/;
46 (my $g = 1.5) =~ tr/1/3/;
47 is($x + $y + $f + $g, 71, 'tr cancels IOK and NOK');
51 is y/dam/ve/rd, 'eve', '/r';
52 is $_, 'adam', '/r leaves param alone';
54 is $g =~ y/bury/repl/r, 'perl', '/r with explicit param';
55 is $g, 'ruby', '/r leaves explicit param alone';
56 is "aaa" =~ y\a\b\r, 'bbb', '/r with constant param';
57 ok !eval '$_ !~ y///r', "!~ y///r is forbidden";
58 like $@, qr\^Using !~ with tr///r doesn't make sense\,
59 "!~ y///r error message";
63 local $SIG{__WARN__} = sub { $w = shift; ++$wc };
66 like $w, qr '^Useless use of non-destructive transliteration \(tr///r\)',
67 '/r warns in void context';
68 is $wc, 1, '/r warns just once';
71 # perlbug [ID 20000511.005]
76 is($_, 'Fred', 'harmless if explicitly not updating');
79 # A variant of the above, added in 5.7.2
82 eval '$1 =~ tr/A-Z/A-Z/;';
84 is($_, 'Fred', 'harmless if implicitly not updating');
85 is($@, '', ' no error');
88 # check tr handles UTF8 correctly
89 ($x = 256.65.258) =~ tr/a/b/;
90 is($x, 256.65.258, 'handles UTF8');
95 if (ord("\t") == 9) { # ASCII
102 # EBCDIC variants of the above tests
103 ($x = 256.193.258) =~ tr/a/b/;
109 if (ord("\t") == 9) { # ASCII
118 my $l = chr(300); my $r = chr(400);
120 $x =~ tr/\x{12c}/\x{190}/;
122 'changing UTF8 chars in a UTF8 string, same length');
126 $x =~ tr/\x{12c}/\x{be8}/;
127 is($x, 200.3048.400, ' more bytes');
131 $x =~ tr/\x{64}/\x{190}/;
132 is($x, 400.125.60, 'Putting UT8 chars into a non-UTF8 string');
136 $x =~ tr/\x{190}/\x{64}/;
137 is($x, 100.125.60, 'Removing UTF8 chars from UTF8 string');
141 $y = $x =~ tr/\x{190}/\x{190}/;
142 is($y, 2, 'Counting UTF8 chars in UTF8 string');
144 $x = 60.400.125.60.400;
145 $y = $x =~ tr/\x{3c}/\x{3c}/;
146 is($y, 2, ' non-UTF8 chars in UTF8 string');
148 # 17 - counting UTF8 chars in non-UTF8 string
150 $y = $x =~ tr/\x{190}/\x{190}/;
151 is($y, 0, ' UTF8 chars in non-UTFs string');
154 $_ = "abcdefghijklmnopqrstuvwxyz";
156 like($@, qr/^Ambiguous range in transliteration operator/, 'tr/a-z-9//');
158 # 19-21: Make sure leading and trailing hyphens still work
161 is($_, '..r.rot9', 'hyphens, leading');
165 is($_, '..r.rot9', ' trailing');
169 is($_, '..r.rot9', ' both');
171 $_ = "abcdefghijklmnop";
173 is($_, '.bcd....ijklm.op');
175 $_ = "abcdefghijklmnop";
177 is($_, '...de......lm...');
179 $_ = "abcdefghijklmnop";
181 is($_, '...d.f...j.l...p');
186 like($@, qr/^Invalid range "m-d" in transliteration operator/,
187 'reversed range check');
190 is(eval '$1 =~ tr/abcd//', 3, 'explicit read-only count');
191 is($@, '', ' no error');
194 is(eval '$1 =~ tr/abcd/abcd/', 3, 'implicit read-only count');
195 is($@, '', ' no error');
197 is(eval '"123" =~ tr/12//', 2, 'LHS of non-updating tr');
199 eval '"123" =~ tr/1/2/';
200 like($@, qr|^Can't modify constant item in transliteration \(tr///\)|,
201 'LHS bad on updating tr');
204 # v300 (0x12c) is UTF-8-encoded as 196 172 (0xc4 0xac)
205 # v400 (0x190) is UTF-8-encoded as 198 144 (0xc6 0x90)
207 # Transliterate a byte to a byte, all four ways.
209 ($a = v300.196.172.300.196.172) =~ tr/\xc4/\xc5/;
210 is($a, v300.197.172.300.197.172, 'byte2byte transliteration');
212 ($a = v300.196.172.300.196.172) =~ tr/\xc4/\x{c5}/;
213 is($a, v300.197.172.300.197.172);
215 ($a = v300.196.172.300.196.172) =~ tr/\x{c4}/\xc5/;
216 is($a, v300.197.172.300.197.172);
218 ($a = v300.196.172.300.196.172) =~ tr/\x{c4}/\x{c5}/;
219 is($a, v300.197.172.300.197.172);
222 ($a = v300.196.172.300.196.172) =~ tr/\xc4/\x{12d}/;
223 is($a, v300.301.172.300.301.172, 'byte2wide transliteration');
225 ($a = v300.196.172.300.196.172) =~ tr/\x{12c}/\xc3/;
226 is($a, v195.196.172.195.196.172, ' wide2byte');
228 ($a = v300.196.172.300.196.172) =~ tr/\x{12c}/\x{12d}/;
229 is($a, v301.196.172.301.196.172, ' wide2wide');
232 ($a = v300.196.172.300.196.172) =~ tr/\xc4\x{12c}/\x{12d}\xc3/;
233 is($a, v195.301.172.195.301.172, 'byte2wide & wide2byte');
236 ($a = v300.196.172.300.196.172.400.198.144) =~
237 tr/\xac\xc4\x{12c}\x{190}/\xad\x{12d}\xc5\x{191}/;
238 is($a, v197.301.173.197.301.173.401.198.144, 'all together now!');
241 is((($a = v300.196.172.300.196.172) =~ tr/\xc4/\xc5/), 2,
242 'transliterate and count');
244 is((($a = v300.196.172.300.196.172) =~ tr/\x{12c}/\x{12d}/), 2);
247 ($a = v300.196.172.300.196.172) =~ tr/\xc4/\x{12d}/c;
248 is($a, v301.196.301.301.196.301, 'translit w/complement');
250 ($a = v300.196.172.300.196.172) =~ tr/\x{12c}/\xc5/c;
251 is($a, v300.197.197.300.197.197);
254 ($a = v300.196.172.300.196.172) =~ tr/\xc4//d;
255 is($a, v300.172.300.172, 'translit w/deletion');
257 ($a = v300.196.172.300.196.172) =~ tr/\x{12c}//d;
258 is($a, v196.172.196.172);
261 ($a = v196.196.172.300.300.196.172) =~ tr/\xc4/\xc5/s;
262 is($a, v197.172.300.300.197.172, 'translit w/squeeze');
264 ($a = v196.172.300.300.196.172.172) =~ tr/\x{12c}/\x{12d}/s;
265 is($a, v196.172.301.196.172.172);
268 # Tricky cases (When Simon Cozens Attacks)
269 ($a = v196.172.200) =~ tr/\x{12c}/a/;
270 is(sprintf("%vd", $a), '196.172.200');
272 ($a = v196.172.200) =~ tr/\x{12c}/\x{12c}/;
273 is(sprintf("%vd", $a), '196.172.200');
275 ($a = v196.172.200) =~ tr/\x{12c}//d;
276 is(sprintf("%vd", $a), '196.172.200');
279 # UTF8 range tests from Inaba Hiroto
281 # Not working in EBCDIC as of 12674.
282 ($a = v300.196.172.302.197.172) =~ tr/\x{12c}-\x{130}/\xc0-\xc4/;
283 is($a, v192.196.172.194.197.172, 'UTF range');
285 ($a = v300.196.172.302.197.172) =~ tr/\xc4-\xc8/\x{12c}-\x{130}/;
286 is($a, v300.300.172.302.301.172);
289 # UTF8 range tests from Karsten Sperling (patch #9008 required)
291 ($a = "\x{0100}") =~ tr/\x00-\x{100}/X/;
294 ($a = "\x{0100}") =~ tr/\x{0000}-\x{00ff}/X/c;
297 ($a = "\x{0100}") =~ tr/\x{0000}-\x{00ff}\x{0101}/X/c;
300 ($a = v256) =~ tr/\x{0000}-\x{00ff}\x{0101}/X/c;
304 # UTF8 range tests from Inaba Hiroto
306 ($a = "\x{200}") =~ tr/\x00-\x{100}/X/c;
309 ($a = "\x{200}") =~ tr/\x00-\x{100}/X/cs;
313 # Tricky on EBCDIC: while [a-z] [A-Z] must not match the gap characters,
314 # (i-j, r-s, I-J, R-S), [\x89-\x91] [\xc9-\xd1] has to match them,
315 # from Karsten Sperling.
317 $c = ($a = "\x89\x8a\x8b\x8c\x8d\x8f\x90\x91") =~ tr/\x89-\x91/X/;
321 $c = ($a = "\xc9\xca\xcb\xcc\xcd\xcf\xd0\xd1") =~ tr/\xc9-\xd1/X/;
326 skip "not EBCDIC", 4 unless $Is_EBCDIC;
328 $c = ($a = "\x89\x8a\x8b\x8c\x8d\x8f\x90\x91") =~ tr/i-j/X/;
330 is($a, "X\x8a\x8b\x8c\x8d\x8f\x90X");
332 $c = ($a = "\xc9\xca\xcb\xcc\xcd\xcf\xd0\xd1") =~ tr/I-J/X/;
334 is($a, "X\xca\xcb\xcc\xcd\xcf\xd0X");
337 ($a = "\x{100}") =~ tr/\x00-\xff/X/c;
338 is(ord($a), ord("X"));
340 ($a = "\x{100}") =~ tr/\x00-\xff/X/cs;
341 is(ord($a), ord("X"));
343 ($a = "\x{100}\x{100}") =~ tr/\x{101}-\x{200}//c;
344 is($a, "\x{100}\x{100}");
346 ($a = "\x{100}\x{100}") =~ tr/\x{101}-\x{200}//cs;
349 $a = "\xfe\xff"; $a =~ tr/\xfe\xff/\x{1ff}\x{1fe}/;
350 is($a, "\x{1ff}\x{1fe}");
354 ($a = "R0_001") =~ tr/R_//d;
358 @a = (1,2); map { y/1/./ for $_ } @a;
361 @a = (1,2); map { y/1/./ for $_.'' } @a;
365 # Additional test for Inaba Hiroto patch (robin@kitsite.com)
366 ($a = "\x{100}\x{102}\x{101}") =~ tr/\x00-\377/XYZ/c;
370 # Used to fail with "Modification of a read-only value attempted"
374 is($_, 'n', 'pp_trans needs to unshare shared hash keys');
375 is($@, '', ' no error');
379 $x = eval '"1213" =~ tr/1/1/';
380 is($x, 2, 'implicit count on constant');
381 is($@, '', ' no error');
385 eval '$foo[-1] =~ tr/N/N/';
386 is( $@, '', 'implicit count outside array bounds, index negative' );
387 is( scalar @foo, 0, " doesn't extend the array");
389 eval '$foo[1] =~ tr/N/N/';
390 is( $@, '', 'implicit count outside array bounds, index positive' );
391 is( scalar @foo, 0, " doesn't extend the array");
395 eval '$foo{bar} =~ tr/N/N/';
396 is( $@, '', 'implicit count outside hash bounds' );
397 is( scalar keys %foo, 0, " doesn't extend the hash");
400 is( $x =~ tr/A/A/, 2, 'non-modifying tr/// on a scalar ref' );
401 is( ref $x, 'SCALAR', " doesn't stringify its argument" );
403 # rt.perl.org 36622. Perl didn't like a y/// at end of file. No trailing
405 fresh_perl_is(q[$_ = "foo"; y/A-Z/a-z/], '');
408 { # [perl #38293] chr(65535) should be allowed in regexes
409 no warnings 'utf8'; # to allow non-characters
411 $s = "\x{d800}\x{ffff}";
413 is($s, "\x{d800}\x{ffff}", "do_trans_simple");
415 $s = "\x{d800}\x{ffff}";
417 is($i, 0, "do_trans_count");
419 $s = "\x{d800}\x{ffff}";
421 is($s, "\x{d800}\x{ffff}", "do_trans_complex, SQUASH");
423 $s = "\x{d800}\x{ffff}";
425 is($s, "AA", "do_trans_complex, COMPLEMENT");
428 $s =~ tr/\x{ffff}/\x{1ffff}/;
429 is($s, "A\x{1ffff}B", "utf8, SEARCHLIST");
431 $s = "\x{fffd}\x{fffe}\x{ffff}";
432 $s =~ tr/\x{fffd}-\x{ffff}/ABC/;
433 is($s, "ABC", "utf8, SEARCHLIST range");
436 $s =~ tr/ABC/\x{ffff}/;
437 is($s, "\x{ffff}"x3, "utf8, REPLACEMENTLIST");
440 $s =~ tr/ABC/\x{fffd}-\x{ffff}/;
441 is($s, "\x{fffd}\x{fffe}\x{ffff}", "utf8, REPLACEMENTLIST range");
443 $s = "A\x{ffff}B\x{100}\0\x{fffe}\x{ffff}";
444 $i = $s =~ tr/\x{ffff}//;
445 is($i, 2, "utf8, count");
447 $s = "A\x{ffff}\x{ffff}C";
448 $s =~ tr/\x{ffff}/\x{100}/s;
449 is($s, "A\x{100}C", "utf8, SQUASH");
451 $s = "A\x{ffff}\x{ffff}\x{fffe}\x{fffe}\x{fffe}C";
452 $s =~ tr/\x{fffe}\x{ffff}//s;
453 is($s, "A\x{ffff}\x{fffe}C", "utf8, SQUASH");
456 $s =~ tr/AB/\x{ffff}/s;
457 is($s, "x\x{ffff}y", "utf8, SQUASH");
460 $s =~ tr/AB/\x{fffe}\x{ffff}/s;
461 is($s, "x\x{fffe}\x{ffff}y", "utf8, SQUASH");
463 $s = "A\x{ffff}B\x{fffe}C";
464 $s =~ tr/\x{fffe}\x{ffff}/x/c;
465 is($s, "x\x{ffff}x\x{fffe}x", "utf8, COMPLEMENT");
467 $s = "A\x{10000}B\x{2abcd}C";
468 $s =~ tr/\0-\x{ffff}/x/c;
469 is($s, "AxBxC", "utf8, COMPLEMENT range");
471 $s = "A\x{fffe}B\x{ffff}C";
472 $s =~ tr/\x{fffe}\x{ffff}/x/d;
473 is($s, "AxBC", "utf8, DELETE");
475 } # non-characters end
477 { # related to [perl #27940]
480 ($c = "\x20\c@\x30\cA\x40\cZ\x50\c_\x60") =~ tr/\c@-\c_//d;
481 is($c, "\x20\x30\x40\x50\x60", "tr/\\c\@-\\c_//d");
483 ($c = "\x20\x00\x30\x01\x40\x1A\x50\x1F\x60") =~ tr/\x00-\x1f//d;
484 is($c, "\x20\x30\x40\x50\x60", "tr/\\x00-\\x1f//d");
487 ($s) = keys %{{pie => 3}};
489 if (!eval { require B }) { skip "no B", 1 }
490 my $wasro = B::svref_2object(\$s)->FLAGS & &B::SVf_READONLY;
491 $wasro or local $TODO = "didn't have a COW";
493 ok( B::svref_2object(\$s)->FLAGS & &B::SVf_READONLY,
494 "count-only tr doesn't deCOW COWs" );
499 # under threads, unicode tr within a cloned closure would SEGV or assert
500 # fail, since the pointer in the pad to the swash was getting zeroed out
506 $x =~ tr[\x{142}][\x{143}];
508 is($x,"\x{143}", "utf8 + closure");
511 # Freeing of trans ops prior to pmtrans() [perl #102858].
512 eval q{ $a ~= tr/a/b/; };
515 skip "no encoding", 1 unless eval { require encoding; 1 };
516 eval q{ use encoding "utf8"; $a ~= tr/a/b/; };
525 is($x, "Perlβ", "Only first of multiple transliterations is used");