• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 /********************************************************************
2  * COPYRIGHT:
3  * Copyright (c) 1997-2008, International Business Machines Corporation and
4  * others. All Rights Reserved.
5  ********************************************************************/
6 
7 #include "unicode/utypes.h"
8 
9 #if !UCONFIG_NO_NORMALIZATION
10 
11 #include "unicode/uchar.h"
12 #include "unicode/normlzr.h"
13 #include "unicode/uniset.h"
14 #include "unicode/usetiter.h"
15 #include "unicode/schriter.h"
16 #include "cstring.h"
17 #include "unormimp.h"
18 #include "tstnorm.h"
19 
20 #define LENGTHOF(array) ((int32_t)(sizeof(array)/sizeof((array)[0])))
21 #define ARRAY_LENGTH(array) LENGTHOF(array)
22 
23 #define CASE(id,test) case id:                          \
24                           name = #test;                 \
25                           if (exec) {                   \
26                               logln(#test "---");       \
27                               logln((UnicodeString)""); \
28                               test();                   \
29                           }                             \
30                           break
31 
32 static UErrorCode status = U_ZERO_ERROR;
33 
runIndexedTest(int32_t index,UBool exec,const char * & name,char *)34 void BasicNormalizerTest::runIndexedTest(int32_t index, UBool exec,
35                                          const char* &name, char* /*par*/) {
36     switch (index) {
37         CASE(0,TestDecomp);
38         CASE(1,TestCompatDecomp);
39         CASE(2,TestCanonCompose);
40         CASE(3,TestCompatCompose);
41         CASE(4,TestPrevious);
42         CASE(5,TestHangulDecomp);
43         CASE(6,TestHangulCompose);
44         CASE(7,TestTibetan);
45         CASE(8,TestCompositionExclusion);
46         CASE(9,TestZeroIndex);
47         CASE(10,TestVerisign);
48         CASE(11,TestPreviousNext);
49         CASE(12,TestNormalizerAPI);
50         CASE(13,TestConcatenate);
51         CASE(14,FindFoldFCDExceptions);
52         CASE(15,TestCompare);
53         CASE(16,TestSkippable);
54         default: name = ""; break;
55     }
56 }
57 
58 /**
59  * Convert Java-style strings with \u Unicode escapes into UnicodeString objects
60  */
str(const char * input)61 static UnicodeString str(const char *input)
62 {
63     UnicodeString str(input, ""); // Invariant conversion
64     return str.unescape();
65 }
66 
67 
BasicNormalizerTest()68 BasicNormalizerTest::BasicNormalizerTest()
69 {
70   // canonTest
71   // Input                    Decomposed                    Composed
72 
73     canonTests[0][0] = str("cat");  canonTests[0][1] = str("cat"); canonTests[0][2] =  str("cat");
74 
75     canonTests[1][0] = str("\\u00e0ardvark");    canonTests[1][1] = str("a\\u0300ardvark");  canonTests[1][2] = str("\\u00e0ardvark");
76 
77     canonTests[2][0] = str("\\u1e0a"); canonTests[2][1] = str("D\\u0307"); canonTests[2][2] = str("\\u1e0a");                 // D-dot_above
78 
79     canonTests[3][0] = str("D\\u0307");  canonTests[3][1] = str("D\\u0307"); canonTests[3][2] = str("\\u1e0a");            // D dot_above
80 
81     canonTests[4][0] = str("\\u1e0c\\u0307"); canonTests[4][1] = str("D\\u0323\\u0307");  canonTests[4][2] = str("\\u1e0c\\u0307");         // D-dot_below dot_above
82 
83     canonTests[5][0] = str("\\u1e0a\\u0323"); canonTests[5][1] = str("D\\u0323\\u0307");  canonTests[5][2] = str("\\u1e0c\\u0307");        // D-dot_above dot_below
84 
85     canonTests[6][0] = str("D\\u0307\\u0323"); canonTests[6][1] = str("D\\u0323\\u0307");  canonTests[6][2] = str("\\u1e0c\\u0307");         // D dot_below dot_above
86 
87     canonTests[7][0] = str("\\u1e10\\u0307\\u0323");  canonTests[7][1] = str("D\\u0327\\u0323\\u0307"); canonTests[7][2] = str("\\u1e10\\u0323\\u0307");     // D dot_below cedilla dot_above
88 
89     canonTests[8][0] = str("D\\u0307\\u0328\\u0323"); canonTests[8][1] = str("D\\u0328\\u0323\\u0307"); canonTests[8][2] = str("\\u1e0c\\u0328\\u0307");     // D dot_above ogonek dot_below
90 
91     canonTests[9][0] = str("\\u1E14"); canonTests[9][1] = str("E\\u0304\\u0300"); canonTests[9][2] = str("\\u1E14");         // E-macron-grave
92 
93     canonTests[10][0] = str("\\u0112\\u0300"); canonTests[10][1] = str("E\\u0304\\u0300");  canonTests[10][2] = str("\\u1E14");            // E-macron + grave
94 
95     canonTests[11][0] = str("\\u00c8\\u0304"); canonTests[11][1] = str("E\\u0300\\u0304");  canonTests[11][2] = str("\\u00c8\\u0304");         // E-grave + macron
96 
97     canonTests[12][0] = str("\\u212b"); canonTests[12][1] = str("A\\u030a"); canonTests[12][2] = str("\\u00c5");             // angstrom_sign
98 
99     canonTests[13][0] = str("\\u00c5");      canonTests[13][1] = str("A\\u030a");  canonTests[13][2] = str("\\u00c5");            // A-ring
100 
101     canonTests[14][0] = str("\\u00C4ffin");  canonTests[14][1] = str("A\\u0308ffin");  canonTests[14][2] = str("\\u00C4ffin");
102 
103     canonTests[15][0] = str("\\u00C4\\uFB03n"); canonTests[15][1] = str("A\\u0308\\uFB03n"); canonTests[15][2] = str("\\u00C4\\uFB03n");
104 
105     canonTests[16][0] = str("Henry IV"); canonTests[16][1] = str("Henry IV"); canonTests[16][2] = str("Henry IV");
106 
107     canonTests[17][0] = str("Henry \\u2163");  canonTests[17][1] = str("Henry \\u2163");  canonTests[17][2] = str("Henry \\u2163");
108 
109     canonTests[18][0] = str("\\u30AC");  canonTests[18][1] = str("\\u30AB\\u3099");  canonTests[18][2] = str("\\u30AC");              // ga (Katakana)
110 
111     canonTests[19][0] = str("\\u30AB\\u3099"); canonTests[19][1] = str("\\u30AB\\u3099");  canonTests[19][2] = str("\\u30AC");            // ka + ten
112 
113     canonTests[20][0] = str("\\uFF76\\uFF9E"); canonTests[20][1] = str("\\uFF76\\uFF9E");  canonTests[20][2] = str("\\uFF76\\uFF9E");       // hw_ka + hw_ten
114 
115     canonTests[21][0] = str("\\u30AB\\uFF9E"); canonTests[21][1] = str("\\u30AB\\uFF9E");  canonTests[21][2] = str("\\u30AB\\uFF9E");         // ka + hw_ten
116 
117     canonTests[22][0] = str("\\uFF76\\u3099"); canonTests[22][1] = str("\\uFF76\\u3099");  canonTests[22][2] = str("\\uFF76\\u3099");         // hw_ka + ten
118 
119     canonTests[23][0] = str("A\\u0300\\u0316"); canonTests[23][1] = str("A\\u0316\\u0300");  canonTests[23][2] = str("\\u00C0\\u0316");
120 
121     /* compatTest */
122   // Input                        Decomposed                        Composed
123   compatTests[0][0] = str("cat"); compatTests[0][1] = str("cat"); compatTests[0][2] = str("cat") ;
124 
125   compatTests[1][0] = str("\\uFB4f");  compatTests[1][1] = str("\\u05D0\\u05DC"); compatTests[1][2] = str("\\u05D0\\u05DC");  // Alef-Lamed vs. Alef, Lamed
126 
127   compatTests[2][0] = str("\\u00C4ffin"); compatTests[2][1] = str("A\\u0308ffin"); compatTests[2][2] = str("\\u00C4ffin") ;
128 
129   compatTests[3][0] = str("\\u00C4\\uFB03n"); compatTests[3][1] = str("A\\u0308ffin"); compatTests[3][2] = str("\\u00C4ffin") ; // ffi ligature -> f + f + i
130 
131   compatTests[4][0] = str("Henry IV"); compatTests[4][1] = str("Henry IV"); compatTests[4][2] = str("Henry IV") ;
132 
133   compatTests[5][0] = str("Henry \\u2163"); compatTests[5][1] = str("Henry IV");  compatTests[5][2] = str("Henry IV") ;
134 
135   compatTests[6][0] = str("\\u30AC"); compatTests[6][1] = str("\\u30AB\\u3099"); compatTests[6][2] = str("\\u30AC") ; // ga (Katakana)
136 
137   compatTests[7][0] = str("\\u30AB\\u3099"); compatTests[7][1] = str("\\u30AB\\u3099"); compatTests[7][2] = str("\\u30AC") ; // ka + ten
138 
139   compatTests[8][0] = str("\\uFF76\\u3099"); compatTests[8][1] = str("\\u30AB\\u3099"); compatTests[8][2] = str("\\u30AC") ; // hw_ka + ten
140 
141   /* These two are broken in Unicode 2.1.2 but fixed in 2.1.5 and later */
142   compatTests[9][0] = str("\\uFF76\\uFF9E"); compatTests[9][1] = str("\\u30AB\\u3099"); compatTests[9][2] = str("\\u30AC") ; // hw_ka + hw_ten
143 
144   compatTests[10][0] = str("\\u30AB\\uFF9E"); compatTests[10][1] = str("\\u30AB\\u3099"); compatTests[10][2] = str("\\u30AC") ; // ka + hw_ten
145 
146   /* Hangul Canonical */
147   // Input                        Decomposed                        Composed
148   hangulCanon[0][0] = str("\\ud4db"); hangulCanon[0][1] = str("\\u1111\\u1171\\u11b6"); hangulCanon[0][2] = str("\\ud4db") ;
149 
150   hangulCanon[1][0] = str("\\u1111\\u1171\\u11b6"), hangulCanon[1][1] = str("\\u1111\\u1171\\u11b6"),   hangulCanon[1][2] = str("\\ud4db");
151 }
152 
~BasicNormalizerTest()153 BasicNormalizerTest::~BasicNormalizerTest()
154 {
155 }
156 
TestPrevious()157 void BasicNormalizerTest::TestPrevious()
158 {
159   Normalizer* norm = new Normalizer("", UNORM_NFD);
160 
161   logln("testing decomp...");
162   uint32_t i;
163   for (i = 0; i < ARRAY_LENGTH(canonTests); i++) {
164     backAndForth(norm, canonTests[i][0]);
165   }
166 
167   logln("testing compose...");
168   norm->setMode(UNORM_NFC);
169   for (i = 0; i < ARRAY_LENGTH(canonTests); i++) {
170     backAndForth(norm, canonTests[i][0]);
171   }
172 
173   delete norm;
174 }
175 
TestDecomp()176 void BasicNormalizerTest::TestDecomp()
177 {
178   Normalizer* norm = new Normalizer("", UNORM_NFD);
179   iterateTest(norm, canonTests, ARRAY_LENGTH(canonTests), 1);
180   staticTest(UNORM_NFD, 0, canonTests, ARRAY_LENGTH(canonTests), 1);
181   delete norm;
182 }
183 
TestCompatDecomp()184 void BasicNormalizerTest::TestCompatDecomp()
185 {
186   Normalizer* norm = new Normalizer("", UNORM_NFKD);
187   iterateTest(norm, compatTests, ARRAY_LENGTH(compatTests), 1);
188 
189   staticTest(UNORM_NFKD, 0,
190          compatTests, ARRAY_LENGTH(compatTests), 1);
191   delete norm;
192 }
193 
TestCanonCompose()194 void BasicNormalizerTest::TestCanonCompose()
195 {
196   Normalizer* norm = new Normalizer("", UNORM_NFC);
197   iterateTest(norm, canonTests, ARRAY_LENGTH(canonTests), 2);
198 
199   staticTest(UNORM_NFC, 0, canonTests,
200          ARRAY_LENGTH(canonTests), 2);
201   delete norm;
202 }
203 
TestCompatCompose()204 void BasicNormalizerTest::TestCompatCompose()
205 {
206   Normalizer* norm = new Normalizer("", UNORM_NFKC);
207   iterateTest(norm, compatTests, ARRAY_LENGTH(compatTests), 2);
208 
209   staticTest(UNORM_NFKC, 0,
210          compatTests, ARRAY_LENGTH(compatTests), 2);
211   delete norm;
212 }
213 
214 
215 //-------------------------------------------------------------------------------
216 
TestHangulCompose()217 void BasicNormalizerTest::TestHangulCompose()
218 {
219   // Make sure that the static composition methods work
220   logln("Canonical composition...");
221   staticTest(UNORM_NFC, 0,                    hangulCanon,  ARRAY_LENGTH(hangulCanon),  2);
222   logln("Compatibility composition...");
223 
224   // Now try iterative composition....
225   logln("Static composition...");
226   Normalizer* norm = new Normalizer("", UNORM_NFC);
227   iterateTest(norm, hangulCanon, ARRAY_LENGTH(hangulCanon), 2);
228   norm->setMode(UNORM_NFKC);
229 
230   // And finally, make sure you can do it in reverse too
231   logln("Reverse iteration...");
232   norm->setMode(UNORM_NFC);
233   for (uint32_t i = 0; i < ARRAY_LENGTH(hangulCanon); i++) {
234     backAndForth(norm, hangulCanon[i][0]);
235   }
236   delete norm;
237 }
238 
TestHangulDecomp()239 void BasicNormalizerTest::TestHangulDecomp()
240 {
241   // Make sure that the static decomposition methods work
242   logln("Canonical decomposition...");
243   staticTest(UNORM_NFD, 0,                     hangulCanon,  ARRAY_LENGTH(hangulCanon),  1);
244   logln("Compatibility decomposition...");
245 
246   // Now the iterative decomposition methods...
247   logln("Iterative decomposition...");
248   Normalizer* norm = new Normalizer("", UNORM_NFD);
249   iterateTest(norm, hangulCanon, ARRAY_LENGTH(hangulCanon), 1);
250   norm->setMode(UNORM_NFKD);
251 
252   // And finally, make sure you can do it in reverse too
253   logln("Reverse iteration...");
254   norm->setMode(UNORM_NFD);
255   for (uint32_t i = 0; i < ARRAY_LENGTH(hangulCanon); i++) {
256     backAndForth(norm, hangulCanon[i][0]);
257   }
258   delete norm;
259 }
260 
261 /**
262  * The Tibetan vowel sign AA, 0f71, was messed up prior to Unicode version 2.1.9.
263  */
TestTibetan(void)264 void BasicNormalizerTest::TestTibetan(void) {
265     UnicodeString decomp[1][3];
266     decomp[0][0] = str("\\u0f77");
267     decomp[0][1] = str("\\u0f77");
268     decomp[0][2] = str("\\u0fb2\\u0f71\\u0f80");
269 
270     UnicodeString compose[1][3];
271     compose[0][0] = str("\\u0fb2\\u0f71\\u0f80");
272     compose[0][1] = str("\\u0fb2\\u0f71\\u0f80");
273     compose[0][2] = str("\\u0fb2\\u0f71\\u0f80");
274 
275     staticTest(UNORM_NFD,         0, decomp, ARRAY_LENGTH(decomp), 1);
276     staticTest(UNORM_NFKD,  0, decomp, ARRAY_LENGTH(decomp), 2);
277     staticTest(UNORM_NFC,        0, compose, ARRAY_LENGTH(compose), 1);
278     staticTest(UNORM_NFKC, 0, compose, ARRAY_LENGTH(compose), 2);
279 }
280 
281 /**
282  * Make sure characters in the CompositionExclusion.txt list do not get
283  * composed to.
284  */
TestCompositionExclusion(void)285 void BasicNormalizerTest::TestCompositionExclusion(void) {
286     // This list is generated from CompositionExclusion.txt.
287     // Update whenever the normalizer tables are updated.  Note
288     // that we test all characters listed, even those that can be
289     // derived from the Unicode DB and are therefore commented
290     // out.
291     // ### TODO read composition exclusion from source/data/unidata file
292     // and test against that
293     UnicodeString EXCLUDED = str(
294         "\\u0340\\u0341\\u0343\\u0344\\u0374\\u037E\\u0387\\u0958"
295         "\\u0959\\u095A\\u095B\\u095C\\u095D\\u095E\\u095F\\u09DC"
296         "\\u09DD\\u09DF\\u0A33\\u0A36\\u0A59\\u0A5A\\u0A5B\\u0A5E"
297         "\\u0B5C\\u0B5D\\u0F43\\u0F4D\\u0F52\\u0F57\\u0F5C\\u0F69"
298         "\\u0F73\\u0F75\\u0F76\\u0F78\\u0F81\\u0F93\\u0F9D\\u0FA2"
299         "\\u0FA7\\u0FAC\\u0FB9\\u1F71\\u1F73\\u1F75\\u1F77\\u1F79"
300         "\\u1F7B\\u1F7D\\u1FBB\\u1FBE\\u1FC9\\u1FCB\\u1FD3\\u1FDB"
301         "\\u1FE3\\u1FEB\\u1FEE\\u1FEF\\u1FF9\\u1FFB\\u1FFD\\u2000"
302         "\\u2001\\u2126\\u212A\\u212B\\u2329\\u232A\\uF900\\uFA10"
303         "\\uFA12\\uFA15\\uFA20\\uFA22\\uFA25\\uFA26\\uFA2A\\uFB1F"
304         "\\uFB2A\\uFB2B\\uFB2C\\uFB2D\\uFB2E\\uFB2F\\uFB30\\uFB31"
305         "\\uFB32\\uFB33\\uFB34\\uFB35\\uFB36\\uFB38\\uFB39\\uFB3A"
306         "\\uFB3B\\uFB3C\\uFB3E\\uFB40\\uFB41\\uFB43\\uFB44\\uFB46"
307         "\\uFB47\\uFB48\\uFB49\\uFB4A\\uFB4B\\uFB4C\\uFB4D\\uFB4E"
308         );
309     for (int32_t i=0; i<EXCLUDED.length(); ++i) {
310         UnicodeString a(EXCLUDED.charAt(i));
311         UnicodeString b;
312         UnicodeString c;
313         Normalizer::normalize(a, UNORM_NFKD, 0, b, status);
314         Normalizer::normalize(b, UNORM_NFC, 0, c, status);
315         if (c == a) {
316             errln("FAIL: " + hex(a) + " x DECOMP_COMPAT => " +
317                   hex(b) + " x COMPOSE => " +
318                   hex(c));
319         } else if (verbose) {
320             logln("Ok: " + hex(a) + " x DECOMP_COMPAT => " +
321                   hex(b) + " x COMPOSE => " +
322                   hex(c));
323         }
324     }
325 }
326 
327 /**
328  * Test for a problem that showed up just before ICU 1.6 release
329  * having to do with combining characters with an index of zero.
330  * Such characters do not participate in any canonical
331  * decompositions.  However, having an index of zero means that
332  * they all share one typeMask[] entry, that is, they all have to
333  * map to the same canonical class, which is not the case, in
334  * reality.
335  */
TestZeroIndex(void)336 void BasicNormalizerTest::TestZeroIndex(void) {
337     const char* DATA[] = {
338         // Expect col1 x COMPOSE_COMPAT => col2
339         // Expect col2 x DECOMP => col3
340         "A\\u0316\\u0300", "\\u00C0\\u0316", "A\\u0316\\u0300",
341         "A\\u0300\\u0316", "\\u00C0\\u0316", "A\\u0316\\u0300",
342         "A\\u0327\\u0300", "\\u00C0\\u0327", "A\\u0327\\u0300",
343         "c\\u0321\\u0327", "c\\u0321\\u0327", "c\\u0321\\u0327",
344         "c\\u0327\\u0321", "\\u00E7\\u0321", "c\\u0327\\u0321",
345     };
346     int32_t DATA_length = (int32_t)(sizeof(DATA) / sizeof(DATA[0]));
347 
348     for (int32_t i=0; i<DATA_length; i+=3) {
349         UErrorCode status = U_ZERO_ERROR;
350         UnicodeString a(DATA[i], "");
351         a = a.unescape();
352         UnicodeString b;
353         Normalizer::normalize(a, UNORM_NFKC, 0, b, status);
354         UnicodeString exp(DATA[i+1], "");
355         exp = exp.unescape();
356         if (b == exp) {
357             logln((UnicodeString)"Ok: " + hex(a) + " x COMPOSE_COMPAT => " + hex(b));
358         } else {
359             errln((UnicodeString)"FAIL: " + hex(a) + " x COMPOSE_COMPAT => " + hex(b) +
360                   ", expect " + hex(exp));
361         }
362         Normalizer::normalize(b, UNORM_NFD, 0, a, status);
363         exp = UnicodeString(DATA[i+2], "").unescape();
364         if (a == exp) {
365             logln((UnicodeString)"Ok: " + hex(b) + " x DECOMP => " + hex(a));
366         } else {
367             errln((UnicodeString)"FAIL: " + hex(b) + " x DECOMP => " + hex(a) +
368                   ", expect " + hex(exp));
369         }
370     }
371 }
372 
373 /**
374  * Run a few specific cases that are failing for Verisign.
375  */
TestVerisign(void)376 void BasicNormalizerTest::TestVerisign(void) {
377     /*
378       > Their input:
379       > 05B8 05B9 05B1 0591 05C3 05B0 05AC 059F
380       > Their output (supposedly from ICU):
381       > 05B8 05B1 05B9 0591 05C3 05B0 05AC 059F
382       > My output from charlint:
383       > 05B1 05B8 05B9 0591 05C3 05B0 05AC 059F
384 
385       05B8 05B9 05B1 0591 05C3 05B0 05AC 059F => 05B1 05B8 05B9 0591 05C3 05B0
386       05AC 059F
387 
388       U+05B8  18  E HEBREW POINT QAMATS
389       U+05B9  19  F HEBREW POINT HOLAM
390       U+05B1  11 HEBREW POINT HATAF SEGOL
391       U+0591 220 HEBREW ACCENT ETNAHTA
392       U+05C3   0 HEBREW PUNCTUATION SOF PASUQ
393       U+05B0  10 HEBREW POINT SHEVA
394       U+05AC 230 HEBREW ACCENT ILUY
395       U+059F 230 HEBREW ACCENT QARNEY PARA
396 
397       U+05B1  11 HEBREW POINT HATAF SEGOL
398       U+05B8  18 HEBREW POINT QAMATS
399       U+05B9  19 HEBREW POINT HOLAM
400       U+0591 220 HEBREW ACCENT ETNAHTA
401       U+05C3   0 HEBREW PUNCTUATION SOF PASUQ
402       U+05B0  10 HEBREW POINT SHEVA
403       U+05AC 230 HEBREW ACCENT ILUY
404       U+059F 230 HEBREW ACCENT QARNEY PARA
405 
406       Wrong result:
407       U+05B8  18 HEBREW POINT QAMATS
408       U+05B1  11 HEBREW POINT HATAF SEGOL
409       U+05B9  19 HEBREW POINT HOLAM
410       U+0591 220 HEBREW ACCENT ETNAHTA
411       U+05C3   0 HEBREW PUNCTUATION SOF PASUQ
412       U+05B0  10 HEBREW POINT SHEVA
413       U+05AC 230 HEBREW ACCENT ILUY
414       U+059F 230 HEBREW ACCENT QARNEY PARA
415 
416 
417       > Their input:
418       >0592 05B7 05BC 05A5 05B0 05C0 05C4 05AD
419       >Their output (supposedly from ICU):
420       >0592 05B0 05B7 05BC 05A5 05C0 05AD 05C4
421       >My output from charlint:
422       >05B0 05B7 05BC 05A5 0592 05C0 05AD 05C4
423 
424       0592 05B7 05BC 05A5 05B0 05C0 05C4 05AD => 05B0 05B7 05BC 05A5 0592 05C0
425       05AD 05C4
426 
427       U+0592 230 HEBREW ACCENT SEGOL
428       U+05B7  17 HEBREW POINT PATAH
429       U+05BC  21 HEBREW POINT DAGESH OR MAPIQ
430       U+05A5 220 HEBREW ACCENT MERKHA
431       U+05B0  10 HEBREW POINT SHEVA
432       U+05C0   0 HEBREW PUNCTUATION PASEQ
433       U+05C4 230 HEBREW MARK UPPER DOT
434       U+05AD 222 HEBREW ACCENT DEHI
435 
436       U+05B0  10 HEBREW POINT SHEVA
437       U+05B7  17 HEBREW POINT PATAH
438       U+05BC  21 HEBREW POINT DAGESH OR MAPIQ
439       U+05A5 220 HEBREW ACCENT MERKHA
440       U+0592 230 HEBREW ACCENT SEGOL
441       U+05C0   0 HEBREW PUNCTUATION PASEQ
442       U+05AD 222 HEBREW ACCENT DEHI
443       U+05C4 230 HEBREW MARK UPPER DOT
444 
445       Wrong result:
446       U+0592 230 HEBREW ACCENT SEGOL
447       U+05B0  10 HEBREW POINT SHEVA
448       U+05B7  17 HEBREW POINT PATAH
449       U+05BC  21 HEBREW POINT DAGESH OR MAPIQ
450       U+05A5 220 HEBREW ACCENT MERKHA
451       U+05C0   0 HEBREW PUNCTUATION PASEQ
452       U+05AD 222 HEBREW ACCENT DEHI
453       U+05C4 230 HEBREW MARK UPPER DOT
454     */
455     UnicodeString data[2][3];
456     data[0][0] = str("\\u05B8\\u05B9\\u05B1\\u0591\\u05C3\\u05B0\\u05AC\\u059F");
457     data[0][1] = str("\\u05B1\\u05B8\\u05B9\\u0591\\u05C3\\u05B0\\u05AC\\u059F");
458     data[0][2] = str("");
459     data[1][0] = str("\\u0592\\u05B7\\u05BC\\u05A5\\u05B0\\u05C0\\u05C4\\u05AD");
460     data[1][1] = str("\\u05B0\\u05B7\\u05BC\\u05A5\\u0592\\u05C0\\u05AD\\u05C4");
461     data[1][2] = str("");
462 
463     staticTest(UNORM_NFD, 0, data, ARRAY_LENGTH(data), 1);
464     staticTest(UNORM_NFC, 0, data, ARRAY_LENGTH(data), 1);
465 }
466 
467 //------------------------------------------------------------------------
468 // Internal utilities
469 //
470 
hex(UChar ch)471 UnicodeString BasicNormalizerTest::hex(UChar ch) {
472     UnicodeString result;
473     return appendHex(ch, 4, result);
474 }
475 
hex(const UnicodeString & s)476 UnicodeString BasicNormalizerTest::hex(const UnicodeString& s) {
477     UnicodeString result;
478     for (int i = 0; i < s.length(); ++i) {
479         if (i != 0) result += (UChar)0x2c/*,*/;
480         appendHex(s[i], 4, result);
481     }
482     return result;
483 }
484 
485 
insert(UnicodeString & dest,int pos,UChar32 ch)486 inline static void insert(UnicodeString& dest, int pos, UChar32 ch)
487 {
488     dest.replace(pos, 0, ch);
489 }
490 
backAndForth(Normalizer * iter,const UnicodeString & input)491 void BasicNormalizerTest::backAndForth(Normalizer* iter, const UnicodeString& input)
492 {
493     UChar32 ch;
494     iter->setText(input, status);
495 
496     // Run through the iterator forwards and stick it into a StringBuffer
497     UnicodeString forward;
498     for (ch = iter->first(); ch != iter->DONE; ch = iter->next()) {
499         forward += ch;
500     }
501 
502     // Now do it backwards
503     UnicodeString reverse;
504     for (ch = iter->last(); ch != iter->DONE; ch = iter->previous()) {
505         insert(reverse, 0, ch);
506     }
507 
508     if (forward != reverse) {
509         errln("Forward/reverse mismatch for input " + hex(input)
510               + ", forward: " + hex(forward) + ", backward: " + hex(reverse));
511     }
512 }
513 
staticTest(UNormalizationMode mode,int options,UnicodeString tests[][3],int length,int outCol)514 void BasicNormalizerTest::staticTest(UNormalizationMode mode, int options,
515                      UnicodeString tests[][3], int length,
516                      int outCol)
517 {
518     for (int i = 0; i < length; i++)
519     {
520         UnicodeString& input = tests[i][0];
521         UnicodeString& expect = tests[i][outCol];
522 
523         logln("Normalizing '" + input + "' (" + hex(input) + ")" );
524 
525         UnicodeString output;
526         Normalizer::normalize(input, mode, options, output, status);
527 
528         if (output != expect) {
529             errln(UnicodeString("ERROR: case ") + i + " normalized " + hex(input) + "\n"
530                 + "                expected " + hex(expect) + "\n"
531                 + "              static got " + hex(output) );
532         }
533     }
534 }
535 
iterateTest(Normalizer * iter,UnicodeString tests[][3],int length,int outCol)536 void BasicNormalizerTest::iterateTest(Normalizer* iter,
537                                       UnicodeString tests[][3], int length,
538                                       int outCol)
539 {
540     for (int i = 0; i < length; i++)
541     {
542         UnicodeString& input = tests[i][0];
543         UnicodeString& expect = tests[i][outCol];
544 
545         logln("Normalizing '" + input + "' (" + hex(input) + ")" );
546 
547         iter->setText(input, status);
548         assertEqual(input, expect, iter, UnicodeString("ERROR: case ") + i + " ");
549     }
550 }
551 
assertEqual(const UnicodeString & input,const UnicodeString & expected,Normalizer * iter,const UnicodeString & errPrefix)552 void BasicNormalizerTest::assertEqual(const UnicodeString&    input,
553                       const UnicodeString&    expected,
554                       Normalizer*        iter,
555                       const UnicodeString&    errPrefix)
556 {
557     UnicodeString result;
558 
559     for (UChar32 ch = iter->first(); ch != iter->DONE; ch = iter->next()) {
560         result += ch;
561     }
562     if (result != expected) {
563         errln(errPrefix + "normalized " + hex(input) + "\n"
564             + "                expected " + hex(expected) + "\n"
565             + "             iterate got " + hex(result) );
566     }
567 }
568 
569 // helper class for TestPreviousNext()
570 // simple UTF-32 character iterator
571 class UChar32Iterator {
572 public:
UChar32Iterator(const UChar32 * text,int32_t len,int32_t index)573     UChar32Iterator(const UChar32 *text, int32_t len, int32_t index) :
574         s(text), length(len), i(index) {}
575 
current()576     UChar32 current() {
577         if(i<length) {
578             return s[i];
579         } else {
580             return 0xffff;
581         }
582     }
583 
next()584     UChar32 next() {
585         if(i<length) {
586             return s[i++];
587         } else {
588             return 0xffff;
589         }
590     }
591 
previous()592     UChar32 previous() {
593         if(i>0) {
594             return s[--i];
595         } else {
596             return 0xffff;
597         }
598     }
599 
getIndex()600     int32_t getIndex() {
601         return i;
602     }
603 private:
604     const UChar32 *s;
605     int32_t length, i;
606 };
607 
608 void
TestPreviousNext(const UChar * src,int32_t srcLength,const UChar32 * expect,int32_t expectLength,const int32_t * expectIndex,int32_t srcMiddle,int32_t expectMiddle,const char * moves,UNormalizationMode mode,const char * name)609 BasicNormalizerTest::TestPreviousNext(const UChar *src, int32_t srcLength,
610                                       const UChar32 *expect, int32_t expectLength,
611                                       const int32_t *expectIndex, // its length=expectLength+1
612                                       int32_t srcMiddle, int32_t expectMiddle,
613                                       const char *moves,
614                                       UNormalizationMode mode,
615                                       const char *name) {
616     // iterators
617     Normalizer iter(src, srcLength, mode);
618 
619     // test getStaticClassID and getDynamicClassID
620     if(iter.getDynamicClassID() != Normalizer::getStaticClassID()) {
621         errln("getStaticClassID != getDynamicClassID for Normalizer.");
622     }
623 
624     UChar32Iterator iter32(expect, expectLength, expectMiddle);
625 
626     UChar32 c1, c2;
627     char m;
628 
629     // initially set the indexes into the middle of the strings
630     iter.setIndexOnly(srcMiddle);
631 
632     // move around and compare the iteration code points with
633     // the expected ones
634     const char *move=moves;
635     while((m=*move++)!=0) {
636         if(m=='-') {
637             c1=iter.previous();
638             c2=iter32.previous();
639         } else if(m=='0') {
640             c1=iter.current();
641             c2=iter32.current();
642         } else /* m=='+' */ {
643             c1=iter.next();
644             c2=iter32.next();
645         }
646 
647         // compare results
648         if(c1!=c2) {
649             // copy the moves until the current (m) move, and terminate
650             char history[64];
651             uprv_strcpy(history, moves);
652             history[move-moves]=0;
653             errln("error: mismatch in Normalizer iteration (%s) at %s: "
654                   "got c1=U+%04lx != expected c2=U+%04lx\n",
655                   name, history, c1, c2);
656             break;
657         }
658 
659         // compare indexes
660         if(iter.getIndex()!=expectIndex[iter32.getIndex()]) {
661             // copy the moves until the current (m) move, and terminate
662             char history[64];
663             uprv_strcpy(history, moves);
664             history[move-moves]=0;
665             errln("error: index mismatch in Normalizer iteration (%s) at %s: "
666                   "Normalizer index %ld expected %ld\n",
667                   name, history, iter.getIndex(), expectIndex[iter32.getIndex()]);
668             break;
669         }
670     }
671 }
672 
673 void
TestPreviousNext()674 BasicNormalizerTest::TestPreviousNext() {
675     // src and expect strings
676     static const UChar src[]={
677         UTF16_LEAD(0x2f999), UTF16_TRAIL(0x2f999),
678         UTF16_LEAD(0x1d15f), UTF16_TRAIL(0x1d15f),
679         0xc4,
680         0x1ed0
681     };
682     static const UChar32 expect[]={
683         0x831d,
684         0x1d158, 0x1d165,
685         0x41, 0x308,
686         0x4f, 0x302, 0x301
687     };
688 
689     // expected src indexes corresponding to expect indexes
690     static const int32_t expectIndex[]={
691         0,
692         2, 2,
693         4, 4,
694         5, 5, 5,
695         6 // behind last character
696     };
697 
698     // src and expect strings for regression test for j2911
699     static const UChar src_j2911[]={
700         UTF16_LEAD(0x2f999), UTF16_TRAIL(0x2f999),
701         0xdd00, 0xd900, // unpaired surrogates - regression test for j2911
702         0xc4,
703         0x4f, 0x302, 0x301
704     };
705     static const UChar32 expect_j2911[]={
706         0x831d,
707         0xdd00, 0xd900, // unpaired surrogates - regression test for j2911
708         0xc4,
709         0x1ed0
710     };
711 
712     // expected src indexes corresponding to expect indexes
713     static const int32_t expectIndex_j2911[]={
714         0,
715         2, 3,
716         4,
717         5,
718         8 // behind last character
719     };
720 
721     // initial indexes into the src and expect strings
722     // for both sets of test data
723     enum {
724         SRC_MIDDLE=4,
725         EXPECT_MIDDLE=3,
726         SRC_MIDDLE_2=2,
727         EXPECT_MIDDLE_2=1
728     };
729 
730     // movement vector
731     // - for previous(), 0 for current(), + for next()
732     // for both sets of test data
733     static const char *const moves="0+0+0--0-0-+++0--+++++++0--------";
734 
735     TestPreviousNext(src, LENGTHOF(src),
736                      expect, LENGTHOF(expect),
737                      expectIndex,
738                      SRC_MIDDLE, EXPECT_MIDDLE,
739                      moves, UNORM_NFD, "basic");
740 
741     TestPreviousNext(src_j2911, LENGTHOF(src_j2911),
742                      expect_j2911, LENGTHOF(expect_j2911),
743                      expectIndex_j2911,
744                      SRC_MIDDLE, EXPECT_MIDDLE,
745                      moves, UNORM_NFKC, "j2911");
746 
747     // try again from different "middle" indexes
748     TestPreviousNext(src, LENGTHOF(src),
749                      expect, LENGTHOF(expect),
750                      expectIndex,
751                      SRC_MIDDLE_2, EXPECT_MIDDLE_2,
752                      moves, UNORM_NFD, "basic_2");
753 
754     TestPreviousNext(src_j2911, LENGTHOF(src_j2911),
755                      expect_j2911, LENGTHOF(expect_j2911),
756                      expectIndex_j2911,
757                      SRC_MIDDLE_2, EXPECT_MIDDLE_2,
758                      moves, UNORM_NFKC, "j2911_2");
759 }
760 
TestConcatenate()761 void BasicNormalizerTest::TestConcatenate() {
762     static const char *const
763     cases[][4]={
764         /* mode, left, right, result */
765         {
766             "C",
767             "re",
768             "\\u0301sum\\u00e9",
769             "r\\u00e9sum\\u00e9"
770         },
771         {
772             "C",
773             "a\\u1100",
774             "\\u1161bcdefghijk",
775             "a\\uac00bcdefghijk"
776         },
777         /* ### TODO: add more interesting cases */
778         {
779             "D",
780             "\\u0340\\u0341\\u0343\\u0344\\u0374\\u037E\\u0387\\u0958"
781             "\\u0959\\u095A\\u095B\\u095C\\u095D\\u095E\\u095F\\u09DC"
782             "\\u09DD\\u09DF\\u0A33\\u0A36\\u0A59\\u0A5A\\u0A5B\\u0A5E"
783             "\\u0B5C\\u0B5D\\u0F43\\u0F4D\\u0F52\\u0F57\\u0F5C\\u0F69"
784             "\\u0F73\\u0F75\\u0F76\\u0F78\\u0F81\\u0F93\\u0F9D\\u0FA2"
785             "\\u0FA7\\u0FAC\\u0FB9\\u1F71\\u1F73\\u1F75\\u1F77\\u1F79"
786             "\\u1F7B\\u1F7D\\u1FBB\\u1FBE\\u1FC9\\u1FCB\\u1FD3\\u1FDB",
787 
788             "\\u1FE3\\u1FEB\\u1FEE\\u1FEF\\u1FF9\\u1FFB\\u1FFD\\u2000"
789             "\\u2001\\u2126\\u212A\\u212B\\u2329\\u232A\\uF900\\uFA10"
790             "\\uFA12\\uFA15\\uFA20\\uFA22\\uFA25\\uFA26\\uFA2A\\uFB1F"
791             "\\uFB2A\\uFB2B\\uFB2C\\uFB2D\\uFB2E\\uFB2F\\uFB30\\uFB31"
792             "\\uFB32\\uFB33\\uFB34\\uFB35\\uFB36\\uFB38\\uFB39\\uFB3A"
793             "\\uFB3B\\uFB3C\\uFB3E\\uFB40\\uFB41\\uFB43\\uFB44\\uFB46"
794             "\\uFB47\\uFB48\\uFB49\\uFB4A\\uFB4B\\uFB4C\\uFB4D\\uFB4E",
795 
796             "\\u0340\\u0341\\u0343\\u0344\\u0374\\u037E\\u0387\\u0958"
797             "\\u0959\\u095A\\u095B\\u095C\\u095D\\u095E\\u095F\\u09DC"
798             "\\u09DD\\u09DF\\u0A33\\u0A36\\u0A59\\u0A5A\\u0A5B\\u0A5E"
799             "\\u0B5C\\u0B5D\\u0F43\\u0F4D\\u0F52\\u0F57\\u0F5C\\u0F69"
800             "\\u0F73\\u0F75\\u0F76\\u0F78\\u0F81\\u0F93\\u0F9D\\u0FA2"
801             "\\u0FA7\\u0FAC\\u0FB9\\u1F71\\u1F73\\u1F75\\u1F77\\u1F79"
802             "\\u1F7B\\u1F7D\\u1FBB\\u1FBE\\u1FC9\\u1FCB\\u1FD3\\u0399"
803             "\\u0301\\u03C5\\u0308\\u0301\\u1FEB\\u1FEE\\u1FEF\\u1FF9"
804             "\\u1FFB\\u1FFD\\u2000\\u2001\\u2126\\u212A\\u212B\\u2329"
805             "\\u232A\\uF900\\uFA10\\uFA12\\uFA15\\uFA20\\uFA22\\uFA25"
806             "\\uFA26\\uFA2A\\uFB1F\\uFB2A\\uFB2B\\uFB2C\\uFB2D\\uFB2E"
807             "\\uFB2F\\uFB30\\uFB31\\uFB32\\uFB33\\uFB34\\uFB35\\uFB36"
808             "\\uFB38\\uFB39\\uFB3A\\uFB3B\\uFB3C\\uFB3E\\uFB40\\uFB41"
809             "\\uFB43\\uFB44\\uFB46\\uFB47\\uFB48\\uFB49\\uFB4A\\uFB4B"
810             "\\uFB4C\\uFB4D\\uFB4E"
811         }
812     };
813 
814     UnicodeString left, right, expect, result, r;
815     UErrorCode errorCode;
816     UNormalizationMode mode;
817     int32_t i;
818 
819     /* test concatenation */
820     for(i=0; i<(int32_t)(sizeof(cases)/sizeof(cases[0])); ++i) {
821         switch(*cases[i][0]) {
822         case 'C': mode=UNORM_NFC; break;
823         case 'D': mode=UNORM_NFD; break;
824         case 'c': mode=UNORM_NFKC; break;
825         case 'd': mode=UNORM_NFKD; break;
826         default: mode=UNORM_NONE; break;
827         }
828 
829         left=UnicodeString(cases[i][1], "").unescape();
830         right=UnicodeString(cases[i][2], "").unescape();
831         expect=UnicodeString(cases[i][3], "").unescape();
832 
833         //result=r=UnicodeString();
834         errorCode=U_ZERO_ERROR;
835 
836         r=Normalizer::concatenate(left, right, result, mode, 0, errorCode);
837         if(U_FAILURE(errorCode) || /*result!=r ||*/ result!=expect) {
838             errln("error in Normalizer::concatenate(), cases[] fails with "+
839                 UnicodeString(u_errorName(errorCode))+", result==expect: expected: "+
840                 hex(expect)+" =========> got: " + hex(result));
841         }
842     }
843 
844     /* test error cases */
845 
846     /* left.getBuffer()==result.getBuffer() */
847     result=r=expect=UnicodeString("zz", "");
848     errorCode=U_UNEXPECTED_TOKEN;
849     r=Normalizer::concatenate(left, right, result, mode, 0, errorCode);
850     if(errorCode!=U_UNEXPECTED_TOKEN || result!=r || !result.isBogus()) {
851         errln("error in Normalizer::concatenate(), violates UErrorCode protocol");
852     }
853 
854     left.setToBogus();
855     errorCode=U_ZERO_ERROR;
856     r=Normalizer::concatenate(left, right, result, mode, 0, errorCode);
857     if(errorCode!=U_ILLEGAL_ARGUMENT_ERROR || result!=r || !result.isBogus()) {
858         errln("error in Normalizer::concatenate(), does not detect left.isBogus()");
859     }
860 }
861 
862 // reference implementation of Normalizer::compare
863 static int32_t
ref_norm_compare(const UnicodeString & s1,const UnicodeString & s2,uint32_t options,UErrorCode & errorCode)864 ref_norm_compare(const UnicodeString &s1, const UnicodeString &s2, uint32_t options, UErrorCode &errorCode) {
865     UnicodeString r1, r2, t1, t2;
866     int32_t normOptions=(int32_t)(options>>UNORM_COMPARE_NORM_OPTIONS_SHIFT);
867 
868     if(options&U_COMPARE_IGNORE_CASE) {
869         Normalizer::decompose(s1, FALSE, normOptions, r1, errorCode);
870         Normalizer::decompose(s2, FALSE, normOptions, r2, errorCode);
871 
872         r1.foldCase(options);
873         r2.foldCase(options);
874     } else {
875         r1=s1;
876         r2=s2;
877     }
878 
879     Normalizer::decompose(r1, FALSE, normOptions, t1, errorCode);
880     Normalizer::decompose(r2, FALSE, normOptions, t2, errorCode);
881 
882     if(options&U_COMPARE_CODE_POINT_ORDER) {
883         return t1.compareCodePointOrder(t2);
884     } else {
885         return t1.compare(t2);
886     }
887 }
888 
889 // test wrapper for Normalizer::compare, sets UNORM_INPUT_IS_FCD appropriately
890 static int32_t
_norm_compare(const UnicodeString & s1,const UnicodeString & s2,uint32_t options,UErrorCode & errorCode)891 _norm_compare(const UnicodeString &s1, const UnicodeString &s2, uint32_t options, UErrorCode &errorCode) {
892     int32_t normOptions=(int32_t)(options>>UNORM_COMPARE_NORM_OPTIONS_SHIFT);
893 
894     if( UNORM_YES==Normalizer::quickCheck(s1, UNORM_FCD, normOptions, errorCode) &&
895         UNORM_YES==Normalizer::quickCheck(s2, UNORM_FCD, normOptions, errorCode)) {
896         options|=UNORM_INPUT_IS_FCD;
897     }
898 
899     return Normalizer::compare(s1, s2, options, errorCode);
900 }
901 
902 // reference implementation of UnicodeString::caseCompare
903 static int32_t
ref_case_compare(const UnicodeString & s1,const UnicodeString & s2,uint32_t options)904 ref_case_compare(const UnicodeString &s1, const UnicodeString &s2, uint32_t options) {
905     UnicodeString t1, t2;
906 
907     t1=s1;
908     t2=s2;
909 
910     t1.foldCase(options);
911     t2.foldCase(options);
912 
913     if(options&U_COMPARE_CODE_POINT_ORDER) {
914         return t1.compareCodePointOrder(t2);
915     } else {
916         return t1.compare(t2);
917     }
918 }
919 
920 // reduce an integer to -1/0/1
921 static inline int32_t
_sign(int32_t value)922 _sign(int32_t value) {
923     if(value==0) {
924         return 0;
925     } else {
926         return (value>>31)|1;
927     }
928 }
929 
930 static const char *
_signString(int32_t value)931 _signString(int32_t value) {
932     if(value<0) {
933         return "<0";
934     } else if(value==0) {
935         return "=0";
936     } else /* value>0 */ {
937         return ">0";
938     }
939 }
940 
941 void
TestCompare()942 BasicNormalizerTest::TestCompare() {
943     // test Normalizer::compare and unorm_compare (thinly wrapped by the former)
944     // by comparing it with its semantic equivalent
945     // since we trust the pieces, this is sufficient
946 
947     // test each string with itself and each other
948     // each time with all options
949     static const char *const
950     strings[]={
951         // some cases from NormalizationTest.txt
952         // 0..3
953         "D\\u031B\\u0307\\u0323",
954         "\\u1E0C\\u031B\\u0307",
955         "D\\u031B\\u0323\\u0307",
956         "d\\u031B\\u0323\\u0307",
957 
958         // 4..6
959         "\\u00E4",
960         "a\\u0308",
961         "A\\u0308",
962 
963         // Angstrom sign = A ring
964         // 7..10
965         "\\u212B",
966         "\\u00C5",
967         "A\\u030A",
968         "a\\u030A",
969 
970         // 11.14
971         "a\\u059A\\u0316\\u302A\\u032Fb",
972         "a\\u302A\\u0316\\u032F\\u059Ab",
973         "a\\u302A\\u0316\\u032F\\u059Ab",
974         "A\\u059A\\u0316\\u302A\\u032Fb",
975 
976         // from ICU case folding tests
977         // 15..20
978         "A\\u00df\\u00b5\\ufb03\\U0001040c\\u0131",
979         "ass\\u03bcffi\\U00010434i",
980         "\\u0061\\u0042\\u0131\\u03a3\\u00df\\ufb03\\ud93f\\udfff",
981         "\\u0041\\u0062\\u0069\\u03c3\\u0073\\u0053\\u0046\\u0066\\u0049\\ud93f\\udfff",
982         "\\u0041\\u0062\\u0131\\u03c3\\u0053\\u0073\\u0066\\u0046\\u0069\\ud93f\\udfff",
983         "\\u0041\\u0062\\u0069\\u03c3\\u0073\\u0053\\u0046\\u0066\\u0049\\ud93f\\udffd",
984 
985         //     U+d800 U+10001   see implementation comment in unorm_cmpEquivFold
986         // vs. U+10000          at bottom - code point order
987         // 21..22
988         "\\ud800\\ud800\\udc01",
989         "\\ud800\\udc00",
990 
991         // other code point order tests from ustrtest.cpp
992         // 23..31
993         "\\u20ac\\ud801",
994         "\\u20ac\\ud800\\udc00",
995         "\\ud800",
996         "\\ud800\\uff61",
997         "\\udfff",
998         "\\uff61\\udfff",
999         "\\uff61\\ud800\\udc02",
1000         "\\ud800\\udc02",
1001         "\\ud84d\\udc56",
1002 
1003         // long strings, see cnormtst.c/TestNormCoverage()
1004         // equivalent if case-insensitive
1005         // 32..33
1006         "\\uAD8B\\uAD8B\\uAD8B\\uAD8B"
1007         "\\U0001d15e\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1008         "\\U0001d15e\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1009         "\\U0001d15e\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1010         "\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1011         "\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1012         "aaaaaaaaaaaaaaaaaazzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"
1013         "bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb"
1014         "ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc"
1015         "ddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddd"
1016         "\\uAD8B\\uAD8B\\uAD8B\\uAD8B"
1017         "d\\u031B\\u0307\\u0323",
1018 
1019         "\\u1100\\u116f\\u11aa\\uAD8B\\uAD8B\\u1100\\u116f\\u11aa"
1020         "\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1021         "\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1022         "\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1023         "\\U0001d15e\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1024         "\\U0001d15e\\U0001d157\\U0001d165\\U0001d15e\\U0001d15e\\U0001d15e\\U0001d15e"
1025         "aaaaaaaaaaAAAAAAAAZZZZZZZZZZZZZZZZzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"
1026         "bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb"
1027         "ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc"
1028         "ddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddd"
1029         "\\u1100\\u116f\\u11aa\\uAD8B\\uAD8B\\u1100\\u116f\\u11aa"
1030         "\\u1E0C\\u031B\\u0307",
1031 
1032         // some strings that may make a difference whether the compare function
1033         // case-folds or decomposes first
1034         // 34..41
1035         "\\u0360\\u0345\\u0334",
1036         "\\u0360\\u03b9\\u0334",
1037 
1038         "\\u0360\\u1f80\\u0334",
1039         "\\u0360\\u03b1\\u0313\\u03b9\\u0334",
1040 
1041         "\\u0360\\u1ffc\\u0334",
1042         "\\u0360\\u03c9\\u03b9\\u0334",
1043 
1044         "a\\u0360\\u0345\\u0360\\u0345b",
1045         "a\\u0345\\u0360\\u0345\\u0360b",
1046 
1047         // interesting cases for canonical caseless match with turkic i handling
1048         // 42..43
1049         "\\u00cc",
1050         "\\u0069\\u0300",
1051 
1052         // strings with post-Unicode 3.2 normalization or normalization corrections
1053         // 44..45
1054         "\\u00e4\\u193b\\U0002f868",
1055         "\\u0061\\u193b\\u0308\\u36fc",
1056 
1057         // empty string
1058         // 46
1059         ""
1060     };
1061 
1062     UnicodeString s[100]; // at least as many items as in strings[] !
1063 
1064     // all combinations of options
1065     // UNORM_INPUT_IS_FCD is set automatically if both input strings fulfill FCD conditions
1066     // set UNORM_UNICODE_3_2 in one additional combination
1067     static const struct {
1068         uint32_t options;
1069         const char *name;
1070     } opt[]={
1071         { 0, "default" },
1072         { U_COMPARE_CODE_POINT_ORDER, "c.p. order" },
1073         { U_COMPARE_IGNORE_CASE, "ignore case" },
1074         { U_COMPARE_CODE_POINT_ORDER|U_COMPARE_IGNORE_CASE, "c.p. order & ignore case" },
1075         { U_COMPARE_IGNORE_CASE|U_FOLD_CASE_EXCLUDE_SPECIAL_I, "ignore case & special i" },
1076         { U_COMPARE_CODE_POINT_ORDER|U_COMPARE_IGNORE_CASE|U_FOLD_CASE_EXCLUDE_SPECIAL_I, "c.p. order & ignore case & special i" },
1077         { UNORM_UNICODE_3_2<<UNORM_COMPARE_NORM_OPTIONS_SHIFT, "Unicode 3.2" }
1078     };
1079 
1080     int32_t i, j, k, count=LENGTHOF(strings);
1081     int32_t result, refResult;
1082 
1083     UErrorCode errorCode;
1084 
1085     // create the UnicodeStrings
1086     for(i=0; i<count; ++i) {
1087         s[i]=UnicodeString(strings[i], "").unescape();
1088     }
1089 
1090     // test them each with each other
1091     for(i=0; i<count; ++i) {
1092         for(j=i; j<count; ++j) {
1093             for(k=0; k<LENGTHOF(opt); ++k) {
1094                 // test Normalizer::compare
1095                 errorCode=U_ZERO_ERROR;
1096                 result=_norm_compare(s[i], s[j], opt[k].options, errorCode);
1097                 refResult=ref_norm_compare(s[i], s[j], opt[k].options, errorCode);
1098                 if(_sign(result)!=_sign(refResult)) {
1099                     errln("Normalizer::compare(%d, %d, %s)%s should be %s %s",
1100                         i, j, opt[k].name, _signString(result), _signString(refResult),
1101                         U_SUCCESS(errorCode) ? "" : u_errorName(errorCode));
1102                 }
1103 
1104                 // test UnicodeString::caseCompare - same internal implementation function
1105                 if(opt[k].options&U_COMPARE_IGNORE_CASE) {
1106                     errorCode=U_ZERO_ERROR;
1107                     result=s[i].caseCompare(s[j], opt[k].options);
1108                     refResult=ref_case_compare(s[i], s[j], opt[k].options);
1109                     if(_sign(result)!=_sign(refResult)) {
1110                         errln("UniStr::caseCompare(%d, %d, %s)%s should be %s %s",
1111                             i, j, opt[k].name, _signString(result), _signString(refResult),
1112                             U_SUCCESS(errorCode) ? "" : u_errorName(errorCode));
1113                     }
1114                 }
1115             }
1116         }
1117     }
1118 
1119     // test cases with i and I to make sure Turkic works
1120     static const UChar iI[]={ 0x49, 0x69, 0x130, 0x131 };
1121     USerializedSet sset;
1122     UnicodeSet set;
1123 
1124     UnicodeString s1, s2;
1125     UChar32 start, end;
1126 
1127     // collect all sets into one for contiguous output
1128     for(i=0; i<LENGTHOF(iI); ++i) {
1129         if(unorm_getCanonStartSet(iI[i], &sset)) {
1130             count=uset_getSerializedRangeCount(&sset);
1131             for(j=0; j<count; ++j) {
1132                 uset_getSerializedRange(&sset, j, &start, &end);
1133                 set.add(start, end);
1134             }
1135         }
1136     }
1137 
1138     // test all of these precomposed characters
1139     UnicodeSetIterator it(set);
1140     while(it.nextRange() && !it.isString()) {
1141         start=it.getCodepoint();
1142         end=it.getCodepointEnd();
1143         while(start<=end) {
1144             s1.setTo(start);
1145             errorCode=U_ZERO_ERROR;
1146             Normalizer::decompose(s1, FALSE, 0, s2, errorCode);
1147             if(U_FAILURE(errorCode)) {
1148                 errln("Normalizer::decompose(U+%04x) failed: %s", start, u_errorName(errorCode));
1149                 return;
1150             }
1151 
1152             for(k=0; k<LENGTHOF(opt); ++k) {
1153                 // test Normalizer::compare
1154                 errorCode=U_ZERO_ERROR;
1155                 result=_norm_compare(s1, s2, opt[k].options, errorCode);
1156                 refResult=ref_norm_compare(s1, s2, opt[k].options, errorCode);
1157                 if(_sign(result)!=_sign(refResult)) {
1158                     errln("Normalizer::compare(U+%04x with its NFD, %s)%s should be %s %s",
1159                         start, opt[k].name, _signString(result), _signString(refResult),
1160                         U_SUCCESS(errorCode) ? "" : u_errorName(errorCode));
1161                 }
1162 
1163                 // test UnicodeString::caseCompare - same internal implementation function
1164                 if(opt[k].options&U_COMPARE_IGNORE_CASE) {
1165                     errorCode=U_ZERO_ERROR;
1166                     result=s1.caseCompare(s2, opt[k].options);
1167                     refResult=ref_case_compare(s1, s2, opt[k].options);
1168                     if(_sign(result)!=_sign(refResult)) {
1169                         errln("UniStr::caseCompare(U+%04x with its NFD, %s)%s should be %s %s",
1170                             start, opt[k].name, _signString(result), _signString(refResult),
1171                             U_SUCCESS(errorCode) ? "" : u_errorName(errorCode));
1172                     }
1173                 }
1174             }
1175 
1176             ++start;
1177         }
1178     }
1179 }
1180 
1181 // verify that case-folding does not un-FCD strings
1182 int32_t
countFoldFCDExceptions(uint32_t foldingOptions)1183 BasicNormalizerTest::countFoldFCDExceptions(uint32_t foldingOptions) {
1184     UnicodeString s, fold, d;
1185     UChar32 c;
1186     int32_t count;
1187     uint8_t cc, trailCC, foldCC, foldTrailCC;
1188     UNormalizationCheckResult qcResult;
1189     int8_t category;
1190     UBool isNFD;
1191     UErrorCode errorCode;
1192 
1193     logln("Test if case folding may un-FCD a string (folding options %04lx)", foldingOptions);
1194 
1195     count=0;
1196     for(c=0; c<=0x10ffff; ++c) {
1197         errorCode = U_ZERO_ERROR;
1198         category=u_charType(c);
1199         if(category==U_UNASSIGNED) {
1200             continue; // skip unassigned code points
1201         }
1202         if(c==0xac00) {
1203             c=0xd7a3; // skip Hangul - no case folding there
1204             continue;
1205         }
1206         // skip Han blocks - no case folding there either
1207         if(c==0x3400) {
1208             c=0x4db5;
1209             continue;
1210         }
1211         if(c==0x4e00) {
1212             c=0x9fa5;
1213             continue;
1214         }
1215         if(c==0x20000) {
1216             c=0x2a6d6;
1217             continue;
1218         }
1219 
1220         s.setTo(c);
1221 
1222         // get leading and trailing cc for c
1223         Normalizer::decompose(s, FALSE, 0, d, errorCode);
1224         isNFD= s==d;
1225         cc=u_getCombiningClass(d.char32At(0));
1226         trailCC=u_getCombiningClass(d.char32At(d.length()-1));
1227 
1228         // get leading and trailing cc for the case-folding of c
1229         s.foldCase(foldingOptions);
1230         Normalizer::decompose(s, FALSE, 0, d, errorCode);
1231         foldCC=u_getCombiningClass(d.char32At(0));
1232         foldTrailCC=u_getCombiningClass(d.char32At(d.length()-1));
1233 
1234         qcResult=Normalizer::quickCheck(s, UNORM_FCD, errorCode);
1235 
1236         if (U_FAILURE(errorCode)) {
1237             ++count;
1238             errln("U+%04lx: Failed with error %s", u_errorName(errorCode));
1239         }
1240 
1241         // bad:
1242         // - character maps to empty string: adjacent characters may then need reordering
1243         // - folding has different leading/trailing cc's, and they don't become just 0
1244         // - folding itself is not FCD
1245         if( qcResult!=UNORM_YES ||
1246             s.isEmpty() ||
1247             (cc!=foldCC && foldCC!=0) || (trailCC!=foldTrailCC && foldTrailCC!=0)
1248         ) {
1249             ++count;
1250             errln("U+%04lx: case-folding may un-FCD a string (folding options %04lx)", c, foldingOptions);
1251             errln("  cc %02x trailCC %02x    foldCC(U+%04lx) %02x foldTrailCC(U+%04lx) %02x   quickCheck(folded)=%d", cc, trailCC, d.char32At(0), foldCC, d.char32At(d.length()-1), foldTrailCC, qcResult);
1252             continue;
1253         }
1254 
1255         // also bad:
1256         // if a code point is in NFD but its case folding is not, then
1257         // unorm_compare will also fail
1258         if(isNFD && UNORM_YES!=Normalizer::quickCheck(s, UNORM_NFD, errorCode)) {
1259             ++count;
1260             errln("U+%04lx: case-folding un-NFDs this character (folding options %04lx)", c, foldingOptions);
1261         }
1262     }
1263 
1264     logln("There are %ld code points for which case-folding may un-FCD a string (folding options %04lx)", count, foldingOptions);
1265     return count;
1266 }
1267 
1268 void
FindFoldFCDExceptions()1269 BasicNormalizerTest::FindFoldFCDExceptions() {
1270     int32_t count;
1271 
1272     count=countFoldFCDExceptions(0);
1273     count+=countFoldFCDExceptions(U_FOLD_CASE_EXCLUDE_SPECIAL_I);
1274     if(count>0) {
1275         /*
1276          * If case-folding un-FCDs any strings, then unorm_compare() must be
1277          * re-implemented.
1278          * It currently assumes that one can check for FCD then case-fold
1279          * and then still have FCD strings for raw decomposition without reordering.
1280          */
1281         errln("error: There are %ld code points for which case-folding may un-FCD a string for all folding options.\n"
1282               "See comment in BasicNormalizerTest::FindFoldFCDExceptions()!", count);
1283     }
1284 }
1285 
1286 /*
1287  * Hardcoded "NF* Skippable" sets, generated from
1288  * Mark Davis' com.ibm.text.UCD.NFSkippable (see ICU4J CVS, module unicodetools).
1289  * Run com.ibm.text.UCD.Main with the option NFSkippable.
1290  *
1291  * Must be updated for each Unicode version.
1292  */
1293 static void
initExpectedSkippables(UnicodeSet skipSets[UNORM_MODE_COUNT])1294 initExpectedSkippables(UnicodeSet skipSets[UNORM_MODE_COUNT]) {
1295     UErrorCode errorCode=U_ZERO_ERROR;
1296 
1297     skipSets[UNORM_NFD].applyPattern(UnicodeString(
1298         "[^\\u00C0-\\u00C5\\u00C7-\\u00CF\\u00D1-\\u00D6\\u00D9-\\u00DD"
1299         "\\u00E0-\\u00E5\\u00E7-\\u00EF\\u00F1-\\u00F6\\u00F9-\\u00FD"
1300         "\\u00FF-\\u010F\\u0112-\\u0125\\u0128-\\u0130\\u0134-\\u0137"
1301         "\\u0139-\\u013E\\u0143-\\u0148\\u014C-\\u0151\\u0154-\\u0165"
1302         "\\u0168-\\u017E\\u01A0\\u01A1\\u01AF\\u01B0\\u01CD-\\u01DC"
1303         "\\u01DE-\\u01E3\\u01E6-\\u01F0\\u01F4\\u01F5\\u01F8-\\u021B"
1304         "\\u021E\\u021F\\u0226-\\u0233\\u0300-\\u034E\\u0350-\\u036F"
1305         "\\u0374\\u037E\\u0385-\\u038A\\u038C\\u038E-\\u0390\\u03AA-"
1306         "\\u03B0\\u03CA-\\u03CE\\u03D3\\u03D4\\u0400\\u0401\\u0403\\u0407"
1307         "\\u040C-\\u040E\\u0419\\u0439\\u0450\\u0451\\u0453\\u0457\\u045C"
1308         "-\\u045E\\u0476\\u0477\\u0483-\\u0487\\u04C1\\u04C2\\u04D0-"
1309         "\\u04D3\\u04D6\\u04D7\\u04DA-\\u04DF\\u04E2-\\u04E7\\u04EA-"
1310         "\\u04F5\\u04F8\\u04F9\\u0591-\\u05BD\\u05BF\\u05C1\\u05C2\\u05C4"
1311         "\\u05C5\\u05C7\\u0610-\\u061A\\u0622-\\u0626\\u064B-\\u065E"
1312         "\\u0670\\u06C0\\u06C2\\u06D3\\u06D6-\\u06DC\\u06DF-\\u06E4"
1313         "\\u06E7\\u06E8\\u06EA-\\u06ED\\u0711\\u0730-\\u074A\\u07EB-"
1314         "\\u07F3\\u0929\\u0931\\u0934\\u093C\\u094D\\u0951-\\u0954\\u0958"
1315         "-\\u095F\\u09BC\\u09CB-\\u09CD\\u09DC\\u09DD\\u09DF\\u0A33"
1316         "\\u0A36\\u0A3C\\u0A4D\\u0A59-\\u0A5B\\u0A5E\\u0ABC\\u0ACD\\u0B3C"
1317         "\\u0B48\\u0B4B-\\u0B4D\\u0B5C\\u0B5D\\u0B94\\u0BCA-\\u0BCD"
1318         "\\u0C48\\u0C4D\\u0C55\\u0C56\\u0CBC\\u0CC0\\u0CC7\\u0CC8\\u0CCA"
1319         "\\u0CCB\\u0CCD\\u0D4A-\\u0D4D\\u0DCA\\u0DDA\\u0DDC-\\u0DDE"
1320         "\\u0E38-\\u0E3A\\u0E48-\\u0E4B\\u0EB8\\u0EB9\\u0EC8-\\u0ECB"
1321         "\\u0F18\\u0F19\\u0F35\\u0F37\\u0F39\\u0F43\\u0F4D\\u0F52\\u0F57"
1322         "\\u0F5C\\u0F69\\u0F71-\\u0F76\\u0F78\\u0F7A-\\u0F7D\\u0F80-"
1323         "\\u0F84\\u0F86\\u0F87\\u0F93\\u0F9D\\u0FA2\\u0FA7\\u0FAC\\u0FB9"
1324         "\\u0FC6\\u1026\\u1037\\u1039\\u103A\\u108D\\u135F\\u1714\\u1734"
1325         "\\u17D2\\u17DD\\u18A9\\u1939-\\u193B\\u1A17\\u1A18\\u1B06\\u1B08"
1326         "\\u1B0A\\u1B0C\\u1B0E\\u1B12\\u1B34\\u1B3B\\u1B3D\\u1B40\\u1B41"
1327         "\\u1B43\\u1B44\\u1B6B-\\u1B73\\u1BAA\\u1C37\\u1DC0-\\u1DE6"
1328         "\\u1DFE-\\u1E99\\u1E9B\\u1EA0-\\u1EF9\\u1F00-\\u1F15\\u1F18-"
1329         "\\u1F1D\\u1F20-\\u1F45\\u1F48-\\u1F4D\\u1F50-\\u1F57\\u1F59"
1330         "\\u1F5B\\u1F5D\\u1F5F-\\u1F7D\\u1F80-\\u1FB4\\u1FB6-\\u1FBC"
1331         "\\u1FBE\\u1FC1-\\u1FC4\\u1FC6-\\u1FD3\\u1FD6-\\u1FDB\\u1FDD-"
1332         "\\u1FEF\\u1FF2-\\u1FF4\\u1FF6-\\u1FFD\\u2000\\u2001\\u20D0-"
1333         "\\u20DC\\u20E1\\u20E5-\\u20F0\\u2126\\u212A\\u212B\\u219A\\u219B"
1334         "\\u21AE\\u21CD-\\u21CF\\u2204\\u2209\\u220C\\u2224\\u2226\\u2241"
1335         "\\u2244\\u2247\\u2249\\u2260\\u2262\\u226D-\\u2271\\u2274\\u2275"
1336         "\\u2278\\u2279\\u2280\\u2281\\u2284\\u2285\\u2288\\u2289\\u22AC-"
1337         "\\u22AF\\u22E0-\\u22E3\\u22EA-\\u22ED\\u2329\\u232A\\u2ADC"
1338         "\\u2DE0-\\u2DFF\\u302A-\\u302F\\u304C\\u304E\\u3050\\u3052"
1339         "\\u3054\\u3056\\u3058\\u305A\\u305C\\u305E\\u3060\\u3062\\u3065"
1340         "\\u3067\\u3069\\u3070\\u3071\\u3073\\u3074\\u3076\\u3077\\u3079"
1341         "\\u307A\\u307C\\u307D\\u3094\\u3099\\u309A\\u309E\\u30AC\\u30AE"
1342         "\\u30B0\\u30B2\\u30B4\\u30B6\\u30B8\\u30BA\\u30BC\\u30BE\\u30C0"
1343         "\\u30C2\\u30C5\\u30C7\\u30C9\\u30D0\\u30D1\\u30D3\\u30D4\\u30D6"
1344         "\\u30D7\\u30D9\\u30DA\\u30DC\\u30DD\\u30F4\\u30F7-\\u30FA\\u30FE"
1345         "\\uA66F\\uA67C\\uA67D\\uA806\\uA8C4\\uA92B-\\uA92D\\uA953\\uAC00"
1346         "-\\uD7A3\\uF900-\\uFA0D\\uFA10\\uFA12\\uFA15-\\uFA1E\\uFA20"
1347         "\\uFA22\\uFA25\\uFA26\\uFA2A-\\uFA2D\\uFA30-\\uFA6A\\uFA70-"
1348         "\\uFAD9\\uFB1D-\\uFB1F\\uFB2A-\\uFB36\\uFB38-\\uFB3C\\uFB3E"
1349         "\\uFB40\\uFB41\\uFB43\\uFB44\\uFB46-\\uFB4E\\uFE20-\\uFE26"
1350         "\\U000101FD\\U00010A0D\\U00010A0F\\U00010A38-\\U00010A3A\\U00010"
1351         "A3F\\U0001D15E-\\U0001D169\\U0001D16D-\\U0001D172\\U0001D17B-"
1352         "\\U0001D182\\U0001D185-\\U0001D18B\\U0001D1AA-\\U0001D1AD\\U0001"
1353         "D1BB-\\U0001D1C0\\U0001D242-\\U0001D244\\U0002F800-\\U0002FA1D]"
1354         , ""), errorCode);
1355 
1356     skipSets[UNORM_NFC].applyPattern(UnicodeString(
1357         "[^<->A-PR-Za-pr-z\\u00A8\\u00C0-\\u00CF\\u00D1-\\u00D6\\u00D8-"
1358         "\\u00DD\\u00E0-\\u00EF\\u00F1-\\u00F6\\u00F8-\\u00FD\\u00FF-"
1359         "\\u0103\\u0106-\\u010F\\u0112-\\u0117\\u011A-\\u0121\\u0124"
1360         "\\u0125\\u0128-\\u012D\\u0130\\u0139\\u013A\\u013D\\u013E\\u0143"
1361         "\\u0144\\u0147\\u0148\\u014C-\\u0151\\u0154\\u0155\\u0158-"
1362         "\\u015D\\u0160\\u0161\\u0164\\u0165\\u0168-\\u0171\\u0174-"
1363         "\\u017F\\u01A0\\u01A1\\u01AF\\u01B0\\u01B7\\u01CD-\\u01DC\\u01DE"
1364         "-\\u01E1\\u01E6-\\u01EB\\u01F4\\u01F5\\u01F8-\\u01FB\\u0200-"
1365         "\\u021B\\u021E\\u021F\\u0226-\\u0233\\u0292\\u0300-\\u034E"
1366         "\\u0350-\\u036F\\u0374\\u037E\\u0387\\u0391\\u0395\\u0397\\u0399"
1367         "\\u039F\\u03A1\\u03A5\\u03A9\\u03AC\\u03AE\\u03B1\\u03B5\\u03B7"
1368         "\\u03B9\\u03BF\\u03C1\\u03C5\\u03C9-\\u03CB\\u03CE\\u03D2\\u0406"
1369         "\\u0410\\u0413\\u0415-\\u0418\\u041A\\u041E\\u0423\\u0427\\u042B"
1370         "\\u042D\\u0430\\u0433\\u0435-\\u0438\\u043A\\u043E\\u0443\\u0447"
1371         "\\u044B\\u044D\\u0456\\u0474\\u0475\\u0483-\\u0487\\u04D8\\u04D9"
1372         "\\u04E8\\u04E9\\u0591-\\u05BD\\u05BF\\u05C1\\u05C2\\u05C4\\u05C5"
1373         "\\u05C7\\u0610-\\u061A\\u0622\\u0623\\u0627\\u0648\\u064A-"
1374         "\\u065E\\u0670\\u06C1\\u06D2\\u06D5-\\u06DC\\u06DF-\\u06E4"
1375         "\\u06E7\\u06E8\\u06EA-\\u06ED\\u0711\\u0730-\\u074A\\u07EB-"
1376         "\\u07F3\\u0928\\u0930\\u0933\\u093C\\u094D\\u0951-\\u0954\\u0958"
1377         "-\\u095F\\u09BC\\u09BE\\u09C7\\u09CD\\u09D7\\u09DC\\u09DD\\u09DF"
1378         "\\u0A33\\u0A36\\u0A3C\\u0A4D\\u0A59-\\u0A5B\\u0A5E\\u0ABC\\u0ACD"
1379         "\\u0B3C\\u0B3E\\u0B47\\u0B4D\\u0B56\\u0B57\\u0B5C\\u0B5D\\u0B92"
1380         "\\u0BBE\\u0BC6\\u0BC7\\u0BCD\\u0BD7\\u0C46\\u0C4D\\u0C55\\u0C56"
1381         "\\u0CBC\\u0CBF\\u0CC2\\u0CC6\\u0CCA\\u0CCD\\u0CD5\\u0CD6\\u0D3E"
1382         "\\u0D46\\u0D47\\u0D4D\\u0D57\\u0DCA\\u0DCF\\u0DD9\\u0DDC\\u0DDF"
1383         "\\u0E38-\\u0E3A\\u0E48-\\u0E4B\\u0EB8\\u0EB9\\u0EC8-\\u0ECB"
1384         "\\u0F18\\u0F19\\u0F35\\u0F37\\u0F39\\u0F43\\u0F4D\\u0F52\\u0F57"
1385         "\\u0F5C\\u0F69\\u0F71-\\u0F76\\u0F78\\u0F7A-\\u0F7D\\u0F80-"
1386         "\\u0F84\\u0F86\\u0F87\\u0F93\\u0F9D\\u0FA2\\u0FA7\\u0FAC\\u0FB9"
1387         "\\u0FC6\\u1025\\u102E\\u1037\\u1039\\u103A\\u108D\\u1100-\\u1112"
1388         "\\u1161-\\u1175\\u11A8-\\u11C2\\u135F\\u1714\\u1734\\u17D2"
1389         "\\u17DD\\u18A9\\u1939-\\u193B\\u1A17\\u1A18\\u1B05\\u1B07\\u1B09"
1390         "\\u1B0B\\u1B0D\\u1B11\\u1B34\\u1B35\\u1B3A\\u1B3C\\u1B3E\\u1B3F"
1391         "\\u1B42\\u1B44\\u1B6B-\\u1B73\\u1BAA\\u1C37\\u1DC0-\\u1DE6"
1392         "\\u1DFE-\\u1E03\\u1E0A-\\u1E0F\\u1E12-\\u1E1B\\u1E20-\\u1E27"
1393         "\\u1E2A-\\u1E41\\u1E44-\\u1E53\\u1E58-\\u1E7D\\u1E80-\\u1E87"
1394         "\\u1E8E-\\u1E91\\u1E96-\\u1E99\\u1EA0-\\u1EF3\\u1EF6-\\u1EF9"
1395         "\\u1F00-\\u1F11\\u1F18\\u1F19\\u1F20-\\u1F31\\u1F38\\u1F39"
1396         "\\u1F40\\u1F41\\u1F48\\u1F49\\u1F50\\u1F51\\u1F59\\u1F60-\\u1F71"
1397         "\\u1F73-\\u1F75\\u1F77\\u1F79\\u1F7B-\\u1F7D\\u1F80\\u1F81"
1398         "\\u1F88\\u1F89\\u1F90\\u1F91\\u1F98\\u1F99\\u1FA0\\u1FA1\\u1FA8"
1399         "\\u1FA9\\u1FB3\\u1FB6\\u1FBB\\u1FBC\\u1FBE\\u1FBF\\u1FC3\\u1FC6"
1400         "\\u1FC9\\u1FCB\\u1FCC\\u1FD3\\u1FDB\\u1FE3\\u1FEB\\u1FEE\\u1FEF"
1401         "\\u1FF3\\u1FF6\\u1FF9\\u1FFB-\\u1FFE\\u2000\\u2001\\u20D0-"
1402         "\\u20DC\\u20E1\\u20E5-\\u20F0\\u2126\\u212A\\u212B\\u2190\\u2192"
1403         "\\u2194\\u21D0\\u21D2\\u21D4\\u2203\\u2208\\u220B\\u2223\\u2225"
1404         "\\u223C\\u2243\\u2245\\u2248\\u224D\\u2261\\u2264\\u2265\\u2272"
1405         "\\u2273\\u2276\\u2277\\u227A-\\u227D\\u2282\\u2283\\u2286\\u2287"
1406         "\\u2291\\u2292\\u22A2\\u22A8\\u22A9\\u22AB\\u22B2-\\u22B5\\u2329"
1407         "\\u232A\\u2ADC\\u2DE0-\\u2DFF\\u302A-\\u302F\\u3046\\u304B"
1408         "\\u304D\\u304F\\u3051\\u3053\\u3055\\u3057\\u3059\\u305B\\u305D"
1409         "\\u305F\\u3061\\u3064\\u3066\\u3068\\u306F\\u3072\\u3075\\u3078"
1410         "\\u307B\\u3099\\u309A\\u309D\\u30A6\\u30AB\\u30AD\\u30AF\\u30B1"
1411         "\\u30B3\\u30B5\\u30B7\\u30B9\\u30BB\\u30BD\\u30BF\\u30C1\\u30C4"
1412         "\\u30C6\\u30C8\\u30CF\\u30D2\\u30D5\\u30D8\\u30DB\\u30EF-\\u30F2"
1413         "\\u30FD\\uA66F\\uA67C\\uA67D\\uA806\\uA8C4\\uA92B-\\uA92D\\uA953"
1414         "\\uAC00\\uAC1C\\uAC38\\uAC54\\uAC70\\uAC8C\\uACA8\\uACC4\\uACE0"
1415         "\\uACFC\\uAD18\\uAD34\\uAD50\\uAD6C\\uAD88\\uADA4\\uADC0\\uADDC"
1416         "\\uADF8\\uAE14\\uAE30\\uAE4C\\uAE68\\uAE84\\uAEA0\\uAEBC\\uAED8"
1417         "\\uAEF4\\uAF10\\uAF2C\\uAF48\\uAF64\\uAF80\\uAF9C\\uAFB8\\uAFD4"
1418         "\\uAFF0\\uB00C\\uB028\\uB044\\uB060\\uB07C\\uB098\\uB0B4\\uB0D0"
1419         "\\uB0EC\\uB108\\uB124\\uB140\\uB15C\\uB178\\uB194\\uB1B0\\uB1CC"
1420         "\\uB1E8\\uB204\\uB220\\uB23C\\uB258\\uB274\\uB290\\uB2AC\\uB2C8"
1421         "\\uB2E4\\uB300\\uB31C\\uB338\\uB354\\uB370\\uB38C\\uB3A8\\uB3C4"
1422         "\\uB3E0\\uB3FC\\uB418\\uB434\\uB450\\uB46C\\uB488\\uB4A4\\uB4C0"
1423         "\\uB4DC\\uB4F8\\uB514\\uB530\\uB54C\\uB568\\uB584\\uB5A0\\uB5BC"
1424         "\\uB5D8\\uB5F4\\uB610\\uB62C\\uB648\\uB664\\uB680\\uB69C\\uB6B8"
1425         "\\uB6D4\\uB6F0\\uB70C\\uB728\\uB744\\uB760\\uB77C\\uB798\\uB7B4"
1426         "\\uB7D0\\uB7EC\\uB808\\uB824\\uB840\\uB85C\\uB878\\uB894\\uB8B0"
1427         "\\uB8CC\\uB8E8\\uB904\\uB920\\uB93C\\uB958\\uB974\\uB990\\uB9AC"
1428         "\\uB9C8\\uB9E4\\uBA00\\uBA1C\\uBA38\\uBA54\\uBA70\\uBA8C\\uBAA8"
1429         "\\uBAC4\\uBAE0\\uBAFC\\uBB18\\uBB34\\uBB50\\uBB6C\\uBB88\\uBBA4"
1430         "\\uBBC0\\uBBDC\\uBBF8\\uBC14\\uBC30\\uBC4C\\uBC68\\uBC84\\uBCA0"
1431         "\\uBCBC\\uBCD8\\uBCF4\\uBD10\\uBD2C\\uBD48\\uBD64\\uBD80\\uBD9C"
1432         "\\uBDB8\\uBDD4\\uBDF0\\uBE0C\\uBE28\\uBE44\\uBE60\\uBE7C\\uBE98"
1433         "\\uBEB4\\uBED0\\uBEEC\\uBF08\\uBF24\\uBF40\\uBF5C\\uBF78\\uBF94"
1434         "\\uBFB0\\uBFCC\\uBFE8\\uC004\\uC020\\uC03C\\uC058\\uC074\\uC090"
1435         "\\uC0AC\\uC0C8\\uC0E4\\uC100\\uC11C\\uC138\\uC154\\uC170\\uC18C"
1436         "\\uC1A8\\uC1C4\\uC1E0\\uC1FC\\uC218\\uC234\\uC250\\uC26C\\uC288"
1437         "\\uC2A4\\uC2C0\\uC2DC\\uC2F8\\uC314\\uC330\\uC34C\\uC368\\uC384"
1438         "\\uC3A0\\uC3BC\\uC3D8\\uC3F4\\uC410\\uC42C\\uC448\\uC464\\uC480"
1439         "\\uC49C\\uC4B8\\uC4D4\\uC4F0\\uC50C\\uC528\\uC544\\uC560\\uC57C"
1440         "\\uC598\\uC5B4\\uC5D0\\uC5EC\\uC608\\uC624\\uC640\\uC65C\\uC678"
1441         "\\uC694\\uC6B0\\uC6CC\\uC6E8\\uC704\\uC720\\uC73C\\uC758\\uC774"
1442         "\\uC790\\uC7AC\\uC7C8\\uC7E4\\uC800\\uC81C\\uC838\\uC854\\uC870"
1443         "\\uC88C\\uC8A8\\uC8C4\\uC8E0\\uC8FC\\uC918\\uC934\\uC950\\uC96C"
1444         "\\uC988\\uC9A4\\uC9C0\\uC9DC\\uC9F8\\uCA14\\uCA30\\uCA4C\\uCA68"
1445         "\\uCA84\\uCAA0\\uCABC\\uCAD8\\uCAF4\\uCB10\\uCB2C\\uCB48\\uCB64"
1446         "\\uCB80\\uCB9C\\uCBB8\\uCBD4\\uCBF0\\uCC0C\\uCC28\\uCC44\\uCC60"
1447         "\\uCC7C\\uCC98\\uCCB4\\uCCD0\\uCCEC\\uCD08\\uCD24\\uCD40\\uCD5C"
1448         "\\uCD78\\uCD94\\uCDB0\\uCDCC\\uCDE8\\uCE04\\uCE20\\uCE3C\\uCE58"
1449         "\\uCE74\\uCE90\\uCEAC\\uCEC8\\uCEE4\\uCF00\\uCF1C\\uCF38\\uCF54"
1450         "\\uCF70\\uCF8C\\uCFA8\\uCFC4\\uCFE0\\uCFFC\\uD018\\uD034\\uD050"
1451         "\\uD06C\\uD088\\uD0A4\\uD0C0\\uD0DC\\uD0F8\\uD114\\uD130\\uD14C"
1452         "\\uD168\\uD184\\uD1A0\\uD1BC\\uD1D8\\uD1F4\\uD210\\uD22C\\uD248"
1453         "\\uD264\\uD280\\uD29C\\uD2B8\\uD2D4\\uD2F0\\uD30C\\uD328\\uD344"
1454         "\\uD360\\uD37C\\uD398\\uD3B4\\uD3D0\\uD3EC\\uD408\\uD424\\uD440"
1455         "\\uD45C\\uD478\\uD494\\uD4B0\\uD4CC\\uD4E8\\uD504\\uD520\\uD53C"
1456         "\\uD558\\uD574\\uD590\\uD5AC\\uD5C8\\uD5E4\\uD600\\uD61C\\uD638"
1457         "\\uD654\\uD670\\uD68C\\uD6A8\\uD6C4\\uD6E0\\uD6FC\\uD718\\uD734"
1458         "\\uD750\\uD76C\\uD788\\uF900-\\uFA0D\\uFA10\\uFA12\\uFA15-"
1459         "\\uFA1E\\uFA20\\uFA22\\uFA25\\uFA26\\uFA2A-\\uFA2D\\uFA30-"
1460         "\\uFA6A\\uFA70-\\uFAD9\\uFB1D-\\uFB1F\\uFB2A-\\uFB36\\uFB38-"
1461         "\\uFB3C\\uFB3E\\uFB40\\uFB41\\uFB43\\uFB44\\uFB46-\\uFB4E\\uFE20"
1462         "-\\uFE26\\U000101FD\\U00010A0D\\U00010A0F\\U00010A38-\\U00010A3A"
1463         "\\U00010A3F\\U0001D15E-\\U0001D169\\U0001D16D-\\U0001D172\\U0001"
1464         "D17B-\\U0001D182\\U0001D185-\\U0001D18B\\U0001D1AA-\\U0001D1AD"
1465         "\\U0001D1BB-\\U0001D1C0\\U0001D242-\\U0001D244\\U0002F800-"
1466         "\\U0002FA1D]"
1467         , ""), errorCode);
1468 
1469     skipSets[UNORM_NFKD].applyPattern(UnicodeString(
1470         "[^\\u00A0\\u00A8\\u00AA\\u00AF\\u00B2-\\u00B5\\u00B8-\\u00BA"
1471         "\\u00BC-\\u00BE\\u00C0-\\u00C5\\u00C7-\\u00CF\\u00D1-\\u00D6"
1472         "\\u00D9-\\u00DD\\u00E0-\\u00E5\\u00E7-\\u00EF\\u00F1-\\u00F6"
1473         "\\u00F9-\\u00FD\\u00FF-\\u010F\\u0112-\\u0125\\u0128-\\u0130"
1474         "\\u0132-\\u0137\\u0139-\\u0140\\u0143-\\u0149\\u014C-\\u0151"
1475         "\\u0154-\\u0165\\u0168-\\u017F\\u01A0\\u01A1\\u01AF\\u01B0"
1476         "\\u01C4-\\u01DC\\u01DE-\\u01E3\\u01E6-\\u01F5\\u01F8-\\u021B"
1477         "\\u021E\\u021F\\u0226-\\u0233\\u02B0-\\u02B8\\u02D8-\\u02DD"
1478         "\\u02E0-\\u02E4\\u0300-\\u034E\\u0350-\\u036F\\u0374\\u037A"
1479         "\\u037E\\u0384-\\u038A\\u038C\\u038E-\\u0390\\u03AA-\\u03B0"
1480         "\\u03CA-\\u03CE\\u03D0-\\u03D6\\u03F0-\\u03F2\\u03F4\\u03F5"
1481         "\\u03F9\\u0400\\u0401\\u0403\\u0407\\u040C-\\u040E\\u0419\\u0439"
1482         "\\u0450\\u0451\\u0453\\u0457\\u045C-\\u045E\\u0476\\u0477\\u0483"
1483         "-\\u0487\\u04C1\\u04C2\\u04D0-\\u04D3\\u04D6\\u04D7\\u04DA-"
1484         "\\u04DF\\u04E2-\\u04E7\\u04EA-\\u04F5\\u04F8\\u04F9\\u0587"
1485         "\\u0591-\\u05BD\\u05BF\\u05C1\\u05C2\\u05C4\\u05C5\\u05C7\\u0610"
1486         "-\\u061A\\u0622-\\u0626\\u064B-\\u065E\\u0670\\u0675-\\u0678"
1487         "\\u06C0\\u06C2\\u06D3\\u06D6-\\u06DC\\u06DF-\\u06E4\\u06E7"
1488         "\\u06E8\\u06EA-\\u06ED\\u0711\\u0730-\\u074A\\u07EB-\\u07F3"
1489         "\\u0929\\u0931\\u0934\\u093C\\u094D\\u0951-\\u0954\\u0958-"
1490         "\\u095F\\u09BC\\u09CB-\\u09CD\\u09DC\\u09DD\\u09DF\\u0A33\\u0A36"
1491         "\\u0A3C\\u0A4D\\u0A59-\\u0A5B\\u0A5E\\u0ABC\\u0ACD\\u0B3C\\u0B48"
1492         "\\u0B4B-\\u0B4D\\u0B5C\\u0B5D\\u0B94\\u0BCA-\\u0BCD\\u0C48"
1493         "\\u0C4D\\u0C55\\u0C56\\u0CBC\\u0CC0\\u0CC7\\u0CC8\\u0CCA\\u0CCB"
1494         "\\u0CCD\\u0D4A-\\u0D4D\\u0DCA\\u0DDA\\u0DDC-\\u0DDE\\u0E33"
1495         "\\u0E38-\\u0E3A\\u0E48-\\u0E4B\\u0EB3\\u0EB8\\u0EB9\\u0EC8-"
1496         "\\u0ECB\\u0EDC\\u0EDD\\u0F0C\\u0F18\\u0F19\\u0F35\\u0F37\\u0F39"
1497         "\\u0F43\\u0F4D\\u0F52\\u0F57\\u0F5C\\u0F69\\u0F71-\\u0F7D\\u0F80"
1498         "-\\u0F84\\u0F86\\u0F87\\u0F93\\u0F9D\\u0FA2\\u0FA7\\u0FAC\\u0FB9"
1499         "\\u0FC6\\u1026\\u1037\\u1039\\u103A\\u108D\\u10FC\\u135F\\u1714"
1500         "\\u1734\\u17D2\\u17DD\\u18A9\\u1939-\\u193B\\u1A17\\u1A18\\u1B06"
1501         "\\u1B08\\u1B0A\\u1B0C\\u1B0E\\u1B12\\u1B34\\u1B3B\\u1B3D\\u1B40"
1502         "\\u1B41\\u1B43\\u1B44\\u1B6B-\\u1B73\\u1BAA\\u1C37\\u1D2C-"
1503         "\\u1D2E\\u1D30-\\u1D3A\\u1D3C-\\u1D4D\\u1D4F-\\u1D6A\\u1D78"
1504         "\\u1D9B-\\u1DE6\\u1DFE-\\u1E9B\\u1EA0-\\u1EF9\\u1F00-\\u1F15"
1505         "\\u1F18-\\u1F1D\\u1F20-\\u1F45\\u1F48-\\u1F4D\\u1F50-\\u1F57"
1506         "\\u1F59\\u1F5B\\u1F5D\\u1F5F-\\u1F7D\\u1F80-\\u1FB4\\u1FB6-"
1507         "\\u1FC4\\u1FC6-\\u1FD3\\u1FD6-\\u1FDB\\u1FDD-\\u1FEF\\u1FF2-"
1508         "\\u1FF4\\u1FF6-\\u1FFE\\u2000-\\u200A\\u2011\\u2017\\u2024-"
1509         "\\u2026\\u202F\\u2033\\u2034\\u2036\\u2037\\u203C\\u203E\\u2047-"
1510         "\\u2049\\u2057\\u205F\\u2070\\u2071\\u2074-\\u208E\\u2090-"
1511         "\\u2094\\u20A8\\u20D0-\\u20DC\\u20E1\\u20E5-\\u20F0\\u2100-"
1512         "\\u2103\\u2105-\\u2107\\u2109-\\u2113\\u2115\\u2116\\u2119-"
1513         "\\u211D\\u2120-\\u2122\\u2124\\u2126\\u2128\\u212A-\\u212D"
1514         "\\u212F-\\u2131\\u2133-\\u2139\\u213B-\\u2140\\u2145-\\u2149"
1515         "\\u2153-\\u217F\\u219A\\u219B\\u21AE\\u21CD-\\u21CF\\u2204"
1516         "\\u2209\\u220C\\u2224\\u2226\\u222C\\u222D\\u222F\\u2230\\u2241"
1517         "\\u2244\\u2247\\u2249\\u2260\\u2262\\u226D-\\u2271\\u2274\\u2275"
1518         "\\u2278\\u2279\\u2280\\u2281\\u2284\\u2285\\u2288\\u2289\\u22AC-"
1519         "\\u22AF\\u22E0-\\u22E3\\u22EA-\\u22ED\\u2329\\u232A\\u2460-"
1520         "\\u24EA\\u2A0C\\u2A74-\\u2A76\\u2ADC\\u2C7C\\u2C7D\\u2D6F\\u2DE0"
1521         "-\\u2DFF\\u2E9F\\u2EF3\\u2F00-\\u2FD5\\u3000\\u302A-\\u302F"
1522         "\\u3036\\u3038-\\u303A\\u304C\\u304E\\u3050\\u3052\\u3054\\u3056"
1523         "\\u3058\\u305A\\u305C\\u305E\\u3060\\u3062\\u3065\\u3067\\u3069"
1524         "\\u3070\\u3071\\u3073\\u3074\\u3076\\u3077\\u3079\\u307A\\u307C"
1525         "\\u307D\\u3094\\u3099-\\u309C\\u309E\\u309F\\u30AC\\u30AE\\u30B0"
1526         "\\u30B2\\u30B4\\u30B6\\u30B8\\u30BA\\u30BC\\u30BE\\u30C0\\u30C2"
1527         "\\u30C5\\u30C7\\u30C9\\u30D0\\u30D1\\u30D3\\u30D4\\u30D6\\u30D7"
1528         "\\u30D9\\u30DA\\u30DC\\u30DD\\u30F4\\u30F7-\\u30FA\\u30FE\\u30FF"
1529         "\\u3131-\\u318E\\u3192-\\u319F\\u3200-\\u321E\\u3220-\\u3243"
1530         "\\u3250-\\u327E\\u3280-\\u32FE\\u3300-\\u33FF\\uA66F\\uA67C"
1531         "\\uA67D\\uA770\\uA806\\uA8C4\\uA92B-\\uA92D\\uA953\\uAC00-"
1532         "\\uD7A3\\uF900-\\uFA0D\\uFA10\\uFA12\\uFA15-\\uFA1E\\uFA20"
1533         "\\uFA22\\uFA25\\uFA26\\uFA2A-\\uFA2D\\uFA30-\\uFA6A\\uFA70-"
1534         "\\uFAD9\\uFB00-\\uFB06\\uFB13-\\uFB17\\uFB1D-\\uFB36\\uFB38-"
1535         "\\uFB3C\\uFB3E\\uFB40\\uFB41\\uFB43\\uFB44\\uFB46-\\uFBB1\\uFBD3"
1536         "-\\uFD3D\\uFD50-\\uFD8F\\uFD92-\\uFDC7\\uFDF0-\\uFDFC\\uFE10-"
1537         "\\uFE19\\uFE20-\\uFE26\\uFE30-\\uFE44\\uFE47-\\uFE52\\uFE54-"
1538         "\\uFE66\\uFE68-\\uFE6B\\uFE70-\\uFE72\\uFE74\\uFE76-\\uFEFC"
1539         "\\uFF01-\\uFFBE\\uFFC2-\\uFFC7\\uFFCA-\\uFFCF\\uFFD2-\\uFFD7"
1540         "\\uFFDA-\\uFFDC\\uFFE0-\\uFFE6\\uFFE8-\\uFFEE\\U000101FD\\U00010"
1541         "A0D\\U00010A0F\\U00010A38-\\U00010A3A\\U00010A3F\\U0001D15E-"
1542         "\\U0001D169\\U0001D16D-\\U0001D172\\U0001D17B-\\U0001D182\\U0001"
1543         "D185-\\U0001D18B\\U0001D1AA-\\U0001D1AD\\U0001D1BB-\\U0001D1C0"
1544         "\\U0001D242-\\U0001D244\\U0001D400-\\U0001D454\\U0001D456-"
1545         "\\U0001D49C\\U0001D49E\\U0001D49F\\U0001D4A2\\U0001D4A5\\U0001D4"
1546         "A6\\U0001D4A9-\\U0001D4AC\\U0001D4AE-\\U0001D4B9\\U0001D4BB"
1547         "\\U0001D4BD-\\U0001D4C3\\U0001D4C5-\\U0001D505\\U0001D507-"
1548         "\\U0001D50A\\U0001D50D-\\U0001D514\\U0001D516-\\U0001D51C\\U0001"
1549         "D51E-\\U0001D539\\U0001D53B-\\U0001D53E\\U0001D540-\\U0001D544"
1550         "\\U0001D546\\U0001D54A-\\U0001D550\\U0001D552-\\U0001D6A5\\U0001"
1551         "D6A8-\\U0001D7CB\\U0001D7CE-\\U0001D7FF\\U0002F800-\\U0002FA1D]"
1552         , ""), errorCode);
1553 
1554     skipSets[UNORM_NFKC].applyPattern(UnicodeString(
1555         "[^<->A-PR-Za-pr-z\\u00A0\\u00A8\\u00AA\\u00AF\\u00B2-\\u00B5"
1556         "\\u00B8-\\u00BA\\u00BC-\\u00BE\\u00C0-\\u00CF\\u00D1-\\u00D6"
1557         "\\u00D8-\\u00DD\\u00E0-\\u00EF\\u00F1-\\u00F6\\u00F8-\\u00FD"
1558         "\\u00FF-\\u0103\\u0106-\\u010F\\u0112-\\u0117\\u011A-\\u0121"
1559         "\\u0124\\u0125\\u0128-\\u012D\\u0130\\u0132\\u0133\\u0139\\u013A"
1560         "\\u013D-\\u0140\\u0143\\u0144\\u0147-\\u0149\\u014C-\\u0151"
1561         "\\u0154\\u0155\\u0158-\\u015D\\u0160\\u0161\\u0164\\u0165\\u0168"
1562         "-\\u0171\\u0174-\\u017F\\u01A0\\u01A1\\u01AF\\u01B0\\u01B7"
1563         "\\u01C4-\\u01DC\\u01DE-\\u01E1\\u01E6-\\u01EB\\u01F1-\\u01F5"
1564         "\\u01F8-\\u01FB\\u0200-\\u021B\\u021E\\u021F\\u0226-\\u0233"
1565         "\\u0292\\u02B0-\\u02B8\\u02D8-\\u02DD\\u02E0-\\u02E4\\u0300-"
1566         "\\u034E\\u0350-\\u036F\\u0374\\u037A\\u037E\\u0384\\u0385\\u0387"
1567         "\\u0391\\u0395\\u0397\\u0399\\u039F\\u03A1\\u03A5\\u03A9\\u03AC"
1568         "\\u03AE\\u03B1\\u03B5\\u03B7\\u03B9\\u03BF\\u03C1\\u03C5\\u03C9-"
1569         "\\u03CB\\u03CE\\u03D0-\\u03D6\\u03F0-\\u03F2\\u03F4\\u03F5"
1570         "\\u03F9\\u0406\\u0410\\u0413\\u0415-\\u0418\\u041A\\u041E\\u0423"
1571         "\\u0427\\u042B\\u042D\\u0430\\u0433\\u0435-\\u0438\\u043A\\u043E"
1572         "\\u0443\\u0447\\u044B\\u044D\\u0456\\u0474\\u0475\\u0483-\\u0487"
1573         "\\u04D8\\u04D9\\u04E8\\u04E9\\u0587\\u0591-\\u05BD\\u05BF\\u05C1"
1574         "\\u05C2\\u05C4\\u05C5\\u05C7\\u0610-\\u061A\\u0622\\u0623\\u0627"
1575         "\\u0648\\u064A-\\u065E\\u0670\\u0675-\\u0678\\u06C1\\u06D2"
1576         "\\u06D5-\\u06DC\\u06DF-\\u06E4\\u06E7\\u06E8\\u06EA-\\u06ED"
1577         "\\u0711\\u0730-\\u074A\\u07EB-\\u07F3\\u0928\\u0930\\u0933"
1578         "\\u093C\\u094D\\u0951-\\u0954\\u0958-\\u095F\\u09BC\\u09BE"
1579         "\\u09C7\\u09CD\\u09D7\\u09DC\\u09DD\\u09DF\\u0A33\\u0A36\\u0A3C"
1580         "\\u0A4D\\u0A59-\\u0A5B\\u0A5E\\u0ABC\\u0ACD\\u0B3C\\u0B3E\\u0B47"
1581         "\\u0B4D\\u0B56\\u0B57\\u0B5C\\u0B5D\\u0B92\\u0BBE\\u0BC6\\u0BC7"
1582         "\\u0BCD\\u0BD7\\u0C46\\u0C4D\\u0C55\\u0C56\\u0CBC\\u0CBF\\u0CC2"
1583         "\\u0CC6\\u0CCA\\u0CCD\\u0CD5\\u0CD6\\u0D3E\\u0D46\\u0D47\\u0D4D"
1584         "\\u0D57\\u0DCA\\u0DCF\\u0DD9\\u0DDC\\u0DDF\\u0E33\\u0E38-\\u0E3A"
1585         "\\u0E48-\\u0E4B\\u0EB3\\u0EB8\\u0EB9\\u0EC8-\\u0ECB\\u0EDC"
1586         "\\u0EDD\\u0F0C\\u0F18\\u0F19\\u0F35\\u0F37\\u0F39\\u0F43\\u0F4D"
1587         "\\u0F52\\u0F57\\u0F5C\\u0F69\\u0F71-\\u0F7D\\u0F80-\\u0F84"
1588         "\\u0F86\\u0F87\\u0F93\\u0F9D\\u0FA2\\u0FA7\\u0FAC\\u0FB9\\u0FC6"
1589         "\\u1025\\u102E\\u1037\\u1039\\u103A\\u108D\\u10FC\\u1100-\\u1112"
1590         "\\u1161-\\u1175\\u11A8-\\u11C2\\u135F\\u1714\\u1734\\u17D2"
1591         "\\u17DD\\u18A9\\u1939-\\u193B\\u1A17\\u1A18\\u1B05\\u1B07\\u1B09"
1592         "\\u1B0B\\u1B0D\\u1B11\\u1B34\\u1B35\\u1B3A\\u1B3C\\u1B3E\\u1B3F"
1593         "\\u1B42\\u1B44\\u1B6B-\\u1B73\\u1BAA\\u1C37\\u1D2C-\\u1D2E"
1594         "\\u1D30-\\u1D3A\\u1D3C-\\u1D4D\\u1D4F-\\u1D6A\\u1D78\\u1D9B-"
1595         "\\u1DE6\\u1DFE-\\u1E03\\u1E0A-\\u1E0F\\u1E12-\\u1E1B\\u1E20-"
1596         "\\u1E27\\u1E2A-\\u1E41\\u1E44-\\u1E53\\u1E58-\\u1E7D\\u1E80-"
1597         "\\u1E87\\u1E8E-\\u1E91\\u1E96-\\u1E9B\\u1EA0-\\u1EF3\\u1EF6-"
1598         "\\u1EF9\\u1F00-\\u1F11\\u1F18\\u1F19\\u1F20-\\u1F31\\u1F38"
1599         "\\u1F39\\u1F40\\u1F41\\u1F48\\u1F49\\u1F50\\u1F51\\u1F59\\u1F60-"
1600         "\\u1F71\\u1F73-\\u1F75\\u1F77\\u1F79\\u1F7B-\\u1F7D\\u1F80"
1601         "\\u1F81\\u1F88\\u1F89\\u1F90\\u1F91\\u1F98\\u1F99\\u1FA0\\u1FA1"
1602         "\\u1FA8\\u1FA9\\u1FB3\\u1FB6\\u1FBB-\\u1FC1\\u1FC3\\u1FC6\\u1FC9"
1603         "\\u1FCB-\\u1FCF\\u1FD3\\u1FDB\\u1FDD-\\u1FDF\\u1FE3\\u1FEB"
1604         "\\u1FED-\\u1FEF\\u1FF3\\u1FF6\\u1FF9\\u1FFB-\\u1FFE\\u2000-"
1605         "\\u200A\\u2011\\u2017\\u2024-\\u2026\\u202F\\u2033\\u2034\\u2036"
1606         "\\u2037\\u203C\\u203E\\u2047-\\u2049\\u2057\\u205F\\u2070\\u2071"
1607         "\\u2074-\\u208E\\u2090-\\u2094\\u20A8\\u20D0-\\u20DC\\u20E1"
1608         "\\u20E5-\\u20F0\\u2100-\\u2103\\u2105-\\u2107\\u2109-\\u2113"
1609         "\\u2115\\u2116\\u2119-\\u211D\\u2120-\\u2122\\u2124\\u2126"
1610         "\\u2128\\u212A-\\u212D\\u212F-\\u2131\\u2133-\\u2139\\u213B-"
1611         "\\u2140\\u2145-\\u2149\\u2153-\\u217F\\u2190\\u2192\\u2194"
1612         "\\u21D0\\u21D2\\u21D4\\u2203\\u2208\\u220B\\u2223\\u2225\\u222C"
1613         "\\u222D\\u222F\\u2230\\u223C\\u2243\\u2245\\u2248\\u224D\\u2261"
1614         "\\u2264\\u2265\\u2272\\u2273\\u2276\\u2277\\u227A-\\u227D\\u2282"
1615         "\\u2283\\u2286\\u2287\\u2291\\u2292\\u22A2\\u22A8\\u22A9\\u22AB"
1616         "\\u22B2-\\u22B5\\u2329\\u232A\\u2460-\\u24EA\\u2A0C\\u2A74-"
1617         "\\u2A76\\u2ADC\\u2C7C\\u2C7D\\u2D6F\\u2DE0-\\u2DFF\\u2E9F\\u2EF3"
1618         "\\u2F00-\\u2FD5\\u3000\\u302A-\\u302F\\u3036\\u3038-\\u303A"
1619         "\\u3046\\u304B\\u304D\\u304F\\u3051\\u3053\\u3055\\u3057\\u3059"
1620         "\\u305B\\u305D\\u305F\\u3061\\u3064\\u3066\\u3068\\u306F\\u3072"
1621         "\\u3075\\u3078\\u307B\\u3099-\\u309D\\u309F\\u30A6\\u30AB\\u30AD"
1622         "\\u30AF\\u30B1\\u30B3\\u30B5\\u30B7\\u30B9\\u30BB\\u30BD\\u30BF"
1623         "\\u30C1\\u30C4\\u30C6\\u30C8\\u30CF\\u30D2\\u30D5\\u30D8\\u30DB"
1624         "\\u30EF-\\u30F2\\u30FD\\u30FF\\u3131-\\u318E\\u3192-\\u319F"
1625         "\\u3200-\\u321E\\u3220-\\u3243\\u3250-\\u327E\\u3280-\\u32FE"
1626         "\\u3300-\\u33FF\\uA66F\\uA67C\\uA67D\\uA770\\uA806\\uA8C4\\uA92B"
1627         "-\\uA92D\\uA953\\uAC00\\uAC1C\\uAC38\\uAC54\\uAC70\\uAC8C\\uACA8"
1628         "\\uACC4\\uACE0\\uACFC\\uAD18\\uAD34\\uAD50\\uAD6C\\uAD88\\uADA4"
1629         "\\uADC0\\uADDC\\uADF8\\uAE14\\uAE30\\uAE4C\\uAE68\\uAE84\\uAEA0"
1630         "\\uAEBC\\uAED8\\uAEF4\\uAF10\\uAF2C\\uAF48\\uAF64\\uAF80\\uAF9C"
1631         "\\uAFB8\\uAFD4\\uAFF0\\uB00C\\uB028\\uB044\\uB060\\uB07C\\uB098"
1632         "\\uB0B4\\uB0D0\\uB0EC\\uB108\\uB124\\uB140\\uB15C\\uB178\\uB194"
1633         "\\uB1B0\\uB1CC\\uB1E8\\uB204\\uB220\\uB23C\\uB258\\uB274\\uB290"
1634         "\\uB2AC\\uB2C8\\uB2E4\\uB300\\uB31C\\uB338\\uB354\\uB370\\uB38C"
1635         "\\uB3A8\\uB3C4\\uB3E0\\uB3FC\\uB418\\uB434\\uB450\\uB46C\\uB488"
1636         "\\uB4A4\\uB4C0\\uB4DC\\uB4F8\\uB514\\uB530\\uB54C\\uB568\\uB584"
1637         "\\uB5A0\\uB5BC\\uB5D8\\uB5F4\\uB610\\uB62C\\uB648\\uB664\\uB680"
1638         "\\uB69C\\uB6B8\\uB6D4\\uB6F0\\uB70C\\uB728\\uB744\\uB760\\uB77C"
1639         "\\uB798\\uB7B4\\uB7D0\\uB7EC\\uB808\\uB824\\uB840\\uB85C\\uB878"
1640         "\\uB894\\uB8B0\\uB8CC\\uB8E8\\uB904\\uB920\\uB93C\\uB958\\uB974"
1641         "\\uB990\\uB9AC\\uB9C8\\uB9E4\\uBA00\\uBA1C\\uBA38\\uBA54\\uBA70"
1642         "\\uBA8C\\uBAA8\\uBAC4\\uBAE0\\uBAFC\\uBB18\\uBB34\\uBB50\\uBB6C"
1643         "\\uBB88\\uBBA4\\uBBC0\\uBBDC\\uBBF8\\uBC14\\uBC30\\uBC4C\\uBC68"
1644         "\\uBC84\\uBCA0\\uBCBC\\uBCD8\\uBCF4\\uBD10\\uBD2C\\uBD48\\uBD64"
1645         "\\uBD80\\uBD9C\\uBDB8\\uBDD4\\uBDF0\\uBE0C\\uBE28\\uBE44\\uBE60"
1646         "\\uBE7C\\uBE98\\uBEB4\\uBED0\\uBEEC\\uBF08\\uBF24\\uBF40\\uBF5C"
1647         "\\uBF78\\uBF94\\uBFB0\\uBFCC\\uBFE8\\uC004\\uC020\\uC03C\\uC058"
1648         "\\uC074\\uC090\\uC0AC\\uC0C8\\uC0E4\\uC100\\uC11C\\uC138\\uC154"
1649         "\\uC170\\uC18C\\uC1A8\\uC1C4\\uC1E0\\uC1FC\\uC218\\uC234\\uC250"
1650         "\\uC26C\\uC288\\uC2A4\\uC2C0\\uC2DC\\uC2F8\\uC314\\uC330\\uC34C"
1651         "\\uC368\\uC384\\uC3A0\\uC3BC\\uC3D8\\uC3F4\\uC410\\uC42C\\uC448"
1652         "\\uC464\\uC480\\uC49C\\uC4B8\\uC4D4\\uC4F0\\uC50C\\uC528\\uC544"
1653         "\\uC560\\uC57C\\uC598\\uC5B4\\uC5D0\\uC5EC\\uC608\\uC624\\uC640"
1654         "\\uC65C\\uC678\\uC694\\uC6B0\\uC6CC\\uC6E8\\uC704\\uC720\\uC73C"
1655         "\\uC758\\uC774\\uC790\\uC7AC\\uC7C8\\uC7E4\\uC800\\uC81C\\uC838"
1656         "\\uC854\\uC870\\uC88C\\uC8A8\\uC8C4\\uC8E0\\uC8FC\\uC918\\uC934"
1657         "\\uC950\\uC96C\\uC988\\uC9A4\\uC9C0\\uC9DC\\uC9F8\\uCA14\\uCA30"
1658         "\\uCA4C\\uCA68\\uCA84\\uCAA0\\uCABC\\uCAD8\\uCAF4\\uCB10\\uCB2C"
1659         "\\uCB48\\uCB64\\uCB80\\uCB9C\\uCBB8\\uCBD4\\uCBF0\\uCC0C\\uCC28"
1660         "\\uCC44\\uCC60\\uCC7C\\uCC98\\uCCB4\\uCCD0\\uCCEC\\uCD08\\uCD24"
1661         "\\uCD40\\uCD5C\\uCD78\\uCD94\\uCDB0\\uCDCC\\uCDE8\\uCE04\\uCE20"
1662         "\\uCE3C\\uCE58\\uCE74\\uCE90\\uCEAC\\uCEC8\\uCEE4\\uCF00\\uCF1C"
1663         "\\uCF38\\uCF54\\uCF70\\uCF8C\\uCFA8\\uCFC4\\uCFE0\\uCFFC\\uD018"
1664         "\\uD034\\uD050\\uD06C\\uD088\\uD0A4\\uD0C0\\uD0DC\\uD0F8\\uD114"
1665         "\\uD130\\uD14C\\uD168\\uD184\\uD1A0\\uD1BC\\uD1D8\\uD1F4\\uD210"
1666         "\\uD22C\\uD248\\uD264\\uD280\\uD29C\\uD2B8\\uD2D4\\uD2F0\\uD30C"
1667         "\\uD328\\uD344\\uD360\\uD37C\\uD398\\uD3B4\\uD3D0\\uD3EC\\uD408"
1668         "\\uD424\\uD440\\uD45C\\uD478\\uD494\\uD4B0\\uD4CC\\uD4E8\\uD504"
1669         "\\uD520\\uD53C\\uD558\\uD574\\uD590\\uD5AC\\uD5C8\\uD5E4\\uD600"
1670         "\\uD61C\\uD638\\uD654\\uD670\\uD68C\\uD6A8\\uD6C4\\uD6E0\\uD6FC"
1671         "\\uD718\\uD734\\uD750\\uD76C\\uD788\\uF900-\\uFA0D\\uFA10\\uFA12"
1672         "\\uFA15-\\uFA1E\\uFA20\\uFA22\\uFA25\\uFA26\\uFA2A-\\uFA2D"
1673         "\\uFA30-\\uFA6A\\uFA70-\\uFAD9\\uFB00-\\uFB06\\uFB13-\\uFB17"
1674         "\\uFB1D-\\uFB36\\uFB38-\\uFB3C\\uFB3E\\uFB40\\uFB41\\uFB43"
1675         "\\uFB44\\uFB46-\\uFBB1\\uFBD3-\\uFD3D\\uFD50-\\uFD8F\\uFD92-"
1676         "\\uFDC7\\uFDF0-\\uFDFC\\uFE10-\\uFE19\\uFE20-\\uFE26\\uFE30-"
1677         "\\uFE44\\uFE47-\\uFE52\\uFE54-\\uFE66\\uFE68-\\uFE6B\\uFE70-"
1678         "\\uFE72\\uFE74\\uFE76-\\uFEFC\\uFF01-\\uFFBE\\uFFC2-\\uFFC7"
1679         "\\uFFCA-\\uFFCF\\uFFD2-\\uFFD7\\uFFDA-\\uFFDC\\uFFE0-\\uFFE6"
1680         "\\uFFE8-\\uFFEE\\U000101FD\\U00010A0D\\U00010A0F\\U00010A38-"
1681         "\\U00010A3A\\U00010A3F\\U0001D15E-\\U0001D169\\U0001D16D-\\U0001"
1682         "D172\\U0001D17B-\\U0001D182\\U0001D185-\\U0001D18B\\U0001D1AA-"
1683         "\\U0001D1AD\\U0001D1BB-\\U0001D1C0\\U0001D242-\\U0001D244\\U0001"
1684         "D400-\\U0001D454\\U0001D456-\\U0001D49C\\U0001D49E\\U0001D49F"
1685         "\\U0001D4A2\\U0001D4A5\\U0001D4A6\\U0001D4A9-\\U0001D4AC\\U0001D"
1686         "4AE-\\U0001D4B9\\U0001D4BB\\U0001D4BD-\\U0001D4C3\\U0001D4C5-"
1687         "\\U0001D505\\U0001D507-\\U0001D50A\\U0001D50D-\\U0001D514\\U0001"
1688         "D516-\\U0001D51C\\U0001D51E-\\U0001D539\\U0001D53B-\\U0001D53E"
1689         "\\U0001D540-\\U0001D544\\U0001D546\\U0001D54A-\\U0001D550\\U0001"
1690         "D552-\\U0001D6A5\\U0001D6A8-\\U0001D7CB\\U0001D7CE-\\U0001D7FF"
1691         "\\U0002F800-\\U0002FA1D]"
1692         , ""), errorCode);
1693 }
1694 
1695 U_CDECL_BEGIN
1696 
1697 // USetAdder implementation
1698 // Does not use uset.h to reduce code dependencies
1699 static void U_CALLCONV
_set_add(USet * set,UChar32 c)1700 _set_add(USet *set, UChar32 c) {
1701     uset_add(set, c);
1702 }
1703 
1704 static void U_CALLCONV
_set_addRange(USet * set,UChar32 start,UChar32 end)1705 _set_addRange(USet *set, UChar32 start, UChar32 end) {
1706     uset_addRange(set, start, end);
1707 }
1708 
1709 static void U_CALLCONV
_set_addString(USet * set,const UChar * str,int32_t length)1710 _set_addString(USet *set, const UChar *str, int32_t length) {
1711     uset_addString(set, str, length);
1712 }
1713 
1714 U_CDECL_END
1715 
1716 void
TestSkippable()1717 BasicNormalizerTest::TestSkippable() {
1718     UnicodeSet starts, diff, skipSets[UNORM_MODE_COUNT], expectSets[UNORM_MODE_COUNT];
1719     UnicodeSet *startsPtr = &starts;
1720     UnicodeString s, pattern;
1721     UChar32 start, limit, rangeStart, rangeEnd;
1722     int32_t i, range, count;
1723 
1724     UErrorCode status;
1725 
1726     /* build NF*Skippable sets from runtime data */
1727     status=U_ZERO_ERROR;
1728     USetAdder sa = {
1729         (USet *)startsPtr,
1730         _set_add,
1731         _set_addRange,
1732         _set_addString,
1733         NULL, // don't need remove()
1734         NULL
1735     };
1736     unorm_addPropertyStarts(&sa, &status);
1737     if(U_FAILURE(status)) {
1738         errln("unable to load normalization data for unorm_addPropertyStarts(() - %s\n", u_errorName(status));
1739         return;
1740     }
1741     count=starts.getRangeCount();
1742 
1743     start=limit=0;
1744     rangeStart=rangeEnd=0;
1745     range=0;
1746     for(;;) {
1747         if(start<limit) {
1748             /* get properties for start and apply them to [start..limit[ */
1749             if(unorm_isNFSkippable(start, UNORM_NFD)) {
1750                 skipSets[UNORM_NFD].add(start, limit-1);
1751             }
1752             if(unorm_isNFSkippable(start, UNORM_NFKD)) {
1753                 skipSets[UNORM_NFKD].add(start, limit-1);
1754             }
1755             if(unorm_isNFSkippable(start, UNORM_NFC)) {
1756                 skipSets[UNORM_NFC].add(start, limit-1);
1757             }
1758             if(unorm_isNFSkippable(start, UNORM_NFKC)) {
1759                 skipSets[UNORM_NFKC].add(start, limit-1);
1760             }
1761         }
1762 
1763         /* go to next range of same properties */
1764         start=limit;
1765         if(++limit>rangeEnd) {
1766             if(range<count) {
1767                 limit=rangeStart=starts.getRangeStart(range);
1768                 rangeEnd=starts.getRangeEnd(range);
1769                 ++range;
1770             } else if(range==count) {
1771                 /* additional range to complete the Unicode code space */
1772                 limit=rangeStart=rangeEnd=0x110000;
1773                 ++range;
1774             } else {
1775                 break;
1776             }
1777         }
1778     }
1779 
1780     /* get expected sets from hardcoded patterns */
1781     initExpectedSkippables(expectSets);
1782 
1783     for(i=UNORM_NONE; i<UNORM_MODE_COUNT; ++i) {
1784         if(skipSets[i]!=expectSets[i]) {
1785             errln("error: TestSkippable skipSets[%d]!=expectedSets[%d]\n"
1786                   "may need to update hardcoded UnicodeSet patterns in\n"
1787                   "tstnorm.cpp/initExpectedSkippables(),\n"
1788                   "see ICU4J - unicodetools.com.ibm.text.UCD.NFSkippable\n",
1789                   i, i);
1790 
1791             s=UNICODE_STRING_SIMPLE("skip-expect=");
1792             (diff=skipSets[i]).removeAll(expectSets[i]).toPattern(pattern, TRUE);
1793             s.append(pattern);
1794 
1795             pattern.remove();
1796             s.append(UNICODE_STRING_SIMPLE("\n\nexpect-skip="));
1797             (diff=expectSets[i]).removeAll(skipSets[i]).toPattern(pattern, TRUE);
1798             s.append(pattern);
1799             s.append(UNICODE_STRING_SIMPLE("\n\n"));
1800 
1801             errln(s);
1802         }
1803     }
1804 }
1805 
1806 #endif /* #if !UCONFIG_NO_NORMALIZATION */
1807