]>
Commit | Line | Data |
---|---|---|
b75a7d8f A |
1 | /* |
2 | ******************************************************************************* | |
3 | * | |
f3c0d7a5 A |
4 | * © 2016 and later: Unicode, Inc. and others. |
5 | * License & terms of use: http://www.unicode.org/copyright.html#License | |
6 | * | |
7 | ******************************************************************************* | |
8 | ******************************************************************************* | |
9 | * | |
b331163b | 10 | * Copyright (C) 2000-2014, International Business Machines |
b75a7d8f A |
11 | * Corporation and others. All Rights Reserved. |
12 | * | |
13 | ******************************************************************************* | |
14 | * file name: ustring.c | |
f3c0d7a5 | 15 | * encoding: UTF-8 |
b75a7d8f A |
16 | * tab size: 8 (not used) |
17 | * indentation:4 | |
18 | * | |
19 | * created on: 2000aug15 | |
20 | * created by: Markus W. Scherer | |
21 | * | |
22 | * This file contains sample code that illustrates the use of Unicode strings | |
23 | * with ICU. | |
24 | */ | |
25 | ||
26 | #include <stdio.h> | |
27 | #include "unicode/utypes.h" | |
28 | #include "unicode/uchar.h" | |
29 | #include "unicode/locid.h" | |
30 | #include "unicode/ustring.h" | |
31 | #include "unicode/ucnv.h" | |
32 | #include "unicode/unistr.h" | |
33 | ||
f3c0d7a5 A |
34 | #ifndef UPRV_LENGTHOF |
35 | #define UPRV_LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0])) | |
36 | #endif | |
37 | ||
b75a7d8f A |
38 | // helper functions -------------------------------------------------------- *** |
39 | ||
40 | // default converter for the platform encoding | |
41 | static UConverter *cnv=NULL; | |
42 | ||
43 | static void | |
44 | printUString(const char *announce, const UChar *s, int32_t length) { | |
45 | static char out[200]; | |
46 | UChar32 c; | |
47 | int32_t i; | |
48 | UErrorCode errorCode=U_ZERO_ERROR; | |
49 | ||
50 | /* | |
51 | * Convert to the "platform encoding". See notes in printUnicodeString(). | |
52 | * ucnv_fromUChars(), like most ICU APIs understands length==-1 | |
53 | * to mean that the string is NUL-terminated. | |
54 | */ | |
55 | ucnv_fromUChars(cnv, out, sizeof(out), s, length, &errorCode); | |
56 | if(U_FAILURE(errorCode) || errorCode==U_STRING_NOT_TERMINATED_WARNING) { | |
57 | printf("%sproblem converting string from Unicode: %s\n", announce, u_errorName(errorCode)); | |
58 | return; | |
59 | } | |
60 | ||
61 | printf("%s%s {", announce, out); | |
62 | ||
63 | /* output the code points (not code units) */ | |
64 | if(length>=0) { | |
65 | /* s is not NUL-terminated */ | |
66 | for(i=0; i<length; /* U16_NEXT post-increments */) { | |
67 | U16_NEXT(s, i, length, c); | |
68 | printf(" %04x", c); | |
69 | } | |
70 | } else { | |
71 | /* s is NUL-terminated */ | |
72 | for(i=0; /* condition in loop body */; /* U16_NEXT post-increments */) { | |
73 | U16_NEXT(s, i, length, c); | |
74 | if(c==0) { | |
75 | break; | |
76 | } | |
77 | printf(" %04x", c); | |
78 | } | |
79 | } | |
80 | printf(" }\n"); | |
81 | } | |
82 | ||
83 | static void | |
84 | printUnicodeString(const char *announce, const UnicodeString &s) { | |
85 | static char out[200]; | |
86 | int32_t i, length; | |
87 | ||
88 | // output the string, converted to the platform encoding | |
89 | ||
90 | // Note for Windows: The "platform encoding" defaults to the "ANSI codepage", | |
91 | // which is different from the "OEM codepage" in the console window. | |
92 | // However, if you pipe the output into a file and look at it with Notepad | |
93 | // or similar, then "ANSI" characters will show correctly. | |
94 | // Production code should be aware of what encoding is required, | |
95 | // and use a UConverter or at least a charset name explicitly. | |
96 | out[s.extract(0, 99, out)]=0; | |
97 | printf("%s%s {", announce, out); | |
98 | ||
99 | // output the code units (not code points) | |
100 | length=s.length(); | |
101 | for(i=0; i<length; ++i) { | |
102 | printf(" %04x", s.charAt(i)); | |
103 | } | |
104 | printf(" }\n"); | |
105 | } | |
106 | ||
107 | // sample code for utf.h macros -------------------------------------------- *** | |
108 | ||
109 | static void | |
110 | demo_utf_h_macros() { | |
111 | static UChar input[]={ 0x0061, 0xd800, 0xdc00, 0xdbff, 0xdfff, 0x0062 }; | |
112 | UChar32 c; | |
113 | int32_t i; | |
114 | UBool isError; | |
115 | ||
116 | printf("\n* demo_utf_h_macros() -------------- ***\n\n"); | |
117 | ||
b331163b A |
118 | printUString("iterate forward through: ", input, UPRV_LENGTHOF(input)); |
119 | for(i=0; i<UPRV_LENGTHOF(input); /* U16_NEXT post-increments */) { | |
b75a7d8f A |
120 | /* Iterating forwards |
121 | Codepoint at offset 0: U+0061 | |
122 | Codepoint at offset 1: U+10000 | |
123 | Codepoint at offset 3: U+10ffff | |
124 | Codepoint at offset 5: U+0062 | |
125 | */ | |
126 | printf("Codepoint at offset %d: U+", i); | |
b331163b | 127 | U16_NEXT(input, i, UPRV_LENGTHOF(input), c); |
b75a7d8f A |
128 | printf("%04x\n", c); |
129 | } | |
130 | ||
131 | puts(""); | |
132 | ||
133 | isError=FALSE; | |
134 | i=1; /* write position, gets post-incremented so needs to be in an l-value */ | |
b331163b | 135 | U16_APPEND(input, i, UPRV_LENGTHOF(input), 0x0062, isError); |
b75a7d8f | 136 | |
b331163b A |
137 | printUString("iterate backward through: ", input, UPRV_LENGTHOF(input)); |
138 | for(i=UPRV_LENGTHOF(input); i>0; /* U16_PREV pre-decrements */) { | |
b75a7d8f A |
139 | U16_PREV(input, 0, i, c); |
140 | /* Iterating backwards | |
141 | Codepoint at offset 5: U+0062 | |
142 | Codepoint at offset 3: U+10ffff | |
143 | Codepoint at offset 2: U+dc00 -- unpaired surrogate because lead surr. overwritten | |
144 | Codepoint at offset 1: U+0062 -- by this BMP code point | |
145 | Codepoint at offset 0: U+0061 | |
146 | */ | |
147 | printf("Codepoint at offset %d: U+%04x\n", i, c); | |
148 | } | |
149 | } | |
150 | ||
151 | // sample code for Unicode strings in C ------------------------------------ *** | |
152 | ||
153 | static void demo_C_Unicode_strings() { | |
154 | printf("\n* demo_C_Unicode_strings() --------- ***\n\n"); | |
155 | ||
156 | static const UChar text[]={ 0x41, 0x42, 0x43, 0 }; /* "ABC" */ | |
157 | static const UChar appendText[]={ 0x61, 0x62, 0x63, 0 }; /* "abc" */ | |
158 | static const UChar cmpText[]={ 0x61, 0x53, 0x73, 0x43, 0 }; /* "aSsC" */ | |
159 | UChar buffer[32]; | |
160 | int32_t compare; | |
161 | int32_t length=u_strlen(text); /* length=3 */ | |
162 | ||
163 | /* simple ANSI C-style functions */ | |
164 | buffer[0]=0; /* empty, NUL-terminated string */ | |
165 | u_strncat(buffer, text, 1); /* append just n=1 character ('A') */ | |
166 | u_strcat(buffer, appendText); /* buffer=="Aabc" */ | |
167 | length=u_strlen(buffer); /* length=4 */ | |
168 | printUString("should be \"Aabc\": ", buffer, -1); | |
169 | ||
170 | /* bitwise comparing buffer with text */ | |
171 | compare=u_strcmp(buffer, text); | |
172 | if(compare<=0) { | |
173 | printf("String comparison error, expected \"Aabc\" > \"ABC\"\n"); | |
174 | } | |
175 | ||
176 | /* Build "A<sharp s>C" in the buffer... */ | |
177 | u_strcpy(buffer, text); | |
178 | buffer[1]=0xdf; /* sharp s, case-compares equal to "ss" */ | |
179 | printUString("should be \"A<sharp s>C\": ", buffer, -1); | |
180 | ||
181 | /* Compare two strings case-insensitively using full case folding */ | |
182 | compare=u_strcasecmp(buffer, cmpText, U_FOLD_CASE_DEFAULT); | |
183 | if(compare!=0) { | |
184 | printf("String case insensitive comparison error, expected \"AbC\" to be equal to \"ABC\"\n"); | |
185 | } | |
186 | } | |
187 | ||
188 | // sample code for case mappings with C APIs -------------------------------- *** | |
189 | ||
190 | static void demoCaseMapInC() { | |
191 | /* | |
192 | * input= | |
193 | * "aB<capital sigma>" | |
194 | * "iI<small dotless i><capital dotted I> " | |
195 | * "<sharp s> <small lig. ffi>" | |
196 | * "<small final sigma><small sigma><capital sigma>" | |
197 | */ | |
198 | static const UChar input[]={ | |
199 | 0x61, 0x42, 0x3a3, | |
200 | 0x69, 0x49, 0x131, 0x130, 0x20, | |
201 | 0xdf, 0x20, 0xfb03, | |
202 | 0x3c2, 0x3c3, 0x3a3, 0 | |
203 | }; | |
204 | UChar buffer[32]; | |
205 | ||
206 | UErrorCode errorCode; | |
207 | UChar32 c; | |
208 | int32_t i, j, length; | |
209 | UBool isError; | |
210 | ||
211 | printf("\n* demoCaseMapInC() ----------------- ***\n\n"); | |
212 | ||
213 | /* | |
214 | * First, use simple case mapping functions which provide | |
215 | * 1:1 code point mappings without context/locale ID. | |
216 | * | |
217 | * Note that some mappings will not be "right" because some "real" | |
218 | * case mappings require context, depend on the locale ID, | |
219 | * and/or result in a change in the number of code points. | |
220 | */ | |
221 | printUString("input string: ", input, -1); | |
222 | ||
223 | /* uppercase */ | |
224 | isError=FALSE; | |
b331163b | 225 | for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) { |
b75a7d8f A |
226 | U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */ |
227 | if(c==0) { | |
228 | break; /* stop at terminating NUL, no need to terminate buffer */ | |
229 | } | |
230 | c=u_toupper(c); | |
b331163b | 231 | U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError); |
b75a7d8f A |
232 | } |
233 | printUString("simple-uppercased: ", buffer, j); | |
234 | /* lowercase */ | |
235 | isError=FALSE; | |
b331163b | 236 | for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) { |
b75a7d8f A |
237 | U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */ |
238 | if(c==0) { | |
239 | break; /* stop at terminating NUL, no need to terminate buffer */ | |
240 | } | |
241 | c=u_tolower(c); | |
b331163b | 242 | U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError); |
b75a7d8f A |
243 | } |
244 | printUString("simple-lowercased: ", buffer, j); | |
245 | /* titlecase */ | |
246 | isError=FALSE; | |
b331163b | 247 | for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) { |
b75a7d8f A |
248 | U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */ |
249 | if(c==0) { | |
250 | break; /* stop at terminating NUL, no need to terminate buffer */ | |
251 | } | |
252 | c=u_totitle(c); | |
b331163b | 253 | U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError); |
b75a7d8f A |
254 | } |
255 | printUString("simple-titlecased: ", buffer, j); | |
256 | /* case-fold/default */ | |
257 | isError=FALSE; | |
b331163b | 258 | for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) { |
b75a7d8f A |
259 | U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */ |
260 | if(c==0) { | |
261 | break; /* stop at terminating NUL, no need to terminate buffer */ | |
262 | } | |
263 | c=u_foldCase(c, U_FOLD_CASE_DEFAULT); | |
b331163b | 264 | U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError); |
b75a7d8f A |
265 | } |
266 | printUString("simple-case-folded/default: ", buffer, j); | |
267 | /* case-fold/Turkic */ | |
268 | isError=FALSE; | |
b331163b | 269 | for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) { |
b75a7d8f A |
270 | U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */ |
271 | if(c==0) { | |
272 | break; /* stop at terminating NUL, no need to terminate buffer */ | |
273 | } | |
274 | c=u_foldCase(c, U_FOLD_CASE_EXCLUDE_SPECIAL_I); | |
b331163b | 275 | U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError); |
b75a7d8f A |
276 | } |
277 | printUString("simple-case-folded/Turkic: ", buffer, j); | |
278 | ||
279 | /* | |
280 | * Second, use full case mapping functions which provide | |
281 | * 1:n code point mappings (n can be 0!) and are sensitive to context and locale ID. | |
282 | * | |
283 | * Note that lower/upper/titlecasing take a locale ID while case-folding | |
284 | * has bit flag options instead, by design of the Unicode SpecialCasing.txt UCD file. | |
285 | * | |
286 | * Also, string titlecasing requires a BreakIterator to find starts of words. | |
287 | * The sample code here passes in a NULL pointer; u_strToTitle() will open and close a default | |
288 | * titlecasing BreakIterator automatically. | |
289 | * For production code where many strings are titlecased it would be more efficient | |
290 | * to open a BreakIterator externally and pass it in. | |
291 | */ | |
292 | printUString("\ninput string: ", input, -1); | |
293 | ||
294 | /* lowercase/English */ | |
295 | errorCode=U_ZERO_ERROR; | |
b331163b | 296 | length=u_strToLower(buffer, UPRV_LENGTHOF(buffer), input, -1, "en", &errorCode); |
b75a7d8f A |
297 | if(U_SUCCESS(errorCode)) { |
298 | printUString("full-lowercased/en: ", buffer, length); | |
299 | } else { | |
300 | printf("error in u_strToLower(en)=%ld error=%s\n", length, u_errorName(errorCode)); | |
301 | } | |
302 | /* lowercase/Turkish */ | |
303 | errorCode=U_ZERO_ERROR; | |
b331163b | 304 | length=u_strToLower(buffer, UPRV_LENGTHOF(buffer), input, -1, "tr", &errorCode); |
b75a7d8f A |
305 | if(U_SUCCESS(errorCode)) { |
306 | printUString("full-lowercased/tr: ", buffer, length); | |
307 | } else { | |
308 | printf("error in u_strToLower(tr)=%ld error=%s\n", length, u_errorName(errorCode)); | |
309 | } | |
310 | /* uppercase/English */ | |
311 | errorCode=U_ZERO_ERROR; | |
b331163b | 312 | length=u_strToUpper(buffer, UPRV_LENGTHOF(buffer), input, -1, "en", &errorCode); |
b75a7d8f A |
313 | if(U_SUCCESS(errorCode)) { |
314 | printUString("full-uppercased/en: ", buffer, length); | |
315 | } else { | |
316 | printf("error in u_strToUpper(en)=%ld error=%s\n", length, u_errorName(errorCode)); | |
317 | } | |
318 | /* uppercase/Turkish */ | |
319 | errorCode=U_ZERO_ERROR; | |
b331163b | 320 | length=u_strToUpper(buffer, UPRV_LENGTHOF(buffer), input, -1, "tr", &errorCode); |
b75a7d8f A |
321 | if(U_SUCCESS(errorCode)) { |
322 | printUString("full-uppercased/tr: ", buffer, length); | |
323 | } else { | |
324 | printf("error in u_strToUpper(tr)=%ld error=%s\n", length, u_errorName(errorCode)); | |
325 | } | |
326 | /* titlecase/English */ | |
327 | errorCode=U_ZERO_ERROR; | |
b331163b | 328 | length=u_strToTitle(buffer, UPRV_LENGTHOF(buffer), input, -1, NULL, "en", &errorCode); |
b75a7d8f A |
329 | if(U_SUCCESS(errorCode)) { |
330 | printUString("full-titlecased/en: ", buffer, length); | |
331 | } else { | |
332 | printf("error in u_strToTitle(en)=%ld error=%s\n", length, u_errorName(errorCode)); | |
333 | } | |
334 | /* titlecase/Turkish */ | |
335 | errorCode=U_ZERO_ERROR; | |
b331163b | 336 | length=u_strToTitle(buffer, UPRV_LENGTHOF(buffer), input, -1, NULL, "tr", &errorCode); |
b75a7d8f A |
337 | if(U_SUCCESS(errorCode)) { |
338 | printUString("full-titlecased/tr: ", buffer, length); | |
339 | } else { | |
340 | printf("error in u_strToTitle(tr)=%ld error=%s\n", length, u_errorName(errorCode)); | |
341 | } | |
342 | /* case-fold/default */ | |
343 | errorCode=U_ZERO_ERROR; | |
b331163b | 344 | length=u_strFoldCase(buffer, UPRV_LENGTHOF(buffer), input, -1, U_FOLD_CASE_DEFAULT, &errorCode); |
b75a7d8f A |
345 | if(U_SUCCESS(errorCode)) { |
346 | printUString("full-case-folded/default: ", buffer, length); | |
347 | } else { | |
348 | printf("error in u_strFoldCase(default)=%ld error=%s\n", length, u_errorName(errorCode)); | |
349 | } | |
350 | /* case-fold/Turkic */ | |
351 | errorCode=U_ZERO_ERROR; | |
b331163b | 352 | length=u_strFoldCase(buffer, UPRV_LENGTHOF(buffer), input, -1, U_FOLD_CASE_EXCLUDE_SPECIAL_I, &errorCode); |
b75a7d8f A |
353 | if(U_SUCCESS(errorCode)) { |
354 | printUString("full-case-folded/Turkic: ", buffer, length); | |
355 | } else { | |
356 | printf("error in u_strFoldCase(Turkic)=%ld error=%s\n", length, u_errorName(errorCode)); | |
357 | } | |
358 | } | |
359 | ||
360 | // sample code for case mappings with C++ APIs ------------------------------ *** | |
361 | ||
362 | static void demoCaseMapInCPlusPlus() { | |
363 | /* | |
364 | * input= | |
365 | * "aB<capital sigma>" | |
366 | * "iI<small dotless i><capital dotted I> " | |
367 | * "<sharp s> <small lig. ffi>" | |
368 | * "<small final sigma><small sigma><capital sigma>" | |
369 | */ | |
370 | static const UChar input[]={ | |
371 | 0x61, 0x42, 0x3a3, | |
372 | 0x69, 0x49, 0x131, 0x130, 0x20, | |
373 | 0xdf, 0x20, 0xfb03, | |
374 | 0x3c2, 0x3c3, 0x3a3, 0 | |
375 | }; | |
376 | ||
377 | printf("\n* demoCaseMapInCPlusPlus() --------- ***\n\n"); | |
378 | ||
379 | UnicodeString s(input), t; | |
380 | const Locale &en=Locale::getEnglish(); | |
381 | Locale tr("tr"); | |
382 | ||
383 | /* | |
384 | * Full case mappings as in demoCaseMapInC(), using UnicodeString functions. | |
385 | * These functions modify the string object itself. | |
386 | * Since we want to keep the input string around, we copy it each time | |
387 | * and case-map the copy. | |
388 | */ | |
389 | printUnicodeString("input string: ", s); | |
390 | ||
391 | /* lowercase/English */ | |
392 | printUnicodeString("full-lowercased/en: ", (t=s).toLower(en)); | |
393 | /* lowercase/Turkish */ | |
394 | printUnicodeString("full-lowercased/tr: ", (t=s).toLower(tr)); | |
395 | /* uppercase/English */ | |
396 | printUnicodeString("full-uppercased/en: ", (t=s).toUpper(en)); | |
397 | /* uppercase/Turkish */ | |
398 | printUnicodeString("full-uppercased/tr: ", (t=s).toUpper(tr)); | |
399 | /* titlecase/English */ | |
400 | printUnicodeString("full-titlecased/en: ", (t=s).toTitle(NULL, en)); | |
401 | /* titlecase/Turkish */ | |
402 | printUnicodeString("full-titlecased/tr: ", (t=s).toTitle(NULL, tr)); | |
403 | /* case-folde/default */ | |
404 | printUnicodeString("full-case-folded/default: ", (t=s).foldCase(U_FOLD_CASE_DEFAULT)); | |
405 | /* case-folde/Turkic */ | |
406 | printUnicodeString("full-case-folded/Turkic: ", (t=s).foldCase(U_FOLD_CASE_EXCLUDE_SPECIAL_I)); | |
407 | } | |
408 | ||
409 | // sample code for UnicodeString storage models ----------------------------- *** | |
410 | ||
411 | static const UChar readonly[]={ | |
412 | 0x61, 0x31, 0x20ac | |
413 | }; | |
414 | static UChar writeable[]={ | |
415 | 0x62, 0x32, 0xdbc0, 0xdc01 // includes a surrogate pair for a supplementary code point | |
416 | }; | |
417 | static char out[100]; | |
418 | ||
419 | static void | |
420 | demoUnicodeStringStorage() { | |
421 | // These sample code lines illustrate how to use UnicodeString, and the | |
422 | // comments tell what happens internally. There are no APIs to observe | |
423 | // most of this programmatically, except for stepping into the code | |
424 | // with a debugger. | |
425 | // This is by design to hide such details from the user. | |
426 | int32_t i; | |
427 | ||
428 | printf("\n* demoUnicodeStringStorage() ------- ***\n\n"); | |
429 | ||
430 | // * UnicodeString with internally stored contents | |
431 | // instantiate a UnicodeString from a single code point | |
432 | // the few (2) UChars will be stored in the object itself | |
433 | UnicodeString one((UChar32)0x24001); | |
434 | // this copies the few UChars into the "two" object | |
435 | UnicodeString two=one; | |
436 | printf("length of short string copy: %d\n", two.length()); | |
437 | // set "one" to contain the 3 UChars from readonly | |
438 | // this setTo() variant copies the characters | |
b331163b | 439 | one.setTo(readonly, UPRV_LENGTHOF(readonly)); |
b75a7d8f A |
440 | |
441 | // * UnicodeString with allocated contents | |
442 | // build a longer string that will not fit into the object's buffer | |
b331163b | 443 | one+=UnicodeString(writeable, UPRV_LENGTHOF(writeable)); |
b75a7d8f A |
444 | one+=one; |
445 | one+=one; | |
446 | printf("length of longer string: %d\n", one.length()); | |
447 | // copying will use the same allocated buffer and increment the reference | |
448 | // counter | |
449 | two=one; | |
450 | printf("length of longer string copy: %d\n", two.length()); | |
451 | ||
452 | // * UnicodeString using readonly-alias to a const UChar array | |
453 | // construct a string that aliases a readonly buffer | |
b331163b | 454 | UnicodeString three(FALSE, readonly, UPRV_LENGTHOF(readonly)); |
b75a7d8f A |
455 | printUnicodeString("readonly-alias string: ", three); |
456 | // copy-on-write: any modification to the string results in | |
457 | // a copy to either the internal buffer or to a newly allocated one | |
458 | three.setCharAt(1, 0x39); | |
459 | printUnicodeString("readonly-aliasing string after modification: ", three); | |
460 | // the aliased array is not modified | |
461 | for(i=0; i<three.length(); ++i) { | |
462 | printf("readonly buffer[%d] after modifying its string: 0x%lx\n", | |
463 | i, readonly[i]); | |
464 | } | |
465 | // setTo() readonly alias | |
b331163b | 466 | one.setTo(FALSE, writeable, UPRV_LENGTHOF(writeable)); |
b75a7d8f A |
467 | // copying the readonly-alias object with fastCopyFrom() (new in ICU 2.4) |
468 | // will readonly-alias the same buffer | |
469 | two.fastCopyFrom(one); | |
470 | printUnicodeString("fastCopyFrom(readonly alias of \"writeable\" array): ", two); | |
471 | printf("verify that a fastCopyFrom(readonly alias) uses the same buffer pointer: %d (should be 1)\n", | |
472 | one.getBuffer()==two.getBuffer()); | |
473 | // a normal assignment will clone the contents (new in ICU 2.4) | |
474 | two=one; | |
475 | printf("verify that a regular copy of a readonly alias uses a different buffer pointer: %d (should be 0)\n", | |
476 | one.getBuffer()==two.getBuffer()); | |
477 | ||
478 | // * UnicodeString using writeable-alias to a non-const UChar array | |
b331163b | 479 | UnicodeString four(writeable, UPRV_LENGTHOF(writeable), UPRV_LENGTHOF(writeable)); |
b75a7d8f A |
480 | printUnicodeString("writeable-alias string: ", four); |
481 | // a modification writes through to the buffer | |
482 | four.setCharAt(1, 0x39); | |
483 | for(i=0; i<four.length(); ++i) { | |
484 | printf("writeable-alias backing buffer[%d]=0x%lx " | |
485 | "after modification\n", i, writeable[i]); | |
486 | } | |
487 | // a copy will not alias any more; | |
488 | // instead, it will get a copy of the contents into allocated memory | |
489 | two=four; | |
490 | two.setCharAt(1, 0x21); | |
491 | for(i=0; i<two.length(); ++i) { | |
492 | printf("writeable-alias backing buffer[%d]=0x%lx after " | |
493 | "modification of string copy\n", i, writeable[i]); | |
494 | } | |
495 | // setTo() writeable alias, capacity==length | |
b331163b | 496 | one.setTo(writeable, UPRV_LENGTHOF(writeable), UPRV_LENGTHOF(writeable)); |
b75a7d8f A |
497 | // grow the string - it will not fit into the backing buffer any more |
498 | // and will get copied before modification | |
499 | one.append((UChar)0x40); | |
500 | // shrink it back so it would fit | |
501 | one.truncate(one.length()-1); | |
502 | // we still operate on the copy | |
503 | one.setCharAt(1, 0x25); | |
504 | printf("string after growing too much and then shrinking[1]=0x%lx\n" | |
505 | " backing store for this[1]=0x%lx\n", | |
506 | one.charAt(1), writeable[1]); | |
507 | // if we need it in the original buffer, then extract() to it | |
508 | // extract() does not do anything if the string aliases that same buffer | |
509 | // i=min(one.length(), length of array) | |
b331163b | 510 | if(one.length()<UPRV_LENGTHOF(writeable)) { |
b75a7d8f A |
511 | i=one.length(); |
512 | } else { | |
b331163b | 513 | i=UPRV_LENGTHOF(writeable); |
b75a7d8f A |
514 | } |
515 | one.extract(0, i, writeable); | |
b331163b | 516 | for(i=0; i<UPRV_LENGTHOF(writeable); ++i) { |
b75a7d8f A |
517 | printf("writeable-alias backing buffer[%d]=0x%lx after re-extract\n", |
518 | i, writeable[i]); | |
519 | } | |
520 | } | |
521 | ||
522 | // sample code for UnicodeString instantiations ----------------------------- *** | |
523 | ||
524 | static void | |
525 | demoUnicodeStringInit() { | |
526 | // *** Make sure to read about invariant characters in utypes.h! *** | |
527 | // Initialization of Unicode strings from C literals works _only_ for | |
528 | // invariant characters! | |
529 | ||
530 | printf("\n* demoUnicodeStringInit() ---------- ***\n\n"); | |
531 | ||
532 | // the string literal is 32 chars long - this must be counted for the macro | |
533 | UnicodeString invariantOnly=UNICODE_STRING("such characters are safe 123 %-.", 32); | |
534 | ||
535 | /* | |
536 | * In C, we need two macros: one to declare the UChar[] array, and | |
537 | * one to populate it; the second one is a noop on platforms where | |
538 | * wchar_t is compatible with UChar and ASCII-based. | |
539 | * The length of the string literal must be counted for both macros. | |
540 | */ | |
541 | /* declare the invString array for the string */ | |
542 | U_STRING_DECL(invString, "such characters are safe 123 %-.", 32); | |
543 | /* populate it with the characters */ | |
544 | U_STRING_INIT(invString, "such characters are safe 123 %-.", 32); | |
545 | ||
546 | // compare the C and C++ strings | |
547 | printf("C and C++ Unicode strings are equal: %d\n", invariantOnly==UnicodeString(TRUE, invString, 32)); | |
548 | ||
549 | /* | |
550 | * convert between char * and UChar * strings that | |
551 | * contain only invariant characters | |
552 | */ | |
553 | static const char *cs1="such characters are safe 123 %-."; | |
554 | static UChar us1[40]; | |
555 | static char cs2[40]; | |
556 | u_charsToUChars(cs1, us1, 33); /* include the terminating NUL */ | |
557 | u_UCharsToChars(us1, cs2, 33); | |
558 | printf("char * -> UChar * -> char * with only " | |
559 | "invariant characters: \"%s\"\n", | |
560 | cs2); | |
561 | ||
562 | // initialize a UnicodeString from a string literal that contains | |
563 | // escape sequences written with invariant characters | |
564 | // do not forget to duplicate the backslashes for ICU to see them | |
565 | // then, count each double backslash only once! | |
566 | UnicodeString german=UNICODE_STRING( | |
567 | "Sch\\u00f6nes Auto: \\u20ac 11240.\\fPrivates Zeichen: \\U00102345\\n", 64). | |
568 | unescape(); | |
569 | printUnicodeString("german UnicodeString from unescaping:\n ", german); | |
570 | ||
571 | /* | |
572 | * C: convert and unescape a char * string with only invariant | |
573 | * characters to fill a UChar * string | |
574 | */ | |
575 | UChar buffer[200]; | |
576 | int32_t length; | |
577 | length=u_unescape( | |
578 | "Sch\\u00f6nes Auto: \\u20ac 11240.\\fPrivates Zeichen: \\U00102345\\n", | |
b331163b | 579 | buffer, UPRV_LENGTHOF(buffer)); |
b75a7d8f A |
580 | printf("german C Unicode string from char * unescaping: (length %d)\n ", length); |
581 | printUnicodeString("", UnicodeString(buffer)); | |
582 | } | |
583 | ||
584 | extern int | |
585 | main(int argc, const char *argv[]) { | |
586 | UErrorCode errorCode=U_ZERO_ERROR; | |
587 | ||
588 | // Note: Using a global variable for any object is not exactly thread-safe... | |
589 | ||
590 | // You can change this call to e.g. ucnv_open("UTF-8", &errorCode) if you pipe | |
591 | // the output to a file and look at it with a Unicode-capable editor. | |
592 | // This will currently affect only the printUString() function, see the code above. | |
593 | // printUnicodeString() could use this, too, by changing to an extract() overload | |
594 | // that takes a UConverter argument. | |
595 | cnv=ucnv_open(NULL, &errorCode); | |
596 | if(U_FAILURE(errorCode)) { | |
597 | fprintf(stderr, "error %s opening the default converter\n", u_errorName(errorCode)); | |
598 | return errorCode; | |
599 | } | |
600 | ||
601 | ucnv_setFromUCallBack(cnv, UCNV_FROM_U_CALLBACK_ESCAPE, UCNV_ESCAPE_C, NULL, NULL, &errorCode); | |
602 | if(U_FAILURE(errorCode)) { | |
603 | fprintf(stderr, "error %s setting the escape callback in the default converter\n", u_errorName(errorCode)); | |
604 | ucnv_close(cnv); | |
605 | return errorCode; | |
606 | } | |
607 | ||
608 | demo_utf_h_macros(); | |
609 | demo_C_Unicode_strings(); | |
610 | demoCaseMapInC(); | |
611 | demoCaseMapInCPlusPlus(); | |
612 | demoUnicodeStringStorage(); | |
613 | demoUnicodeStringInit(); | |
614 | ||
615 | ucnv_close(cnv); | |
616 | return 0; | |
617 | } |