1 // © 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /*
4 **********************************************************************
5 * Copyright (C) 2011-2015, International Business Machines Corporation
6 * and others. All Rights Reserved.
7 **********************************************************************
8 */
9 /**
10 * IntlTestSpoof tests for USpoofDetector
11 */
12
13 #include "unicode/utypes.h"
14
15 #if !UCONFIG_NO_REGULAR_EXPRESSIONS && !UCONFIG_NO_NORMALIZATION && !UCONFIG_NO_FILE_IO
16
17 #include "itspoof.h"
18
19 #include "unicode/normlzr.h"
20 #include "unicode/regex.h"
21 #include "unicode/unistr.h"
22 #include "unicode/uscript.h"
23 #include "unicode/uspoof.h"
24
25 #include "cstring.h"
26 #include "scriptset.h"
27 #include "uhash.h"
28
29 #include <stdlib.h>
30 #include <stdio.h>
31
32 #define TEST_ASSERT_SUCCESS(status) UPRV_BLOCK_MACRO_BEGIN { \
33 if (U_FAILURE(status)) { \
34 errcheckln(status, "Failure at file %s, line %d, error = %s", __FILE__, __LINE__, u_errorName(status)); \
35 } \
36 } UPRV_BLOCK_MACRO_END
37
38 #define TEST_ASSERT(expr) UPRV_BLOCK_MACRO_BEGIN { \
39 if ((expr)==FALSE) { \
40 errln("Test Failure at file %s, line %d: \"%s\" is false.", __FILE__, __LINE__, #expr); \
41 } \
42 } UPRV_BLOCK_MACRO_END
43
44 #define TEST_ASSERT_MSG(expr, msg) UPRV_BLOCK_MACRO_BEGIN { \
45 if ((expr)==FALSE) { \
46 dataerrln("Test Failure at file %s, line %d, %s: \"%s\" is false.", __FILE__, __LINE__, msg, #expr); \
47 } \
48 } UPRV_BLOCK_MACRO_END
49
50 #define TEST_ASSERT_EQ(a, b) UPRV_BLOCK_MACRO_BEGIN { \
51 if ((a) != (b)) { \
52 errln("Test Failure at file %s, line %d: \"%s\" (%d) != \"%s\" (%d)", \
53 __FILE__, __LINE__, #a, (a), #b, (b)); \
54 } \
55 } UPRV_BLOCK_MACRO_END
56
57 #define TEST_ASSERT_NE(a, b) UPRV_BLOCK_MACRO_BEGIN { \
58 if ((a) == (b)) { \
59 errln("Test Failure at file %s, line %d: \"%s\" (%d) == \"%s\" (%d)", \
60 __FILE__, __LINE__, #a, (a), #b, (b)); \
61 } \
62 } UPRV_BLOCK_MACRO_END
63
64 /*
65 * TEST_SETUP and TEST_TEARDOWN
66 * macros to handle the boilerplate around setting up test case.
67 * Put arbitrary test code between SETUP and TEARDOWN.
68 * "sc" is the ready-to-go SpoofChecker for use in the tests.
69 */
70 #define TEST_SETUP UPRV_BLOCK_MACRO_BEGIN { \
71 UErrorCode status = U_ZERO_ERROR; \
72 USpoofChecker *sc; \
73 sc = uspoof_open(&status); \
74 TEST_ASSERT_SUCCESS(status); \
75 USpoofCheckResult *checkResult; \
76 checkResult = uspoof_openCheckResult(&status); \
77 TEST_ASSERT_SUCCESS(status); \
78 if (U_SUCCESS(status)){
79
80 #define TEST_TEARDOWN \
81 } \
82 TEST_ASSERT_SUCCESS(status); \
83 uspoof_closeCheckResult(checkResult); \
84 uspoof_close(sc); \
85 } UPRV_BLOCK_MACRO_END
86
87
88
89
runIndexedTest(int32_t index,UBool exec,const char * & name,char *)90 void IntlTestSpoof::runIndexedTest( int32_t index, UBool exec, const char* &name, char* /*par*/ )
91 {
92 if (exec) {
93 logln("TestSuite spoof: ");
94 }
95 TESTCASE_AUTO_BEGIN;
96 TESTCASE_AUTO(testSpoofAPI);
97 TESTCASE_AUTO(testSkeleton);
98 TESTCASE_AUTO(testAreConfusable);
99 TESTCASE_AUTO(testInvisible);
100 TESTCASE_AUTO(testConfData);
101 TESTCASE_AUTO(testBug8654);
102 TESTCASE_AUTO(testScriptSet);
103 TESTCASE_AUTO(testRestrictionLevel);
104 TESTCASE_AUTO(testMixedNumbers);
105 TESTCASE_AUTO(testBug12153);
106 TESTCASE_AUTO(testBug12825);
107 TESTCASE_AUTO(testBug12815);
108 TESTCASE_AUTO(testBug13314_MixedNumbers);
109 TESTCASE_AUTO(testBug13328_MixedCombiningMarks);
110 TESTCASE_AUTO(testCombiningDot);
111 TESTCASE_AUTO_END;
112 }
113
testSpoofAPI()114 void IntlTestSpoof::testSpoofAPI() {
115
116 TEST_SETUP
117 UnicodeString s("xyz"); // Many latin ranges are whole-script confusable with other scripts.
118 // If this test starts failing, consult confusablesWholeScript.txt
119 int32_t position = 666;
120 int32_t checkResults = uspoof_checkUnicodeString(sc, s, &position, &status);
121 TEST_ASSERT_SUCCESS(status);
122 TEST_ASSERT_EQ(0, checkResults);
123 TEST_ASSERT_EQ(0, position);
124 TEST_TEARDOWN;
125
126 TEST_SETUP
127 UnicodeString s1("cxs");
128 UnicodeString s2 = UnicodeString("\\u0441\\u0445\\u0455").unescape(); // Cyrillic "cxs"
129 int32_t checkResults = uspoof_areConfusableUnicodeString(sc, s1, s2, &status);
130 TEST_ASSERT_SUCCESS(status);
131 TEST_ASSERT_EQ(USPOOF_MIXED_SCRIPT_CONFUSABLE | USPOOF_WHOLE_SCRIPT_CONFUSABLE, checkResults);
132
133 TEST_TEARDOWN;
134
135 TEST_SETUP
136 UnicodeString s("I1l0O");
137 UnicodeString dest;
138 UnicodeString &retStr = uspoof_getSkeletonUnicodeString(sc, USPOOF_ANY_CASE, s, dest, &status);
139 TEST_ASSERT_SUCCESS(status);
140 TEST_ASSERT(UnicodeString("lllOO") == dest);
141 TEST_ASSERT(&dest == &retStr);
142 TEST_TEARDOWN;
143 }
144
145
146 #define CHECK_SKELETON(type, input, expected) UPRV_BLOCK_MACRO_BEGIN { \
147 checkSkeleton(sc, type, input, expected, __LINE__); \
148 } UPRV_BLOCK_MACRO_END
149
150
151 // testSkeleton. Spot check a number of confusable skeleton substitutions from the
152 // Unicode data file confusables.txt
153 // Test cases chosen for substitutions of various lengths, and
154 // membership in different mapping tables.
155 // Note: for ICU 55, all tables collapsed to the MA table data.
156 // TODO: for ICU 56 with Unicode 8, revisit this test.
157 //
testSkeleton()158 void IntlTestSpoof::testSkeleton() {
159 const uint32_t ML = 0;
160 const uint32_t SL = USPOOF_SINGLE_SCRIPT_CONFUSABLE;
161 const uint32_t MA = USPOOF_ANY_CASE;
162 const uint32_t SA = USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_ANY_CASE;
163
164 TEST_SETUP
165 CHECK_SKELETON(SL, "nochange", "nochange");
166 CHECK_SKELETON(SA, "nochange", "nochange");
167 CHECK_SKELETON(ML, "nochange", "nochange");
168 CHECK_SKELETON(MA, "nochange", "nochange");
169 CHECK_SKELETON(MA, "love", "love");
170 CHECK_SKELETON(MA, "1ove", "love"); // Digit 1 to letter l
171 CHECK_SKELETON(ML, "OOPS", "OOPS");
172 CHECK_SKELETON(ML, "00PS", "OOPS");
173 CHECK_SKELETON(MA, "OOPS", "OOPS");
174 CHECK_SKELETON(MA, "00PS", "OOPS"); // Digit 0 to letter O in any case mode only
175 CHECK_SKELETON(SL, "\\u059c", "\\u0301");
176 CHECK_SKELETON(SL, "\\u2A74", "\\u003A\\u003A\\u003D");
177 CHECK_SKELETON(SL, "\\u247E", "\\u0028\\u006C\\u006C\\u0029"); // "(ll)"
178 CHECK_SKELETON(SL, "\\uFDFB", "\\u062C\\u0644\\u0020\\u062C\\u0644\\u006c\\u0644\\u006f");
179
180 // This mapping exists in the ML and MA tables, does not exist in SL, SA
181 // 0C83 ; 0983 ; ML
182 // 0C83 ; 0983 ; MA
183 //
184
185 CHECK_SKELETON(SL, "\\u0C83", "\\u0983");
186 CHECK_SKELETON(SA, "\\u0C83", "\\u0983");
187 CHECK_SKELETON(ML, "\\u0C83", "\\u0983");
188 CHECK_SKELETON(MA, "\\u0C83", "\\u0983");
189
190 // 0391 mappings exist only in MA and SA tables.
191 CHECK_SKELETON(MA, "\\u0391", "A");
192 CHECK_SKELETON(SA, "\\u0391", "A");
193 CHECK_SKELETON(ML, "\\u0391", "A");
194 CHECK_SKELETON(SL, "\\u0391", "A");
195
196 // 13CF Mappings in all four tables, different in MA.
197 CHECK_SKELETON(ML, "\\u13CF", "b");
198 CHECK_SKELETON(MA, "\\u13CF", "b");
199 CHECK_SKELETON(SL, "\\u13CF", "b");
200 CHECK_SKELETON(SA, "\\u13CF", "b");
201
202 // 0022 ; 0027 0027 ;
203 // all tables.
204 CHECK_SKELETON(SL, "\\u0022", "\\u0027\\u0027");
205 CHECK_SKELETON(SA, "\\u0022", "\\u0027\\u0027");
206 CHECK_SKELETON(ML, "\\u0022", "\\u0027\\u0027");
207 CHECK_SKELETON(MA, "\\u0022", "\\u0027\\u0027");
208
209 // 017F mappings exist only in MA and SA tables.
210 CHECK_SKELETON(MA, "\\u017F", "f");
211 CHECK_SKELETON(SA, "\\u017F", "f");
212 CHECK_SKELETON(ML, "\\u017F", "f");
213 CHECK_SKELETON(SL, "\\u017F", "f");
214
215 TEST_TEARDOWN;
216 }
217
218
219 //
220 // Run a single confusable skeleton transformation test case.
221 //
checkSkeleton(const USpoofChecker * sc,uint32_t type,const char * input,const char * expected,int32_t lineNum)222 void IntlTestSpoof::checkSkeleton(const USpoofChecker *sc, uint32_t type,
223 const char *input, const char *expected, int32_t lineNum) {
224 UnicodeString uInput = UnicodeString(input).unescape();
225 UnicodeString uExpected = UnicodeString(expected).unescape();
226
227 UErrorCode status = U_ZERO_ERROR;
228 UnicodeString actual;
229 uspoof_getSkeletonUnicodeString(sc, type, uInput, actual, &status);
230 if (U_FAILURE(status)) {
231 errln("File %s, Line %d, Test case from line %d, status is %s", __FILE__, __LINE__, lineNum,
232 u_errorName(status));
233 return;
234 }
235 if (uExpected != actual) {
236 errln("File %s, Line %d, Test case from line %d, Actual and Expected skeletons differ.",
237 __FILE__, __LINE__, lineNum);
238 errln(UnicodeString(" Actual Skeleton: \"") + actual + UnicodeString("\"\n") +
239 UnicodeString(" Expected Skeleton: \"") + uExpected + UnicodeString("\""));
240 }
241 }
242
testAreConfusable()243 void IntlTestSpoof::testAreConfusable() {
244 TEST_SETUP
245 UnicodeString s1("A long string that will overflow stack buffers. A long string that will overflow stack buffers. "
246 "A long string that will overflow stack buffers. A long string that will overflow stack buffers. ");
247 UnicodeString s2("A long string that wi11 overflow stack buffers. A long string that will overflow stack buffers. "
248 "A long string that wi11 overflow stack buffers. A long string that will overflow stack buffers. ");
249 int32_t result = uspoof_areConfusableUnicodeString(sc, s1, s2, &status);
250 TEST_ASSERT_SUCCESS(status);
251 TEST_ASSERT_EQ(USPOOF_SINGLE_SCRIPT_CONFUSABLE, result);
252
253 TEST_TEARDOWN;
254 }
255
testInvisible()256 void IntlTestSpoof::testInvisible() {
257 TEST_SETUP
258 UnicodeString s = UnicodeString("abcd\\u0301ef").unescape();
259 int32_t position = -42;
260 TEST_ASSERT_EQ(0, uspoof_checkUnicodeString(sc, s, &position, &status));
261 TEST_ASSERT_SUCCESS(status);
262 TEST_ASSERT(0 == position);
263
264 UnicodeString s2 = UnicodeString("abcd\\u0301\\u0302\\u0301ef").unescape();
265 TEST_ASSERT_EQ(USPOOF_INVISIBLE, uspoof_checkUnicodeString(sc, s2, &position, &status));
266 TEST_ASSERT_SUCCESS(status);
267 TEST_ASSERT_EQ(0, position);
268
269 // Two acute accents, one from the composed a with acute accent, \u00e1,
270 // and one separate.
271 position = -42;
272 UnicodeString s3 = UnicodeString("abcd\\u00e1\\u0301xyz").unescape();
273 TEST_ASSERT_EQ(USPOOF_INVISIBLE, uspoof_checkUnicodeString(sc, s3, &position, &status));
274 TEST_ASSERT_SUCCESS(status);
275 TEST_ASSERT_EQ(0, position);
276 TEST_TEARDOWN;
277 }
278
testBug8654()279 void IntlTestSpoof::testBug8654() {
280 TEST_SETUP
281 UnicodeString s = UnicodeString("B\\u00c1\\u0301").unescape();
282 int32_t position = -42;
283 TEST_ASSERT_EQ(USPOOF_INVISIBLE, uspoof_checkUnicodeString(sc, s, &position, &status) & USPOOF_INVISIBLE );
284 TEST_ASSERT_SUCCESS(status);
285 TEST_ASSERT_EQ(0, position);
286 TEST_TEARDOWN;
287 }
288
parseHex(const UnicodeString & in)289 static UnicodeString parseHex(const UnicodeString &in) {
290 // Convert a series of hex numbers in a Unicode String to a string with the
291 // corresponding characters.
292 // The conversion is _really_ annoying. There must be some function to just do it.
293 UnicodeString result;
294 UChar32 cc = 0;
295 for (int32_t i=0; i<in.length(); i++) {
296 UChar c = in.charAt(i);
297 if (c == 0x20) { // Space
298 if (cc > 0) {
299 result.append(cc);
300 cc = 0;
301 }
302 } else if (c>=0x30 && c<=0x39) {
303 cc = (cc<<4) + (c - 0x30);
304 } else if ((c>=0x41 && c<=0x46) || (c>=0x61 && c<=0x66)) {
305 cc = (cc<<4) + (c & 0x0f)+9;
306 }
307 // else do something with bad input.
308 }
309 if (cc > 0) {
310 result.append(cc);
311 }
312 return result;
313 }
314
315
316 //
317 // Append the hex form of a UChar32 to a UnicodeString.
318 // Used in formatting error messages.
319 // Match the formatting of numbers in confusables.txt
320 // Minimum of 4 digits, no leading zeroes for positions 5 and up.
321 //
appendHexUChar(UnicodeString & dest,UChar32 c)322 static void appendHexUChar(UnicodeString &dest, UChar32 c) {
323 UBool doZeroes = FALSE;
324 for (int bitNum=28; bitNum>=0; bitNum-=4) {
325 if (bitNum <= 12) {
326 doZeroes = TRUE;
327 }
328 int hexDigit = (c>>bitNum) & 0x0f;
329 if (hexDigit != 0 || doZeroes) {
330 doZeroes = TRUE;
331 dest.append((UChar)(hexDigit<=9? hexDigit + 0x30: hexDigit -10 + 0x41));
332 }
333 }
334 dest.append((UChar)0x20);
335 }
336
337 U_DEFINE_LOCAL_OPEN_POINTER(LocalStdioFilePointer, FILE, fclose);
338
339 // testConfData - Check each data item from the Unicode confusables.txt file,
340 // verify that it transforms correctly in a skeleton.
341 //
testConfData()342 void IntlTestSpoof::testConfData() {
343 char buffer[2000];
344 if (getUnidataPath(buffer) == NULL) {
345 errln("Skipping test spoof/testConfData. Unable to find path to source/data/unidata/.");
346 return;
347 }
348 uprv_strcat(buffer, "confusables.txt");
349
350 LocalStdioFilePointer f(fopen(buffer, "rb"));
351 if (f.isNull()) {
352 errln("Skipping test spoof/testConfData. File confusables.txt not accessible.");
353 return;
354 }
355 fseek(f.getAlias(), 0, SEEK_END);
356 int32_t fileSize = ftell(f.getAlias());
357 LocalArray<char> fileBuf(new char[fileSize]);
358 fseek(f.getAlias(), 0, SEEK_SET);
359 int32_t amt_read = static_cast<int32_t>(fread(fileBuf.getAlias(), 1, fileSize, f.getAlias()));
360 TEST_ASSERT_EQ(amt_read, fileSize);
361 TEST_ASSERT(fileSize>0);
362 if (amt_read != fileSize || fileSize <=0) {
363 return;
364 }
365 UnicodeString confusablesTxt = UnicodeString::fromUTF8(StringPiece(fileBuf.getAlias(), fileSize));
366
367 UErrorCode status = U_ZERO_ERROR;
368 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
369 TEST_ASSERT_SUCCESS(status);
370
371 // Parse lines from the confusables.txt file. Example Line:
372 // FF44 ; 0064 ; SL # ( d -> d ) FULLWIDTH ....
373 // Three fields. The hex fields can contain more than one character,
374 // and each character may be more than 4 digits (for supplemntals)
375 // This regular expression matches lines and splits the fields into capture groups.
376 RegexMatcher parseLine("(?m)^([0-9A-F]{4}[^#;]*?);([^#;]*?);([^#]*)", confusablesTxt, 0, status);
377 TEST_ASSERT_SUCCESS(status);
378 while (parseLine.find()) {
379 UnicodeString from = parseHex(parseLine.group(1, status));
380 if (!Normalizer::isNormalized(from, UNORM_NFD, status)) {
381 // The source character was not NFD.
382 // Skip this case; the first step in obtaining a skeleton is to NFD the input,
383 // so the mapping in this line of confusables.txt will never be applied.
384 continue;
385 }
386
387 UnicodeString rawExpected = parseHex(parseLine.group(2, status));
388 UnicodeString expected;
389 Normalizer::decompose(rawExpected, FALSE /*NFD*/, 0, expected, status);
390 TEST_ASSERT_SUCCESS(status);
391
392 int32_t skeletonType = 0;
393 UnicodeString tableType = parseLine.group(3, status);
394 TEST_ASSERT_SUCCESS(status);
395 if (tableType.indexOf("SL") >= 0) {
396 skeletonType = USPOOF_SINGLE_SCRIPT_CONFUSABLE;
397 } else if (tableType.indexOf("SA") >= 0) {
398 skeletonType = USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_ANY_CASE;
399 } else if (tableType.indexOf("ML") >= 0) {
400 skeletonType = 0;
401 } else if (tableType.indexOf("MA") >= 0) {
402 skeletonType = USPOOF_ANY_CASE;
403 }
404
405 UnicodeString actual;
406 uspoof_getSkeletonUnicodeString(sc.getAlias(), skeletonType, from, actual, &status);
407 TEST_ASSERT_SUCCESS(status);
408 TEST_ASSERT(actual == expected);
409 if (actual != expected) {
410 errln(parseLine.group(0, status));
411 UnicodeString line = "Actual: ";
412 int i = 0;
413 while (i < actual.length()) {
414 appendHexUChar(line, actual.char32At(i));
415 i = actual.moveIndex32(i, 1);
416 }
417 errln(line);
418 }
419 if (U_FAILURE(status)) {
420 break;
421 }
422 }
423 }
424
425
testScriptSet()426 void IntlTestSpoof::testScriptSet() {
427 // ScriptSet::SCRIPT_LIMIT is hardcoded.
428 // Increase it by multiples of 32 if there are too many script codes.
429 TEST_ASSERT(USCRIPT_CODE_LIMIT <= ScriptSet::SCRIPT_LIMIT);
430 // USCRIPT_CODE_LIMIT should include all script codes,
431 // but theoretically the data may define more.
432 TEST_ASSERT(u_getIntPropertyMaxValue(UCHAR_SCRIPT) < ScriptSet::SCRIPT_LIMIT);
433
434 ScriptSet s1;
435 ScriptSet s2;
436 UErrorCode status = U_ZERO_ERROR;
437
438 TEST_ASSERT(s1 == s2);
439 s1.set(USCRIPT_ARABIC,status);
440 TEST_ASSERT_SUCCESS(status);
441 TEST_ASSERT(!(s1 == s2));
442 TEST_ASSERT(s1.test(USCRIPT_ARABIC, status));
443 TEST_ASSERT(s1.test(USCRIPT_GREEK, status) == FALSE);
444
445 status = U_ZERO_ERROR;
446 s1.reset(USCRIPT_ARABIC, status);
447 TEST_ASSERT(s1 == s2);
448
449 static constexpr UScriptCode LAST_SCRIPT_CODE = (UScriptCode)(USCRIPT_CODE_LIMIT - 1);
450 status = U_ZERO_ERROR;
451 s1.setAll();
452 TEST_ASSERT(s1.test(USCRIPT_COMMON, status));
453 TEST_ASSERT(s1.test(USCRIPT_ETHIOPIC, status));
454 TEST_ASSERT(s1.test(LAST_SCRIPT_CODE, status));
455 s1.resetAll();
456 TEST_ASSERT(!s1.test(USCRIPT_COMMON, status));
457 TEST_ASSERT(!s1.test(USCRIPT_ETHIOPIC, status));
458 TEST_ASSERT(!s1.test(LAST_SCRIPT_CODE, status));
459
460 status = U_ZERO_ERROR;
461 s1.set(USCRIPT_TAKRI, status);
462 s1.set(USCRIPT_BLISSYMBOLS, status);
463 s2.setAll();
464 TEST_ASSERT(s2.contains(s1));
465 TEST_ASSERT(!s1.contains(s2));
466 TEST_ASSERT(s2.intersects(s1));
467 TEST_ASSERT(s1.intersects(s2));
468 s2.reset(USCRIPT_TAKRI, status);
469 TEST_ASSERT(!s2.contains(s1));
470 TEST_ASSERT(!s1.contains(s2));
471 TEST_ASSERT(s1.intersects(s2));
472 TEST_ASSERT(s2.intersects(s1));
473 TEST_ASSERT_SUCCESS(status);
474
475 status = U_ZERO_ERROR;
476 s1.resetAll();
477 s1.set(USCRIPT_NKO, status);
478 s1.set(USCRIPT_COMMON, status);
479 s2 = s1;
480 TEST_ASSERT(s2 == s1);
481 TEST_ASSERT_EQ(2, s2.countMembers());
482 s2.intersect(s1);
483 TEST_ASSERT(s2 == s1);
484 s2.setAll();
485 TEST_ASSERT(!(s2 == s1));
486 TEST_ASSERT(s2.countMembers() >= USCRIPT_CODE_LIMIT);
487 s2.intersect(s1);
488 TEST_ASSERT(s2 == s1);
489
490 s2.setAll();
491 s2.reset(USCRIPT_COMMON, status);
492 s2.intersect(s1);
493 TEST_ASSERT(s2.countMembers() == 1);
494
495 s1.resetAll();
496 TEST_ASSERT(s1.isEmpty());
497 s1.set(USCRIPT_LATIN, status);
498 TEST_ASSERT(!s1.isEmpty());
499 s1.setAll();
500 TEST_ASSERT(!s1.isEmpty());
501 TEST_ASSERT_SUCCESS(status);
502
503 s1.resetAll();
504 s1.set(USCRIPT_AFAKA, status);
505 s1.set(USCRIPT_VAI, status);
506 s1.set(USCRIPT_INHERITED, status);
507 int32_t n = -1;
508 for (int32_t i=0; i<4; i++) {
509 n = s1.nextSetBit(n+1);
510 switch (i) {
511 case 0: TEST_ASSERT_EQ(USCRIPT_INHERITED, n); break;
512 case 1: TEST_ASSERT_EQ(USCRIPT_VAI, n); break;
513 case 2: TEST_ASSERT_EQ(USCRIPT_AFAKA, n); break;
514 case 3: TEST_ASSERT_EQ(-1, (int32_t)n); break;
515 default: TEST_ASSERT(FALSE);
516 }
517 }
518 TEST_ASSERT_SUCCESS(status);
519
520 // Script extensions. Depends on data.
521 s1.resetAll();
522 s1.setScriptExtensions(0x67, status);
523 TEST_ASSERT(s1.countMembers() == 1);
524 TEST_ASSERT(s1.test(USCRIPT_LATIN, status));
525 TEST_ASSERT_SUCCESS(status);
526
527 s1.resetAll();
528 s1.setScriptExtensions(0x303C, status);
529 TEST_ASSERT(s1.countMembers() == 3);
530 TEST_ASSERT(s1.test(USCRIPT_HAN, status));
531 TEST_ASSERT(s1.test(USCRIPT_HIRAGANA, status));
532 TEST_ASSERT(s1.test(USCRIPT_KATAKANA, status));
533 TEST_ASSERT_SUCCESS(status);
534
535 // Additional tests
536 ScriptSet bitset12; bitset12.set(USCRIPT_LATIN, status).set(USCRIPT_HANGUL, status);
537 ScriptSet bitset2; bitset2.set(USCRIPT_HANGUL, status);
538 TEST_ASSERT(bitset12.contains(bitset2));
539 TEST_ASSERT(bitset12.contains(bitset12));
540 TEST_ASSERT(!bitset2.contains(bitset12));
541
542 ScriptSet arabSet; arabSet.set(USCRIPT_ARABIC, status);
543 ScriptSet latinSet; latinSet.set(USCRIPT_LATIN, status);
544 UElement arabEl; arabEl.pointer = &arabSet;
545 UElement latinEl; latinEl.pointer = &latinSet;
546 TEST_ASSERT(uhash_compareScriptSet(arabEl, latinEl) < 0);
547 TEST_ASSERT(uhash_compareScriptSet(latinEl, arabEl) > 0);
548
549 UnicodeString scriptString;
550 bitset12.displayScripts(scriptString);
551 TEST_ASSERT(UNICODE_STRING_SIMPLE("Hang Latn") == scriptString);
552 }
553
554
testRestrictionLevel()555 void IntlTestSpoof::testRestrictionLevel() {
556 struct Test {
557 const char *fId;
558 URestrictionLevel fExpectedRestrictionLevel;
559 } tests[] = {
560 {"\\u0061\\u03B3\\u2665", USPOOF_UNRESTRICTIVE},
561 {"a", USPOOF_ASCII},
562 {"\\u03B3", USPOOF_SINGLE_SCRIPT_RESTRICTIVE},
563 {"\\u0061\\u30A2\\u30FC", USPOOF_HIGHLY_RESTRICTIVE},
564 {"\\u0061\\u0904", USPOOF_MODERATELY_RESTRICTIVE},
565 {"\\u0061\\u03B3", USPOOF_MINIMALLY_RESTRICTIVE},
566 {"\\u0061\\u2665", USPOOF_UNRESTRICTIVE},
567 {"\\u0061\\u303C", USPOOF_HIGHLY_RESTRICTIVE},
568 {"\\u0061\\u30FC\\u303C", USPOOF_HIGHLY_RESTRICTIVE},
569 {"\\u0061\\u30FC\\u303C\\u30A2", USPOOF_HIGHLY_RESTRICTIVE},
570 {"\\u30A2\\u0061\\u30FC\\u303C", USPOOF_HIGHLY_RESTRICTIVE},
571 {"\\u0061\\u0031\\u0661", USPOOF_MODERATELY_RESTRICTIVE},
572 {"\\u0061\\u0031\\u0661\\u06F1", USPOOF_MODERATELY_RESTRICTIVE},
573 {"\\u0661\\u30FC\\u303C\\u0061\\u30A2\\u0031\\u0967\\u06F1", USPOOF_MINIMALLY_RESTRICTIVE},
574 {"\\u0061\\u30A2\\u30FC\\u303C\\u0031\\u0967\\u0661\\u06F1", USPOOF_MINIMALLY_RESTRICTIVE}
575 };
576 char msgBuffer[100];
577 URestrictionLevel restrictionLevels[] = { USPOOF_ASCII, USPOOF_SINGLE_SCRIPT_RESTRICTIVE,
578 USPOOF_HIGHLY_RESTRICTIVE, USPOOF_MODERATELY_RESTRICTIVE, USPOOF_MINIMALLY_RESTRICTIVE,
579 USPOOF_UNRESTRICTIVE};
580
581 UErrorCode status = U_ZERO_ERROR;
582 UnicodeSet allowedChars;
583 // Allowed Identifier Characters. In addition to the Recommended Set,
584 // allow u303c, which has an interesting script extension of Hani Hira Kana.
585 allowedChars.addAll(*uspoof_getRecommendedUnicodeSet(&status)).add(0x303C);
586
587 for (int32_t testNum=0; testNum < UPRV_LENGTHOF(tests); testNum++) {
588 status = U_ZERO_ERROR;
589 const Test &test = tests[testNum];
590 UnicodeString testString = UnicodeString(test.fId).unescape();
591 URestrictionLevel expectedLevel = test.fExpectedRestrictionLevel;
592 for (int levelIndex=0; levelIndex<UPRV_LENGTHOF(restrictionLevels); levelIndex++) {
593 status = U_ZERO_ERROR;
594 URestrictionLevel levelSetInSpoofChecker = restrictionLevels[levelIndex];
595 USpoofChecker *sc = uspoof_open(&status);
596 uspoof_setAllowedChars(sc, allowedChars.toUSet(), &status);
597 uspoof_setRestrictionLevel(sc, levelSetInSpoofChecker);
598 uspoof_setChecks(sc, USPOOF_RESTRICTION_LEVEL, &status);
599 int32_t actualValue = uspoof_checkUnicodeString(sc, testString, NULL, &status);
600
601 // we want to fail if the text is (say) MODERATE and the testLevel is ASCII
602 int32_t expectedValue = 0;
603 if (expectedLevel > levelSetInSpoofChecker) {
604 expectedValue |= USPOOF_RESTRICTION_LEVEL;
605 }
606 sprintf(msgBuffer, "testNum = %d, levelIndex = %d, expected = %#x, actual = %#x",
607 testNum, levelIndex, expectedValue, actualValue);
608 TEST_ASSERT_MSG(expectedValue == actualValue, msgBuffer);
609 TEST_ASSERT_SUCCESS(status);
610
611 // Run the same check again, with the Spoof Checker configured to return
612 // the actual restriction level.
613 uspoof_setAllowedChars(sc, allowedChars.toUSet(), &status);
614 uspoof_setRestrictionLevel(sc, levelSetInSpoofChecker);
615 uspoof_setChecks(sc, USPOOF_AUX_INFO | USPOOF_RESTRICTION_LEVEL, &status);
616 int32_t result = uspoof_checkUnicodeString(sc, testString, NULL, &status);
617 TEST_ASSERT_SUCCESS(status);
618 if (U_SUCCESS(status)) {
619 TEST_ASSERT_EQ(expectedLevel, result & USPOOF_RESTRICTION_LEVEL_MASK);
620 TEST_ASSERT_EQ(expectedValue, result & USPOOF_ALL_CHECKS);
621 }
622 uspoof_close(sc);
623 }
624 }
625
626 }
627
testMixedNumbers()628 void IntlTestSpoof::testMixedNumbers() {
629 struct Test {
630 const char *fTestString;
631 const char *fExpectedSet;
632 } tests[] = {
633 {"1", "[0]"},
634 {"\\u0967", "[\\u0966]"},
635 {"1\\u0967", "[0\\u0966]"},
636 {"\\u0661\\u06F1", "[\\u0660\\u06F0]"},
637 {"\\u0061\\u2665", "[]"},
638 {"\\u0061\\u303C", "[]"},
639 {"\\u0061\\u30FC\\u303C", "[]"},
640 {"\\u0061\\u30FC\\u303C\\u30A2", "[]"},
641 {"\\u30A2\\u0061\\u30FC\\u303C", "[]"},
642 {"\\u0061\\u0031\\u0661", "[\\u0030\\u0660]"},
643 {"\\u0061\\u0031\\u0661\\u06F1", "[\\u0030\\u0660\\u06F0]"},
644 {"\\u0661\\u30FC\\u303C\\u0061\\u30A2\\u0031\\u0967\\u06F1", "[\\u0030\\u0660\\u06F0\\u0966]"},
645 {"\\u0061\\u30A2\\u30FC\\u303C\\u0031\\u0967\\u0661\\u06F1", "[\\u0030\\u0660\\u06F0\\u0966]"}
646 };
647 UErrorCode status = U_ZERO_ERROR;
648 for (int32_t testNum=0; testNum < UPRV_LENGTHOF(tests); testNum++) {
649 char msgBuf[100];
650 sprintf(msgBuf, "testNum = %d ", testNum);
651 Test &test = tests[testNum];
652
653 status = U_ZERO_ERROR;
654 UnicodeString testString = UnicodeString(test.fTestString).unescape();
655 UnicodeSet expectedSet(UnicodeString(test.fExpectedSet).unescape(), status);
656
657 status = U_ZERO_ERROR;
658 TEST_SETUP
659 uspoof_setChecks(sc, USPOOF_MIXED_NUMBERS, &status); // only check this
660 uspoof_check2UnicodeString(sc, testString, checkResult, &status);
661 UBool mixedNumberFailure = ((uspoof_getCheckResultChecks(checkResult, &status) & USPOOF_MIXED_NUMBERS) != 0);
662 TEST_ASSERT_MSG((expectedSet.size() > 1) == mixedNumberFailure, msgBuf);
663 const UnicodeSet* actualSet = UnicodeSet::fromUSet(uspoof_getCheckResultNumerics(checkResult, &status));
664 TEST_ASSERT_MSG(expectedSet == *actualSet, msgBuf);
665 TEST_TEARDOWN;
666 }
667 }
668
669 // Bug #12153 - uspoof_setRestrictionLevel() should enable restriction level testing.
670 //
testBug12153()671 void IntlTestSpoof::testBug12153() {
672 UErrorCode status = U_ZERO_ERROR;
673 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
674 if (!assertSuccess("", status, true, __FILE__, __LINE__)) { return; }
675 int32_t checks = uspoof_getChecks(sc.getAlias(), &status);
676 TEST_ASSERT((checks & USPOOF_RESTRICTION_LEVEL) != 0);
677 checks &= ~USPOOF_RESTRICTION_LEVEL;
678 uspoof_setChecks(sc.getAlias(), checks, &status);
679 checks = uspoof_getChecks(sc.getAlias(), &status);
680 TEST_ASSERT((checks & USPOOF_RESTRICTION_LEVEL) == 0);
681
682 uspoof_setRestrictionLevel(sc.getAlias(), USPOOF_MODERATELY_RESTRICTIVE);
683 checks = uspoof_getChecks(sc.getAlias(), &status);
684 TEST_ASSERT((checks & USPOOF_RESTRICTION_LEVEL) != 0);
685 TEST_ASSERT_SUCCESS(status);
686 }
687
688 // uspoof_checkUnicodeString should NOT have an infinite loop.
testBug12825()689 void IntlTestSpoof::testBug12825() {
690 UErrorCode status = U_ZERO_ERROR;
691 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
692 TEST_ASSERT_SUCCESS(status);
693 uspoof_setChecks(sc.getAlias(), USPOOF_ALL_CHECKS | USPOOF_AUX_INFO, &status);
694 TEST_ASSERT_SUCCESS(status);
695 uspoof_checkUnicodeString(sc.getAlias(), UnicodeString("\\u30FB").unescape(), NULL, &status);
696 TEST_ASSERT_SUCCESS(status);
697 }
698
699 // uspoof_getSkeleton should NOT set an ILLEGAL_ARGUMENT_EXCEPTION.
testBug12815()700 void IntlTestSpoof::testBug12815() {
701 UErrorCode status = U_ZERO_ERROR;
702 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
703 TEST_ASSERT_SUCCESS(status);
704 uspoof_setChecks(sc.getAlias(), USPOOF_RESTRICTION_LEVEL, &status);
705 TEST_ASSERT_SUCCESS(status);
706 UnicodeString result;
707 uspoof_getSkeletonUnicodeString(sc.getAlias(), 0, UnicodeString("hello world"), result, &status);
708 TEST_ASSERT_SUCCESS(status);
709 }
710
testBug13314_MixedNumbers()711 void IntlTestSpoof::testBug13314_MixedNumbers() {
712 UErrorCode status = U_ZERO_ERROR;
713 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
714 if (!assertSuccess("", status, true, __FILE__, __LINE__)) { return; }
715 uspoof_setChecks(sc.getAlias(), USPOOF_ALL_CHECKS, &status);
716 TEST_ASSERT_SUCCESS(status);
717 int32_t failedChecks = uspoof_areConfusableUnicodeString(sc.getAlias(), u"列", u"列", &status);
718 TEST_ASSERT_SUCCESS(status);
719 assertEquals("The CJK strings should be confusable", USPOOF_SINGLE_SCRIPT_CONFUSABLE, failedChecks);
720 failedChecks = uspoof_check2UnicodeString(sc.getAlias(), u"3Ȝ", nullptr, &status);
721 TEST_ASSERT_SUCCESS(status);
722 assertEquals("The '33' string does not fail spoof", 0, failedChecks);
723 }
724
testBug13328_MixedCombiningMarks()725 void IntlTestSpoof::testBug13328_MixedCombiningMarks() {
726 UErrorCode status = U_ZERO_ERROR;
727 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
728 if (!assertSuccess("", status, true, __FILE__, __LINE__)) { return; }
729 int32_t failedChecks = uspoof_check2UnicodeString(sc.getAlias(), u"\u0061\u0F84", nullptr, &status);
730 TEST_ASSERT_SUCCESS(status);
731 assertEquals(
732 "The mismatched combining marks string fails spoof",
733 USPOOF_RESTRICTION_LEVEL,
734 failedChecks);
735 }
736
testCombiningDot()737 void IntlTestSpoof::testCombiningDot() {
738 UErrorCode status = U_ZERO_ERROR;
739 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
740 if (!assertSuccess("", status, true, __FILE__, __LINE__)) { return; }
741 uspoof_setChecks(sc.getAlias(), USPOOF_HIDDEN_OVERLAY, &status);
742 TEST_ASSERT_SUCCESS(status);
743
744 static const struct TestCase {
745 bool shouldFail;
746 const char16_t* input;
747 } cases[] = {
748 {false, u"i"},
749 {false, u"j"},
750 {false, u"l"},
751 {true, u"i\u0307"},
752 {true, u"j\u0307"},
753 {true, u"l\u0307"},
754 {true, u"ı\u0307"},
755 {true, u"ȷ\u0307"},
756 {true, u"��\u0307"},
757 {true, u"��\u0307"},
758 {false, u"m\u0307"},
759 {true, u"1\u0307"},
760 {true, u"ij\u0307"},
761 {true, u"i\u0307\u0307"},
762 {true, u"abci\u0307def"},
763 {false, u"i\u0301\u0307"}, // U+0301 has combining class ABOVE (230)
764 {true, u"i\u0320\u0307"}, // U+0320 has combining class BELOW
765 {true, u"i\u0320\u0321\u0307"}, // U+0321 also has combining class BELOW
766 {false, u"i\u0320\u0301\u0307"},
767 {false, u"iz\u0307"},
768 };
769
770 for (auto& cas : cases) {
771 int32_t failedChecks = uspoof_check2(sc.getAlias(), cas.input, -1, nullptr, &status);
772 TEST_ASSERT_SUCCESS(status);
773 int32_t expected = cas.shouldFail ? USPOOF_HIDDEN_OVERLAY : 0;
774 assertEquals(cas.input, expected, failedChecks);
775 }
776 }
777
778 #endif /* !UCONFIG_NO_REGULAR_EXPRESSIONS && !UCONFIG_NO_NORMALIZATION && !UCONFIG_NO_FILE_IO */
779