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(testBidiSkeleton);
99 TESTCASE_AUTO(testAreConfusable);
100 TESTCASE_AUTO(testInvisible);
101 TESTCASE_AUTO(testConfData);
102 TESTCASE_AUTO(testBug8654);
103 TESTCASE_AUTO(testScriptSet);
104 TESTCASE_AUTO(testRestrictionLevel);
105 TESTCASE_AUTO(testMixedNumbers);
106 TESTCASE_AUTO(testBug12153);
107 TESTCASE_AUTO(testBug12825);
108 TESTCASE_AUTO(testBug12815);
109 TESTCASE_AUTO(testBug13314_MixedNumbers);
110 TESTCASE_AUTO(testBug13328_MixedCombiningMarks);
111 TESTCASE_AUTO(testCombiningDot);
112 TESTCASE_AUTO_END;
113 }
114
testSpoofAPI()115 void IntlTestSpoof::testSpoofAPI() {
116
117 TEST_SETUP
118 UnicodeString s("xyz"); // Many latin ranges are whole-script confusable with other scripts.
119 // If this test starts failing, consult confusablesWholeScript.txt
120 int32_t position = 666;
121 int32_t checkResults = uspoof_checkUnicodeString(sc, s, &position, &status);
122 TEST_ASSERT_SUCCESS(status);
123 TEST_ASSERT_EQ(0, checkResults);
124 TEST_ASSERT_EQ(0, position);
125 TEST_TEARDOWN;
126
127 TEST_SETUP
128 UnicodeString s1("cxs");
129 UnicodeString s2 = UnicodeString("\\u0441\\u0445\\u0455").unescape(); // Cyrillic "cxs"
130 int32_t checkResults = uspoof_areConfusableUnicodeString(sc, s1, s2, &status);
131 TEST_ASSERT_SUCCESS(status);
132 TEST_ASSERT_EQ(USPOOF_MIXED_SCRIPT_CONFUSABLE | USPOOF_WHOLE_SCRIPT_CONFUSABLE, checkResults);
133
134 TEST_TEARDOWN;
135
136 TEST_SETUP
137 UnicodeString s("I1l0O");
138 UnicodeString dest;
139 UnicodeString &retStr = uspoof_getSkeletonUnicodeString(sc, USPOOF_ANY_CASE, s, dest, &status);
140 TEST_ASSERT_SUCCESS(status);
141 TEST_ASSERT(UnicodeString("lllOO") == dest);
142 TEST_ASSERT(&dest == &retStr);
143 TEST_TEARDOWN;
144
145 TEST_SETUP
146 // Example from UTS #55, Section 5.1.3 https://www.unicode.org/reports/tr55/#General-Security-Profile
147 // of a minimal pair with a ZWNJ in Persian.
148 const UnicodeString behrooz(u"بهروز");
149 const UnicodeString update(u"بهروز");
150 // These strings differ only by a ZWNJ.
151 TEST_ASSERT(UnicodeString(update).findAndReplace(u"\u200C", u"") == behrooz);
152 int32_t checkResults = uspoof_areConfusableUnicodeString(sc, behrooz, update, &status);
153 TEST_ASSERT_SUCCESS(status);
154 TEST_ASSERT_EQ(USPOOF_SINGLE_SCRIPT_CONFUSABLE, checkResults);
155 TEST_TEARDOWN;
156 }
157
158 #define CHECK_SKELETON(type, input, expected) \
159 UPRV_BLOCK_MACRO_BEGIN { checkSkeleton(sc, type, input, expected, __LINE__); } \
160 UPRV_BLOCK_MACRO_END
161
162 #define CHECK_BIDI_SKELETON(type, input, expected) \
163 UPRV_BLOCK_MACRO_BEGIN { checkBidiSkeleton(sc, type, input, expected, __LINE__); } \
164 UPRV_BLOCK_MACRO_END
165
166
167 // testSkeleton. Spot check a number of confusable skeleton substitutions from the
168 // Unicode data file confusables.txt
169 // Test cases chosen for substitutions of various lengths, and
170 // membership in different mapping tables.
171 // Note: for ICU 55, all tables collapsed to the MA table data.
172 // TODO: for ICU 56 with Unicode 8, revisit this test.
173 //
testSkeleton()174 void IntlTestSpoof::testSkeleton() {
175 const uint32_t ML = 0;
176 const uint32_t SL = USPOOF_SINGLE_SCRIPT_CONFUSABLE;
177 const uint32_t MA = USPOOF_ANY_CASE;
178 const uint32_t SA = USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_ANY_CASE;
179
180 TEST_SETUP
181 CHECK_SKELETON(SL, "nochange", "nochange");
182 CHECK_SKELETON(SA, "nochange", "nochange");
183 CHECK_SKELETON(ML, "nochange", "nochange");
184 CHECK_SKELETON(MA, "nochange", "nochange");
185 CHECK_SKELETON(MA, "love", "love");
186 CHECK_SKELETON(MA, "1ove", "love"); // Digit 1 to letter l
187 CHECK_SKELETON(ML, "OOPS", "OOPS");
188 CHECK_SKELETON(ML, "00PS", "OOPS");
189 CHECK_SKELETON(MA, "OOPS", "OOPS");
190 CHECK_SKELETON(MA, "00PS", "OOPS"); // Digit 0 to letter O in any case mode only
191 CHECK_SKELETON(SL, "\\u059c", "\\u0301");
192 CHECK_SKELETON(SL, "\\u2A74", "\\u003A\\u003A\\u003D");
193 CHECK_SKELETON(SL, "\\u247E", "\\u0028\\u006C\\u006C\\u0029"); // "(ll)"
194 CHECK_SKELETON(SL, "\\uFDFB", "\\u062C\\u0644\\u0020\\u062C\\u0644\\u006c\\u0644\\u006f");
195
196 // This mapping exists in the ML and MA tables, does not exist in SL, SA
197 // 0C83 ; 0983 ; ML
198 // 0C83 ; 0983 ; MA
199 //
200
201 CHECK_SKELETON(SL, "\\u0C83", "\\u0983");
202 CHECK_SKELETON(SA, "\\u0C83", "\\u0983");
203 CHECK_SKELETON(ML, "\\u0C83", "\\u0983");
204 CHECK_SKELETON(MA, "\\u0C83", "\\u0983");
205
206 // 0391 mappings exist only in MA and SA tables.
207 CHECK_SKELETON(MA, "\\u0391", "A");
208 CHECK_SKELETON(SA, "\\u0391", "A");
209 CHECK_SKELETON(ML, "\\u0391", "A");
210 CHECK_SKELETON(SL, "\\u0391", "A");
211
212 // 13CF Mappings in all four tables, different in MA.
213 CHECK_SKELETON(ML, "\\u13CF", "b");
214 CHECK_SKELETON(MA, "\\u13CF", "b");
215 CHECK_SKELETON(SL, "\\u13CF", "b");
216 CHECK_SKELETON(SA, "\\u13CF", "b");
217
218 // 0022 ; 0027 0027 ;
219 // all tables.
220 CHECK_SKELETON(SL, "\\u0022", "\\u0027\\u0027");
221 CHECK_SKELETON(SA, "\\u0022", "\\u0027\\u0027");
222 CHECK_SKELETON(ML, "\\u0022", "\\u0027\\u0027");
223 CHECK_SKELETON(MA, "\\u0022", "\\u0027\\u0027");
224
225 // 017F mappings exist only in MA and SA tables.
226 CHECK_SKELETON(MA, "\\u017F", "f");
227 CHECK_SKELETON(SA, "\\u017F", "f");
228 CHECK_SKELETON(ML, "\\u017F", "f");
229 CHECK_SKELETON(SL, "\\u017F", "f");
230
231 TEST_TEARDOWN;
232 }
233
testBidiSkeleton()234 void IntlTestSpoof::testBidiSkeleton() {
235 TEST_SETUP
236 CHECK_BIDI_SKELETON(u"A1<שׂ", UBIDI_LTR, u"Al<ש\u0307");
237 CHECK_BIDI_SKELETON(u"Αשֺ>1", UBIDI_LTR, u"Al<ש\u0307");
238 CHECK_BIDI_SKELETON(u"A1<שׂ", UBIDI_RTL, u"ש\u0307>Al");
239 CHECK_BIDI_SKELETON(u"Αשֺ>1", UBIDI_RTL, u"l<ש\u0307A");
240 TEST_TEARDOWN;
241 }
242
243
244 //
245 // Run a single confusable skeleton transformation test case.
246 //
checkSkeleton(const USpoofChecker * sc,uint32_t type,const char * input,const char * expected,int32_t lineNum)247 void IntlTestSpoof::checkSkeleton(const USpoofChecker *sc, uint32_t type,
248 const char *input, const char *expected, int32_t lineNum) {
249 UnicodeString uInput = UnicodeString(input).unescape();
250 UnicodeString uExpected = UnicodeString(expected).unescape();
251
252 UErrorCode status = U_ZERO_ERROR;
253 UnicodeString actual;
254 uspoof_getSkeletonUnicodeString(sc, type, uInput, actual, &status);
255 if (U_FAILURE(status)) {
256 errln("File %s, Line %d, Test case from line %d, status is %s", __FILE__, __LINE__, lineNum,
257 u_errorName(status));
258 return;
259 }
260 if (uExpected != actual) {
261 errln("File %s, Line %d, Test case from line %d, Actual and Expected skeletons differ.",
262 __FILE__, __LINE__, lineNum);
263 errln(UnicodeString(" Actual Skeleton: \"") + actual + UnicodeString("\"\n") +
264 UnicodeString(" Expected Skeleton: \"") + uExpected + UnicodeString("\""));
265 }
266 }
267
268 //
269 // Run a single confusable bidiSkeleton transformation test case.
270 //
checkBidiSkeleton(const USpoofChecker * sc,const UnicodeString & input,UBiDiDirection direction,const UnicodeString & expected,int32_t lineNum)271 void IntlTestSpoof::checkBidiSkeleton(const USpoofChecker *sc, const UnicodeString &input,
272 UBiDiDirection direction, const UnicodeString &expected,
273 int32_t lineNum) {
274 UnicodeString uInput = input.unescape();
275 UnicodeString uExpected = expected.unescape();
276
277 UErrorCode status = U_ZERO_ERROR;
278 UnicodeString actual;
279 uspoof_getBidiSkeletonUnicodeString(sc, direction, uInput, actual, &status);
280 if (U_FAILURE(status)) {
281 errln("File %s, Line %d, Test case from line %d, status is %s", __FILE__, __LINE__, lineNum,
282 u_errorName(status));
283 return;
284 }
285 if (uExpected != actual) {
286 errln("File %s, Line %d, Test case from line %d, Actual and Expected skeletons differ.",
287 __FILE__, __LINE__, lineNum);
288 errln(UnicodeString(" Actual Skeleton: \"") + actual + UnicodeString("\"\n") +
289 UnicodeString(" Expected Skeleton: \"") + uExpected + UnicodeString("\""));
290 }
291 }
292
testAreConfusable()293 void IntlTestSpoof::testAreConfusable() {
294 TEST_SETUP
295 UnicodeString s1("A long string that will overflow stack buffers. A long string that will overflow stack buffers. "
296 "A long string that will overflow stack buffers. A long string that will overflow stack buffers. ");
297 UnicodeString s2("A long string that wi11 overflow stack buffers. A long string that will overflow stack buffers. "
298 "A long string that wi11 overflow stack buffers. A long string that will overflow stack buffers. ");
299 int32_t result = uspoof_areConfusableUnicodeString(sc, s1, s2, &status);
300 TEST_ASSERT_SUCCESS(status);
301 TEST_ASSERT_EQ(USPOOF_SINGLE_SCRIPT_CONFUSABLE, result);
302
303 TEST_TEARDOWN;
304 }
305
testAreBidiConfusable()306 void IntlTestSpoof::testAreBidiConfusable() {
307 TEST_SETUP
308 const UnicodeString jHyphen2(u"J-2");
309 // The following string has RLMs around the 2–, flipping it; it uses an
310 // EN DASH instead of the HYPHEN-MINUS above.
311 const UnicodeString j2Dash(u"J\u200F2\u2013\u200F");
312 TEST_ASSERT(j2Dash == u"J2–");
313 int32_t result = uspoof_areBidiConfusableUnicodeString(sc, UBIDI_LTR, jHyphen2, j2Dash, &status);
314 TEST_ASSERT_SUCCESS(status);
315 TEST_ASSERT_EQ(USPOOF_SINGLE_SCRIPT_CONFUSABLE, result);
316
317 TEST_TEARDOWN;
318 }
319
testInvisible()320 void IntlTestSpoof::testInvisible() {
321 TEST_SETUP
322 UnicodeString s = UnicodeString("abcd\\u0301ef").unescape();
323 int32_t position = -42;
324 TEST_ASSERT_EQ(0, uspoof_checkUnicodeString(sc, s, &position, &status));
325 TEST_ASSERT_SUCCESS(status);
326 TEST_ASSERT(0 == position);
327
328 UnicodeString s2 = UnicodeString("abcd\\u0301\\u0302\\u0301ef").unescape();
329 TEST_ASSERT_EQ(USPOOF_INVISIBLE, uspoof_checkUnicodeString(sc, s2, &position, &status));
330 TEST_ASSERT_SUCCESS(status);
331 TEST_ASSERT_EQ(0, position);
332
333 // Two acute accents, one from the composed a with acute accent, \u00e1,
334 // and one separate.
335 position = -42;
336 UnicodeString s3 = UnicodeString("abcd\\u00e1\\u0301xyz").unescape();
337 TEST_ASSERT_EQ(USPOOF_INVISIBLE, uspoof_checkUnicodeString(sc, s3, &position, &status));
338 TEST_ASSERT_SUCCESS(status);
339 TEST_ASSERT_EQ(0, position);
340 TEST_TEARDOWN;
341 }
342
testBug8654()343 void IntlTestSpoof::testBug8654() {
344 TEST_SETUP
345 UnicodeString s = UnicodeString("B\\u00c1\\u0301").unescape();
346 int32_t position = -42;
347 TEST_ASSERT_EQ(USPOOF_INVISIBLE, uspoof_checkUnicodeString(sc, s, &position, &status) & USPOOF_INVISIBLE );
348 TEST_ASSERT_SUCCESS(status);
349 TEST_ASSERT_EQ(0, position);
350 TEST_TEARDOWN;
351 }
352
parseHex(const UnicodeString & in)353 static UnicodeString parseHex(const UnicodeString &in) {
354 // Convert a series of hex numbers in a Unicode String to a string with the
355 // corresponding characters.
356 // The conversion is _really_ annoying. There must be some function to just do it.
357 UnicodeString result;
358 UChar32 cc = 0;
359 for (int32_t i=0; i<in.length(); i++) {
360 char16_t c = in.charAt(i);
361 if (c == 0x20) { // Space
362 if (cc > 0) {
363 result.append(cc);
364 cc = 0;
365 }
366 } else if (c>=0x30 && c<=0x39) {
367 cc = (cc<<4) + (c - 0x30);
368 } else if ((c>=0x41 && c<=0x46) || (c>=0x61 && c<=0x66)) {
369 cc = (cc<<4) + (c & 0x0f)+9;
370 }
371 // else do something with bad input.
372 }
373 if (cc > 0) {
374 result.append(cc);
375 }
376 return result;
377 }
378
379
380 //
381 // Append the hex form of a UChar32 to a UnicodeString.
382 // Used in formatting error messages.
383 // Match the formatting of numbers in confusables.txt
384 // Minimum of 4 digits, no leading zeroes for positions 5 and up.
385 //
appendHexUChar(UnicodeString & dest,UChar32 c)386 static void appendHexUChar(UnicodeString &dest, UChar32 c) {
387 UBool doZeroes = false;
388 for (int bitNum=28; bitNum>=0; bitNum-=4) {
389 if (bitNum <= 12) {
390 doZeroes = true;
391 }
392 int hexDigit = (c>>bitNum) & 0x0f;
393 if (hexDigit != 0 || doZeroes) {
394 doZeroes = true;
395 dest.append(static_cast<char16_t>(hexDigit <= 9 ? hexDigit + 0x30 : hexDigit - 10 + 0x41));
396 }
397 }
398 dest.append(static_cast<char16_t>(0x20));
399 }
400
401 U_DEFINE_LOCAL_OPEN_POINTER(LocalStdioFilePointer, FILE, fclose);
402
403 // testConfData - Check each data item from the Unicode confusables.txt file,
404 // verify that it transforms correctly in a skeleton.
405 //
testConfData()406 void IntlTestSpoof::testConfData() {
407 char buffer[2000];
408 if (getUnidataPath(buffer) == nullptr) {
409 errln("Skipping test spoof/testConfData. Unable to find path to source/data/unidata/.");
410 return;
411 }
412 uprv_strcat(buffer, "confusables.txt");
413
414 LocalStdioFilePointer f(fopen(buffer, "rb"));
415 if (f.isNull()) {
416 errln("Skipping test spoof/testConfData. File confusables.txt not accessible.");
417 return;
418 }
419 fseek(f.getAlias(), 0, SEEK_END);
420 int32_t fileSize = ftell(f.getAlias());
421 LocalArray<char> fileBuf(new char[fileSize]);
422 fseek(f.getAlias(), 0, SEEK_SET);
423 int32_t amt_read = static_cast<int32_t>(fread(fileBuf.getAlias(), 1, fileSize, f.getAlias()));
424 TEST_ASSERT_EQ(amt_read, fileSize);
425 TEST_ASSERT(fileSize>0);
426 if (amt_read != fileSize || fileSize <=0) {
427 return;
428 }
429 UnicodeString confusablesTxt = UnicodeString::fromUTF8(StringPiece(fileBuf.getAlias(), fileSize));
430
431 UErrorCode status = U_ZERO_ERROR;
432 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
433 TEST_ASSERT_SUCCESS(status);
434
435 // Parse lines from the confusables.txt file. Example Line:
436 // FF44 ; 0064 ; SL # ( d -> d ) FULLWIDTH ....
437 // Three fields. The hex fields can contain more than one character,
438 // and each character may be more than 4 digits (for supplemntals)
439 // This regular expression matches lines and splits the fields into capture groups.
440 RegexMatcher parseLine("(?m)^([0-9A-F]{4}[^#;]*?);([^#;]*?);([^#]*)", confusablesTxt, 0, status);
441 TEST_ASSERT_SUCCESS(status);
442 while (parseLine.find()) {
443 UnicodeString from = parseHex(parseLine.group(1, status));
444 if (!Normalizer::isNormalized(from, UNORM_NFD, status)) {
445 // The source character was not NFD.
446 // Skip this case; the first step in obtaining a skeleton is to NFD the input,
447 // so the mapping in this line of confusables.txt will never be applied.
448 continue;
449 }
450
451 if (u_hasBinaryProperty(from.char32At(0), UCHAR_DEFAULT_IGNORABLE_CODE_POINT)) {
452 // The source character is a default ignorable code point.
453 // Skip this case; the second step in obtaining a skeleton is to remove DIs,
454 // so the mapping in this line of confusables.txt will never be applied.
455 continue;
456 }
457
458 UnicodeString rawExpected = parseHex(parseLine.group(2, status));
459 UnicodeString expected;
460 Normalizer::decompose(rawExpected, false /*NFD*/, 0, expected, status);
461 TEST_ASSERT_SUCCESS(status);
462
463 int32_t skeletonType = 0;
464 UnicodeString tableType = parseLine.group(3, status);
465 TEST_ASSERT_SUCCESS(status);
466 if (tableType.indexOf("SL") >= 0) {
467 skeletonType = USPOOF_SINGLE_SCRIPT_CONFUSABLE;
468 } else if (tableType.indexOf("SA") >= 0) {
469 skeletonType = USPOOF_SINGLE_SCRIPT_CONFUSABLE | USPOOF_ANY_CASE;
470 } else if (tableType.indexOf("ML") >= 0) {
471 skeletonType = 0;
472 } else if (tableType.indexOf("MA") >= 0) {
473 skeletonType = USPOOF_ANY_CASE;
474 }
475
476 UnicodeString actual;
477 uspoof_getSkeletonUnicodeString(sc.getAlias(), skeletonType, from, actual, &status);
478 TEST_ASSERT_SUCCESS(status);
479 TEST_ASSERT(actual == expected);
480 if (actual != expected) {
481 errln(parseLine.group(0, status));
482 UnicodeString line = "Actual: ";
483 int i = 0;
484 while (i < actual.length()) {
485 appendHexUChar(line, actual.char32At(i));
486 i = actual.moveIndex32(i, 1);
487 }
488 errln(line);
489 }
490 if (U_FAILURE(status)) {
491 break;
492 }
493 }
494 }
495
496
testScriptSet()497 void IntlTestSpoof::testScriptSet() {
498 // ScriptSet::SCRIPT_LIMIT is hardcoded.
499 // Increase it by multiples of 32 if there are too many script codes.
500 TEST_ASSERT(USCRIPT_CODE_LIMIT <= ScriptSet::SCRIPT_LIMIT);
501 // USCRIPT_CODE_LIMIT should include all script codes,
502 // but theoretically the data may define more.
503 TEST_ASSERT(u_getIntPropertyMaxValue(UCHAR_SCRIPT) < ScriptSet::SCRIPT_LIMIT);
504
505 ScriptSet s1;
506 ScriptSet s2;
507 UErrorCode status = U_ZERO_ERROR;
508
509 TEST_ASSERT(s1 == s2);
510 s1.set(USCRIPT_ARABIC,status);
511 TEST_ASSERT_SUCCESS(status);
512 TEST_ASSERT(!(s1 == s2));
513 TEST_ASSERT(s1.test(USCRIPT_ARABIC, status));
514 TEST_ASSERT(s1.test(USCRIPT_GREEK, status) == false);
515
516 status = U_ZERO_ERROR;
517 s1.reset(USCRIPT_ARABIC, status);
518 TEST_ASSERT(s1 == s2);
519
520 static constexpr UScriptCode LAST_SCRIPT_CODE = static_cast<UScriptCode>(USCRIPT_CODE_LIMIT - 1);
521 status = U_ZERO_ERROR;
522 s1.setAll();
523 TEST_ASSERT(s1.test(USCRIPT_COMMON, status));
524 TEST_ASSERT(s1.test(USCRIPT_ETHIOPIC, status));
525 TEST_ASSERT(s1.test(LAST_SCRIPT_CODE, status));
526 s1.resetAll();
527 TEST_ASSERT(!s1.test(USCRIPT_COMMON, status));
528 TEST_ASSERT(!s1.test(USCRIPT_ETHIOPIC, status));
529 TEST_ASSERT(!s1.test(LAST_SCRIPT_CODE, status));
530
531 status = U_ZERO_ERROR;
532 s1.set(USCRIPT_TAKRI, status);
533 s1.set(USCRIPT_BLISSYMBOLS, status);
534 s2.setAll();
535 TEST_ASSERT(s2.contains(s1));
536 TEST_ASSERT(!s1.contains(s2));
537 TEST_ASSERT(s2.intersects(s1));
538 TEST_ASSERT(s1.intersects(s2));
539 s2.reset(USCRIPT_TAKRI, status);
540 TEST_ASSERT(!s2.contains(s1));
541 TEST_ASSERT(!s1.contains(s2));
542 TEST_ASSERT(s1.intersects(s2));
543 TEST_ASSERT(s2.intersects(s1));
544 TEST_ASSERT_SUCCESS(status);
545
546 status = U_ZERO_ERROR;
547 s1.resetAll();
548 s1.set(USCRIPT_NKO, status);
549 s1.set(USCRIPT_COMMON, status);
550 s2 = s1;
551 TEST_ASSERT(s2 == s1);
552 TEST_ASSERT_EQ(2, s2.countMembers());
553 s2.intersect(s1);
554 TEST_ASSERT(s2 == s1);
555 s2.setAll();
556 TEST_ASSERT(!(s2 == s1));
557 TEST_ASSERT(s2.countMembers() >= USCRIPT_CODE_LIMIT);
558 s2.intersect(s1);
559 TEST_ASSERT(s2 == s1);
560
561 s2.setAll();
562 s2.reset(USCRIPT_COMMON, status);
563 s2.intersect(s1);
564 TEST_ASSERT(s2.countMembers() == 1);
565
566 s1.resetAll();
567 TEST_ASSERT(s1.isEmpty());
568 s1.set(USCRIPT_LATIN, status);
569 TEST_ASSERT(!s1.isEmpty());
570 s1.setAll();
571 TEST_ASSERT(!s1.isEmpty());
572 TEST_ASSERT_SUCCESS(status);
573
574 s1.resetAll();
575 s1.set(USCRIPT_AFAKA, status);
576 s1.set(USCRIPT_VAI, status);
577 s1.set(USCRIPT_INHERITED, status);
578 int32_t n = -1;
579 for (int32_t i=0; i<4; i++) {
580 n = s1.nextSetBit(n+1);
581 switch (i) {
582 case 0: TEST_ASSERT_EQ(USCRIPT_INHERITED, n); break;
583 case 1: TEST_ASSERT_EQ(USCRIPT_VAI, n); break;
584 case 2: TEST_ASSERT_EQ(USCRIPT_AFAKA, n); break;
585 case 3: TEST_ASSERT_EQ(-1, (int32_t)n); break;
586 default: TEST_ASSERT(false);
587 }
588 }
589 TEST_ASSERT_SUCCESS(status);
590
591 // Script extensions. Depends on data.
592 s1.resetAll();
593 s1.setScriptExtensions(0x67, status);
594 TEST_ASSERT(s1.countMembers() == 1);
595 TEST_ASSERT(s1.test(USCRIPT_LATIN, status));
596 TEST_ASSERT_SUCCESS(status);
597
598 s1.resetAll();
599 s1.setScriptExtensions(0x303C, status);
600 TEST_ASSERT(s1.countMembers() == 3);
601 TEST_ASSERT(s1.test(USCRIPT_HAN, status));
602 TEST_ASSERT(s1.test(USCRIPT_HIRAGANA, status));
603 TEST_ASSERT(s1.test(USCRIPT_KATAKANA, status));
604 TEST_ASSERT_SUCCESS(status);
605
606 // Additional tests
607 ScriptSet bitset12; bitset12.set(USCRIPT_LATIN, status).set(USCRIPT_HANGUL, status);
608 ScriptSet bitset2; bitset2.set(USCRIPT_HANGUL, status);
609 TEST_ASSERT(bitset12.contains(bitset2));
610 TEST_ASSERT(bitset12.contains(bitset12));
611 TEST_ASSERT(!bitset2.contains(bitset12));
612
613 ScriptSet arabSet; arabSet.set(USCRIPT_ARABIC, status);
614 ScriptSet latinSet; latinSet.set(USCRIPT_LATIN, status);
615 UElement arabEl; arabEl.pointer = &arabSet;
616 UElement latinEl; latinEl.pointer = &latinSet;
617 TEST_ASSERT(uhash_compareScriptSet(arabEl, latinEl) < 0);
618 TEST_ASSERT(uhash_compareScriptSet(latinEl, arabEl) > 0);
619
620 UnicodeString scriptString;
621 bitset12.displayScripts(scriptString);
622 TEST_ASSERT(UNICODE_STRING_SIMPLE("Hang Latn") == scriptString);
623 }
624
625
testRestrictionLevel()626 void IntlTestSpoof::testRestrictionLevel() {
627 struct Test {
628 const char *fId;
629 URestrictionLevel fExpectedRestrictionLevel;
630 } tests[] = {
631 {"\\u0061\\u03B3\\u2665", USPOOF_UNRESTRICTIVE},
632 {"a", USPOOF_ASCII},
633 {"\\u03B3", USPOOF_SINGLE_SCRIPT_RESTRICTIVE},
634 {"\\u0061\\u30A2\\u30FC", USPOOF_HIGHLY_RESTRICTIVE},
635 {"\\u0061\\u0904", USPOOF_MODERATELY_RESTRICTIVE},
636 {"\\u0061\\u03B3", USPOOF_MINIMALLY_RESTRICTIVE},
637 {"\\u0061\\u2665", USPOOF_UNRESTRICTIVE},
638 {"\\u0061\\u303C", USPOOF_HIGHLY_RESTRICTIVE},
639 {"\\u0061\\u30FC\\u303C", USPOOF_HIGHLY_RESTRICTIVE},
640 {"\\u0061\\u30FC\\u303C\\u30A2", USPOOF_HIGHLY_RESTRICTIVE},
641 {"\\u30A2\\u0061\\u30FC\\u303C", USPOOF_HIGHLY_RESTRICTIVE},
642 {"\\u0061\\u0031\\u0661", USPOOF_MODERATELY_RESTRICTIVE},
643 {"\\u0061\\u0031\\u0661\\u06F1", USPOOF_MODERATELY_RESTRICTIVE},
644 {"\\u0661\\u30FC\\u303C\\u0061\\u30A2\\u0031\\u0967\\u06F1", USPOOF_MINIMALLY_RESTRICTIVE},
645 {"\\u0061\\u30A2\\u30FC\\u303C\\u0031\\u0967\\u0661\\u06F1", USPOOF_MINIMALLY_RESTRICTIVE}
646 };
647 char msgBuffer[100];
648 URestrictionLevel restrictionLevels[] = { USPOOF_ASCII, USPOOF_SINGLE_SCRIPT_RESTRICTIVE,
649 USPOOF_HIGHLY_RESTRICTIVE, USPOOF_MODERATELY_RESTRICTIVE, USPOOF_MINIMALLY_RESTRICTIVE,
650 USPOOF_UNRESTRICTIVE};
651
652 UErrorCode status = U_ZERO_ERROR;
653 UnicodeSet allowedChars;
654 // Allowed Identifier Characters. In addition to the Recommended Set,
655 // allow u303c, which has an interesting script extension of Hani Hira Kana.
656 allowedChars.addAll(*uspoof_getRecommendedUnicodeSet(&status)).add(0x303C);
657
658 for (int32_t testNum=0; testNum < UPRV_LENGTHOF(tests); testNum++) {
659 status = U_ZERO_ERROR;
660 const Test &test = tests[testNum];
661 UnicodeString testString = UnicodeString(test.fId).unescape();
662 URestrictionLevel expectedLevel = test.fExpectedRestrictionLevel;
663 for (int levelIndex=0; levelIndex<UPRV_LENGTHOF(restrictionLevels); levelIndex++) {
664 status = U_ZERO_ERROR;
665 URestrictionLevel levelSetInSpoofChecker = restrictionLevels[levelIndex];
666 USpoofChecker *sc = uspoof_open(&status);
667 uspoof_setAllowedChars(sc, allowedChars.toUSet(), &status);
668 uspoof_setRestrictionLevel(sc, levelSetInSpoofChecker);
669 uspoof_setChecks(sc, USPOOF_RESTRICTION_LEVEL, &status);
670 int32_t actualValue = uspoof_checkUnicodeString(sc, testString, nullptr, &status);
671
672 // we want to fail if the text is (say) MODERATE and the testLevel is ASCII
673 int32_t expectedValue = 0;
674 if (expectedLevel > levelSetInSpoofChecker) {
675 expectedValue |= USPOOF_RESTRICTION_LEVEL;
676 }
677 snprintf(msgBuffer, sizeof(msgBuffer), "testNum = %d, levelIndex = %d, expected = %#x, actual = %#x",
678 testNum, levelIndex, expectedValue, actualValue);
679 TEST_ASSERT_MSG(expectedValue == actualValue, msgBuffer);
680 TEST_ASSERT_SUCCESS(status);
681
682 // Run the same check again, with the Spoof Checker configured to return
683 // the actual restriction level.
684 uspoof_setAllowedChars(sc, allowedChars.toUSet(), &status);
685 uspoof_setRestrictionLevel(sc, levelSetInSpoofChecker);
686 uspoof_setChecks(sc, USPOOF_AUX_INFO | USPOOF_RESTRICTION_LEVEL, &status);
687 int32_t result = uspoof_checkUnicodeString(sc, testString, nullptr, &status);
688 TEST_ASSERT_SUCCESS(status);
689 if (U_SUCCESS(status)) {
690 TEST_ASSERT_EQ(expectedLevel, result & USPOOF_RESTRICTION_LEVEL_MASK);
691 TEST_ASSERT_EQ(expectedValue, result & USPOOF_ALL_CHECKS);
692 }
693 uspoof_close(sc);
694 }
695 }
696
697 }
698
testMixedNumbers()699 void IntlTestSpoof::testMixedNumbers() {
700 struct Test {
701 const char *fTestString;
702 const char *fExpectedSet;
703 } tests[] = {
704 {"1", "[0]"},
705 {"\\u0967", "[\\u0966]"},
706 {"1\\u0967", "[0\\u0966]"},
707 {"\\u0661\\u06F1", "[\\u0660\\u06F0]"},
708 {"\\u0061\\u2665", "[]"},
709 {"\\u0061\\u303C", "[]"},
710 {"\\u0061\\u30FC\\u303C", "[]"},
711 {"\\u0061\\u30FC\\u303C\\u30A2", "[]"},
712 {"\\u30A2\\u0061\\u30FC\\u303C", "[]"},
713 {"\\u0061\\u0031\\u0661", "[\\u0030\\u0660]"},
714 {"\\u0061\\u0031\\u0661\\u06F1", "[\\u0030\\u0660\\u06F0]"},
715 {"\\u0661\\u30FC\\u303C\\u0061\\u30A2\\u0031\\u0967\\u06F1", "[\\u0030\\u0660\\u06F0\\u0966]"},
716 {"\\u0061\\u30A2\\u30FC\\u303C\\u0031\\u0967\\u0661\\u06F1", "[\\u0030\\u0660\\u06F0\\u0966]"}
717 };
718 UErrorCode status = U_ZERO_ERROR;
719 for (int32_t testNum=0; testNum < UPRV_LENGTHOF(tests); testNum++) {
720 char msgBuf[100];
721 snprintf(msgBuf, sizeof(msgBuf), "testNum = %d ", testNum);
722 Test &test = tests[testNum];
723
724 status = U_ZERO_ERROR;
725 UnicodeString testString = UnicodeString(test.fTestString).unescape();
726 UnicodeSet expectedSet(UnicodeString(test.fExpectedSet).unescape(), status);
727
728 status = U_ZERO_ERROR;
729 TEST_SETUP
730 uspoof_setChecks(sc, USPOOF_MIXED_NUMBERS, &status); // only check this
731 uspoof_check2UnicodeString(sc, testString, checkResult, &status);
732 UBool mixedNumberFailure = ((uspoof_getCheckResultChecks(checkResult, &status) & USPOOF_MIXED_NUMBERS) != 0);
733 TEST_ASSERT_MSG((expectedSet.size() > 1) == mixedNumberFailure, msgBuf);
734 const UnicodeSet* actualSet = UnicodeSet::fromUSet(uspoof_getCheckResultNumerics(checkResult, &status));
735 TEST_ASSERT_MSG(expectedSet == *actualSet, msgBuf);
736 TEST_TEARDOWN;
737 }
738 }
739
740 // Bug #12153 - uspoof_setRestrictionLevel() should enable restriction level testing.
741 //
testBug12153()742 void IntlTestSpoof::testBug12153() {
743 UErrorCode status = U_ZERO_ERROR;
744 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
745 if (!assertSuccess("", status, true, __FILE__, __LINE__)) { return; }
746 int32_t checks = uspoof_getChecks(sc.getAlias(), &status);
747 TEST_ASSERT((checks & USPOOF_RESTRICTION_LEVEL) != 0);
748 checks &= ~USPOOF_RESTRICTION_LEVEL;
749 uspoof_setChecks(sc.getAlias(), checks, &status);
750 checks = uspoof_getChecks(sc.getAlias(), &status);
751 TEST_ASSERT((checks & USPOOF_RESTRICTION_LEVEL) == 0);
752
753 uspoof_setRestrictionLevel(sc.getAlias(), USPOOF_MODERATELY_RESTRICTIVE);
754 checks = uspoof_getChecks(sc.getAlias(), &status);
755 TEST_ASSERT((checks & USPOOF_RESTRICTION_LEVEL) != 0);
756 TEST_ASSERT_SUCCESS(status);
757 }
758
759 // uspoof_checkUnicodeString should NOT have an infinite loop.
testBug12825()760 void IntlTestSpoof::testBug12825() {
761 UErrorCode status = U_ZERO_ERROR;
762 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
763 TEST_ASSERT_SUCCESS(status);
764 uspoof_setChecks(sc.getAlias(), USPOOF_ALL_CHECKS | USPOOF_AUX_INFO, &status);
765 TEST_ASSERT_SUCCESS(status);
766 uspoof_checkUnicodeString(sc.getAlias(), UnicodeString("\\u30FB").unescape(), nullptr, &status);
767 TEST_ASSERT_SUCCESS(status);
768 }
769
770 // uspoof_getSkeleton should NOT set an ILLEGAL_ARGUMENT_EXCEPTION.
testBug12815()771 void IntlTestSpoof::testBug12815() {
772 UErrorCode status = U_ZERO_ERROR;
773 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
774 TEST_ASSERT_SUCCESS(status);
775 uspoof_setChecks(sc.getAlias(), USPOOF_RESTRICTION_LEVEL, &status);
776 TEST_ASSERT_SUCCESS(status);
777 UnicodeString result;
778 uspoof_getSkeletonUnicodeString(sc.getAlias(), 0, UnicodeString("hello world"), result, &status);
779 TEST_ASSERT_SUCCESS(status);
780 }
781
testBug13314_MixedNumbers()782 void IntlTestSpoof::testBug13314_MixedNumbers() {
783 UErrorCode status = U_ZERO_ERROR;
784 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
785 if (!assertSuccess("", status, true, __FILE__, __LINE__)) { return; }
786 uspoof_setChecks(sc.getAlias(), USPOOF_ALL_CHECKS, &status);
787 TEST_ASSERT_SUCCESS(status);
788 int32_t failedChecks = uspoof_areConfusableUnicodeString(sc.getAlias(), u"列", u"列", &status);
789 TEST_ASSERT_SUCCESS(status);
790 assertEquals("The CJK strings should be confusable", USPOOF_SINGLE_SCRIPT_CONFUSABLE, failedChecks);
791 failedChecks = uspoof_check2UnicodeString(sc.getAlias(), u"3Ȝ", nullptr, &status);
792 TEST_ASSERT_SUCCESS(status);
793 assertEquals("The '33' string does not fail spoof", 0, failedChecks);
794 }
795
testBug13328_MixedCombiningMarks()796 void IntlTestSpoof::testBug13328_MixedCombiningMarks() {
797 UErrorCode status = U_ZERO_ERROR;
798 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
799 if (!assertSuccess("", status, true, __FILE__, __LINE__)) { return; }
800 int32_t failedChecks = uspoof_check2UnicodeString(sc.getAlias(), u"\u0061\u0F84", nullptr, &status);
801 TEST_ASSERT_SUCCESS(status);
802 assertEquals(
803 "The mismatched combining marks string fails spoof",
804 USPOOF_RESTRICTION_LEVEL,
805 failedChecks);
806 }
807
testCombiningDot()808 void IntlTestSpoof::testCombiningDot() {
809 UErrorCode status = U_ZERO_ERROR;
810 LocalUSpoofCheckerPointer sc(uspoof_open(&status));
811 if (!assertSuccess("", status, true, __FILE__, __LINE__)) { return; }
812 uspoof_setChecks(sc.getAlias(), USPOOF_HIDDEN_OVERLAY, &status);
813 TEST_ASSERT_SUCCESS(status);
814
815 static const struct TestCase {
816 bool shouldFail;
817 const char16_t* input;
818 } cases[] = {
819 {false, u"i"},
820 {false, u"j"},
821 {false, u"l"},
822 {true, u"i\u0307"},
823 {true, u"j\u0307"},
824 {true, u"l\u0307"},
825 {true, u"ı\u0307"},
826 {true, u"ȷ\u0307"},
827 {true, u"��\u0307"},
828 {true, u"��\u0307"},
829 {false, u"m\u0307"},
830 {true, u"1\u0307"},
831 {true, u"ij\u0307"},
832 {true, u"i\u0307\u0307"},
833 {true, u"abci\u0307def"},
834 {false, u"i\u0301\u0307"}, // U+0301 has combining class ABOVE (230)
835 {true, u"i\u0320\u0307"}, // U+0320 has combining class BELOW
836 {true, u"i\u0320\u0321\u0307"}, // U+0321 also has combining class BELOW
837 {false, u"i\u0320\u0301\u0307"},
838 {false, u"iz\u0307"},
839 };
840
841 for (const auto& cas : cases) {
842 int32_t failedChecks = uspoof_check2(sc.getAlias(), cas.input, -1, nullptr, &status);
843 TEST_ASSERT_SUCCESS(status);
844 int32_t expected = cas.shouldFail ? USPOOF_HIDDEN_OVERLAY : 0;
845 assertEquals(cas.input, expected, failedChecks);
846 }
847 }
848
849 #endif /* !UCONFIG_NO_REGULAR_EXPRESSIONS && !UCONFIG_NO_NORMALIZATION && !UCONFIG_NO_FILE_IO */
850