1 // © 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /*
4 *******************************************************************************
5 *
6 * Copyright (C) 2003-2014, International Business Machines
7 * Corporation and others. All Rights Reserved.
8 *
9 *******************************************************************************
10 * file name: uidna.cpp
11 * encoding: UTF-8
12 * tab size: 8 (not used)
13 * indentation:4
14 *
15 * created on: 2003feb1
16 * created by: Ram Viswanadha
17 */
18
19 #include "unicode/utypes.h"
20
21 #if !UCONFIG_NO_IDNA
22
23 #include "unicode/uidna.h"
24 #include "unicode/ustring.h"
25 #include "unicode/usprep.h"
26 #include "punycode.h"
27 #include "ustr_imp.h"
28 #include "cmemory.h"
29 #include "uassert.h"
30 #include "sprpimpl.h"
31
32 /* it is official IDNA ACE Prefix is "xn--" */
33 static const UChar ACE_PREFIX[] ={ 0x0078,0x006E,0x002d,0x002d } ;
34 #define ACE_PREFIX_LENGTH 4
35
36 #define MAX_LABEL_LENGTH 63
37 /* The Max length of the labels should not be more than MAX_LABEL_LENGTH */
38 #define MAX_LABEL_BUFFER_SIZE 100
39
40 #define MAX_DOMAIN_NAME_LENGTH 255
41 /* The Max length of the domain names should not be more than MAX_DOMAIN_NAME_LENGTH */
42 #define MAX_IDN_BUFFER_SIZE MAX_DOMAIN_NAME_LENGTH+1
43
44 #define LOWER_CASE_DELTA 0x0020
45 #define HYPHEN 0x002D
46 #define FULL_STOP 0x002E
47 #define CAPITAL_A 0x0041
48 #define CAPITAL_Z 0x005A
49
50 inline static UChar
toASCIILower(UChar ch)51 toASCIILower(UChar ch){
52 if(CAPITAL_A <= ch && ch <= CAPITAL_Z){
53 return ch + LOWER_CASE_DELTA;
54 }
55 return ch;
56 }
57
58 inline static UBool
startsWithPrefix(const UChar * src,int32_t srcLength)59 startsWithPrefix(const UChar* src , int32_t srcLength){
60 UBool startsWithPrefix = TRUE;
61
62 if(srcLength < ACE_PREFIX_LENGTH){
63 return FALSE;
64 }
65
66 for(int8_t i=0; i< ACE_PREFIX_LENGTH; i++){
67 if(toASCIILower(src[i]) != ACE_PREFIX[i]){
68 startsWithPrefix = FALSE;
69 }
70 }
71 return startsWithPrefix;
72 }
73
74
75 inline static int32_t
compareCaseInsensitiveASCII(const UChar * s1,int32_t s1Len,const UChar * s2,int32_t s2Len)76 compareCaseInsensitiveASCII(const UChar* s1, int32_t s1Len,
77 const UChar* s2, int32_t s2Len){
78
79 int32_t minLength;
80 int32_t lengthResult;
81
82 // are we comparing different lengths?
83 if(s1Len != s2Len) {
84 if(s1Len < s2Len) {
85 minLength = s1Len;
86 lengthResult = -1;
87 } else {
88 minLength = s2Len;
89 lengthResult = 1;
90 }
91 } else {
92 // ok the lengths are equal
93 minLength = s1Len;
94 lengthResult = 0;
95 }
96
97 UChar c1,c2;
98 int32_t rc;
99
100 for(int32_t i =0;/* no condition */;i++) {
101
102 /* If we reach the ends of both strings then they match */
103 if(i == minLength) {
104 return lengthResult;
105 }
106
107 c1 = s1[i];
108 c2 = s2[i];
109
110 /* Case-insensitive comparison */
111 if(c1!=c2) {
112 rc=(int32_t)toASCIILower(c1)-(int32_t)toASCIILower(c2);
113 if(rc!=0) {
114 lengthResult=rc;
115 break;
116 }
117 }
118 }
119 return lengthResult;
120 }
121
122
123 /**
124 * Ascertain if the given code point is a label separator as
125 * defined by the IDNA RFC
126 *
127 * @param ch The code point to be ascertained
128 * @return true if the char is a label separator
129 * @stable ICU 2.8
130 */
isLabelSeparator(UChar ch)131 static inline UBool isLabelSeparator(UChar ch){
132 switch(ch){
133 case 0x002e:
134 case 0x3002:
135 case 0xFF0E:
136 case 0xFF61:
137 return TRUE;
138 default:
139 return FALSE;
140 }
141 }
142
143 // returns the length of the label excluding the separator
144 // if *limit == separator then the length returned does not include
145 // the separtor.
146 static inline int32_t
getNextSeparator(UChar * src,int32_t srcLength,UChar ** limit,UBool * done)147 getNextSeparator(UChar *src, int32_t srcLength,
148 UChar **limit, UBool *done){
149 if(srcLength == -1){
150 int32_t i;
151 for(i=0 ; ;i++){
152 if(src[i] == 0){
153 *limit = src + i; // point to null
154 *done = TRUE;
155 return i;
156 }
157 if(isLabelSeparator(src[i])){
158 *limit = src + (i+1); // go past the delimiter
159 return i;
160
161 }
162 }
163 }else{
164 int32_t i;
165 for(i=0;i<srcLength;i++){
166 if(isLabelSeparator(src[i])){
167 *limit = src + (i+1); // go past the delimiter
168 return i;
169 }
170 }
171 // we have not found the delimiter
172 // if(i==srcLength)
173 *limit = src+srcLength;
174 *done = TRUE;
175
176 return i;
177 }
178 }
isLDHChar(UChar ch)179 static inline UBool isLDHChar(UChar ch){
180 // high runner case
181 if(ch>0x007A){
182 return FALSE;
183 }
184 //[\\u002D \\u0030-\\u0039 \\u0041-\\u005A \\u0061-\\u007A]
185 if( (ch==0x002D) ||
186 (0x0030 <= ch && ch <= 0x0039) ||
187 (0x0041 <= ch && ch <= 0x005A) ||
188 (0x0061 <= ch && ch <= 0x007A)
189 ){
190 return TRUE;
191 }
192 return FALSE;
193 }
194
195 static int32_t
_internal_toASCII(const UChar * src,int32_t srcLength,UChar * dest,int32_t destCapacity,int32_t options,UStringPrepProfile * nameprep,UParseError * parseError,UErrorCode * status)196 _internal_toASCII(const UChar* src, int32_t srcLength,
197 UChar* dest, int32_t destCapacity,
198 int32_t options,
199 UStringPrepProfile* nameprep,
200 UParseError* parseError,
201 UErrorCode* status)
202 {
203
204 // TODO Revisit buffer handling. The label should not be over 63 ASCII characters. ICU4J may need to be updated too.
205 UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE];
206 //initialize pointers to stack buffers
207 UChar *b1 = b1Stack, *b2 = b2Stack;
208 int32_t b1Len=0, b2Len,
209 b1Capacity = MAX_LABEL_BUFFER_SIZE,
210 b2Capacity = MAX_LABEL_BUFFER_SIZE ,
211 reqLength=0;
212
213 int32_t namePrepOptions = ((options & UIDNA_ALLOW_UNASSIGNED) != 0) ? USPREP_ALLOW_UNASSIGNED: 0;
214 UBool* caseFlags = NULL;
215
216 // the source contains all ascii codepoints
217 UBool srcIsASCII = TRUE;
218 // assume the source contains all LDH codepoints
219 UBool srcIsLDH = TRUE;
220
221 int32_t j=0;
222
223 //get the options
224 UBool useSTD3ASCIIRules = (UBool)((options & UIDNA_USE_STD3_RULES) != 0);
225
226 int32_t failPos = -1;
227
228 if(srcLength == -1){
229 srcLength = u_strlen(src);
230 }
231
232 if(srcLength > b1Capacity){
233 b1 = (UChar*) uprv_malloc(srcLength * U_SIZEOF_UCHAR);
234 if(b1==NULL){
235 *status = U_MEMORY_ALLOCATION_ERROR;
236 goto CLEANUP;
237 }
238 b1Capacity = srcLength;
239 }
240
241 // step 1
242 for( j=0;j<srcLength;j++){
243 if(src[j] > 0x7F){
244 srcIsASCII = FALSE;
245 }
246 b1[b1Len++] = src[j];
247 }
248
249 // step 2 is performed only if the source contains non ASCII
250 if(srcIsASCII == FALSE){
251
252 // step 2
253 b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Capacity, namePrepOptions, parseError, status);
254
255 if(*status == U_BUFFER_OVERFLOW_ERROR){
256 // redo processing of string
257 // we do not have enough room so grow the buffer
258 if(b1 != b1Stack){
259 uprv_free(b1);
260 }
261 b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
262 if(b1==NULL){
263 *status = U_MEMORY_ALLOCATION_ERROR;
264 goto CLEANUP;
265 }
266
267 *status = U_ZERO_ERROR; // reset error
268
269 b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Len, namePrepOptions, parseError, status);
270 }
271 }
272 // error bail out
273 if(U_FAILURE(*status)){
274 goto CLEANUP;
275 }
276 if(b1Len == 0){
277 *status = U_IDNA_ZERO_LENGTH_LABEL_ERROR;
278 goto CLEANUP;
279 }
280
281 // for step 3 & 4
282 srcIsASCII = TRUE;
283 for( j=0;j<b1Len;j++){
284 // check if output of usprep_prepare is all ASCII
285 if(b1[j] > 0x7F){
286 srcIsASCII = FALSE;
287 }else if(isLDHChar(b1[j])==FALSE){ // if the char is in ASCII range verify that it is an LDH character
288 srcIsLDH = FALSE;
289 failPos = j;
290 }
291 }
292 if(useSTD3ASCIIRules == TRUE){
293 // verify 3a and 3b
294 // 3(a) Verify the absence of non-LDH ASCII code points; that is, the
295 // absence of 0..2C, 2E..2F, 3A..40, 5B..60, and 7B..7F.
296 // 3(b) Verify the absence of leading and trailing hyphen-minus; that
297 // is, the absence of U+002D at the beginning and end of the
298 // sequence.
299 if( srcIsLDH == FALSE /* source at this point should not contain anyLDH characters */
300 || b1[0] == HYPHEN || b1[b1Len-1] == HYPHEN){
301 *status = U_IDNA_STD3_ASCII_RULES_ERROR;
302
303 /* populate the parseError struct */
304 if(srcIsLDH==FALSE){
305 // failPos is always set the index of failure
306 uprv_syntaxError(b1,failPos, b1Len,parseError);
307 }else if(b1[0] == HYPHEN){
308 // fail position is 0
309 uprv_syntaxError(b1,0,b1Len,parseError);
310 }else{
311 // the last index in the source is always length-1
312 uprv_syntaxError(b1, (b1Len>0) ? b1Len-1 : b1Len, b1Len,parseError);
313 }
314
315 goto CLEANUP;
316 }
317 }
318 // Step 4: if the source is ASCII then proceed to step 8
319 if(srcIsASCII){
320 if(b1Len <= destCapacity){
321 u_memmove(dest, b1, b1Len);
322 reqLength = b1Len;
323 }else{
324 reqLength = b1Len;
325 goto CLEANUP;
326 }
327 }else{
328 // step 5 : verify the sequence does not begin with ACE prefix
329 if(!startsWithPrefix(b1,b1Len)){
330
331 //step 6: encode the sequence with punycode
332
333 // do not preserve the case flags for now!
334 // TODO: Preserve the case while implementing the RFE
335 // caseFlags = (UBool*) uprv_malloc(b1Len * sizeof(UBool));
336 // uprv_memset(caseFlags,TRUE,b1Len);
337
338 b2Len = u_strToPunycode(b1,b1Len,b2,b2Capacity,caseFlags, status);
339
340 if(*status == U_BUFFER_OVERFLOW_ERROR){
341 // redo processing of string
342 /* we do not have enough room so grow the buffer*/
343 b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR);
344 if(b2 == NULL){
345 *status = U_MEMORY_ALLOCATION_ERROR;
346 goto CLEANUP;
347 }
348
349 *status = U_ZERO_ERROR; // reset error
350
351 b2Len = u_strToPunycode(b1,b1Len,b2,b2Len,caseFlags, status);
352 }
353 //error bail out
354 if(U_FAILURE(*status)){
355 goto CLEANUP;
356 }
357 // TODO : Reconsider while implementing the case preserve RFE
358 // convert all codepoints to lower case ASCII
359 // toASCIILower(b2,b2Len);
360 reqLength = b2Len+ACE_PREFIX_LENGTH;
361
362 if(reqLength > destCapacity){
363 *status = U_BUFFER_OVERFLOW_ERROR;
364 goto CLEANUP;
365 }
366 //Step 7: prepend the ACE prefix
367 u_memcpy(dest, ACE_PREFIX, ACE_PREFIX_LENGTH);
368 //Step 6: copy the contents in b2 into dest
369 u_memcpy(dest+ACE_PREFIX_LENGTH, b2, b2Len);
370
371 }else{
372 *status = U_IDNA_ACE_PREFIX_ERROR;
373 //position of failure is 0
374 uprv_syntaxError(b1,0,b1Len,parseError);
375 goto CLEANUP;
376 }
377 }
378 // step 8: verify the length of label
379 if(reqLength > MAX_LABEL_LENGTH){
380 *status = U_IDNA_LABEL_TOO_LONG_ERROR;
381 }
382
383 CLEANUP:
384 if(b1 != b1Stack){
385 uprv_free(b1);
386 }
387 if(b2 != b2Stack){
388 uprv_free(b2);
389 }
390 uprv_free(caseFlags);
391
392 return u_terminateUChars(dest, destCapacity, reqLength, status);
393 }
394
395 static int32_t
_internal_toUnicode(const UChar * src,int32_t srcLength,UChar * dest,int32_t destCapacity,int32_t options,UStringPrepProfile * nameprep,UParseError * parseError,UErrorCode * status)396 _internal_toUnicode(const UChar* src, int32_t srcLength,
397 UChar* dest, int32_t destCapacity,
398 int32_t options,
399 UStringPrepProfile* nameprep,
400 UParseError* parseError,
401 UErrorCode* status)
402 {
403
404 //get the options
405 //UBool useSTD3ASCIIRules = (UBool)((options & UIDNA_USE_STD3_RULES) != 0);
406 int32_t namePrepOptions = ((options & UIDNA_ALLOW_UNASSIGNED) != 0) ? USPREP_ALLOW_UNASSIGNED: 0;
407
408 // TODO Revisit buffer handling. The label should not be over 63 ASCII characters. ICU4J may need to be updated too.
409 UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE], b3Stack[MAX_LABEL_BUFFER_SIZE];
410
411 //initialize pointers to stack buffers
412 UChar *b1 = b1Stack, *b2 = b2Stack, *b1Prime=NULL, *b3=b3Stack;
413 int32_t b1Len = 0, b2Len, b1PrimeLen, b3Len,
414 b1Capacity = MAX_LABEL_BUFFER_SIZE,
415 b2Capacity = MAX_LABEL_BUFFER_SIZE,
416 b3Capacity = MAX_LABEL_BUFFER_SIZE,
417 reqLength=0;
418
419 UBool* caseFlags = NULL;
420
421 UBool srcIsASCII = TRUE;
422 /*UBool srcIsLDH = TRUE;
423 int32_t failPos =0;*/
424
425 // step 1: find out if all the codepoints in src are ASCII
426 if(srcLength==-1){
427 srcLength = 0;
428 for(;src[srcLength]!=0;){
429 if(src[srcLength]> 0x7f){
430 srcIsASCII = FALSE;
431 }/*else if(isLDHChar(src[srcLength])==FALSE){
432 // here we do not assemble surrogates
433 // since we know that LDH code points
434 // are in the ASCII range only
435 srcIsLDH = FALSE;
436 failPos = srcLength;
437 }*/
438 srcLength++;
439 }
440 }else if(srcLength > 0){
441 for(int32_t j=0; j<srcLength; j++){
442 if(src[j]> 0x7f){
443 srcIsASCII = FALSE;
444 }/*else if(isLDHChar(src[j])==FALSE){
445 // here we do not assemble surrogates
446 // since we know that LDH code points
447 // are in the ASCII range only
448 srcIsLDH = FALSE;
449 failPos = j;
450 }*/
451 }
452 }else{
453 return 0;
454 }
455
456 if(srcIsASCII == FALSE){
457 // step 2: process the string
458 b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Capacity, namePrepOptions, parseError, status);
459 if(*status == U_BUFFER_OVERFLOW_ERROR){
460 // redo processing of string
461 /* we do not have enough room so grow the buffer*/
462 b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
463 if(b1==NULL){
464 *status = U_MEMORY_ALLOCATION_ERROR;
465 goto CLEANUP;
466 }
467
468 *status = U_ZERO_ERROR; // reset error
469
470 b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Len, namePrepOptions, parseError, status);
471 }
472 //bail out on error
473 if(U_FAILURE(*status)){
474 goto CLEANUP;
475 }
476 }else{
477
478 //just point src to b1
479 b1 = (UChar*) src;
480 b1Len = srcLength;
481 }
482
483 // The RFC states that
484 // <quote>
485 // ToUnicode never fails. If any step fails, then the original input
486 // is returned immediately in that step.
487 // </quote>
488
489 //step 3: verify ACE Prefix
490 if(startsWithPrefix(b1,b1Len)){
491
492 //step 4: Remove the ACE Prefix
493 b1Prime = b1 + ACE_PREFIX_LENGTH;
494 b1PrimeLen = b1Len - ACE_PREFIX_LENGTH;
495
496 //step 5: Decode using punycode
497 b2Len = u_strFromPunycode(b1Prime, b1PrimeLen, b2, b2Capacity, caseFlags,status);
498
499 if(*status == U_BUFFER_OVERFLOW_ERROR){
500 // redo processing of string
501 /* we do not have enough room so grow the buffer*/
502 b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR);
503 if(b2==NULL){
504 *status = U_MEMORY_ALLOCATION_ERROR;
505 goto CLEANUP;
506 }
507
508 *status = U_ZERO_ERROR; // reset error
509
510 b2Len = u_strFromPunycode(b1Prime, b1PrimeLen, b2, b2Len, caseFlags, status);
511 }
512
513
514 //step 6:Apply toASCII
515 b3Len = uidna_toASCII(b2, b2Len, b3, b3Capacity, options, parseError, status);
516
517 if(*status == U_BUFFER_OVERFLOW_ERROR){
518 // redo processing of string
519 /* we do not have enough room so grow the buffer*/
520 b3 = (UChar*) uprv_malloc(b3Len * U_SIZEOF_UCHAR);
521 if(b3==NULL){
522 *status = U_MEMORY_ALLOCATION_ERROR;
523 goto CLEANUP;
524 }
525
526 *status = U_ZERO_ERROR; // reset error
527
528 b3Len = uidna_toASCII(b2,b2Len,b3,b3Len,options,parseError, status);
529
530 }
531 //bail out on error
532 if(U_FAILURE(*status)){
533 goto CLEANUP;
534 }
535
536 //step 7: verify
537 if(compareCaseInsensitiveASCII(b1, b1Len, b3, b3Len) !=0){
538 // Cause the original to be returned.
539 *status = U_IDNA_VERIFICATION_ERROR;
540 goto CLEANUP;
541 }
542
543 //step 8: return output of step 5
544 reqLength = b2Len;
545 if(b2Len <= destCapacity) {
546 u_memmove(dest, b2, b2Len);
547 }
548 }
549 else{
550 // See the start of this if statement for why this is commented out.
551 // verify that STD3 ASCII rules are satisfied
552 /*if(useSTD3ASCIIRules == TRUE){
553 if( srcIsLDH == FALSE // source contains some non-LDH characters
554 || src[0] == HYPHEN || src[srcLength-1] == HYPHEN){
555 *status = U_IDNA_STD3_ASCII_RULES_ERROR;
556
557 // populate the parseError struct
558 if(srcIsLDH==FALSE){
559 // failPos is always set the index of failure
560 uprv_syntaxError(src,failPos, srcLength,parseError);
561 }else if(src[0] == HYPHEN){
562 // fail position is 0
563 uprv_syntaxError(src,0,srcLength,parseError);
564 }else{
565 // the last index in the source is always length-1
566 uprv_syntaxError(src, (srcLength>0) ? srcLength-1 : srcLength, srcLength,parseError);
567 }
568
569 goto CLEANUP;
570 }
571 }*/
572 // just return the source
573 //copy the source to destination
574 if(srcLength <= destCapacity){
575 u_memmove(dest, src, srcLength);
576 }
577 reqLength = srcLength;
578 }
579
580
581 CLEANUP:
582
583 if(b1 != b1Stack && b1!=src){
584 uprv_free(b1);
585 }
586 if(b2 != b2Stack){
587 uprv_free(b2);
588 }
589 uprv_free(caseFlags);
590
591 // The RFC states that
592 // <quote>
593 // ToUnicode never fails. If any step fails, then the original input
594 // is returned immediately in that step.
595 // </quote>
596 // So if any step fails lets copy source to destination
597 if(U_FAILURE(*status)){
598 //copy the source to destination
599 if(dest && srcLength <= destCapacity){
600 // srcLength should have already been set earlier.
601 U_ASSERT(srcLength >= 0);
602 u_memmove(dest, src, srcLength);
603 }
604 reqLength = srcLength;
605 *status = U_ZERO_ERROR;
606 }
607
608 return u_terminateUChars(dest, destCapacity, reqLength, status);
609 }
610
611 U_CAPI int32_t U_EXPORT2
uidna_toASCII(const UChar * src,int32_t srcLength,UChar * dest,int32_t destCapacity,int32_t options,UParseError * parseError,UErrorCode * status)612 uidna_toASCII(const UChar* src, int32_t srcLength,
613 UChar* dest, int32_t destCapacity,
614 int32_t options,
615 UParseError* parseError,
616 UErrorCode* status){
617
618 if(status == NULL || U_FAILURE(*status)){
619 return 0;
620 }
621 if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
622 *status = U_ILLEGAL_ARGUMENT_ERROR;
623 return 0;
624 }
625
626 UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
627
628 if(U_FAILURE(*status)){
629 return -1;
630 }
631
632 int32_t retLen = _internal_toASCII(src, srcLength, dest, destCapacity, options, nameprep, parseError, status);
633
634 /* close the profile*/
635 usprep_close(nameprep);
636
637 return retLen;
638 }
639
640 U_CAPI int32_t U_EXPORT2
uidna_toUnicode(const UChar * src,int32_t srcLength,UChar * dest,int32_t destCapacity,int32_t options,UParseError * parseError,UErrorCode * status)641 uidna_toUnicode(const UChar* src, int32_t srcLength,
642 UChar* dest, int32_t destCapacity,
643 int32_t options,
644 UParseError* parseError,
645 UErrorCode* status){
646
647 if(status == NULL || U_FAILURE(*status)){
648 return 0;
649 }
650 if( (src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
651 *status = U_ILLEGAL_ARGUMENT_ERROR;
652 return 0;
653 }
654
655 UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
656
657 if(U_FAILURE(*status)){
658 return -1;
659 }
660
661 int32_t retLen = _internal_toUnicode(src, srcLength, dest, destCapacity, options, nameprep, parseError, status);
662
663 usprep_close(nameprep);
664
665 return retLen;
666 }
667
668
669 U_CAPI int32_t U_EXPORT2
uidna_IDNToASCII(const UChar * src,int32_t srcLength,UChar * dest,int32_t destCapacity,int32_t options,UParseError * parseError,UErrorCode * status)670 uidna_IDNToASCII( const UChar *src, int32_t srcLength,
671 UChar* dest, int32_t destCapacity,
672 int32_t options,
673 UParseError *parseError,
674 UErrorCode *status){
675
676 if(status == NULL || U_FAILURE(*status)){
677 return 0;
678 }
679 if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
680 *status = U_ILLEGAL_ARGUMENT_ERROR;
681 return 0;
682 }
683
684 int32_t reqLength = 0;
685
686 UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
687
688 if(U_FAILURE(*status)){
689 return 0;
690 }
691
692 //initialize pointers
693 UChar *delimiter = (UChar*)src;
694 UChar *labelStart = (UChar*)src;
695 UChar *currentDest = (UChar*) dest;
696 int32_t remainingLen = srcLength;
697 int32_t remainingDestCapacity = destCapacity;
698 int32_t labelLen = 0, labelReqLength = 0;
699 UBool done = FALSE;
700
701
702 for(;;){
703
704 labelLen = getNextSeparator(labelStart,remainingLen, &delimiter,&done);
705 labelReqLength = 0;
706 if(!(labelLen==0 && done)){// make sure this is not a root label separator.
707
708 labelReqLength = _internal_toASCII( labelStart, labelLen,
709 currentDest, remainingDestCapacity,
710 options, nameprep,
711 parseError, status);
712
713 if(*status == U_BUFFER_OVERFLOW_ERROR){
714
715 *status = U_ZERO_ERROR; // reset error
716 remainingDestCapacity = 0;
717 }
718 }
719
720
721 if(U_FAILURE(*status)){
722 break;
723 }
724
725 reqLength +=labelReqLength;
726 // adjust the destination pointer
727 if(labelReqLength < remainingDestCapacity){
728 currentDest = currentDest + labelReqLength;
729 remainingDestCapacity -= labelReqLength;
730 }else{
731 // should never occur
732 remainingDestCapacity = 0;
733 }
734
735 if(done == TRUE){
736 break;
737 }
738
739 // add the label separator
740 if(remainingDestCapacity > 0){
741 *currentDest++ = FULL_STOP;
742 remainingDestCapacity--;
743 }
744 reqLength++;
745
746 labelStart = delimiter;
747 if(remainingLen >0 ){
748 remainingLen = (int32_t)(srcLength - (delimiter - src));
749 }
750
751 }
752
753 if(reqLength > MAX_DOMAIN_NAME_LENGTH){
754 *status = U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR;
755 }
756
757 usprep_close(nameprep);
758
759 return u_terminateUChars(dest, destCapacity, reqLength, status);
760 }
761
762 U_CAPI int32_t U_EXPORT2
uidna_IDNToUnicode(const UChar * src,int32_t srcLength,UChar * dest,int32_t destCapacity,int32_t options,UParseError * parseError,UErrorCode * status)763 uidna_IDNToUnicode( const UChar* src, int32_t srcLength,
764 UChar* dest, int32_t destCapacity,
765 int32_t options,
766 UParseError* parseError,
767 UErrorCode* status){
768
769 if(status == NULL || U_FAILURE(*status)){
770 return 0;
771 }
772 if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){
773 *status = U_ILLEGAL_ARGUMENT_ERROR;
774 return 0;
775 }
776
777 int32_t reqLength = 0;
778
779 UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status);
780
781 if(U_FAILURE(*status)){
782 return 0;
783 }
784
785 //initialize pointers
786 UChar *delimiter = (UChar*)src;
787 UChar *labelStart = (UChar*)src;
788 UChar *currentDest = (UChar*) dest;
789 int32_t remainingLen = srcLength;
790 int32_t remainingDestCapacity = destCapacity;
791 int32_t labelLen = 0, labelReqLength = 0;
792 UBool done = FALSE;
793
794 for(;;){
795
796 labelLen = getNextSeparator(labelStart,remainingLen, &delimiter,&done);
797
798 // The RFC states that
799 // <quote>
800 // ToUnicode never fails. If any step fails, then the original input
801 // is returned immediately in that step.
802 // </quote>
803 // _internal_toUnicode will copy the label.
804 /*if(labelLen==0 && done==FALSE){
805 *status = U_IDNA_ZERO_LENGTH_LABEL_ERROR;
806 break;
807 }*/
808
809 labelReqLength = _internal_toUnicode(labelStart, labelLen,
810 currentDest, remainingDestCapacity,
811 options, nameprep,
812 parseError, status);
813
814 if(*status == U_BUFFER_OVERFLOW_ERROR){
815 *status = U_ZERO_ERROR; // reset error
816 remainingDestCapacity = 0;
817 }
818
819 if(U_FAILURE(*status)){
820 break;
821 }
822
823 reqLength +=labelReqLength;
824 // adjust the destination pointer
825 if(labelReqLength < remainingDestCapacity){
826 currentDest = currentDest + labelReqLength;
827 remainingDestCapacity -= labelReqLength;
828 }else{
829 // should never occur
830 remainingDestCapacity = 0;
831 }
832
833 if(done == TRUE){
834 break;
835 }
836
837 // add the label separator
838 // Unlike the ToASCII operation we don't normalize the label separators
839 if(remainingDestCapacity > 0){
840 *currentDest++ = *(labelStart + labelLen);
841 remainingDestCapacity--;
842 }
843 reqLength++;
844
845 labelStart = delimiter;
846 if(remainingLen >0 ){
847 remainingLen = (int32_t)(srcLength - (delimiter - src));
848 }
849
850 }
851
852 if(reqLength > MAX_DOMAIN_NAME_LENGTH){
853 *status = U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR;
854 }
855
856 usprep_close(nameprep);
857
858 return u_terminateUChars(dest, destCapacity, reqLength, status);
859 }
860
861 U_CAPI int32_t U_EXPORT2
uidna_compare(const UChar * s1,int32_t length1,const UChar * s2,int32_t length2,int32_t options,UErrorCode * status)862 uidna_compare( const UChar *s1, int32_t length1,
863 const UChar *s2, int32_t length2,
864 int32_t options,
865 UErrorCode* status){
866
867 if(status == NULL || U_FAILURE(*status)){
868 return -1;
869 }
870
871 UChar b1Stack[MAX_IDN_BUFFER_SIZE], b2Stack[MAX_IDN_BUFFER_SIZE];
872 UChar *b1 = b1Stack, *b2 = b2Stack;
873 int32_t b1Len, b2Len, b1Capacity = MAX_IDN_BUFFER_SIZE, b2Capacity = MAX_IDN_BUFFER_SIZE;
874 int32_t result=-1;
875
876 UParseError parseError;
877
878 b1Len = uidna_IDNToASCII(s1, length1, b1, b1Capacity, options, &parseError, status);
879 if(*status == U_BUFFER_OVERFLOW_ERROR){
880 // redo processing of string
881 b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR);
882 if(b1==NULL){
883 *status = U_MEMORY_ALLOCATION_ERROR;
884 goto CLEANUP;
885 }
886
887 *status = U_ZERO_ERROR; // reset error
888
889 b1Len = uidna_IDNToASCII(s1,length1,b1,b1Len, options, &parseError, status);
890
891 }
892
893 b2Len = uidna_IDNToASCII(s2,length2, b2,b2Capacity, options, &parseError, status);
894 if(*status == U_BUFFER_OVERFLOW_ERROR){
895 // redo processing of string
896 b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR);
897 if(b2==NULL){
898 *status = U_MEMORY_ALLOCATION_ERROR;
899 goto CLEANUP;
900 }
901
902 *status = U_ZERO_ERROR; // reset error
903
904 b2Len = uidna_IDNToASCII(s2, length2, b2, b2Len, options, &parseError, status);
905
906 }
907 // when toASCII is applied all label separators are replaced with FULL_STOP
908 result = compareCaseInsensitiveASCII(b1,b1Len,b2,b2Len);
909
910 CLEANUP:
911 if(b1 != b1Stack){
912 uprv_free(b1);
913 }
914
915 if(b2 != b2Stack){
916 uprv_free(b2);
917 }
918
919 return result;
920 }
921
922 #endif /* #if !UCONFIG_NO_IDNA */
923