icuSources/common/ustring.c

   1 /*
   2 ******************************************************************************
   3 *
   4 *   Copyright (C) 1998-2004, International Business Machines
   5 *   Corporation and others.  All Rights Reserved.
   6 *
   7 ******************************************************************************
   8 *
   9 * File ustring.h
  10 *
  11 * Modification History:
  12 *
  13 *   Date        Name        Description
  14 *   12/07/98    bertrand    Creation.
  15 ******************************************************************************
  16 */
  17
  18 #include "unicode/utypes.h"
  19 #include "unicode/putil.h"
  20 #include "unicode/ustring.h"
  21 #include "cstring.h"
  22 #include "cwchar.h"
  23 #include "cmemory.h"
  24 #include "ustr_imp.h"
  25
  26 /* ANSI string.h - style functions ------------------------------------------ */
  27
  28 /* U+ffff is the highest BMP code point, the highest one that fits into a 16-bit UChar */
  29 #define U_BMP_MAX 0xffff
  30
  31 /* Forward binary string search functions ----------------------------------- */
  32
  33 /*
  34  * Test if a substring match inside a string is at code point boundaries.
  35  * All pointers refer to the same buffer.
  36  * The limit pointer may be NULL, all others must be real pointers.
  37  */
  38 static U_INLINE UBool
  39 isMatchAtCPBoundary(const UChar *start, const UChar *match, const UChar *matchLimit, const UChar *limit) {
  40     if(U16_IS_TRAIL(*match) && start!=match && U16_IS_LEAD(*(match-1))) {
  41         /* the leading edge of the match is in the middle of a surrogate pair */
  42         return FALSE;
  43     }
  44     if(U16_IS_LEAD(*(matchLimit-1)) && match!=limit && U16_IS_TRAIL(*matchLimit)) {
  45         /* the trailing edge of the match is in the middle of a surrogate pair */
  46         return FALSE;
  47     }
  48     return TRUE;
  49 }
  50
  51 U_CAPI UChar * U_EXPORT2
  52 u_strFindFirst(const UChar *s, int32_t length,
  53                const UChar *sub, int32_t subLength) {
  54     const UChar *start, *p, *q, *subLimit;
  55     UChar c, cs, cq;
  56
  57     if(sub==NULL || subLength<-1) {
  58         return (UChar *)s;
  59     }
  60     if(s==NULL || length<-1) {
  61         return NULL;
  62     }
  63
  64     start=s;
  65
  66     if(length<0 && subLength<0) {
  67         /* both strings are NUL-terminated */
  68         if((cs=*sub++)==0) {
  69             return (UChar *)s;
  70         }
  71         if(*sub==0 && !U16_IS_SURROGATE(cs)) {
  72             /* the substring consists of a single, non-surrogate BMP code point */
  73             return u_strchr(s, cs);
  74         }
  75
  76         while((c=*s++)!=0) {
  77             if(c==cs) {
  78                 /* found first substring UChar, compare rest */
  79                 p=s;
  80                 q=sub;
  81                 for(;;) {
  82                     if((cq=*q)==0) {
  83                         if(isMatchAtCPBoundary(start, s-1, p, NULL)) {
  84                             return (UChar *)(s-1); /* well-formed match */
  85                         } else {
  86                             break; /* no match because surrogate pair is split */
  87                         }
  88                     }
  89                     if((c=*p)==0) {
  90                         return NULL; /* no match, and none possible after s */
  91                     }
  92                     if(c!=cq) {
  93                         break; /* no match */
  94                     }
  95                     ++p;
  96                     ++q;
  97                 }
  98             }
  99         }
 100
 101         /* not found */
 102         return NULL;
 103     }
 104
 105     if(subLength<0) {
 106         subLength=u_strlen(sub);
 107     }
 108     if(subLength==0) {
 109         return (UChar *)s;
 110     }
 111
 112     /* get sub[0] to search for it fast */
 113     cs=*sub++;
 114     --subLength;
 115     subLimit=sub+subLength;
 116
 117     if(subLength==0 && !U16_IS_SURROGATE(cs)) {
 118         /* the substring consists of a single, non-surrogate BMP code point */
 119         return length<0 ? u_strchr(s, cs) : u_memchr(s, cs, length);
 120     }
 121
 122     if(length<0) {
 123         /* s is NUL-terminated */
 124         while((c=*s++)!=0) {
 125             if(c==cs) {
 126                 /* found first substring UChar, compare rest */
 127                 p=s;
 128                 q=sub;
 129                 for(;;) {
 130                     if(q==subLimit) {
 131                         if(isMatchAtCPBoundary(start, s-1, p, NULL)) {
 132                             return (UChar *)(s-1); /* well-formed match */
 133                         } else {
 134                             break; /* no match because surrogate pair is split */
 135                         }
 136                     }
 137                     if((c=*p)==0) {
 138                         return NULL; /* no match, and none possible after s */
 139                     }
 140                     if(c!=*q) {
 141                         break; /* no match */
 142                     }
 143                     ++p;
 144                     ++q;
 145                 }
 146             }
 147         }
 148     } else {
 149         const UChar *limit, *preLimit;
 150
 151         /* subLength was decremented above */
 152         if(length<=subLength) {
 153             return NULL; /* s is shorter than sub */
 154         }
 155
 156         limit=s+length;
 157
 158         /* the substring must start before preLimit */
 159         preLimit=limit-subLength;
 160
 161         while(s!=preLimit) {
 162             c=*s++;
 163             if(c==cs) {
 164                 /* found first substring UChar, compare rest */
 165                 p=s;
 166                 q=sub;
 167                 for(;;) {
 168                     if(q==subLimit) {
 169                         if(isMatchAtCPBoundary(start, s-1, p, limit)) {
 170                             return (UChar *)(s-1); /* well-formed match */
 171                         } else {
 172                             break; /* no match because surrogate pair is split */
 173                         }
 174                     }
 175                     if(*p!=*q) {
 176                         break; /* no match */
 177                     }
 178                     ++p;
 179                     ++q;
 180                 }
 181             }
 182         }
 183     }
 184
 185     /* not found */
 186     return NULL;
 187 }
 188
 189 U_CAPI UChar * U_EXPORT2
 190 u_strstr(const UChar *s, const UChar *substring) {
 191     return u_strFindFirst(s, -1, substring, -1);
 192 }
 193
 194 U_CAPI UChar * U_EXPORT2
 195 u_strchr(const UChar *s, UChar c) {
 196     if(U16_IS_SURROGATE(c)) {
 197         /* make sure to not find half of a surrogate pair */
 198         return u_strFindFirst(s, -1, &c, 1);
 199     } else {
 200         UChar cs;
 201
 202         /* trivial search for a BMP code point */
 203         for(;;) {
 204             if((cs=*s)==c) {
 205                 return (UChar *)s;
 206             }
 207             if(cs==0) {
 208                 return NULL;
 209             }
 210             ++s;
 211         }
 212     }
 213 }
 214
 215 U_CAPI UChar * U_EXPORT2
 216 u_strchr32(const UChar *s, UChar32 c) {
 217     if((uint32_t)c<=U_BMP_MAX) {
 218         /* find BMP code point */
 219         return u_strchr(s, (UChar)c);
 220     } else if((uint32_t)c<=UCHAR_MAX_VALUE) {
 221         /* find supplementary code point as surrogate pair */
 222         UChar cs, lead=U16_LEAD(c), trail=U16_TRAIL(c);
 223
 224         while((cs=*s++)!=0) {
 225             if(cs==lead && *s==trail) {
 226                 return (UChar *)(s-1);
 227             }
 228         }
 229         return NULL;
 230     } else {
 231         /* not a Unicode code point, not findable */
 232         return NULL;
 233     }
 234 }
 235
 236 U_CAPI UChar * U_EXPORT2
 237 u_memchr(const UChar *s, UChar c, int32_t count) {
 238     if(count<=0) {
 239         return NULL; /* no string */
 240     } else if(U16_IS_SURROGATE(c)) {
 241         /* make sure to not find half of a surrogate pair */
 242         return u_strFindFirst(s, count, &c, 1);
 243     } else {
 244         /* trivial search for a BMP code point */
 245         const UChar *limit=s+count;
 246         do {
 247             if(*s==c) {
 248                 return (UChar *)s;
 249             }
 250         } while(++s!=limit);
 251         return NULL;
 252     }
 253 }
 254
 255 U_CAPI UChar * U_EXPORT2
 256 u_memchr32(const UChar *s, UChar32 c, int32_t count) {
 257     if((uint32_t)c<=U_BMP_MAX) {
 258         /* find BMP code point */
 259         return u_memchr(s, (UChar)c, count);
 260     } else if(count<2) {
 261         /* too short for a surrogate pair */
 262         return NULL;
 263     } else if((uint32_t)c<=UCHAR_MAX_VALUE) {
 264         /* find supplementary code point as surrogate pair */
 265         const UChar *limit=s+count-1; /* -1 so that we do not need a separate check for the trail unit */
 266         UChar lead=U16_LEAD(c), trail=U16_TRAIL(c);
 267
 268         do {
 269             if(*s==lead && *(s+1)==trail) {
 270                 return (UChar *)s;
 271             }
 272         } while(++s!=limit);
 273         return NULL;
 274     } else {
 275         /* not a Unicode code point, not findable */
 276         return NULL;
 277     }
 278 }
 279
 280 /* Backward binary string search functions ---------------------------------- */
 281
 282 U_CAPI UChar * U_EXPORT2
 283 u_strFindLast(const UChar *s, int32_t length,
 284               const UChar *sub, int32_t subLength) {
 285     const UChar *start, *limit, *p, *q, *subLimit;
 286     UChar c, cs;
 287
 288     if(sub==NULL || subLength<-1) {
 289         return (UChar *)s;
 290     }
 291     if(s==NULL || length<-1) {
 292         return NULL;
 293     }
 294
 295     /*
 296      * This implementation is more lazy than the one for u_strFindFirst():
 297      * There is no special search code for NUL-terminated strings.
 298      * It does not seem to be worth it for searching substrings to
 299      * search forward and find all matches like in u_strrchr() and similar.
 300      * Therefore, we simply get both string lengths and search backward.
 301      *
 302      * markus 2002oct23
 303      */
 304
 305     if(subLength<0) {
 306         subLength=u_strlen(sub);
 307     }
 308     if(subLength==0) {
 309         return (UChar *)s;
 310     }
 311
 312     /* get sub[subLength-1] to search for it fast */
 313     subLimit=sub+subLength;
 314     cs=*(--subLimit);
 315     --subLength;
 316
 317     if(subLength==0 && !U16_IS_SURROGATE(cs)) {
 318         /* the substring consists of a single, non-surrogate BMP code point */
 319         return length<0 ? u_strrchr(s, cs) : u_memrchr(s, cs, length);
 320     }
 321
 322     if(length<0) {
 323         length=u_strlen(s);
 324     }
 325
 326     /* subLength was decremented above */
 327     if(length<=subLength) {
 328         return NULL; /* s is shorter than sub */
 329     }
 330
 331     start=s;
 332     limit=s+length;
 333
 334     /* the substring must start no later than s+subLength */
 335     s+=subLength;
 336
 337     while(s!=limit) {
 338         c=*(--limit);
 339         if(c==cs) {
 340             /* found last substring UChar, compare rest */
 341             p=limit;
 342             q=subLimit;
 343             for(;;) {
 344                 if(q==sub) {
 345                     if(isMatchAtCPBoundary(start, p, limit+1, start+length)) {
 346                         return (UChar *)p; /* well-formed match */
 347                     } else {
 348                         break; /* no match because surrogate pair is split */
 349                     }
 350                 }
 351                 if(*(--p)!=*(--q)) {
 352                     break; /* no match */
 353                 }
 354             }
 355         }
 356     }
 357
 358     /* not found */
 359     return NULL;
 360 }
 361
 362 U_CAPI UChar * U_EXPORT2
 363 u_strrstr(const UChar *s, const UChar *substring) {
 364     return u_strFindLast(s, -1, substring, -1);
 365 }
 366
 367 U_CAPI UChar * U_EXPORT2
 368 u_strrchr(const UChar *s, UChar c) {
 369     if(U16_IS_SURROGATE(c)) {
 370         /* make sure to not find half of a surrogate pair */
 371         return u_strFindLast(s, -1, &c, 1);
 372     } else {
 373         const UChar *result=NULL;
 374         UChar cs;
 375
 376         /* trivial search for a BMP code point */
 377         for(;;) {
 378             if((cs=*s)==c) {
 379                 result=s;
 380             }
 381             if(cs==0) {
 382                 return (UChar *)result;
 383             }
 384             ++s;
 385         }
 386     }
 387 }
 388
 389 U_CAPI UChar * U_EXPORT2
 390 u_strrchr32(const UChar *s, UChar32 c) {
 391     if((uint32_t)c<=U_BMP_MAX) {
 392         /* find BMP code point */
 393         return u_strrchr(s, (UChar)c);
 394     } else if((uint32_t)c<=UCHAR_MAX_VALUE) {
 395         /* find supplementary code point as surrogate pair */
 396         const UChar *result=NULL;
 397         UChar cs, lead=U16_LEAD(c), trail=U16_TRAIL(c);
 398
 399         while((cs=*s++)!=0) {
 400             if(cs==lead && *s==trail) {
 401                 result=s-1;
 402             }
 403         }
 404         return (UChar *)result;
 405     } else {
 406         /* not a Unicode code point, not findable */
 407         return NULL;
 408     }
 409 }
 410
 411 U_CAPI UChar * U_EXPORT2
 412 u_memrchr(const UChar *s, UChar c, int32_t count) {
 413     if(count<=0) {
 414         return NULL; /* no string */
 415     } else if(U16_IS_SURROGATE(c)) {
 416         /* make sure to not find half of a surrogate pair */
 417         return u_strFindLast(s, count, &c, 1);
 418     } else {
 419         /* trivial search for a BMP code point */
 420         const UChar *limit=s+count;
 421         do {
 422             if(*(--limit)==c) {
 423                 return (UChar *)limit;
 424             }
 425         } while(s!=limit);
 426         return NULL;
 427     }
 428 }
 429
 430 U_CAPI UChar * U_EXPORT2
 431 u_memrchr32(const UChar *s, UChar32 c, int32_t count) {
 432     if((uint32_t)c<=U_BMP_MAX) {
 433         /* find BMP code point */
 434         return u_memrchr(s, (UChar)c, count);
 435     } else if(count<2) {
 436         /* too short for a surrogate pair */
 437         return NULL;
 438     } else if((uint32_t)c<=UCHAR_MAX_VALUE) {
 439         /* find supplementary code point as surrogate pair */
 440         const UChar *limit=s+count-1;
 441         UChar lead=U16_LEAD(c), trail=U16_TRAIL(c);
 442
 443         do {
 444             if(*limit==trail && *(limit-1)==lead) {
 445                 return (UChar *)(limit-1);
 446             }
 447         } while(s!=--limit);
 448         return NULL;
 449     } else {
 450         /* not a Unicode code point, not findable */
 451         return NULL;
 452     }
 453 }
 454
 455 /* Tokenization functions --------------------------------------------------- */
 456
 457 /*
 458  * Match each code point in a string against each code point in the matchSet.
 459  * Return the index of the first string code point that
 460  * is (polarity==TRUE) or is not (FALSE) contained in the matchSet.
 461  * Return -(string length)-1 if there is no such code point.
 462  */
 463 static int32_t
 464 _matchFromSet(const UChar *string, const UChar *matchSet, UBool polarity) {
 465     int32_t matchLen, matchBMPLen, strItr, matchItr;
 466     UChar32 stringCh, matchCh;
 467     UChar c, c2;
 468
 469     /* first part of matchSet contains only BMP code points */
 470     matchBMPLen = 0;
 471     while((c = matchSet[matchBMPLen]) != 0 && U16_IS_SINGLE(c)) {
 472         ++matchBMPLen;
 473     }
 474
 475     /* second part of matchSet contains BMP and supplementary code points */
 476     matchLen = matchBMPLen;
 477     while(matchSet[matchLen] != 0) {
 478         ++matchLen;
 479     }
 480
 481     for(strItr = 0; (c = string[strItr]) != 0;) {
 482         ++strItr;
 483         if(U16_IS_SINGLE(c)) {
 484             if(polarity) {
 485                 for(matchItr = 0; matchItr < matchLen; ++matchItr) {
 486                     if(c == matchSet[matchItr]) {
 487                         return strItr - 1; /* one matches */
 488                     }
 489                 }
 490             } else {
 491                 for(matchItr = 0; matchItr < matchLen; ++matchItr) {
 492                     if(c == matchSet[matchItr]) {
 493                         goto endloop;
 494                     }
 495                 }
 496                 return strItr - 1; /* none matches */
 497             }
 498         } else {
 499             /*
 500              * No need to check for string length before U16_IS_TRAIL
 501              * because c2 could at worst be the terminating NUL.
 502              */
 503             if(U16_IS_SURROGATE_LEAD(c) && U16_IS_TRAIL(c2 = string[strItr])) {
 504                 ++strItr;
 505                 stringCh = U16_GET_SUPPLEMENTARY(c, c2);
 506             } else {
 507                 stringCh = c; /* unpaired trail surrogate */
 508             }
 509
 510             if(polarity) {
 511                 for(matchItr = matchBMPLen; matchItr < matchLen;) {
 512                     U16_NEXT(matchSet, matchItr, matchLen, matchCh);
 513                     if(stringCh == matchCh) {
 514                         return strItr - U16_LENGTH(stringCh); /* one matches */
 515                     }
 516                 }
 517             } else {
 518                 for(matchItr = matchBMPLen; matchItr < matchLen;) {
 519                     U16_NEXT(matchSet, matchItr, matchLen, matchCh);
 520                     if(stringCh == matchCh) {
 521                         goto endloop;
 522                     }
 523                 }
 524                 return strItr - U16_LENGTH(stringCh); /* none matches */
 525             }
 526         }
 527 endloop:
 528         /* wish C had continue with labels like Java... */;
 529     }
 530
 531     /* Didn't find it. */
 532     return -strItr-1;
 533 }
 534
 535 /* Search for a codepoint in a string that matches one of the matchSet codepoints. */
 536 U_CAPI UChar * U_EXPORT2
 537 u_strpbrk(const UChar *string, const UChar *matchSet)
 538 {
 539     int32_t index = _matchFromSet(string, matchSet, TRUE);
 540     if(index >= 0) {
 541         return (UChar *)string + index;
 542     } else {
 543         return NULL;
 544     }
 545 }
 546
 547 /* Search for a codepoint in a string that matches one of the matchSet codepoints. */
 548 U_CAPI int32_t U_EXPORT2
 549 u_strcspn(const UChar *string, const UChar *matchSet)
 550 {
 551     int32_t index = _matchFromSet(string, matchSet, TRUE);
 552     if(index >= 0) {
 553         return index;
 554     } else {
 555         return -index - 1; /* == u_strlen(string) */
 556     }
 557 }
 558
 559 /* Search for a codepoint in a string that does not match one of the matchSet codepoints. */
 560 U_CAPI int32_t U_EXPORT2
 561 u_strspn(const UChar *string, const UChar *matchSet)
 562 {
 563     int32_t index = _matchFromSet(string, matchSet, FALSE);
 564     if(index >= 0) {
 565         return index;
 566     } else {
 567         return -index - 1; /* == u_strlen(string) */
 568     }
 569 }
 570
 571 /* ----- Text manipulation functions --- */
 572
 573 U_CAPI UChar* U_EXPORT2
 574 u_strtok_r(UChar    *src,
 575      const UChar    *delim,
 576            UChar   **saveState)
 577 {
 578     UChar *tokSource;
 579     UChar *nextToken;
 580     uint32_t nonDelimIdx;
 581
 582     /* If saveState is NULL, the user messed up. */
 583     if (src != NULL) {
 584         tokSource = src;
 585         *saveState = src; /* Set to "src" in case there are no delimiters */
 586     }
 587     else if (*saveState) {
 588         tokSource = *saveState;
 589     }
 590     else {
 591         /* src == NULL && *saveState == NULL */
 592         /* This shouldn't happen. We already finished tokenizing. */
 593         return NULL;
 594     }
 595
 596     /* Skip initial delimiters */
 597     nonDelimIdx = u_strspn(tokSource, delim);
 598     tokSource = &tokSource[nonDelimIdx];
 599
 600     if (*tokSource) {
 601         nextToken = u_strpbrk(tokSource, delim);
 602         if (nextToken != NULL) {
 603             /* Create a token */
 604             *(nextToken++) = 0;
 605             *saveState = nextToken;
 606             return tokSource;
 607         }
 608         else if (*saveState) {
 609             /* Return the last token */
 610             *saveState = NULL;
 611             return tokSource;
 612         }
 613     }
 614     else {
 615         /* No tokens were found. Only delimiters were left. */
 616         *saveState = NULL;
 617     }
 618     return NULL;
 619 }
 620
 621 /* Miscellaneous functions -------------------------------------------------- */
 622
 623 U_CAPI UChar* U_EXPORT2
 624 u_strcat(UChar     *dst,
 625     const UChar     *src)
 626 {
 627     UChar *anchor = dst;            /* save a pointer to start of dst */
 628
 629     while(*dst != 0) {              /* To end of first string          */
 630         ++dst;
 631     }
 632     while((*(dst++) = *(src++)) != 0) {     /* copy string 2 over              */
 633     }
 634
 635     return anchor;
 636 }
 637
 638 U_CAPI UChar*  U_EXPORT2
 639 u_strncat(UChar     *dst,
 640      const UChar     *src,
 641      int32_t     n )
 642 {
 643     if(n > 0) {
 644         UChar *anchor = dst;            /* save a pointer to start of dst */
 645
 646         while(*dst != 0) {              /* To end of first string          */
 647             ++dst;
 648         }
 649         while((*dst = *src) != 0) {     /* copy string 2 over              */
 650             ++dst;
 651             if(--n == 0) {
 652                 *dst = 0;
 653                 break;
 654             }
 655             ++src;
 656         }
 657
 658         return anchor;
 659     } else {
 660         return dst;
 661     }
 662 }
 663
 664 /* ----- Text property functions --- */
 665
 666 U_CAPI int32_t   U_EXPORT2
 667 u_strcmp(const UChar *s1,
 668     const UChar *s2)
 669 {
 670     UChar  c1, c2;
 671
 672     for(;;) {
 673         c1=*s1++;
 674         c2=*s2++;
 675         if (c1 != c2 || c1 == 0) {
 676             break;
 677         }
 678     }
 679     return (int32_t)c1 - (int32_t)c2;
 680 }
 681
 682 U_CAPI int32_t U_EXPORT2
 683 uprv_strCompare(const UChar *s1, int32_t length1,
 684                 const UChar *s2, int32_t length2,
 685                 UBool strncmpStyle, UBool codePointOrder) {
 686     const UChar *start1, *start2, *limit1, *limit2;
 687     UChar c1, c2;
 688
 689     /* setup for fix-up */
 690     start1=s1;
 691     start2=s2;
 692
 693     /* compare identical prefixes - they do not need to be fixed up */
 694     if(length1<0 && length2<0) {
 695         /* strcmp style, both NUL-terminated */
 696         if(s1==s2) {
 697             return 0;
 698         }
 699
 700         for(;;) {
 701             c1=*s1;
 702             c2=*s2;
 703             if(c1!=c2) {
 704                 break;
 705             }
 706             if(c1==0) {
 707                 return 0;
 708             }
 709             ++s1;
 710             ++s2;
 711         }
 712
 713         /* setup for fix-up */
 714         limit1=limit2=NULL;
 715     } else if(strncmpStyle) {
 716         /* special handling for strncmp, assume length1==length2>=0 but also check for NUL */
 717         if(s1==s2) {
 718             return 0;
 719         }
 720
 721         limit1=start1+length1;
 722
 723         for(;;) {
 724             /* both lengths are same, check only one limit */
 725             if(s1==limit1) {
 726                 return 0;
 727             }
 728
 729             c1=*s1;
 730             c2=*s2;
 731             if(c1!=c2) {
 732                 break;
 733             }
 734             if(c1==0) {
 735                 return 0;
 736             }
 737             ++s1;
 738             ++s2;
 739         }
 740
 741         /* setup for fix-up */
 742         limit2=start2+length1; /* use length1 here, too, to enforce assumption */
 743     } else {
 744         /* memcmp/UnicodeString style, both length-specified */
 745         int32_t lengthResult;
 746
 747         if(length1<0) {
 748             length1=u_strlen(s1);
 749         }
 750         if(length2<0) {
 751             length2=u_strlen(s2);
 752         }
 753
 754         /* limit1=start1+min(lenght1, length2) */
 755         if(length1<length2) {
 756             lengthResult=-1;
 757             limit1=start1+length1;
 758         } else if(length1==length2) {
 759             lengthResult=0;
 760             limit1=start1+length1;
 761         } else /* length1>length2 */ {
 762             lengthResult=1;
 763             limit1=start1+length2;
 764         }
 765
 766         if(s1==s2) {
 767             return lengthResult;
 768         }
 769
 770         for(;;) {
 771             /* check pseudo-limit */
 772             if(s1==limit1) {
 773                 return lengthResult;
 774             }
 775
 776             c1=*s1;
 777             c2=*s2;
 778             if(c1!=c2) {
 779                 break;
 780             }
 781             ++s1;
 782             ++s2;
 783         }
 784
 785         /* setup for fix-up */
 786         limit1=start1+length1;
 787         limit2=start2+length2;
 788     }
 789
 790     /* if both values are in or above the surrogate range, fix them up */
 791     if(c1>=0xd800 && c2>=0xd800 && codePointOrder) {
 792         /* subtract 0x2800 from BMP code points to make them smaller than supplementary ones */
 793         if(
 794             (c1<=0xdbff && (s1+1)!=limit1 && UTF_IS_TRAIL(*(s1+1))) ||
 795             (UTF_IS_TRAIL(c1) && start1!=s1 && UTF_IS_LEAD(*(s1-1)))
 796         ) {
 797             /* part of a surrogate pair, leave >=d800 */
 798         } else {
 799             /* BMP code point - may be surrogate code point - make <d800 */
 800             c1-=0x2800;
 801         }
 802
 803         if(
 804             (c2<=0xdbff && (s2+1)!=limit2 && UTF_IS_TRAIL(*(s2+1))) ||
 805             (UTF_IS_TRAIL(c2) && start2!=s2 && UTF_IS_LEAD(*(s2-1)))
 806         ) {
 807             /* part of a surrogate pair, leave >=d800 */
 808         } else {
 809             /* BMP code point - may be surrogate code point - make <d800 */
 810             c2-=0x2800;
 811         }
 812     }
 813
 814     /* now c1 and c2 are in the requested (code unit or code point) order */
 815     return (int32_t)c1-(int32_t)c2;
 816 }
 817
 818 /*
 819  * Compare two strings as presented by UCharIterators.
 820  * Use code unit or code point order.
 821  * When the function returns, it is undefined where the iterators
 822  * have stopped.
 823  */
 824 U_CAPI int32_t U_EXPORT2
 825 u_strCompareIter(UCharIterator *iter1, UCharIterator *iter2, UBool codePointOrder) {
 826     UChar32 c1, c2;
 827
 828     /* argument checking */
 829     if(iter1==NULL || iter2==NULL) {
 830         return 0; /* bad arguments */
 831     }
 832     if(iter1==iter2) {
 833         return 0; /* identical iterators */
 834     }
 835
 836     /* reset iterators to start? */
 837     iter1->move(iter1, 0, UITER_START);
 838     iter2->move(iter2, 0, UITER_START);
 839
 840     /* compare identical prefixes - they do not need to be fixed up */
 841     for(;;) {
 842         c1=iter1->next(iter1);
 843         c2=iter2->next(iter2);
 844         if(c1!=c2) {
 845             break;
 846         }
 847         if(c1==-1) {
 848             return 0;
 849         }
 850     }
 851
 852     /* if both values are in or above the surrogate range, fix them up */
 853     if(c1>=0xd800 && c2>=0xd800 && codePointOrder) {
 854         /* subtract 0x2800 from BMP code points to make them smaller than supplementary ones */
 855         if(
 856             (c1<=0xdbff && UTF_IS_TRAIL(iter1->current(iter1))) ||
 857             (UTF_IS_TRAIL(c1) && (iter1->previous(iter1), UTF_IS_LEAD(iter1->previous(iter1))))
 858         ) {
 859             /* part of a surrogate pair, leave >=d800 */
 860         } else {
 861             /* BMP code point - may be surrogate code point - make <d800 */
 862             c1-=0x2800;
 863         }
 864
 865         if(
 866             (c2<=0xdbff && UTF_IS_TRAIL(iter2->current(iter2))) ||
 867             (UTF_IS_TRAIL(c2) && (iter2->previous(iter2), UTF_IS_LEAD(iter2->previous(iter2))))
 868         ) {
 869             /* part of a surrogate pair, leave >=d800 */
 870         } else {
 871             /* BMP code point - may be surrogate code point - make <d800 */
 872             c2-=0x2800;
 873         }
 874     }
 875
 876     /* now c1 and c2 are in the requested (code unit or code point) order */
 877     return (int32_t)c1-(int32_t)c2;
 878 }
 879
 880 #if 0
 881 /*
 882  * u_strCompareIter() does not leave the iterators _on_ the different units.
 883  * This is possible but would cost a few extra indirect function calls to back
 884  * up if the last unit (c1 or c2 respectively) was >=0.
 885  *
 886  * Consistently leaving them _behind_ the different units is not an option
 887  * because the current "unit" is the end of the string if that is reached,
 888  * and in such a case the iterator does not move.
 889  * For example, when comparing "ab" with "abc", both iterators rest _on_ the end
 890  * of their strings. Calling previous() on each does not move them to where
 891  * the comparison fails.
 892  *
 893  * So the simplest semantics is to not define where the iterators end up.
 894  *
 895  * The following fragment is part of what would need to be done for backing up.
 896  */
 897 void fragment {
 898         /* iff a surrogate is part of a surrogate pair, leave >=d800 */
 899         if(c1<=0xdbff) {
 900             if(!UTF_IS_TRAIL(iter1->current(iter1))) {
 901                 /* lead surrogate code point - make <d800 */
 902                 c1-=0x2800;
 903             }
 904         } else if(c1<=0xdfff) {
 905             int32_t index=iter1->getIndex(iter1, UITER_CURRENT);
 906             iter1->previous(iter1); /* ==c1 */
 907             if(!UTF_IS_LEAD(iter1->previous(iter1))) {
 908                 /* trail surrogate code point - make <d800 */
 909                 c1-=0x2800;
 910             }
 911             /* go back to behind where the difference is */
 912             iter1->move(iter1, index, UITER_ZERO);
 913         } else /* 0xe000<=c1<=0xffff */ {
 914             /* BMP code point - make <d800 */
 915             c1-=0x2800;
 916         }
 917 }
 918 #endif
 919
 920 U_CAPI int32_t U_EXPORT2
 921 u_strCompare(const UChar *s1, int32_t length1,
 922              const UChar *s2, int32_t length2,
 923              UBool codePointOrder) {
 924     /* argument checking */
 925     if(s1==NULL || length1<-1 || s2==NULL || length2<-1) {
 926         return 0;
 927     }
 928     return uprv_strCompare(s1, length1, s2, length2, FALSE, codePointOrder);
 929 }
 930
 931 /* String compare in code point order - u_strcmp() compares in code unit order. */
 932 U_CAPI int32_t U_EXPORT2
 933 u_strcmpCodePointOrder(const UChar *s1, const UChar *s2) {
 934     return uprv_strCompare(s1, -1, s2, -1, FALSE, TRUE);
 935 }
 936
 937 U_CAPI int32_t   U_EXPORT2
 938 u_strncmp(const UChar     *s1,
 939      const UChar     *s2,
 940      int32_t     n)
 941 {
 942     if(n > 0) {
 943         int32_t rc;
 944         for(;;) {
 945             rc = (int32_t)*s1 - (int32_t)*s2;
 946             if(rc != 0 || *s1 == 0 || --n == 0) {
 947                 return rc;
 948             }
 949             ++s1;
 950             ++s2;
 951         }
 952     } else {
 953         return 0;
 954     }
 955 }
 956
 957 U_CAPI int32_t U_EXPORT2
 958 u_strncmpCodePointOrder(const UChar *s1, const UChar *s2, int32_t n) {
 959     return uprv_strCompare(s1, n, s2, n, TRUE, TRUE);
 960 }
 961
 962 U_CAPI UChar* U_EXPORT2
 963 u_strcpy(UChar     *dst,
 964     const UChar     *src)
 965 {
 966     UChar *anchor = dst;            /* save a pointer to start of dst */
 967
 968     while((*(dst++) = *(src++)) != 0) {     /* copy string 2 over              */
 969     }
 970
 971     return anchor;
 972 }
 973
 974 U_CAPI UChar*  U_EXPORT2
 975 u_strncpy(UChar     *dst,
 976      const UChar     *src,
 977      int32_t     n)
 978 {
 979     UChar *anchor = dst;            /* save a pointer to start of dst */
 980
 981     /* copy string 2 over */
 982     while(n > 0 && (*(dst++) = *(src++)) != 0) {
 983         --n;
 984     }
 985
 986     return anchor;
 987 }
 988
 989 U_CAPI int32_t   U_EXPORT2
 990 u_strlen(const UChar *s)
 991 {
 992 #if U_SIZEOF_WCHAR_T == U_SIZEOF_UCHAR
 993     return uprv_wcslen(s);
 994 #else
 995     const UChar *t = s;
 996     while(*t != 0) {
 997       ++t;
 998     }
 999     return t - s;
1000 #endif
1001 }
1002
1003 U_CAPI int32_t U_EXPORT2
1004 u_countChar32(const UChar *s, int32_t length) {
1005     int32_t count;
1006
1007     if(s==NULL || length<-1) {
1008         return 0;
1009     }
1010
1011     count=0;
1012     if(length>=0) {
1013         while(length>0) {
1014             ++count;
1015             if(UTF_IS_LEAD(*s) && length>=2 && UTF_IS_TRAIL(*(s+1))) {
1016                 s+=2;
1017                 length-=2;
1018             } else {
1019                 ++s;
1020                 --length;
1021             }
1022         }
1023     } else /* length==-1 */ {
1024         UChar c;
1025
1026         for(;;) {
1027             if((c=*s++)==0) {
1028                 break;
1029             }
1030             ++count;
1031
1032             /*
1033              * sufficient to look ahead one because of UTF-16;
1034              * safe to look ahead one because at worst that would be the terminating NUL
1035              */
1036             if(UTF_IS_LEAD(c) && UTF_IS_TRAIL(*s)) {
1037                 ++s;
1038             }
1039         }
1040     }
1041     return count;
1042 }
1043
1044 U_CAPI UBool U_EXPORT2
1045 u_strHasMoreChar32Than(const UChar *s, int32_t length, int32_t number) {
1046
1047     if(number<0) {
1048         return TRUE;
1049     }
1050     if(s==NULL || length<-1) {
1051         return FALSE;
1052     }
1053
1054     if(length==-1) {
1055         /* s is NUL-terminated */
1056         UChar c;
1057
1058         /* count code points until they exceed */
1059         for(;;) {
1060             if((c=*s++)==0) {
1061                 return FALSE;
1062             }
1063             if(number==0) {
1064                 return TRUE;
1065             }
1066             if(U16_IS_LEAD(c) && U16_IS_TRAIL(*s)) {
1067                 ++s;
1068             }
1069             --number;
1070         }
1071     } else {
1072         /* length>=0 known */
1073         const UChar *limit;
1074         int32_t maxSupplementary;
1075
1076         /* s contains at least (length+1)/2 code points: <=2 UChars per cp */
1077         if(((length+1)/2)>number) {
1078             return TRUE;
1079         }
1080
1081         /* check if s does not even contain enough UChars */
1082         maxSupplementary=length-number;
1083         if(maxSupplementary<=0) {
1084             return FALSE;
1085         }
1086         /* there are maxSupplementary=length-number more UChars than asked-for code points */
1087
1088         /*
1089          * count code points until they exceed and also check that there are
1090          * no more than maxSupplementary supplementary code points (UChar pairs)
1091          */
1092         limit=s+length;
1093         for(;;) {
1094             if(s==limit) {
1095                 return FALSE;
1096             }
1097             if(number==0) {
1098                 return TRUE;
1099             }
1100             if(U16_IS_LEAD(*s++) && s!=limit && U16_IS_TRAIL(*s)) {
1101                 ++s;
1102                 if(--maxSupplementary<=0) {
1103                     /* too many pairs - too few code points */
1104                     return FALSE;
1105                 }
1106             }
1107             --number;
1108         }
1109     }
1110 }
1111
1112 U_CAPI UChar * U_EXPORT2
1113 u_memcpy(UChar *dest, const UChar *src, int32_t count) {
1114     return (UChar *)uprv_memcpy(dest, src, count*U_SIZEOF_UCHAR);
1115 }
1116
1117 U_CAPI UChar * U_EXPORT2
1118 u_memmove(UChar *dest, const UChar *src, int32_t count) {
1119     return (UChar *)uprv_memmove(dest, src, count*U_SIZEOF_UCHAR);
1120 }
1121
1122 U_CAPI UChar * U_EXPORT2
1123 u_memset(UChar *dest, UChar c, int32_t count) {
1124     if(count > 0) {
1125         UChar *ptr = dest;
1126         UChar *limit = dest + count;
1127
1128         while (ptr < limit) {
1129             *(ptr++) = c;
1130         }
1131     }
1132     return dest;
1133 }
1134
1135 U_CAPI int32_t U_EXPORT2
1136 u_memcmp(const UChar *buf1, const UChar *buf2, int32_t count) {
1137     if(count > 0) {
1138         const UChar *limit = buf1 + count;
1139         int32_t result;
1140
1141         while (buf1 < limit) {
1142             result = (int32_t)(uint16_t)*buf1 - (int32_t)(uint16_t)*buf2;
1143             if (result != 0) {
1144                 return result;
1145             }
1146             buf1++;
1147             buf2++;
1148         }
1149     }
1150     return 0;
1151 }
1152
1153 U_CAPI int32_t U_EXPORT2
1154 u_memcmpCodePointOrder(const UChar *s1, const UChar *s2, int32_t count) {
1155     return uprv_strCompare(s1, count, s2, count, FALSE, TRUE);
1156 }
1157
1158 /* u_unescape & support fns ------------------------------------------------- */
1159
1160 /* This map must be in ASCENDING ORDER OF THE ESCAPE CODE */
1161 static const UChar UNESCAPE_MAP[] = {
1162     /*"   0x22, 0x22 */
1163     /*'   0x27, 0x27 */
1164     /*?   0x3F, 0x3F */
1165     /*\   0x5C, 0x5C */
1166     /*a*/ 0x61, 0x07,
1167     /*b*/ 0x62, 0x08,
1168     /*e*/ 0x65, 0x1b,
1169     /*f*/ 0x66, 0x0c,
1170     /*n*/ 0x6E, 0x0a,
1171     /*r*/ 0x72, 0x0d,
1172     /*t*/ 0x74, 0x09,
1173     /*v*/ 0x76, 0x0b
1174 };
1175 enum { UNESCAPE_MAP_LENGTH = sizeof(UNESCAPE_MAP) / sizeof(UNESCAPE_MAP[0]) };
1176
1177 /* Convert one octal digit to a numeric value 0..7, or -1 on failure */
1178 static int8_t _digit8(UChar c) {
1179     if (c >= 0x0030 && c <= 0x0037) {
1180         return (int8_t)(c - 0x0030);
1181     }
1182     return -1;
1183 }
1184
1185 /* Convert one hex digit to a numeric value 0..F, or -1 on failure */
1186 static int8_t _digit16(UChar c) {
1187     if (c >= 0x0030 && c <= 0x0039) {
1188         return (int8_t)(c - 0x0030);
1189     }
1190     if (c >= 0x0041 && c <= 0x0046) {
1191         return (int8_t)(c - (0x0041 - 10));
1192     }
1193     if (c >= 0x0061 && c <= 0x0066) {
1194         return (int8_t)(c - (0x0061 - 10));
1195     }
1196     return -1;
1197 }
1198
1199 /* Parse a single escape sequence.  Although this method deals in
1200  * UChars, it does not use C++ or UnicodeString.  This allows it to
1201  * be used from C contexts. */
1202 U_CAPI UChar32 U_EXPORT2
1203 u_unescapeAt(UNESCAPE_CHAR_AT charAt,
1204              int32_t *offset,
1205              int32_t length,
1206              void *context) {
1207
1208     int32_t start = *offset;
1209     UChar c;
1210     UChar32 result = 0;
1211     int8_t n = 0;
1212     int8_t minDig = 0;
1213     int8_t maxDig = 0;
1214     int8_t bitsPerDigit = 4;
1215     int8_t dig;
1216     int32_t i;
1217     UBool braces = FALSE;
1218
1219     /* Check that offset is in range */
1220     if (*offset < 0 || *offset >= length) {
1221         goto err;
1222     }
1223
1224     /* Fetch first UChar after '\\' */
1225     c = charAt((*offset)++, context);
1226
1227     /* Convert hexadecimal and octal escapes */
1228     switch (c) {
1229     case 0x0075 /*'u'*/:
1230         minDig = maxDig = 4;
1231         break;
1232     case 0x0055 /*'U'*/:
1233         minDig = maxDig = 8;
1234         break;
1235     case 0x0078 /*'x'*/:
1236         minDig = 1;
1237         if (*offset < length && charAt(*offset, context) == 0x7B /*{*/) {
1238             ++(*offset);
1239             braces = TRUE;
1240             maxDig = 8;
1241         } else {
1242             maxDig = 2;
1243         }
1244         break;
1245     default:
1246         dig = _digit8(c);
1247         if (dig >= 0) {
1248             minDig = 1;
1249             maxDig = 3;
1250             n = 1; /* Already have first octal digit */
1251             bitsPerDigit = 3;
1252             result = dig;
1253         }
1254         break;
1255     }
1256     if (minDig != 0) {
1257         while (*offset < length && n < maxDig) {
1258             c = charAt(*offset, context);
1259             dig = (int8_t)((bitsPerDigit == 3) ? _digit8(c) : _digit16(c));
1260             if (dig < 0) {
1261                 break;
1262             }
1263             result = (result << bitsPerDigit) | dig;
1264             ++(*offset);
1265             ++n;
1266         }
1267         if (n < minDig) {
1268             goto err;
1269         }
1270         if (braces) {
1271             if (c != 0x7D /*}*/) {
1272                 goto err;
1273             }
1274             ++(*offset);
1275         }
1276         if (result < 0 || result >= 0x110000) {
1277             goto err;
1278         }
1279         /* If an escape sequence specifies a lead surrogate, see if
1280          * there is a trail surrogate after it, either as an escape or
1281          * as a literal.  If so, join them up into a supplementary.
1282          */
1283         if (*offset < length && U16_IS_LEAD(result)) {
1284             int32_t ahead = *offset + 1;
1285             c = charAt(*offset, context);
1286             if (c == 0x5C /*'\\'*/ && ahead < length) {
1287                 c = (UChar) u_unescapeAt(charAt, &ahead, length, context);
1288             }
1289             if (U16_IS_TRAIL(c)) {
1290                 *offset = ahead;
1291                 result = U16_GET_SUPPLEMENTARY(result, c);
1292             }
1293         }
1294         return result;
1295     }
1296
1297     /* Convert C-style escapes in table */
1298     for (i=0; i<UNESCAPE_MAP_LENGTH; i+=2) {
1299         if (c == UNESCAPE_MAP[i]) {
1300             return UNESCAPE_MAP[i+1];
1301         } else if (c < UNESCAPE_MAP[i]) {
1302             break;
1303         }
1304     }
1305
1306     /* Map \cX to control-X: X & 0x1F */
1307     if (c == 0x0063 /*'c'*/ && *offset < length) {
1308         c = charAt((*offset)++, context);
1309         if (UTF_IS_FIRST_SURROGATE(c) && *offset < length) {
1310             UChar c2 = charAt(*offset, context);
1311             if (UTF_IS_SECOND_SURROGATE(c2)) {
1312                 ++(*offset);
1313                 c = (UChar) UTF16_GET_PAIR_VALUE(c, c2); /* [sic] */
1314             }
1315         }
1316         return 0x1F & c;
1317     }
1318
1319     /* If no special forms are recognized, then consider
1320      * the backslash to generically escape the next character.
1321      * Deal with surrogate pairs. */
1322     if (UTF_IS_FIRST_SURROGATE(c) && *offset < length) {
1323         UChar c2 = charAt(*offset, context);
1324         if (UTF_IS_SECOND_SURROGATE(c2)) {
1325             ++(*offset);
1326             return UTF16_GET_PAIR_VALUE(c, c2);
1327         }
1328     }
1329     return c;
1330
1331  err:
1332     /* Invalid escape sequence */
1333     *offset = start; /* Reset to initial value */
1334     return (UChar32)0xFFFFFFFF;
1335 }
1336
1337 /* u_unescapeAt() callback to return a UChar from a char* */
1338 static UChar U_CALLCONV
1339 _charPtr_charAt(int32_t offset, void *context) {
1340     UChar c16;
1341     /* It would be more efficient to access the invariant tables
1342      * directly but there is no API for that. */
1343     u_charsToUChars(((char*) context) + offset, &c16, 1);
1344     return c16;
1345 }
1346
1347 /* Append an escape-free segment of the text; used by u_unescape() */
1348 static void _appendUChars(UChar *dest, int32_t destCapacity,
1349                           const char *src, int32_t srcLen) {
1350     if (destCapacity < 0) {
1351         destCapacity = 0;
1352     }
1353     if (srcLen > destCapacity) {
1354         srcLen = destCapacity;
1355     }
1356     u_charsToUChars(src, dest, srcLen);
1357 }
1358
1359 /* Do an invariant conversion of char* -> UChar*, with escape parsing */
1360 U_CAPI int32_t U_EXPORT2
1361 u_unescape(const char *src, UChar *dest, int32_t destCapacity) {
1362     const char *segment = src;
1363     int32_t i = 0;
1364     char c;
1365
1366     while ((c=*src) != 0) {
1367         /* '\\' intentionally written as compiler-specific
1368          * character constant to correspond to compiler-specific
1369          * char* constants. */
1370         if (c == '\\') {
1371             int32_t lenParsed = 0;
1372             UChar32 c32;
1373             if (src != segment) {
1374                 if (dest != NULL) {
1375                     _appendUChars(dest + i, destCapacity - i,
1376                                   segment, src - segment);
1377                 }
1378                 i += src - segment;
1379             }
1380             ++src; /* advance past '\\' */
1381             c32 = u_unescapeAt(_charPtr_charAt, &lenParsed, uprv_strlen(src), (void*)src);
1382             if (lenParsed == 0) {
1383                 goto err;
1384             }
1385             src += lenParsed; /* advance past escape seq. */
1386             if (dest != NULL && UTF_CHAR_LENGTH(c32) <= (destCapacity - i)) {
1387                 UTF_APPEND_CHAR_UNSAFE(dest, i, c32);
1388             } else {
1389                 i += UTF_CHAR_LENGTH(c32);
1390             }
1391             segment = src;
1392         } else {
1393             ++src;
1394         }
1395     }
1396     if (src != segment) {
1397         if (dest != NULL) {
1398             _appendUChars(dest + i, destCapacity - i,
1399                           segment, src - segment);
1400         }
1401         i += src - segment;
1402     }
1403     if (dest != NULL && i < destCapacity) {
1404         dest[i] = 0;
1405     }
1406     return i;
1407
1408  err:
1409     if (dest != NULL && destCapacity > 0) {
1410         *dest = 0;
1411     }
1412     return 0;
1413 }
1414
1415 /* C UGrowBuffer implementation --------------------------------------------- */
1416
1417 U_CAPI UBool /* U_CALLCONV U_EXPORT2 */
1418 u_growBufferFromStatic(void *context,
1419                        UChar **pBuffer, int32_t *pCapacity, int32_t reqCapacity,
1420                        int32_t length) {
1421     UChar *newBuffer=(UChar *)uprv_malloc(reqCapacity*U_SIZEOF_UCHAR);
1422     if(newBuffer!=NULL) {
1423         if(length>0) {
1424             uprv_memcpy(newBuffer, *pBuffer, length*U_SIZEOF_UCHAR);
1425         }
1426         *pCapacity=reqCapacity;
1427     } else {
1428         *pCapacity=0;
1429     }
1430
1431     /* release the old pBuffer if it was not statically allocated */
1432     if(*pBuffer!=(UChar *)context) {
1433         uprv_free(*pBuffer);
1434     }
1435
1436     *pBuffer=newBuffer;
1437     return (UBool)(newBuffer!=NULL);
1438 }
1439
1440 /* NUL-termination of strings ----------------------------------------------- */
1441
1442 /**
1443  * NUL-terminate a string no matter what its type.
1444  * Set warning and error codes accordingly.
1445  */
1446 #define __TERMINATE_STRING(dest, destCapacity, length, pErrorCode)      \
1447     if(pErrorCode!=NULL && U_SUCCESS(*pErrorCode)) {                    \
1448         /* not a public function, so no complete argument checking */   \
1449                                                                         \
1450         if(length<0) {                                                  \
1451             /* assume that the caller handles this */                   \
1452         } else if(length<destCapacity) {                                \
1453             /* NUL-terminate the string, the NUL fits */                \
1454             dest[length]=0;                                             \
1455             /* unset the not-terminated warning but leave all others */ \
1456             if(*pErrorCode==U_STRING_NOT_TERMINATED_WARNING) {          \
1457                 *pErrorCode=U_ZERO_ERROR;                               \
1458             }                                                           \
1459         } else if(length==destCapacity) {                               \
1460             /* unable to NUL-terminate, but the string itself fit - set a warning code */ \
1461             *pErrorCode=U_STRING_NOT_TERMINATED_WARNING;                \
1462         } else /* length>destCapacity */ {                              \
1463             /* even the string itself did not fit - set an error code */ \
1464             *pErrorCode=U_BUFFER_OVERFLOW_ERROR;                        \
1465         }                                                               \
1466     }
1467
1468 U_CAPI int32_t U_EXPORT2
1469 u_terminateUChars(UChar *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode) {
1470     __TERMINATE_STRING(dest, destCapacity, length, pErrorCode);
1471     return length;
1472 }
1473
1474 U_CAPI int32_t U_EXPORT2
1475 u_terminateChars(char *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode) {
1476     __TERMINATE_STRING(dest, destCapacity, length, pErrorCode);
1477     return length;
1478 }
1479
1480 U_CAPI int32_t U_EXPORT2
1481 u_terminateUChar32s(UChar32 *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode) {
1482     __TERMINATE_STRING(dest, destCapacity, length, pErrorCode);
1483     return length;
1484 }
1485
1486 U_CAPI int32_t U_EXPORT2
1487 u_terminateWChars(wchar_t *dest, int32_t destCapacity, int32_t length, UErrorCode *pErrorCode) {
1488     __TERMINATE_STRING(dest, destCapacity, length, pErrorCode);
1489     return length;
1490 }