/* ******************************************************************************* * * © 2016 and later: Unicode, Inc. and others. * License & terms of use: http://www.unicode.org/copyright.html#License * ******************************************************************************* ******************************************************************************* * * Copyright (C) 2000-2014, International Business Machines * Corporation and others. All Rights Reserved. * ******************************************************************************* * file name: ustring.c * encoding: UTF-8 * tab size: 8 (not used) * indentation:4 * * created on: 2000aug15 * created by: Markus W. Scherer * * This file contains sample code that illustrates the use of Unicode strings * with ICU. */ #include #include "unicode/utypes.h" #include "unicode/uchar.h" #include "unicode/locid.h" #include "unicode/ustring.h" #include "unicode/ucnv.h" #include "unicode/unistr.h" using namespace icu; #ifndef UPRV_LENGTHOF #define UPRV_LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0])) #endif // helper functions -------------------------------------------------------- *** // default converter for the platform encoding static UConverter *cnv=NULL; static void printUString(const char *announce, const UChar *s, int32_t length) { static char out[200]; UChar32 c; int32_t i; UErrorCode errorCode=U_ZERO_ERROR; /* * Convert to the "platform encoding". See notes in printUnicodeString(). * ucnv_fromUChars(), like most ICU APIs understands length==-1 * to mean that the string is NUL-terminated. */ ucnv_fromUChars(cnv, out, sizeof(out), s, length, &errorCode); if(U_FAILURE(errorCode) || errorCode==U_STRING_NOT_TERMINATED_WARNING) { printf("%sproblem converting string from Unicode: %s\n", announce, u_errorName(errorCode)); return; } printf("%s%s {", announce, out); /* output the code points (not code units) */ if(length>=0) { /* s is not NUL-terminated */ for(i=0; i0; /* U16_PREV pre-decrements */) { U16_PREV(input, 0, i, c); /* Iterating backwards Codepoint at offset 5: U+0062 Codepoint at offset 3: U+10ffff Codepoint at offset 2: U+dc00 -- unpaired surrogate because lead surr. overwritten Codepoint at offset 1: U+0062 -- by this BMP code point Codepoint at offset 0: U+0061 */ printf("Codepoint at offset %d: U+%04x\n", i, c); } } // sample code for Unicode strings in C ------------------------------------ *** static void demo_C_Unicode_strings() { printf("\n* demo_C_Unicode_strings() --------- ***\n\n"); static const UChar text[]={ 0x41, 0x42, 0x43, 0 }; /* "ABC" */ static const UChar appendText[]={ 0x61, 0x62, 0x63, 0 }; /* "abc" */ static const UChar cmpText[]={ 0x61, 0x53, 0x73, 0x43, 0 }; /* "aSsC" */ UChar buffer[32]; int32_t compare; int32_t length=u_strlen(text); /* length=3 */ /* simple ANSI C-style functions */ buffer[0]=0; /* empty, NUL-terminated string */ u_strncat(buffer, text, 1); /* append just n=1 character ('A') */ u_strcat(buffer, appendText); /* buffer=="Aabc" */ length=u_strlen(buffer); /* length=4 */ printUString("should be \"Aabc\": ", buffer, -1); /* bitwise comparing buffer with text */ compare=u_strcmp(buffer, text); if(compare<=0) { printf("String comparison error, expected \"Aabc\" > \"ABC\"\n"); } /* Build "AC" in the buffer... */ u_strcpy(buffer, text); buffer[1]=0xdf; /* sharp s, case-compares equal to "ss" */ printUString("should be \"AC\": ", buffer, -1); /* Compare two strings case-insensitively using full case folding */ compare=u_strcasecmp(buffer, cmpText, U_FOLD_CASE_DEFAULT); if(compare!=0) { printf("String case insensitive comparison error, expected \"AbC\" to be equal to \"ABC\"\n"); } } // sample code for case mappings with C APIs -------------------------------- *** static void demoCaseMapInC() { /* * input= * "aB" * "iI " * " " * "" */ static const UChar input[]={ 0x61, 0x42, 0x3a3, 0x69, 0x49, 0x131, 0x130, 0x20, 0xdf, 0x20, 0xfb03, 0x3c2, 0x3c3, 0x3a3, 0 }; UChar buffer[32]; UErrorCode errorCode; UChar32 c; int32_t i, j, length; UBool isError; printf("\n* demoCaseMapInC() ----------------- ***\n\n"); /* * First, use simple case mapping functions which provide * 1:1 code point mappings without context/locale ID. * * Note that some mappings will not be "right" because some "real" * case mappings require context, depend on the locale ID, * and/or result in a change in the number of code points. */ printUString("input string: ", input, -1); /* uppercase */ isError=FALSE; for(i=j=0; j" * "iI " * " " * "" */ static const UChar input[]={ 0x61, 0x42, 0x3a3, 0x69, 0x49, 0x131, 0x130, 0x20, 0xdf, 0x20, 0xfb03, 0x3c2, 0x3c3, 0x3a3, 0 }; printf("\n* demoCaseMapInCPlusPlus() --------- ***\n\n"); UnicodeString s(input), t; const Locale &en=Locale::getEnglish(); Locale tr("tr"); /* * Full case mappings as in demoCaseMapInC(), using UnicodeString functions. * These functions modify the string object itself. * Since we want to keep the input string around, we copy it each time * and case-map the copy. */ printUnicodeString("input string: ", s); /* lowercase/English */ printUnicodeString("full-lowercased/en: ", (t=s).toLower(en)); /* lowercase/Turkish */ printUnicodeString("full-lowercased/tr: ", (t=s).toLower(tr)); /* uppercase/English */ printUnicodeString("full-uppercased/en: ", (t=s).toUpper(en)); /* uppercase/Turkish */ printUnicodeString("full-uppercased/tr: ", (t=s).toUpper(tr)); /* titlecase/English */ printUnicodeString("full-titlecased/en: ", (t=s).toTitle(NULL, en)); /* titlecase/Turkish */ printUnicodeString("full-titlecased/tr: ", (t=s).toTitle(NULL, tr)); /* case-folde/default */ printUnicodeString("full-case-folded/default: ", (t=s).foldCase(U_FOLD_CASE_DEFAULT)); /* case-folde/Turkic */ printUnicodeString("full-case-folded/Turkic: ", (t=s).foldCase(U_FOLD_CASE_EXCLUDE_SPECIAL_I)); } // sample code for UnicodeString storage models ----------------------------- *** static const UChar readonly[]={ 0x61, 0x31, 0x20ac }; static UChar writeable[]={ 0x62, 0x32, 0xdbc0, 0xdc01 // includes a surrogate pair for a supplementary code point }; static char out[100]; static void demoUnicodeStringStorage() { // These sample code lines illustrate how to use UnicodeString, and the // comments tell what happens internally. There are no APIs to observe // most of this programmatically, except for stepping into the code // with a debugger. // This is by design to hide such details from the user. int32_t i; printf("\n* demoUnicodeStringStorage() ------- ***\n\n"); // * UnicodeString with internally stored contents // instantiate a UnicodeString from a single code point // the few (2) UChars will be stored in the object itself UnicodeString one((UChar32)0x24001); // this copies the few UChars into the "two" object UnicodeString two=one; printf("length of short string copy: %d\n", two.length()); // set "one" to contain the 3 UChars from readonly // this setTo() variant copies the characters one.setTo(readonly, UPRV_LENGTHOF(readonly)); // * UnicodeString with allocated contents // build a longer string that will not fit into the object's buffer one+=UnicodeString(writeable, UPRV_LENGTHOF(writeable)); one+=one; one+=one; printf("length of longer string: %d\n", one.length()); // copying will use the same allocated buffer and increment the reference // counter two=one; printf("length of longer string copy: %d\n", two.length()); // * UnicodeString using readonly-alias to a const UChar array // construct a string that aliases a readonly buffer UnicodeString three(FALSE, readonly, UPRV_LENGTHOF(readonly)); printUnicodeString("readonly-alias string: ", three); // copy-on-write: any modification to the string results in // a copy to either the internal buffer or to a newly allocated one three.setCharAt(1, 0x39); printUnicodeString("readonly-aliasing string after modification: ", three); // the aliased array is not modified for(i=0; i UChar * -> char * with only " "invariant characters: \"%s\"\n", cs2); // initialize a UnicodeString from a string literal that contains // escape sequences written with invariant characters // do not forget to duplicate the backslashes for ICU to see them // then, count each double backslash only once! UnicodeString german=UNICODE_STRING( "Sch\\u00f6nes Auto: \\u20ac 11240.\\fPrivates Zeichen: \\U00102345\\n", 64). unescape(); printUnicodeString("german UnicodeString from unescaping:\n ", german); /* * C: convert and unescape a char * string with only invariant * characters to fill a UChar * string */ UChar buffer[200]; int32_t length; length=u_unescape( "Sch\\u00f6nes Auto: \\u20ac 11240.\\fPrivates Zeichen: \\U00102345\\n", buffer, UPRV_LENGTHOF(buffer)); printf("german C Unicode string from char * unescaping: (length %d)\n ", length); printUnicodeString("", UnicodeString(buffer)); } extern int main(int argc, const char *argv[]) { UErrorCode errorCode=U_ZERO_ERROR; // Note: Using a global variable for any object is not exactly thread-safe... // You can change this call to e.g. ucnv_open("UTF-8", &errorCode) if you pipe // the output to a file and look at it with a Unicode-capable editor. // This will currently affect only the printUString() function, see the code above. // printUnicodeString() could use this, too, by changing to an extract() overload // that takes a UConverter argument. cnv=ucnv_open(NULL, &errorCode); if(U_FAILURE(errorCode)) { fprintf(stderr, "error %s opening the default converter\n", u_errorName(errorCode)); return errorCode; } ucnv_setFromUCallBack(cnv, UCNV_FROM_U_CALLBACK_ESCAPE, UCNV_ESCAPE_C, NULL, NULL, &errorCode); if(U_FAILURE(errorCode)) { fprintf(stderr, "error %s setting the escape callback in the default converter\n", u_errorName(errorCode)); ucnv_close(cnv); return errorCode; } demo_utf_h_macros(); demo_C_Unicode_strings(); demoCaseMapInC(); demoCaseMapInCPlusPlus(); demoUnicodeStringStorage(); demoUnicodeStringInit(); ucnv_close(cnv); return 0; }