summaryrefslogtreecommitdiff
path: root/locale.c
blob: 7a4a4186c9e1e1f3498c27b8c9d1aa7fcbc24ecf (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
/*    locale.c
 *
 *    Copyright (C) 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001,
 *    2002, 2003, 2005, 2006, 2007, 2008 by Larry Wall and others
 *
 *    You may distribute under the terms of either the GNU General Public
 *    License or the Artistic License, as specified in the README file.
 *
 */

/*
 *      A Elbereth Gilthoniel,
 *      silivren penna míriel
 *      o menel aglar elenath!
 *      Na-chaered palan-díriel
 *      o galadhremmin ennorath,
 *      Fanuilos, le linnathon
 *      nef aear, si nef aearon!
 *
 *     [p.238 of _The Lord of the Rings_, II/i: "Many Meetings"]
 */

/* utility functions for handling locale-specific stuff like what
 * character represents the decimal point.
 *
 * All C programs have an underlying locale.  Perl generally doesn't pay any
 * attention to it except within the scope of a 'use locale'.  For most
 * categories, it accomplishes this by just using different operations if it is
 * in such scope than if not.  However, various libc functions called by Perl
 * are affected by the LC_NUMERIC category, so there are macros in perl.h that
 * are used to toggle between the current locale and the C locale depending on
 * the desired behavior of those functions at the moment.
 */

#include "EXTERN.h"
#define PERL_IN_LOCALE_C
#include "perl.h"

#ifdef I_LANGINFO
#   include <langinfo.h>
#endif

#include "reentr.h"

#ifdef USE_LOCALE

/*
 * Standardize the locale name from a string returned by 'setlocale', possibly
 * modifying that string.
 *
 * The typical return value of setlocale() is either
 * (1) "xx_YY" if the first argument of setlocale() is not LC_ALL
 * (2) "xa_YY xb_YY ..." if the first argument of setlocale() is LC_ALL
 *     (the space-separated values represent the various sublocales,
 *      in some unspecified order).  This is not handled by this function.
 *
 * In some platforms it has a form like "LC_SOMETHING=Lang_Country.866\n",
 * which is harmful for further use of the string in setlocale().  This
 * function removes the trailing new line and everything up through the '='
 *
 */
STATIC char *
S_stdize_locale(pTHX_ char *locs)
{
    const char * const s = strchr(locs, '=');
    bool okay = TRUE;

    PERL_ARGS_ASSERT_STDIZE_LOCALE;

    if (s) {
	const char * const t = strchr(s, '.');
	okay = FALSE;
	if (t) {
	    const char * const u = strchr(t, '\n');
	    if (u && (u[1] == 0)) {
		const STRLEN len = u - s;
		Move(s + 1, locs, len, char);
		locs[len] = 0;
		okay = TRUE;
	    }
	}
    }

    if (!okay)
	Perl_croak(aTHX_ "Can't fix broken locale name \"%s\"", locs);

    return locs;
}

#endif

void
Perl_set_numeric_radix(pTHX)
{
#ifdef USE_LOCALE_NUMERIC
# ifdef HAS_LOCALECONV
    const struct lconv* const lc = localeconv();

    if (lc && lc->decimal_point) {
	if (lc->decimal_point[0] == '.' && lc->decimal_point[1] == 0) {
	    SvREFCNT_dec(PL_numeric_radix_sv);
	    PL_numeric_radix_sv = NULL;
	}
	else {
	    if (PL_numeric_radix_sv)
		sv_setpv(PL_numeric_radix_sv, lc->decimal_point);
	    else
		PL_numeric_radix_sv = newSVpv(lc->decimal_point, 0);
            if (! is_invariant_string((U8 *) lc->decimal_point, 0)
                && is_utf8_string((U8 *) lc->decimal_point, 0)
                && _is_cur_LC_category_utf8(LC_NUMERIC))
            {
		SvUTF8_on(PL_numeric_radix_sv);
            }
	}
    }
    else
	PL_numeric_radix_sv = NULL;

    DEBUG_L(PerlIO_printf(Perl_debug_log, "Locale radix is %s\n",
                                          (PL_numeric_radix_sv)
                                          ? lc->decimal_point
                                          : "NULL"));

# endif /* HAS_LOCALECONV */
#endif /* USE_LOCALE_NUMERIC */
}

/* Is the C string input 'name' "C" or "POSIX"?  If so, and 'name' is the
 * return of setlocale(), then this is extremely likely to be the C or POSIX
 * locale.  However, the output of setlocale() is documented to be opaque, but
 * the odds are extremely small that it would return these two strings for some
 * other locale.  Note that VMS in these two locales includes many non-ASCII
 * characters as controls and punctuation (below are hex bytes):
 *   cntrl:  00-1F 7F 84-97 9B-9F
 *   punct:  21-2F 3A-40 5B-60 7B-7E A1-A3 A5 A7-AB B0-B3 B5-B7 B9-BD BF-CF D1-DD DF-EF F1-FD
 * Oddly, none there are listed as alphas, though some represent alphabetics
 * http://www.nntp.perl.org/group/perl.perl5.porters/2013/02/msg198753.html */
#define isNAME_C_OR_POSIX(name) ((name) != NULL                                 \
                                  && ((*(name) == 'C' && (*(name + 1)) == '\0') \
                                       || strEQ((name), "POSIX")))

void
Perl_new_numeric(pTHX_ const char *newnum)
{
#ifdef USE_LOCALE_NUMERIC

    /* Called after all libc setlocale() calls affecting LC_NUMERIC, to tell
     * core Perl this and that 'newnum' is the name of the new locale.
     * It installs this locale as the current underlying default.
     *
     * The default locale and the C locale can be toggled between by use of the
     * set_numeric_local() and set_numeric_standard() functions, which should
     * probably not be called directly, but only via macros like
     * SET_NUMERIC_STANDARD() in perl.h.
     *
     * The toggling is necessary mainly so that a non-dot radix decimal point
     * character can be output, while allowing internal calculations to use a
     * dot.
     *
     * This sets several interpreter-level variables:
     * PL_numeric_name  The underlying locale's name: a copy of 'newnum'
     * PL_numeric_local A boolean indicating if the toggled state is such
     *                  that the current locale is the program's underlying
     *                  locale
     * PL_numeric_standard An int indicating if the toggled state is such
     *                  that the current locale is the C locale.  If non-zero,
     *                  it is in C; if > 1, it means it may not be toggled away
     *                  from C.
     * Note that both of the last two variables can be true at the same time,
     * if the underlying locale is C.  (Toggling is a no-op under these
     * circumstances.)
     *
     * Any code changing the locale (outside this file) should use
     * POSIX::setlocale, which calls this function.  Therefore this function
     * should be called directly only from this file and from
     * POSIX::setlocale() */

    char *save_newnum;

    if (! newnum) {
	Safefree(PL_numeric_name);
	PL_numeric_name = NULL;
	PL_numeric_standard = TRUE;
	PL_numeric_local = TRUE;
	return;
    }

    save_newnum = stdize_locale(savepv(newnum));
    if (! PL_numeric_name || strNE(PL_numeric_name, save_newnum)) {
	Safefree(PL_numeric_name);
	PL_numeric_name = save_newnum;
    }

    PL_numeric_standard = isNAME_C_OR_POSIX(save_newnum);
    PL_numeric_local = TRUE;

    /* Keep LC_NUMERIC in the C locale.  This is for XS modules, so they don't
     * have to worry about the radix being a non-dot.  (Core operations that
     * need the underlying locale change to it temporarily). */
    set_numeric_standard();

    set_numeric_radix();

#else
    PERL_UNUSED_ARG(newnum);
#endif /* USE_LOCALE_NUMERIC */
}

void
Perl_set_numeric_standard(pTHX)
{
#ifdef USE_LOCALE_NUMERIC
    /* Toggle the LC_NUMERIC locale to C.  Most code should use the macros like
     * SET_NUMERIC_STANDARD() in perl.h instead of calling this directly.  The
     * macro avoids calling this routine if toggling isn't necessary according
     * to our records (which could be wrong if some XS code has changed the
     * locale behind our back) */

    setlocale(LC_NUMERIC, "C");
    PL_numeric_standard = TRUE;
    PL_numeric_local = isNAME_C_OR_POSIX(PL_numeric_name);
    set_numeric_radix();
    DEBUG_L(PerlIO_printf(Perl_debug_log,
                          "Underlying LC_NUMERIC locale now is C\n"));

#endif /* USE_LOCALE_NUMERIC */
}

void
Perl_set_numeric_local(pTHX)
{
#ifdef USE_LOCALE_NUMERIC
    /* Toggle the LC_NUMERIC locale to the current underlying default.  Most
     * code should use the macros like SET_NUMERIC_LOCAL() in perl.h instead of
     * calling this directly.  The macro avoids calling this routine if
     * toggling isn't necessary according to our records (which could be wrong
     * if some XS code has changed the locale behind our back) */

    setlocale(LC_NUMERIC, PL_numeric_name);
    PL_numeric_standard = isNAME_C_OR_POSIX(PL_numeric_name);
    PL_numeric_local = TRUE;
    set_numeric_radix();
    DEBUG_L(PerlIO_printf(Perl_debug_log,
                          "Underlying LC_NUMERIC locale now is %s\n",
                          PL_numeric_name));

#endif /* USE_LOCALE_NUMERIC */
}

/*
 * Set up for a new ctype locale.
 */
void
Perl_new_ctype(pTHX_ const char *newctype)
{
#ifdef USE_LOCALE_CTYPE

    /* Called after all libc setlocale() calls affecting LC_CTYPE, to tell
     * core Perl this and that 'newctype' is the name of the new locale.
     *
     * This function sets up the folding arrays for all 256 bytes, assuming
     * that tofold() is tolc() since fold case is not a concept in POSIX,
     *
     * Any code changing the locale (outside this file) should use
     * POSIX::setlocale, which calls this function.  Therefore this function
     * should be called directly only from this file and from
     * POSIX::setlocale() */

    dVAR;
    UV i;

    PERL_ARGS_ASSERT_NEW_CTYPE;

    /* We will replace any bad locale warning with 1) nothing if the new one is
     * ok; or 2) a new warning for the bad new locale */
    if (PL_warn_locale) {
        SvREFCNT_dec_NN(PL_warn_locale);
        PL_warn_locale = NULL;
    }

    PL_in_utf8_CTYPE_locale = _is_cur_LC_category_utf8(LC_CTYPE);

    /* A UTF-8 locale gets standard rules.  But note that code still has to
     * handle this specially because of the three problematic code points */
    if (PL_in_utf8_CTYPE_locale) {
        Copy(PL_fold_latin1, PL_fold_locale, 256, U8);
    }
    else {
        /* Assume enough space for every character being bad.  4 spaces each
         * for the 94 printable characters that are output like "'x' "; and 5
         * spaces each for "'\\' ", "'\t' ", and "'\n' "; plus a terminating
         * NUL */
        char bad_chars_list[ (94 * 4) + (3 * 5) + 1 ];

        bool check_for_problems = ckWARN_d(WARN_LOCALE); /* No warnings means
                                                            no check */
        bool multi_byte_locale = FALSE;     /* Assume is a single-byte locale
                                               to start */
        unsigned int bad_count = 0;         /* Count of bad characters */

        for (i = 0; i < 256; i++) {
            if (isUPPER_LC((U8) i))
                PL_fold_locale[i] = (U8) toLOWER_LC((U8) i);
            else if (isLOWER_LC((U8) i))
                PL_fold_locale[i] = (U8) toUPPER_LC((U8) i);
            else
                PL_fold_locale[i] = (U8) i;

            /* If checking for locale problems, see if the native ASCII-range
             * printables plus \n and \t are in their expected categories in
             * the new locale.  If not, this could mean big trouble, upending
             * Perl's and most programs' assumptions, like having a
             * metacharacter with special meaning become a \w.  Fortunately,
             * it's very rare to find locales that aren't supersets of ASCII
             * nowadays.  It isn't a problem for most controls to be changed
             * into something else; we check only \n and \t, though perhaps \r
             * could be an issue as well. */
            if (check_for_problems
                && (isGRAPH_A(i) || isBLANK_A(i) || i == '\n'))
            {
                if ((isALPHANUMERIC_A(i) && ! isALPHANUMERIC_LC(i))
                     || (isPUNCT_A(i) && ! isPUNCT_LC(i))
                     || (isBLANK_A(i) && ! isBLANK_LC(i))
                     || (i == '\n' && ! isCNTRL_LC(i)))
                {
                    if (bad_count) {    /* Separate multiple entries with a
                                           blank */
                        bad_chars_list[bad_count++] = ' ';
                    }
                    bad_chars_list[bad_count++] = '\'';
                    if (isPRINT_A(i)) {
                        bad_chars_list[bad_count++] = (char) i;
                    }
                    else {
                        bad_chars_list[bad_count++] = '\\';
                        if (i == '\n') {
                            bad_chars_list[bad_count++] = 'n';
                        }
                        else {
                            assert(i == '\t');
                            bad_chars_list[bad_count++] = 't';
                        }
                    }
                    bad_chars_list[bad_count++] = '\'';
                    bad_chars_list[bad_count] = '\0';
                }
            }
        }

#ifdef MB_CUR_MAX
        /* We only handle single-byte locales (outside of UTF-8 ones; so if
         * this locale requires than one byte, there are going to be
         * problems. */
        if (check_for_problems && MB_CUR_MAX > 1

               /* Some platforms return MB_CUR_MAX > 1 for even the "C"
                * locale.  Just assume that the implementation for them (plus
                * for POSIX) is correct and the > 1 value is spurious.  (Since
                * these are specially handled to never be considered UTF-8
                * locales, as long as this is the only problem, everything
                * should work fine */
            && strNE(newctype, "C") && strNE(newctype, "POSIX"))
        {
            multi_byte_locale = TRUE;
        }
#endif

        if (bad_count || multi_byte_locale) {
            PL_warn_locale = Perl_newSVpvf(aTHX_
                             "Locale '%s' may not work well.%s%s%s\n",
                             newctype,
                             (multi_byte_locale)
                              ? "  Some characters in it are not recognized by"
                                " Perl."
                              : "",
                             (bad_count)
                              ? "\nThe following characters (and maybe others)"
                                " may not have the same meaning as the Perl"
                                " program expects:\n"
                              : "",
                             (bad_count)
                              ? bad_chars_list
                              : ""
                            );
            /* If we are actually in the scope of the locale, output the
             * message now.  Otherwise we save it to be output at the first
             * operation using this locale, if that actually happens.  Most
             * programs don't use locales, so they are immune to bad ones */
            if (IN_LC(LC_CTYPE)) {

                /* We have to save 'newctype' because the setlocale() just
                 * below may destroy it.  The next setlocale() further down
                 * should restore it properly so that the intermediate change
                 * here is transparent to this function's caller */
                const char * const badlocale = savepv(newctype);

                setlocale(LC_CTYPE, "C");

                /* The '0' below suppresses a bogus gcc compiler warning */
                Perl_warner(aTHX_ packWARN(WARN_LOCALE), SvPVX(PL_warn_locale), 0);
                setlocale(LC_CTYPE, badlocale);
                Safefree(badlocale);
                SvREFCNT_dec_NN(PL_warn_locale);
                PL_warn_locale = NULL;
            }
        }
    }

#endif /* USE_LOCALE_CTYPE */
    PERL_ARGS_ASSERT_NEW_CTYPE;
    PERL_UNUSED_ARG(newctype);
    PERL_UNUSED_CONTEXT;
}

void
Perl__warn_problematic_locale()
{

#ifdef USE_LOCALE_CTYPE

    dTHX;

    /* Internal-to-core function that outputs the message in PL_warn_locale,
     * and then NULLS it.  Should be called only through the macro
     * _CHECK_AND_WARN_PROBLEMATIC_LOCALE */

    if (PL_warn_locale) {
        /*GCC_DIAG_IGNORE(-Wformat-security);   Didn't work */
        Perl_ck_warner(aTHX_ packWARN(WARN_LOCALE),
                             SvPVX(PL_warn_locale),
                             0 /* dummy to avoid compiler warning */ );
        /* GCC_DIAG_RESTORE; */
        SvREFCNT_dec_NN(PL_warn_locale);
        PL_warn_locale = NULL;
    }

#endif

}

void
Perl_new_collate(pTHX_ const char *newcoll)
{
#ifdef USE_LOCALE_COLLATE

    /* Called after all libc setlocale() calls affecting LC_COLLATE, to tell
     * core Perl this and that 'newcoll' is the name of the new locale.
     *
     * Any code changing the locale (outside this file) should use
     * POSIX::setlocale, which calls this function.  Therefore this function
     * should be called directly only from this file and from
     * POSIX::setlocale() */

    if (! newcoll) {
	if (PL_collation_name) {
	    ++PL_collation_ix;
	    Safefree(PL_collation_name);
	    PL_collation_name = NULL;
	}
	PL_collation_standard = TRUE;
	PL_collxfrm_base = 0;
	PL_collxfrm_mult = 2;
	return;
    }

    if (! PL_collation_name || strNE(PL_collation_name, newcoll)) {
	++PL_collation_ix;
	Safefree(PL_collation_name);
	PL_collation_name = stdize_locale(savepv(newcoll));
	PL_collation_standard = isNAME_C_OR_POSIX(newcoll);

	{
	  /*  2: at most so many chars ('a', 'b'). */
	  /* 50: surely no system expands a char more. */
#define XFRMBUFSIZE  (2 * 50)
	  char xbuf[XFRMBUFSIZE];
	  const Size_t fa = strxfrm(xbuf, "a",  XFRMBUFSIZE);
	  const Size_t fb = strxfrm(xbuf, "ab", XFRMBUFSIZE);
	  const SSize_t mult = fb - fa;
	  if (mult < 1 && !(fa == 0 && fb == 0))
	      Perl_croak(aTHX_ "panic: strxfrm() gets absurd - a => %"UVuf", ab => %"UVuf,
			 (UV) fa, (UV) fb);
	  PL_collxfrm_base = (fa > (Size_t)mult) ? (fa - mult) : 0;
	  PL_collxfrm_mult = mult;
	}
    }

#else
    PERL_UNUSED_ARG(newcoll);
#endif /* USE_LOCALE_COLLATE */
}

#ifdef WIN32

char *
Perl_my_setlocale(pTHX_ int category, const char* locale)
{
    /* This, for Windows, emulates POSIX setlocale() behavior.  There is no
     * difference unless the input locale is "", which means on Windows to get
     * the machine default, which is set via the computer's "Regional and
     * Language Options" (or its current equivalent).  In POSIX, it instead
     * means to find the locale from the user's environment.  This routine
     * looks in the environment, and, if anything is found, uses that instead
     * of going to the machine default.  If there is no environment override,
     * the machine default is used, as normal, by calling the real setlocale()
     * with "".  The POSIX behavior is to use the LC_ALL variable if set;
     * otherwise to use the particular category's variable if set; otherwise to
     * use the LANG variable. */

    bool override_LC_ALL = FALSE;
    char * result;

    if (locale && strEQ(locale, "")) {
#   ifdef LC_ALL
        locale = PerlEnv_getenv("LC_ALL");
        if (! locale) {
#endif
            switch (category) {
#   ifdef LC_ALL
                case LC_ALL:
                    override_LC_ALL = TRUE;
                    break;  /* We already know its variable isn't set */
#   endif
#   ifdef USE_LOCALE_TIME
                case LC_TIME:
                    locale = PerlEnv_getenv("LC_TIME");
                    break;
#   endif
#   ifdef USE_LOCALE_CTYPE
                case LC_CTYPE:
                    locale = PerlEnv_getenv("LC_CTYPE");
                    break;
#   endif
#   ifdef USE_LOCALE_COLLATE
                case LC_COLLATE:
                    locale = PerlEnv_getenv("LC_COLLATE");
                    break;
#   endif
#   ifdef USE_LOCALE_MONETARY
                case LC_MONETARY:
                    locale = PerlEnv_getenv("LC_MONETARY");
                    break;
#   endif
#   ifdef USE_LOCALE_NUMERIC
                case LC_NUMERIC:
                    locale = PerlEnv_getenv("LC_NUMERIC");
                    break;
#   endif
#   ifdef USE_LOCALE_MESSAGES
                case LC_MESSAGES:
                    locale = PerlEnv_getenv("LC_MESSAGES");
                    break;
#   endif
                default:
                    /* This is a category, like PAPER_SIZE that we don't
                     * know about; and so can't provide a wrapper. */
                    break;
            }
            if (! locale) {
                locale = PerlEnv_getenv("LANG");
                if (! locale) {
                    locale = "";
                }
            }
#   ifdef LC_ALL
        }
#   endif
    }

    result = setlocale(category, locale);

    if (! override_LC_ALL)  {
        return result;
    }

    /* Here the input category was LC_ALL, and we have set it to what is in the
     * LANG variable or the system default if there is no LANG.  But these have
     * lower priority than the other LC_foo variables, so override it for each
     * one that is set.  (If they are set to "", it means to use the same thing
     * we just set LC_ALL to, so can skip) */
#   ifdef USE_LOCALE_TIME
    result = PerlEnv_getenv("LC_TIME");
    if (result && strNE(result, "")) {
        setlocale(LC_TIME, result);
    }
#   endif
#   ifdef USE_LOCALE_CTYPE
    result = PerlEnv_getenv("LC_CTYPE");
    if (result && strNE(result, "")) {
        setlocale(LC_CTYPE, result);
    }
#   endif
#   ifdef USE_LOCALE_COLLATE
    result = PerlEnv_getenv("LC_COLLATE");
    if (result && strNE(result, "")) {
        setlocale(LC_COLLATE, result);
    }
#   endif
#   ifdef USE_LOCALE_MONETARY
    result = PerlEnv_getenv("LC_MONETARY");
    if (result && strNE(result, "")) {
        setlocale(LC_MONETARY, result);
    }
#   endif
#   ifdef USE_LOCALE_NUMERIC
    result = PerlEnv_getenv("LC_NUMERIC");
    if (result && strNE(result, "")) {
        setlocale(LC_NUMERIC, result);
    }
#   endif
#   ifdef USE_LOCALE_MESSAGES
    result = PerlEnv_getenv("LC_MESSAGES");
    if (result && strNE(result, "")) {
        setlocale(LC_MESSAGES, result);
    }
#   endif

    return setlocale(LC_ALL, NULL);

}

#endif


/*
 * Initialize locale awareness.
 */
int
Perl_init_i18nl10n(pTHX_ int printwarn)
{
    /* printwarn is
     *
     *    0 if not to output warning when setup locale is bad
     *    1 if to output warning based on value of PERL_BADLANG
     *    >1 if to output regardless of PERL_BADLANG
     *
     * returns
     *    1 = set ok or not applicable,
     *    0 = fallback to a locale of lower priority
     *   -1 = fallback to all locales failed, not even to the C locale
     */

    int ok = 1;

#if defined(USE_LOCALE)
#ifdef USE_LOCALE_CTYPE
    char *curctype   = NULL;
#endif /* USE_LOCALE_CTYPE */
#ifdef USE_LOCALE_COLLATE
    char *curcoll    = NULL;
#endif /* USE_LOCALE_COLLATE */
#ifdef USE_LOCALE_NUMERIC
    char *curnum     = NULL;
#endif /* USE_LOCALE_NUMERIC */
#ifdef __GLIBC__
    const char * const language   = savepv(PerlEnv_getenv("LANGUAGE"));
#endif

    /* NULL uses the existing already set up locale */
    const char * const setlocale_init = (PerlEnv_getenv("PERL_SKIP_LOCALE_INIT"))
                                        ? NULL
                                        : "";
    const char* trial_locales[5];   /* 5 = 1 each for "", LC_ALL, LANG, "", C */
    unsigned int trial_locales_count;
    const char * const lc_all     = savepv(PerlEnv_getenv("LC_ALL"));
    const char * const lang       = savepv(PerlEnv_getenv("LANG"));
    bool setlocale_failure = FALSE;
    unsigned int i;
    char *p;

    /* A later getenv() could zap this, so only use here */
    const char * const bad_lang_use_once = PerlEnv_getenv("PERL_BADLANG");

    const bool locwarn = (printwarn > 1
                          || (printwarn
                              && (! bad_lang_use_once
                                  || (
                                    /* disallow with "" or "0" */
                                    *bad_lang_use_once
                                    && strNE("0", bad_lang_use_once)))));
    bool done = FALSE;
#ifdef WIN32
    /* In some systems you can find out the system default locale
     * and use that as the fallback locale. */
#   define SYSTEM_DEFAULT_LOCALE
#endif
#ifdef SYSTEM_DEFAULT_LOCALE
    const char *system_default_locale = NULL;
#endif

#ifndef LOCALE_ENVIRON_REQUIRED
    PERL_UNUSED_VAR(done);
#else

    /*
     * Ultrix setlocale(..., "") fails if there are no environment
     * variables from which to get a locale name.
     */

#   ifdef LC_ALL
    if (lang) {
	if (my_setlocale(LC_ALL, setlocale_init))
	    done = TRUE;
	else
	    setlocale_failure = TRUE;
    }
    if (!setlocale_failure) {
#       ifdef USE_LOCALE_CTYPE
	Safefree(curctype);
	if (! (curctype =
	       my_setlocale(LC_CTYPE,
			 (!done && (lang || PerlEnv_getenv("LC_CTYPE")))
				    ? setlocale_init : NULL)))
	    setlocale_failure = TRUE;
	else
	    curctype = savepv(curctype);
#       endif /* USE_LOCALE_CTYPE */
#       ifdef USE_LOCALE_COLLATE
	Safefree(curcoll);
	if (! (curcoll =
	       my_setlocale(LC_COLLATE,
			 (!done && (lang || PerlEnv_getenv("LC_COLLATE")))
				   ? setlocale_init : NULL)))
	    setlocale_failure = TRUE;
	else
	    curcoll = savepv(curcoll);
#       endif /* USE_LOCALE_COLLATE */
#       ifdef USE_LOCALE_NUMERIC
	Safefree(curnum);
	if (! (curnum =
	       my_setlocale(LC_NUMERIC,
			 (!done && (lang || PerlEnv_getenv("LC_NUMERIC")))
				  ? setlocale_init : NULL)))
	    setlocale_failure = TRUE;
	else
	    curnum = savepv(curnum);
#       endif /* USE_LOCALE_NUMERIC */
#       ifdef USE_LOCALE_MESSAGES
	if (! my_setlocale(LC_MESSAGES,
			 (!done && (lang || PerlEnv_getenv("LC_MESSAGES")))
				  ? setlocale_init : NULL))
        {
	    setlocale_failure = TRUE;
        }
#       endif /* USE_LOCALE_MESSAGES */
#       ifdef USE_LOCALE_MONETARY
	if (! my_setlocale(LC_MONETARY,
			 (!done && (lang || PerlEnv_getenv("LC_MONETARY")))
				  ? setlocale_init : NULL))
        {
	    setlocale_failure = TRUE;
        }
#       endif /* USE_LOCALE_MONETARY */
    }

#   endif /* LC_ALL */

#endif /* !LOCALE_ENVIRON_REQUIRED */

    /* We try each locale in the list until we get one that works, or exhaust
     * the list */
    trial_locales[0] = setlocale_init;
    trial_locales_count = 1;
    for (i= 0; i < trial_locales_count; i++) {
        const char * trial_locale = trial_locales[i];

        if (i > 0) {

            /* XXX This is to preserve old behavior for LOCALE_ENVIRON_REQUIRED
             * when i==0, but I (khw) don't think that behavior makes much
             * sense */
            setlocale_failure = FALSE;

#ifdef SYSTEM_DEFAULT_LOCALE
#  ifdef WIN32
            /* On Windows machines, an entry of "" after the 0th means to use
             * the system default locale, which we now proceed to get. */
            if (strEQ(trial_locale, "")) {
                unsigned int j;

                /* Note that this may change the locale, but we are going to do
                 * that anyway just below */
                system_default_locale = setlocale(LC_ALL, "");

                /* Skip if invalid or it's already on the list of locales to
                 * try */
                if (! system_default_locale) {
                    goto next_iteration;
                }
                for (j = 0; j < trial_locales_count; j++) {
                    if (strEQ(system_default_locale, trial_locales[j])) {
                        goto next_iteration;
                    }
                }

                trial_locale = system_default_locale;
            }
#  endif /* WIN32 */
#endif /* SYSTEM_DEFAULT_LOCALE */
        }

#ifdef LC_ALL
        if (! my_setlocale(LC_ALL, trial_locale)) {
            setlocale_failure = TRUE;
        }
        else {
            /* Since LC_ALL succeeded, it should have changed all the other
             * categories it can to its value; so we massage things so that the
             * setlocales below just return their category's current values.
             * This adequately handles the case in NetBSD where LC_COLLATE may
             * not be defined for a locale, and setting it individually will
             * fail, whereas setting LC_ALL suceeds, leaving LC_COLLATE set to
             * the POSIX locale. */
            trial_locale = NULL;
        }
#endif /* LC_ALL */

        if (!setlocale_failure) {
#ifdef USE_LOCALE_CTYPE
            Safefree(curctype);
            if (! (curctype = my_setlocale(LC_CTYPE, trial_locale)))
                setlocale_failure = TRUE;
            else
                curctype = savepv(curctype);
#endif /* USE_LOCALE_CTYPE */
#ifdef USE_LOCALE_COLLATE
            Safefree(curcoll);
            if (! (curcoll = my_setlocale(LC_COLLATE, trial_locale)))
                setlocale_failure = TRUE;
            else
                curcoll = savepv(curcoll);
#endif /* USE_LOCALE_COLLATE */
#ifdef USE_LOCALE_NUMERIC
            Safefree(curnum);
            if (! (curnum = my_setlocale(LC_NUMERIC, trial_locale)))
                setlocale_failure = TRUE;
            else
                curnum = savepv(curnum);
#endif /* USE_LOCALE_NUMERIC */
#ifdef USE_LOCALE_MESSAGES
            if (! (my_setlocale(LC_MESSAGES, trial_locale)))
                setlocale_failure = TRUE;
#endif /* USE_LOCALE_MESSAGES */
#ifdef USE_LOCALE_MONETARY
            if (! (my_setlocale(LC_MONETARY, trial_locale)))
                setlocale_failure = TRUE;
#endif /* USE_LOCALE_MONETARY */

            if (! setlocale_failure) {  /* Success */
                break;
            }
        }

        /* Here, something failed; will need to try a fallback. */
        ok = 0;

        if (i == 0) {
            unsigned int j;

            if (locwarn) { /* Output failure info only on the first one */
#ifdef LC_ALL

                PerlIO_printf(Perl_error_log,
                "perl: warning: Setting locale failed.\n");

#else /* !LC_ALL */

                PerlIO_printf(Perl_error_log,
                "perl: warning: Setting locale failed for the categories:\n\t");
#ifdef USE_LOCALE_CTYPE
                if (! curctype)
                    PerlIO_printf(Perl_error_log, "LC_CTYPE ");
#endif /* USE_LOCALE_CTYPE */
#ifdef USE_LOCALE_COLLATE
                if (! curcoll)
                    PerlIO_printf(Perl_error_log, "LC_COLLATE ");
#endif /* USE_LOCALE_COLLATE */
#ifdef USE_LOCALE_NUMERIC
                if (! curnum)
                    PerlIO_printf(Perl_error_log, "LC_NUMERIC ");
#endif /* USE_LOCALE_NUMERIC */
                PerlIO_printf(Perl_error_log, "and possibly others\n");

#endif /* LC_ALL */

                PerlIO_printf(Perl_error_log,
                    "perl: warning: Please check that your locale settings:\n");

#ifdef __GLIBC__
                PerlIO_printf(Perl_error_log,
                            "\tLANGUAGE = %c%s%c,\n",
                            language ? '"' : '(',
                            language ? language : "unset",
                            language ? '"' : ')');
#endif

                PerlIO_printf(Perl_error_log,
                            "\tLC_ALL = %c%s%c,\n",
                            lc_all ? '"' : '(',
                            lc_all ? lc_all : "unset",
                            lc_all ? '"' : ')');

#if defined(USE_ENVIRON_ARRAY)
                {
                char **e;
                for (e = environ; *e; e++) {
                    if (strnEQ(*e, "LC_", 3)
                            && strnNE(*e, "LC_ALL=", 7)
                            && (p = strchr(*e, '=')))
                        PerlIO_printf(Perl_error_log, "\t%.*s = \"%s\",\n",
                                        (int)(p - *e), *e, p + 1);
                }
                }
#else
                PerlIO_printf(Perl_error_log,
                            "\t(possibly more locale environment variables)\n");
#endif

                PerlIO_printf(Perl_error_log,
                            "\tLANG = %c%s%c\n",
                            lang ? '"' : '(',
                            lang ? lang : "unset",
                            lang ? '"' : ')');

                PerlIO_printf(Perl_error_log,
                            "    are supported and installed on your system.\n");
            }

            /* Calculate what fallback locales to try.  We have avoided this
             * until we have to, because failure is quite unlikely.  This will
             * usually change the upper bound of the loop we are in.
             *
             * Since the system's default way of setting the locale has not
             * found one that works, We use Perl's defined ordering: LC_ALL,
             * LANG, and the C locale.  We don't try the same locale twice, so
             * don't add to the list if already there.  (On POSIX systems, the
             * LC_ALL element will likely be a repeat of the 0th element "",
             * but there's no harm done by doing it explicitly */
            if (lc_all) {
                for (j = 0; j < trial_locales_count; j++) {
                    if (strEQ(lc_all, trial_locales[j])) {
                        goto done_lc_all;
                    }
                }
                trial_locales[trial_locales_count++] = lc_all;
            }
          done_lc_all:

            if (lang) {
                for (j = 0; j < trial_locales_count; j++) {
                    if (strEQ(lang, trial_locales[j])) {
                        goto done_lang;
                    }
                }
                trial_locales[trial_locales_count++] = lang;
            }
          done_lang:

#if defined(WIN32) && defined(LC_ALL)
            /* For Windows, we also try the system default locale before "C".
             * (If there exists a Windows without LC_ALL we skip this because
             * it gets too complicated.  For those, the "C" is the next
             * fallback possibility).  The "" is the same as the 0th element of
             * the array, but the code at the loop above knows to treat it
             * differently when not the 0th */
            trial_locales[trial_locales_count++] = "";
#endif

            for (j = 0; j < trial_locales_count; j++) {
                if (strEQ("C", trial_locales[j])) {
                    goto done_C;
                }
            }
            trial_locales[trial_locales_count++] = "C";

          done_C: ;
        }   /* end of first time through the loop */

#ifdef WIN32
      next_iteration: ;
#endif

    }   /* end of looping through the trial locales */

    if (ok < 1) {   /* If we tried to fallback */
        const char* msg;
        if (! setlocale_failure) {  /* fallback succeeded */
           msg = "Falling back to";
        }
        else {  /* fallback failed */

            /* We dropped off the end of the loop, so have to decrement i to
             * get back to the value the last time through */
            i--;

            ok = -1;
            msg = "Failed to fall back to";

            /* To continue, we should use whatever values we've got */
#ifdef USE_LOCALE_CTYPE
            Safefree(curctype);
            curctype = savepv(setlocale(LC_CTYPE, NULL));
#endif /* USE_LOCALE_CTYPE */
#ifdef USE_LOCALE_COLLATE
            Safefree(curcoll);
            curcoll = savepv(setlocale(LC_COLLATE, NULL));
#endif /* USE_LOCALE_COLLATE */
#ifdef USE_LOCALE_NUMERIC
            Safefree(curnum);
            curnum = savepv(setlocale(LC_NUMERIC, NULL));
#endif /* USE_LOCALE_NUMERIC */
        }

        if (locwarn) {
            const char * description;
            const char * name = "";
            if (strEQ(trial_locales[i], "C")) {
                description = "the standard locale";
                name = "C";
            }
#ifdef SYSTEM_DEFAULT_LOCALE
            else if (strEQ(trial_locales[i], "")) {
                description = "the system default locale";
                if (system_default_locale) {
                    name = system_default_locale;
                }
            }
#endif /* SYSTEM_DEFAULT_LOCALE */
            else {
                description = "a fallback locale";
                name = trial_locales[i];
            }
            if (name && strNE(name, "")) {
                PerlIO_printf(Perl_error_log,
                    "perl: warning: %s %s (\"%s\").\n", msg, description, name);
            }
            else {
                PerlIO_printf(Perl_error_log,
                                   "perl: warning: %s %s.\n", msg, description);
            }
        }
    } /* End of tried to fallback */

#ifdef USE_LOCALE_CTYPE
    new_ctype(curctype);
#endif /* USE_LOCALE_CTYPE */

#ifdef USE_LOCALE_COLLATE
    new_collate(curcoll);
#endif /* USE_LOCALE_COLLATE */

#ifdef USE_LOCALE_NUMERIC
    new_numeric(curnum);
#endif /* USE_LOCALE_NUMERIC */

#if defined(USE_PERLIO) && defined(USE_LOCALE_CTYPE)
    /* Set PL_utf8locale to TRUE if using PerlIO _and_ the current LC_CTYPE
     * locale is UTF-8.  If PL_utf8locale and PL_unicode (set by -C or by
     * $ENV{PERL_UNICODE}) are true, perl.c:S_parse_body() will turn on the
     * PerlIO :utf8 layer on STDIN, STDOUT, STDERR, _and_ the default open
     * discipline.  */
    PL_utf8locale = _is_cur_LC_category_utf8(LC_CTYPE);

    /* Set PL_unicode to $ENV{PERL_UNICODE} if using PerlIO.
       This is an alternative to using the -C command line switch
       (the -C if present will override this). */
    {
	 const char *p = PerlEnv_getenv("PERL_UNICODE");
	 PL_unicode = p ? parse_unicode_opts(&p) : 0;
	 if (PL_unicode & PERL_UNICODE_UTF8CACHEASSERT_FLAG)
	     PL_utf8cache = -1;
    }
#endif

#ifdef USE_LOCALE_CTYPE
    Safefree(curctype);
#endif /* USE_LOCALE_CTYPE */
#ifdef USE_LOCALE_COLLATE
    Safefree(curcoll);
#endif /* USE_LOCALE_COLLATE */
#ifdef USE_LOCALE_NUMERIC
    Safefree(curnum);
#endif /* USE_LOCALE_NUMERIC */

#ifdef __GLIBC__
    Safefree(language);
#endif

    Safefree(lc_all);
    Safefree(lang);

#else  /* !USE_LOCALE */
    PERL_UNUSED_ARG(printwarn);
#endif /* USE_LOCALE */

    return ok;
}


#ifdef USE_LOCALE_COLLATE

/*
 * mem_collxfrm() is a bit like strxfrm() but with two important
 * differences. First, it handles embedded NULs. Second, it allocates
 * a bit more memory than needed for the transformed data itself.
 * The real transformed data begins at offset sizeof(collationix).
 * Please see sv_collxfrm() to see how this is used.
 */

char *
Perl_mem_collxfrm(pTHX_ const char *s, STRLEN len, STRLEN *xlen)
{
    char *xbuf;
    STRLEN xAlloc, xin, xout; /* xalloc is a reserved word in VC */

    PERL_ARGS_ASSERT_MEM_COLLXFRM;

    /* the first sizeof(collationix) bytes are used by sv_collxfrm(). */
    /* the +1 is for the terminating NUL. */

    xAlloc = sizeof(PL_collation_ix) + PL_collxfrm_base + (PL_collxfrm_mult * len) + 1;
    Newx(xbuf, xAlloc, char);
    if (! xbuf)
	goto bad;

    *(U32*)xbuf = PL_collation_ix;
    xout = sizeof(PL_collation_ix);
    for (xin = 0; xin < len; ) {
	Size_t xused;

	for (;;) {
	    xused = strxfrm(xbuf + xout, s + xin, xAlloc - xout);
	    if (xused >= PERL_INT_MAX)
		goto bad;
	    if ((STRLEN)xused < xAlloc - xout)
		break;
	    xAlloc = (2 * xAlloc) + 1;
	    Renew(xbuf, xAlloc, char);
	    if (! xbuf)
		goto bad;
	}

	xin += strlen(s + xin) + 1;
	xout += xused;

	/* Embedded NULs are understood but silently skipped
	 * because they make no sense in locale collation. */
    }

    xbuf[xout] = '\0';
    *xlen = xout - sizeof(PL_collation_ix);
    return xbuf;

  bad:
    Safefree(xbuf);
    *xlen = 0;
    return NULL;
}

#endif /* USE_LOCALE_COLLATE */

#ifdef USE_LOCALE

bool
Perl__is_cur_LC_category_utf8(pTHX_ int category)
{
    /* Returns TRUE if the current locale for 'category' is UTF-8; FALSE
     * otherwise. 'category' may not be LC_ALL.  If the platform doesn't have
     * nl_langinfo(), nor MB_CUR_MAX, this employs a heuristic, which hence
     * could give the wrong result.  The result will very likely be correct for
     * languages that have commonly used non-ASCII characters, but for notably
     * English, it comes down to if the locale's name ends in something like
     * "UTF-8".  It errs on the side of not being a UTF-8 locale. */

    char *save_input_locale = NULL;
    STRLEN final_pos;

#ifdef LC_ALL
    assert(category != LC_ALL);
#endif

    /* First dispose of the trivial cases */
    save_input_locale = setlocale(category, NULL);
    if (! save_input_locale) {
        DEBUG_L(PerlIO_printf(Perl_debug_log,
                              "Could not find current locale for category %d\n",
                              category));
        return FALSE;   /* XXX maybe should croak */
    }
    save_input_locale = stdize_locale(savepv(save_input_locale));
    if (isNAME_C_OR_POSIX(save_input_locale)) {
        DEBUG_L(PerlIO_printf(Perl_debug_log,
                              "Current locale for category %d is %s\n",
                              category, save_input_locale));
        Safefree(save_input_locale);
        return FALSE;
    }

#if defined(USE_LOCALE_CTYPE)    \
    && (defined(MB_CUR_MAX) || (defined(HAS_NL_LANGINFO) && defined(CODESET)))

    { /* Next try nl_langinfo or MB_CUR_MAX if available */

        char *save_ctype_locale = NULL;
        bool is_utf8;

        if (category != LC_CTYPE) { /* These work only on LC_CTYPE */

            /* Get the current LC_CTYPE locale */
            save_ctype_locale = setlocale(LC_CTYPE, NULL);
            if (! save_ctype_locale) {
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                               "Could not find current locale for LC_CTYPE\n"));
                goto cant_use_nllanginfo;
            }
            save_ctype_locale = stdize_locale(savepv(save_ctype_locale));

            /* If LC_CTYPE and the desired category use the same locale, this
             * means that finding the value for LC_CTYPE is the same as finding
             * the value for the desired category.  Otherwise, switch LC_CTYPE
             * to the desired category's locale */
            if (strEQ(save_ctype_locale, save_input_locale)) {
                Safefree(save_ctype_locale);
                save_ctype_locale = NULL;
            }
            else if (! setlocale(LC_CTYPE, save_input_locale)) {
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                                    "Could not change LC_CTYPE locale to %s\n",
                                    save_input_locale));
                Safefree(save_ctype_locale);
                goto cant_use_nllanginfo;
            }
        }

        DEBUG_L(PerlIO_printf(Perl_debug_log, "Current LC_CTYPE locale=%s\n",
                                              save_input_locale));

        /* Here the current LC_CTYPE is set to the locale of the category whose
         * information is desired.  This means that nl_langinfo() and MB_CUR_MAX
         * should give the correct results */

#   if defined(HAS_NL_LANGINFO) && defined(CODESET)
        {
            char *codeset = nl_langinfo(CODESET);
            if (codeset && strNE(codeset, "")) {
                codeset = savepv(codeset);

                /* If we switched LC_CTYPE, switch back */
                if (save_ctype_locale) {
                    setlocale(LC_CTYPE, save_ctype_locale);
                    Safefree(save_ctype_locale);
                }

                is_utf8 = foldEQ(codeset, STR_WITH_LEN("UTF-8"))
                        || foldEQ(codeset, STR_WITH_LEN("UTF8"));

                DEBUG_L(PerlIO_printf(Perl_debug_log,
                       "\tnllanginfo returned CODESET '%s'; ?UTF8 locale=%d\n",
                                                     codeset,         is_utf8));
                Safefree(codeset);
                Safefree(save_input_locale);
                return is_utf8;
            }
        }

#   endif
#   ifdef MB_CUR_MAX

        /* Here, either we don't have nl_langinfo, or it didn't return a
         * codeset.  Try MB_CUR_MAX */

        /* Standard UTF-8 needs at least 4 bytes to represent the maximum
         * Unicode code point.  Since UTF-8 is the only non-single byte
         * encoding we handle, we just say any such encoding is UTF-8, and if
         * turns out to be wrong, other things will fail */
        is_utf8 = MB_CUR_MAX >= 4;

        DEBUG_L(PerlIO_printf(Perl_debug_log,
                              "\tMB_CUR_MAX=%d; ?UTF8 locale=%d\n",
                                   (int) MB_CUR_MAX,      is_utf8));

        Safefree(save_input_locale);

#       ifdef HAS_MBTOWC

        /* ... But, most system that have MB_CUR_MAX will also have mbtowc(),
         * since they are both in the C99 standard.  We can feed a known byte
         * string to the latter function, and check that it gives the expected
         * result */
        if (is_utf8) {
            wchar_t wc;
            PERL_UNUSED_RESULT(mbtowc(&wc, NULL, 0));/* Reset any shift state */
            errno = 0;
            if ((size_t)mbtowc(&wc, HYPHEN_UTF8, strlen(HYPHEN_UTF8))
                                                        != strlen(HYPHEN_UTF8)
                || wc != (wchar_t) 0x2010)
            {
                is_utf8 = FALSE;
                DEBUG_L(PerlIO_printf(Perl_debug_log, "\thyphen=U+%x\n", (unsigned int)wc));
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                        "\treturn from mbtowc=%d; errno=%d; ?UTF8 locale=0\n",
                        mbtowc(&wc, HYPHEN_UTF8, strlen(HYPHEN_UTF8)), errno));
            }
        }
#       endif

        /* If we switched LC_CTYPE, switch back */
        if (save_ctype_locale) {
            setlocale(LC_CTYPE, save_ctype_locale);
            Safefree(save_ctype_locale);
        }

        return is_utf8;
#   endif
    }

  cant_use_nllanginfo:

#else   /* nl_langinfo should work if available, so don't bother compiling this
           fallback code.  The final fallback of looking at the name is
           compiled, and will be executed if nl_langinfo fails */

    /* nl_langinfo not available or failed somehow.  Next try looking at the
     * currency symbol to see if it disambiguates things.  Often that will be
     * in the native script, and if the symbol isn't in UTF-8, we know that the
     * locale isn't.  If it is non-ASCII UTF-8, we infer that the locale is
     * too, as the odds of a non-UTF8 string being valid UTF-8 are quite small
     * */

#ifdef HAS_LOCALECONV
#   ifdef USE_LOCALE_MONETARY
    {
        char *save_monetary_locale = NULL;
        bool only_ascii = FALSE;
        bool is_utf8 = FALSE;
        struct lconv* lc;

        /* Like above for LC_CTYPE, we first set LC_MONETARY to the locale of
         * the desired category, if it isn't that locale already */

        if (category != LC_MONETARY) {

            save_monetary_locale = setlocale(LC_MONETARY, NULL);
            if (! save_monetary_locale) {
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                            "Could not find current locale for LC_MONETARY\n"));
                goto cant_use_monetary;
            }
            save_monetary_locale = stdize_locale(savepv(save_monetary_locale));

            if (strEQ(save_monetary_locale, save_input_locale)) {
                Safefree(save_monetary_locale);
                save_monetary_locale = NULL;
            }
            else if (! setlocale(LC_MONETARY, save_input_locale)) {
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                            "Could not change LC_MONETARY locale to %s\n",
                                                        save_input_locale));
                Safefree(save_monetary_locale);
                goto cant_use_monetary;
            }
        }

        /* Here the current LC_MONETARY is set to the locale of the category
         * whose information is desired. */

        lc = localeconv();
        if (! lc
            || ! lc->currency_symbol
            || is_invariant_string((U8 *) lc->currency_symbol, 0))
        {
            DEBUG_L(PerlIO_printf(Perl_debug_log, "Couldn't get currency symbol for %s, or contains only ASCII; can't use for determining if UTF-8 locale\n", save_input_locale));
            only_ascii = TRUE;
        }
        else {
            is_utf8 = is_utf8_string((U8 *) lc->currency_symbol, 0);
        }

        /* If we changed it, restore LC_MONETARY to its original locale */
        if (save_monetary_locale) {
            setlocale(LC_MONETARY, save_monetary_locale);
            Safefree(save_monetary_locale);
        }

        if (! only_ascii) {

            /* It isn't a UTF-8 locale if the symbol is not legal UTF-8;
             * otherwise assume the locale is UTF-8 if and only if the symbol
             * is non-ascii UTF-8. */
            DEBUG_L(PerlIO_printf(Perl_debug_log, "\t?Currency symbol for %s is UTF-8=%d\n",
                                    save_input_locale, is_utf8));
            Safefree(save_input_locale);
            return is_utf8;
        }
    }
  cant_use_monetary:

#   endif /* USE_LOCALE_MONETARY */
#endif /* HAS_LOCALECONV */

#if defined(HAS_STRFTIME) && defined(USE_LOCALE_TIME)

/* Still haven't found a non-ASCII string to disambiguate UTF-8 or not.  Try
 * the names of the months and weekdays, timezone, and am/pm indicator */
    {
        char *save_time_locale = NULL;
        int hour = 10;
        bool is_dst = FALSE;
        int dom = 1;
        int month = 0;
        int i;
        char * formatted_time;


        /* Like above for LC_MONETARY, we set LC_TIME to the locale of the
         * desired category, if it isn't that locale already */

        if (category != LC_TIME) {

            save_time_locale = setlocale(LC_TIME, NULL);
            if (! save_time_locale) {
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                            "Could not find current locale for LC_TIME\n"));
                goto cant_use_time;
            }
            save_time_locale = stdize_locale(savepv(save_time_locale));

            if (strEQ(save_time_locale, save_input_locale)) {
                Safefree(save_time_locale);
                save_time_locale = NULL;
            }
            else if (! setlocale(LC_TIME, save_input_locale)) {
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                            "Could not change LC_TIME locale to %s\n",
                                                        save_input_locale));
                Safefree(save_time_locale);
                goto cant_use_time;
            }
        }

        /* Here the current LC_TIME is set to the locale of the category
         * whose information is desired.  Look at all the days of the week and
         * month names, and the timezone and am/pm indicator for UTF-8 variant
         * characters.  The first such a one found will tell us if the locale
         * is UTF-8 or not */

        for (i = 0; i < 7 + 12; i++) {  /* 7 days; 12 months */
            formatted_time = my_strftime("%A %B %Z %p",
                                    0, 0, hour, dom, month, 112, 0, 0, is_dst);
            if (! formatted_time || is_invariant_string((U8 *) formatted_time, 0)) {

                /* Here, we didn't find a non-ASCII.  Try the next time through
                 * with the complemented dst and am/pm, and try with the next
                 * weekday.  After we have gotten all weekdays, try the next
                 * month */
                is_dst = ! is_dst;
                hour = (hour + 12) % 24;
                dom++;
                if (i > 6) {
                    month++;
                }
                continue;
            }

            /* Here, we have a non-ASCII.  Return TRUE is it is valid UTF8;
             * false otherwise.  But first, restore LC_TIME to its original
             * locale if we changed it */
            if (save_time_locale) {
                setlocale(LC_TIME, save_time_locale);
                Safefree(save_time_locale);
            }

            DEBUG_L(PerlIO_printf(Perl_debug_log, "\t?time-related strings for %s are UTF-8=%d\n",
                                save_input_locale,
                                is_utf8_string((U8 *) formatted_time, 0)));
            Safefree(save_input_locale);
            return is_utf8_string((U8 *) formatted_time, 0);
        }

        /* Falling off the end of the loop indicates all the names were just
         * ASCII.  Go on to the next test.  If we changed it, restore LC_TIME
         * to its original locale */
        if (save_time_locale) {
            setlocale(LC_TIME, save_time_locale);
            Safefree(save_time_locale);
        }
        DEBUG_L(PerlIO_printf(Perl_debug_log, "All time-related words for %s contain only ASCII; can't use for determining if UTF-8 locale\n", save_input_locale));
    }
  cant_use_time:

#endif

#if 0 && defined(USE_LOCALE_MESSAGES) && defined(HAS_SYS_ERRLIST)

/* This code is ifdefd out because it was found to not be necessary in testing
 * on our dromedary test machine, which has over 700 locales.  There, this
 * added no value to looking at the currency symbol and the time strings.  I
 * left it in so as to avoid rewriting it if real-world experience indicates
 * that dromedary is an outlier.  Essentially, instead of returning abpve if we
 * haven't found illegal utf8, we continue on and examine all the strerror()
 * messages on the platform for utf8ness.  If all are ASCII, we still don't
 * know the answer; but otherwise we have a pretty good indication of the
 * utf8ness.  The reason this doesn't help much is that the messages may not
 * have been translated into the locale.  The currency symbol and time strings
 * are much more likely to have been translated.  */
    {
        int e;
        bool is_utf8 = FALSE;
        bool non_ascii = FALSE;
        char *save_messages_locale = NULL;
        const char * errmsg = NULL;

        /* Like above, we set LC_MESSAGES to the locale of the desired
         * category, if it isn't that locale already */

        if (category != LC_MESSAGES) {

            save_messages_locale = setlocale(LC_MESSAGES, NULL);
            if (! save_messages_locale) {
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                            "Could not find current locale for LC_MESSAGES\n"));
                goto cant_use_messages;
            }
            save_messages_locale = stdize_locale(savepv(save_messages_locale));

            if (strEQ(save_messages_locale, save_input_locale)) {
                Safefree(save_messages_locale);
                save_messages_locale = NULL;
            }
            else if (! setlocale(LC_MESSAGES, save_input_locale)) {
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                            "Could not change LC_MESSAGES locale to %s\n",
                                                        save_input_locale));
                Safefree(save_messages_locale);
                goto cant_use_messages;
            }
        }

        /* Here the current LC_MESSAGES is set to the locale of the category
         * whose information is desired.  Look through all the messages.  We
         * can't use Strerror() here because it may expand to code that
         * segfaults in miniperl */

        for (e = 0; e <= sys_nerr; e++) {
            errno = 0;
            errmsg = sys_errlist[e];
            if (errno || !errmsg) {
                break;
            }
            errmsg = savepv(errmsg);
            if (! is_invariant_string((U8 *) errmsg, 0)) {
                non_ascii = TRUE;
                is_utf8 = is_utf8_string((U8 *) errmsg, 0);
                break;
            }
        }
        Safefree(errmsg);

        /* And, if we changed it, restore LC_MESSAGES to its original locale */
        if (save_messages_locale) {
            setlocale(LC_MESSAGES, save_messages_locale);
            Safefree(save_messages_locale);
        }

        if (non_ascii) {

            /* Any non-UTF-8 message means not a UTF-8 locale; if all are valid,
             * any non-ascii means it is one; otherwise we assume it isn't */
            DEBUG_L(PerlIO_printf(Perl_debug_log, "\t?error messages for %s are UTF-8=%d\n",
                                save_input_locale,
                                is_utf8));
            Safefree(save_input_locale);
            return is_utf8;
        }

        DEBUG_L(PerlIO_printf(Perl_debug_log, "All error messages for %s contain only ASCII; can't use for determining if UTF-8 locale\n", save_input_locale));
    }
  cant_use_messages:

#endif

#endif /* the code that is compiled when no nl_langinfo */

#ifndef EBCDIC  /* On os390, even if the name ends with "UTF-8', it isn't a
                   UTF-8 locale */
    /* As a last resort, look at the locale name to see if it matches
     * qr/UTF -?  * 8 /ix, or some other common locale names.  This "name", the
     * return of setlocale(), is actually defined to be opaque, so we can't
     * really rely on the absence of various substrings in the name to indicate
     * its UTF-8ness, but if it has UTF8 in the name, it is extremely likely to
     * be a UTF-8 locale.  Similarly for the other common names */

    final_pos = strlen(save_input_locale) - 1;
    if (final_pos >= 3) {
        char *name = save_input_locale;

        /* Find next 'U' or 'u' and look from there */
        while ((name += strcspn(name, "Uu") + 1)
                                            <= save_input_locale + final_pos - 2)
        {
            if (!isALPHA_FOLD_NE(*name, 't')
                || isALPHA_FOLD_NE(*(name + 1), 'f'))
            {
                continue;
            }
            name += 2;
            if (*(name) == '-') {
                if ((name > save_input_locale + final_pos - 1)) {
                    break;
                }
                name++;
            }
            if (*(name) == '8') {
                DEBUG_L(PerlIO_printf(Perl_debug_log,
                                      "Locale %s ends with UTF-8 in name\n",
                                      save_input_locale));
                Safefree(save_input_locale);
                return TRUE;
            }
        }
        DEBUG_L(PerlIO_printf(Perl_debug_log,
                              "Locale %s doesn't end with UTF-8 in name\n",
                                save_input_locale));
    }
#endif

#ifdef WIN32
    /* http://msdn.microsoft.com/en-us/library/windows/desktop/dd317756.aspx */
    if (final_pos >= 4
        && *(save_input_locale + final_pos - 0) == '1'
        && *(save_input_locale + final_pos - 1) == '0'
        && *(save_input_locale + final_pos - 2) == '0'
        && *(save_input_locale + final_pos - 3) == '5'
        && *(save_input_locale + final_pos - 4) == '6')
    {
        DEBUG_L(PerlIO_printf(Perl_debug_log,
                        "Locale %s ends with 10056 in name, is UTF-8 locale\n",
                        save_input_locale));
        Safefree(save_input_locale);
        return TRUE;
    }
#endif

    /* Other common encodings are the ISO 8859 series, which aren't UTF-8.  But
     * since we are about to return FALSE anyway, there is no point in doing
     * this extra work */
#if 0
    if (instr(save_input_locale, "8859")) {
        DEBUG_L(PerlIO_printf(Perl_debug_log,
                             "Locale %s has 8859 in name, not UTF-8 locale\n",
                             save_input_locale));
        Safefree(save_input_locale);
        return FALSE;
    }
#endif

    DEBUG_L(PerlIO_printf(Perl_debug_log,
                          "Assuming locale %s is not a UTF-8 locale\n",
                                    save_input_locale));
    Safefree(save_input_locale);
    return FALSE;
}

#endif


bool
Perl__is_in_locale_category(pTHX_ const bool compiling, const int category)
{
    dVAR;
    /* Internal function which returns if we are in the scope of a pragma that
     * enables the locale category 'category'.  'compiling' should indicate if
     * this is during the compilation phase (TRUE) or not (FALSE). */

    const COP * const cop = (compiling) ? &PL_compiling : PL_curcop;

    SV *categories = cop_hints_fetch_pvs(cop, "locale", 0);
    if (! categories || categories == &PL_sv_placeholder) {
        return FALSE;
    }

    /* The pseudo-category 'not_characters' is -1, so just add 1 to each to get
     * a valid unsigned */
    assert(category >= -1);
    return cBOOL(SvUV(categories) & (1U << (category + 1)));
}

char *
Perl_my_strerror(pTHX_ const int errnum) {

    /* Uses C locale for the error text unless within scope of 'use locale' for
     * LC_MESSAGES */

#ifdef USE_LOCALE_MESSAGES
    if (! IN_LC(LC_MESSAGES)) {
        char * save_locale = setlocale(LC_MESSAGES, NULL);
        if (! isNAME_C_OR_POSIX(save_locale)) {
            char *errstr;

            /* The next setlocale likely will zap this, so create a copy */
            save_locale = savepv(save_locale);

            setlocale(LC_MESSAGES, "C");

            /* This points to the static space in Strerror, with all its
             * limitations */
            errstr = Strerror(errnum);

            setlocale(LC_MESSAGES, save_locale);
            Safefree(save_locale);
            return errstr;
        }
    }
#endif

    return Strerror(errnum);
}

/*

=head1 Locale-related functions and macros

=for apidoc sync_locale

Changing the program's locale should be avoided by XS code.  Nevertheless,
certain non-Perl libraries called from XS, such as C<Gtk> do so.  When this
happens, Perl needs to be told that the locale has changed.  Use this function
to do so, before returning to Perl.

=cut
*/

void
Perl_sync_locale(pTHX)
{

#ifdef USE_LOCALE_CTYPE
    new_ctype(setlocale(LC_CTYPE, NULL));
#endif /* USE_LOCALE_CTYPE */

#ifdef USE_LOCALE_COLLATE
    new_collate(setlocale(LC_COLLATE, NULL));
#endif

#ifdef USE_LOCALE_NUMERIC
    set_numeric_local();    /* Switch from "C" to underlying LC_NUMERIC */
    new_numeric(setlocale(LC_NUMERIC, NULL));
#endif /* USE_LOCALE_NUMERIC */

}



/*
 * ex: set ts=8 sts=4 sw=4 et:
 */