• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 // © 2017 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 
4 #include "unicode/utypes.h"
5 
6 #if !UCONFIG_NO_FORMATTING
7 
8 #include "unicode/dcfmtsym.h"
9 
10 #include "cstr.h"
11 #include "numbertest.h"
12 #include "number_utils.h"
13 #include "number_skeletons.h"
14 #include "putilimp.h"
15 
16 using namespace icu::number::impl;
17 
18 
runIndexedTest(int32_t index,UBool exec,const char * & name,char *)19 void NumberSkeletonTest::runIndexedTest(int32_t index, UBool exec, const char*& name, char*) {
20     if (exec) {
21         logln("TestSuite AffixUtilsTest: ");
22     }
23     TESTCASE_AUTO_BEGIN;
24         TESTCASE_AUTO(validTokens);
25         TESTCASE_AUTO(invalidTokens);
26         TESTCASE_AUTO(unknownTokens);
27         TESTCASE_AUTO(unexpectedTokens);
28         TESTCASE_AUTO(duplicateValues);
29         TESTCASE_AUTO(stemsRequiringOption);
30         TESTCASE_AUTO(defaultTokens);
31         TESTCASE_AUTO(flexibleSeparators);
32         TESTCASE_AUTO(wildcardCharacters);
33         TESTCASE_AUTO(perUnitInArabic);
34         TESTCASE_AUTO(perUnitToSkeleton);
35     TESTCASE_AUTO_END;
36 }
37 
validTokens()38 void NumberSkeletonTest::validTokens() {
39     IcuTestErrorCode status(*this, "validTokens");
40 
41     // This tests only if the tokens are valid, not their behavior.
42     // Most of these are from the design doc.
43     static const char16_t* cases[] = {
44             u"precision-integer",
45             u"precision-unlimited",
46             u"@@@##",
47             u"@@*",
48             u"@@+",
49             u"@@+/w",
50             u".000##",
51             u".00*",
52             u".00+",
53             u".",
54             u"./w",
55             u".*",
56             u".+",
57             u".+/w",
58             u".######",
59             u".00/@@*",
60             u".00/@@+",
61             u".00/@##",
62             u".00/@##/w",
63             u".00/@",
64             u".00/@r",
65             u".00/@@s",
66             u".00/@@#r",
67             u"precision-increment/3.14",
68             u"precision-increment/3.14/w",
69             u"precision-currency-standard",
70             u"precision-currency-standard/w",
71             u"precision-integer rounding-mode-half-up",
72             u".00# rounding-mode-ceiling",
73             u".00/@@* rounding-mode-floor",
74             u".00/@@+ rounding-mode-floor",
75             u"scientific",
76             u"scientific/*ee",
77             u"scientific/+ee",
78             u"scientific/sign-always",
79             u"scientific/*ee/sign-always",
80             u"scientific/+ee/sign-always",
81             u"scientific/sign-always/*ee",
82             u"scientific/sign-always/+ee",
83             u"scientific/sign-except-zero",
84             u"engineering",
85             u"engineering/*eee",
86             u"engineering/+eee",
87             u"compact-short",
88             u"compact-long",
89             u"notation-simple",
90             u"percent",
91             u"permille",
92             u"measure-unit/length-meter",
93             u"measure-unit/area-square-meter",
94             u"measure-unit/energy-joule per-measure-unit/length-meter",
95             u"unit/square-meter-per-square-meter",
96             u"currency/XXX",
97             u"currency/ZZZ",
98             u"currency/usd",
99             u"group-off",
100             u"group-min2",
101             u"group-auto",
102             u"group-on-aligned",
103             u"group-thousands",
104             u"integer-width/00",
105             u"integer-width/#0",
106             u"integer-width/*00",
107             u"integer-width/+00",
108             u"sign-always",
109             u"sign-auto",
110             u"sign-never",
111             u"sign-accounting",
112             u"sign-accounting-always",
113             u"sign-except-zero",
114             u"sign-accounting-except-zero",
115             u"unit-width-narrow",
116             u"unit-width-short",
117             u"unit-width-iso-code",
118             u"unit-width-full-name",
119             u"unit-width-hidden",
120             u"decimal-auto",
121             u"decimal-always",
122             u"scale/5.2",
123             u"scale/-5.2",
124             u"scale/100",
125             u"scale/1E2",
126             u"scale/1",
127             u"latin",
128             u"numbering-system/arab",
129             u"numbering-system/latn",
130             u"precision-integer/@##",
131             u"precision-integer rounding-mode-ceiling",
132             u"precision-currency-cash rounding-mode-ceiling",
133             u"0",
134             u"00",
135             u"000",
136             u"E0",
137             u"E00",
138             u"E000",
139             u"EE0",
140             u"EE00",
141             u"EE+?0",
142             u"EE+?00",
143             u"EE+!0",
144             u"EE+!00",
145     };
146 
147     for (auto& cas : cases) {
148         UnicodeString skeletonString(cas);
149         status.setScope(skeletonString);
150         UParseError perror;
151         NumberFormatter::forSkeleton(skeletonString, perror, status);
152         assertSuccess(CStr(skeletonString)(), status, true);
153         assertEquals(skeletonString, -1, perror.offset);
154         status.errIfFailureAndReset();
155     }
156 }
157 
invalidTokens()158 void NumberSkeletonTest::invalidTokens() {
159     static const char16_t* cases[] = {
160             u".00x",
161             u".00i",
162             u".00/x",
163             u".00/ww",
164             u".00##0",
165             u".##*",
166             u".00##*",
167             u".0#*",
168             u"@#*",
169             u".##+",
170             u".00##+",
171             u".0#+",
172             u"@#+",
173             u"@@x",
174             u"@@##0",
175             u".00/@@",
176             u".00/@@x",
177             u".00/@@#",
178             u".00/@@#*",
179             u".00/floor/@@*", // wrong order
180             u".00/@@#+",
181             u".00/@@@+r",
182             u".00/floor/@@+", // wrong order
183             u"precision-increment/français", // non-invariant characters for C++
184             u"scientific/ee",
185             u"precision-increment/xxx",
186             u"precision-increment/NaN",
187             u"precision-increment/0.1.2",
188             u"scale/xxx",
189             u"scale/NaN",
190             u"scale/0.1.2",
191             u"scale/français", // non-invariant characters for C++
192             u"currency/dummy",
193             u"currency/ççç", // three characters but not ASCII
194             u"measure-unit/foo",
195             u"integer-width/xxx",
196             u"integer-width/0*",
197             u"integer-width/*0#",
198             u"integer-width/*#",
199             u"integer-width/*#0",
200             u"integer-width/0+",
201             u"integer-width/+0#",
202             u"integer-width/+#",
203             u"integer-width/+#0",
204             u"scientific/foo",
205             u"E",
206             u"E1",
207             u"E+",
208             u"E+?",
209             u"E+!",
210             u"E+0",
211             u"EE",
212             u"EE+",
213             u"EEE",
214             u"EEE0",
215             u"001",
216             u"00*",
217             u"00+",
218     };
219 
220     expectedErrorSkeleton(cases, UPRV_LENGTHOF(cases));
221 }
222 
unknownTokens()223 void NumberSkeletonTest::unknownTokens() {
224     static const char16_t* cases[] = {
225             u"maesure-unit",
226             u"measure-unit/foo-bar",
227             u"numbering-system/dummy",
228             u"français",
229             u"measure-unit/français-français", // non-invariant characters for C++
230             u"numbering-system/français", // non-invariant characters for C++
231             u"currency-USD"};
232 
233     expectedErrorSkeleton(cases, UPRV_LENGTHOF(cases));
234 }
235 
unexpectedTokens()236 void NumberSkeletonTest::unexpectedTokens() {
237     static const char16_t* cases[] = {
238             u".00/w/w",
239             u"group-thousands/foo",
240             u"precision-integer//@## group-off",
241             u"precision-integer//@##  group-off",
242             u"precision-integer/ group-off",
243             u"precision-integer// group-off"};
244 
245     expectedErrorSkeleton(cases, UPRV_LENGTHOF(cases));
246 }
247 
duplicateValues()248 void NumberSkeletonTest::duplicateValues() {
249     static const char16_t* cases[] = {
250             u"precision-integer precision-integer",
251             u"precision-integer .00+",
252             u"precision-integer precision-unlimited",
253             u"precision-integer @@@",
254             u"scientific engineering",
255             u"engineering compact-long",
256             u"sign-auto sign-always"};
257 
258     expectedErrorSkeleton(cases, UPRV_LENGTHOF(cases));
259 }
260 
stemsRequiringOption()261 void NumberSkeletonTest::stemsRequiringOption() {
262     static const char16_t* stems[] = {
263             u"precision-increment",
264             u"measure-unit",
265             u"per-measure-unit",
266             u"currency",
267             u"integer-width",
268             u"numbering-system",
269             u"scale"};
270     static const char16_t* suffixes[] = {u"", u"/@##", u" scientific", u"/@## scientific"};
271 
272     for (auto& stem : stems) {
273         for (auto& suffix : suffixes) {
274             UnicodeString skeletonString = UnicodeString(stem) + suffix;
275             UErrorCode status = U_ZERO_ERROR;
276             UParseError perror;
277             NumberFormatter::forSkeleton(skeletonString, perror, status);
278             assertEquals(skeletonString, U_NUMBER_SKELETON_SYNTAX_ERROR, status);
279 
280             // Check the UParseError for integrity.
281             // If an option is present, the option is wrong; error offset is at the start of the option
282             // If an option is not present, the error offset is at the token separator (end of stem)
283             int32_t expectedOffset = u_strlen(stem) + ((suffix[0] == u'/') ? 1 : 0);
284             assertEquals(skeletonString, expectedOffset, perror.offset);
285             UnicodeString expectedPreContext = skeletonString.tempSubString(0, expectedOffset);
286             if (expectedPreContext.length() >= U_PARSE_CONTEXT_LEN - 1) {
287                 expectedPreContext = expectedPreContext.tempSubString(expectedOffset - U_PARSE_CONTEXT_LEN + 1);
288             }
289             assertEquals(skeletonString, expectedPreContext, perror.preContext);
290             UnicodeString expectedPostContext = skeletonString.tempSubString(expectedOffset);
291             // None of the postContext strings in this test exceed U_PARSE_CONTEXT_LEN
292             assertEquals(skeletonString, expectedPostContext, perror.postContext);
293         }
294     }
295 }
296 
defaultTokens()297 void NumberSkeletonTest::defaultTokens() {
298     IcuTestErrorCode status(*this, "defaultTokens");
299 
300     static const char16_t* cases[] = {
301             u"notation-simple",
302             u"base-unit",
303             u"group-auto",
304             u"integer-width/+0",
305             u"sign-auto",
306             u"unit-width-short",
307             u"decimal-auto"};
308 
309     for (auto& cas : cases) {
310         UnicodeString skeletonString(cas);
311         status.setScope(skeletonString);
312         UnicodeString normalized = NumberFormatter::forSkeleton(
313                 skeletonString, status).toSkeleton(status);
314         // Skeleton should become empty when normalized
315         assertEquals(skeletonString, u"", normalized);
316         status.errIfFailureAndReset();
317     }
318 }
319 
flexibleSeparators()320 void NumberSkeletonTest::flexibleSeparators() {
321     IcuTestErrorCode status(*this, "flexibleSeparators");
322 
323     static struct TestCase {
324         const char16_t* skeleton;
325         const char16_t* expected;
326     } cases[] = {{u"precision-integer group-off", u"5142"},
327                  {u"precision-integer  group-off", u"5142"},
328                  {u"precision-integer/@## group-off", u"5140"},
329                  {u"precision-integer/@##  group-off", u"5140"}};
330 
331     for (auto& cas : cases) {
332         UnicodeString skeletonString(cas.skeleton);
333         UnicodeString expected(cas.expected);
334         status.setScope(skeletonString);
335         UnicodeString actual = NumberFormatter::forSkeleton(skeletonString, status).locale("en")
336                                .formatDouble(5142.3, status)
337                                .toString(status);
338         if (!status.errDataIfFailureAndReset()) {
339             assertEquals(skeletonString, expected, actual);
340         }
341         status.errIfFailureAndReset();
342     }
343 }
344 
wildcardCharacters()345 void NumberSkeletonTest::wildcardCharacters() {
346     IcuTestErrorCode status(*this, "wildcardCharacters");
347 
348     struct TestCase {
349         const char16_t* star;
350         const char16_t* plus;
351     } cases[] = {
352         { u".00*", u".00+" },
353         { u"@@*", u"@@+" },
354         { u"scientific/*ee", u"scientific/+ee" },
355         { u"integer-width/*00", u"integer-width/+00" },
356     };
357 
358     for (const auto& cas : cases) {
359         UnicodeString star(cas.star);
360         UnicodeString plus(cas.plus);
361         status.setScope(star);
362 
363         UnicodeString normalized = NumberFormatter::forSkeleton(plus, status)
364             .toSkeleton(status);
365         assertEquals("Plus should normalize to star", star, normalized);
366         status.errIfFailureAndReset();
367     }
368 }
369 
370 // In C++, there is no distinguishing between "invalid", "unknown", and "unexpected" tokens.
expectedErrorSkeleton(const char16_t ** cases,int32_t casesLen)371 void NumberSkeletonTest::expectedErrorSkeleton(const char16_t** cases, int32_t casesLen) {
372     for (int32_t i = 0; i < casesLen; i++) {
373         UnicodeString skeletonString(cases[i]);
374         UErrorCode status = U_ZERO_ERROR;
375         NumberFormatter::forSkeleton(skeletonString, status);
376         assertEquals(skeletonString, U_NUMBER_SKELETON_SYNTAX_ERROR, status);
377     }
378 }
379 
perUnitInArabic()380 void NumberSkeletonTest::perUnitInArabic() {
381     IcuTestErrorCode status(*this, "perUnitInArabic");
382 
383     struct TestCase {
384         const char16_t* type;
385         const char16_t* subtype;
386     } cases[] = {
387         {u"area", u"acre"},
388         {u"digital", u"bit"},
389         {u"digital", u"byte"},
390         {u"temperature", u"celsius"},
391         {u"length", u"centimeter"},
392         {u"duration", u"day"},
393         {u"angle", u"degree"},
394         {u"temperature", u"fahrenheit"},
395         {u"volume", u"fluid-ounce"},
396         {u"length", u"foot"},
397         {u"volume", u"gallon"},
398         {u"digital", u"gigabit"},
399         {u"digital", u"gigabyte"},
400         {u"mass", u"gram"},
401         {u"area", u"hectare"},
402         {u"duration", u"hour"},
403         {u"length", u"inch"},
404         {u"digital", u"kilobit"},
405         {u"digital", u"kilobyte"},
406         {u"mass", u"kilogram"},
407         {u"length", u"kilometer"},
408         {u"volume", u"liter"},
409         {u"digital", u"megabit"},
410         {u"digital", u"megabyte"},
411         {u"length", u"meter"},
412         {u"length", u"mile"},
413         {u"length", u"mile-scandinavian"},
414         {u"volume", u"milliliter"},
415         {u"length", u"millimeter"},
416         {u"duration", u"millisecond"},
417         {u"duration", u"minute"},
418         {u"duration", u"month"},
419         {u"mass", u"ounce"},
420         {u"concentr", u"percent"},
421         {u"digital", u"petabyte"},
422         {u"mass", u"pound"},
423         {u"duration", u"second"},
424         {u"mass", u"stone"},
425         {u"digital", u"terabit"},
426         {u"digital", u"terabyte"},
427         {u"duration", u"week"},
428         {u"length", u"yard"},
429         {u"duration", u"year"},
430     };
431 
432     for (const auto& cas1 : cases) {
433         for (const auto& cas2 : cases) {
434             UnicodeString skeleton(u"measure-unit/");
435             skeleton += cas1.type;
436             skeleton += u"-";
437             skeleton += cas1.subtype;
438             skeleton += u" ";
439             skeleton += u"per-measure-unit/";
440             skeleton += cas2.type;
441             skeleton += u"-";
442             skeleton += cas2.subtype;
443 
444             status.setScope(skeleton);
445             UnicodeString actual = NumberFormatter::forSkeleton(skeleton, status).locale("ar")
446                                    .formatDouble(5142.3, status)
447                                    .toString(status);
448             status.errIfFailureAndReset();
449         }
450     }
451 }
452 
perUnitToSkeleton()453 void NumberSkeletonTest::perUnitToSkeleton() {
454     IcuTestErrorCode status(*this, "perUnitToSkeleton");
455     struct TestCase {
456         const char16_t* type;
457         const char16_t* subtype;
458     } cases[] = {
459         {u"area", u"acre"},
460         {u"concentr", u"percent"},
461         {u"concentr", u"permille"},
462         {u"concentr", u"permillion"},
463         {u"concentr", u"permyriad"},
464         {u"digital", u"bit"},
465         {u"length", u"yard"},
466     };
467 
468     for (const auto& cas1 : cases) {
469         for (const auto& cas2 : cases) {
470             UnicodeString skeleton(u"measure-unit/");
471             skeleton += cas1.type;
472             skeleton += u"-";
473             skeleton += cas1.subtype;
474             skeleton += u" ";
475             skeleton += u"per-measure-unit/";
476             skeleton += cas2.type;
477             skeleton += u"-";
478             skeleton += cas2.subtype;
479 
480             status.setScope(skeleton);
481             if (cas1.type != cas2.type && cas1.subtype != cas2.subtype) {
482                 UnicodeString toSkeleton = NumberFormatter::forSkeleton(
483                     skeleton, status).toSkeleton(status);
484                 if (status.errIfFailureAndReset()) {
485                     continue;
486                 }
487                 // Ensure both subtype are in the toSkeleton.
488                 UnicodeString msg;
489                 msg.append(toSkeleton)
490                     .append(" should contain '")
491                     .append(UnicodeString(cas1.subtype))
492                     .append("' when constructed from ")
493                     .append(skeleton);
494                 assertTrue(msg, toSkeleton.indexOf(cas1.subtype) >= 0);
495 
496                 msg.remove();
497                 msg.append(toSkeleton)
498                     .append(" should contain '")
499                     .append(UnicodeString(cas2.subtype))
500                     .append("' when constructed from ")
501                     .append(skeleton);
502                 assertTrue(msg, toSkeleton.indexOf(cas2.subtype) >= 0);
503             }
504         }
505     }
506 }
507 
508 #endif /* #if !UCONFIG_NO_FORMATTING */
509