Update ICU to 54.1 step 1

source/i18n/identifier_info.cpp

 /*
 **********************************************************************
-*   Copyright (C) 2012-2013, International Business Machines
+*   Copyright (C) 2012-2014, International Business Machines
 *   Corporation and others.  All Rights Reserved.
 **********************************************************************
 */
 
 #include "unicode/utypes.h"
 
 #include "unicode/uchar.h"
 #include "unicode/utf16.h"
 
 #include "identifier_info.h"
 #include "mutex.h"
 #include "scriptset.h"
 #include "ucln_in.h"
 #include "uvector.h"
 
 U_NAMESPACE_BEGIN
 
-#define LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0]))
+static UnicodeSet *ASCII;
+static ScriptSet *JAPANESE;
+static ScriptSet *CHINESE;
+static ScriptSet *KOREAN;
+static ScriptSet *CONFUSABLE_WITH_LATIN;
+static UInitOnce gIdentifierInfoInitOnce = U_INITONCE_INITIALIZER;
 
-static UMutex gInitMutex = U_MUTEX_INITIALIZER;
-static UBool gStaticsAreInitialized = FALSE;
 
-UnicodeSet *IdentifierInfo::ASCII;
-ScriptSet *IdentifierInfo::JAPANESE;
-ScriptSet *IdentifierInfo::CHINESE;
-ScriptSet *IdentifierInfo::KOREAN;
-ScriptSet *IdentifierInfo::CONFUSABLE_WITH_LATIN;
-
-UBool IdentifierInfo::cleanup() {
+U_CDECL_BEGIN
+static UBool U_CALLCONV
+IdentifierInfo_cleanup(void) {
     delete ASCII;
     ASCII = NULL;
     delete JAPANESE;
     JAPANESE = NULL;
     delete CHINESE;
     CHINESE = NULL;
     delete KOREAN;
     KOREAN = NULL;
     delete CONFUSABLE_WITH_LATIN;
     CONFUSABLE_WITH_LATIN = NULL;
-    gStaticsAreInitialized = FALSE;
+    gIdentifierInfoInitOnce.reset(); 
     return TRUE;
 }
 
-U_CDECL_BEGIN
-static UBool U_CALLCONV
-IdentifierInfo_cleanup(void) {
-    return IdentifierInfo::cleanup();
+static void U_CALLCONV
+IdentifierInfo_init(UErrorCode &status) {
+    ASCII    = new UnicodeSet(0, 0x7f);
+    JAPANESE = new ScriptSet();
+    CHINESE  = new ScriptSet();
+    KOREAN   = new ScriptSet();
+    CONFUSABLE_WITH_LATIN = new ScriptSet();
+    if (ASCII == NULL || JAPANESE == NULL || CHINESE == NULL || KOREAN == NULL 
+            || CONFUSABLE_WITH_LATIN == NULL) {
+        status = U_MEMORY_ALLOCATION_ERROR;
+        return;
+    }
+    ASCII->freeze();
+    JAPANESE->set(USCRIPT_LATIN, status).set(USCRIPT_HAN, status).set(USCRIPT_HIRAGANA, status)
+             .set(USCRIPT_KATAKANA, status);
+    CHINESE->set(USCRIPT_LATIN, status).set(USCRIPT_HAN, status).set(USCRIPT_BOPOMOFO, status);
+    KOREAN->set(USCRIPT_LATIN, status).set(USCRIPT_HAN, status).set(USCRIPT_HANGUL, status);
+    CONFUSABLE_WITH_LATIN->set(USCRIPT_CYRILLIC, status).set(USCRIPT_GREEK, status)
+              .set(USCRIPT_CHEROKEE, status);
+    ucln_i18n_registerCleanup(UCLN_I18N_IDENTIFIER_INFO, IdentifierInfo_cleanup);
 }
 U_CDECL_END
 
 
 IdentifierInfo::IdentifierInfo(UErrorCode &status):
          fIdentifier(NULL), fRequiredScripts(NULL), fScriptSetSet(NULL), 
          fCommonAmongAlternates(NULL), fNumerics(NULL), fIdentifierProfile(NULL) {
+    umtx_initOnce(gIdentifierInfoInitOnce, &IdentifierInfo_init, status);
     if (U_FAILURE(status)) {
         return;
     }
-    {
-        Mutex lock(&gInitMutex);
-        if (!gStaticsAreInitialized) {
-            ASCII    = new UnicodeSet(0, 0x7f);
-            JAPANESE = new ScriptSet();
-            CHINESE  = new ScriptSet();
-            KOREAN   = new ScriptSet();
-            CONFUSABLE_WITH_LATIN = new ScriptSet();
-            if (ASCII == NULL || JAPANESE == NULL || CHINESE == NULL || KOREAN == NULL 
-                    || CONFUSABLE_WITH_LATIN == NULL) {
-                status = U_MEMORY_ALLOCATION_ERROR;
-                return;
-            }
-            ASCII->freeze();
-            JAPANESE->set(USCRIPT_LATIN, status).set(USCRIPT_HAN, status).set(USCRIPT_HIRAGANA, status)
-                     .set(USCRIPT_KATAKANA, status);
-            CHINESE->set(USCRIPT_LATIN, status).set(USCRIPT_HAN, status).set(USCRIPT_BOPOMOFO, status);
-            KOREAN->set(USCRIPT_LATIN, status).set(USCRIPT_HAN, status).set(USCRIPT_HANGUL, status);
-            CONFUSABLE_WITH_LATIN->set(USCRIPT_CYRILLIC, status).set(USCRIPT_GREEK, status)
-                      .set(USCRIPT_CHEROKEE, status);
-            ucln_i18n_registerCleanup(UCLN_I18N_IDENTIFIER_INFO, IdentifierInfo_cleanup);
-            gStaticsAreInitialized = TRUE;
-        }
-    }
+    
     fIdentifier = new UnicodeString();
     fRequiredScripts = new ScriptSet();
     fScriptSetSet = uhash_open(uhash_hashScriptSet, uhash_compareScriptSet, NULL, &status);
     uhash_setKeyDeleter(fScriptSetSet, uhash_deleteScriptSet);
     fCommonAmongAlternates = new ScriptSet();
     fNumerics = new UnicodeSet();
     fIdentifierProfile = new UnicodeSet(0, 0x10FFFF);
 
     if (U_SUCCESS(status) && (fIdentifier == NULL || fRequiredScripts == NULL || fScriptSetSet == NULL ||
                               fCommonAmongAlternates == NULL || fNumerics == NULL || fIdentifierProfile == NULL)) {
@@ skipping 40 matching lines @@
     ScriptSet scriptsForCP;
     UChar32 cp;
     for (int32_t i = 0; i < identifier.length(); i += U16_LENGTH(cp)) {
         cp = identifier.char32At(i);
         // Store a representative character for each kind of decimal digit
         if (u_charType(cp) == U_DECIMAL_DIGIT_NUMBER) {
             // Just store the zero character as a representative for comparison. Unicode guarantees it is cp - value
             fNumerics->add(cp - (UChar32)u_getNumericValue(cp));
         }
         UScriptCode extensions[500];
-        int32_t extensionsCount = uscript_getScriptExtensions(cp, extensions, LENGTHOF(extensions), &status);
+        int32_t extensionsCount = uscript_getScriptExtensions(cp, extensions, UPRV_LENGTHOF(extensions), &status);
         if (U_FAILURE(status)) {
             return *this;
         }
         scriptsForCP.resetAll();
         for (int32_t j=0; j<extensionsCount; j++) {
             scriptsForCP.set(extensions[j], status);
         }
         scriptsForCP.reset(USCRIPT_COMMON, status);
         scriptsForCP.reset(USCRIPT_INHERITED, status);
         switch (scriptsForCP.countMembers()) {
@@ skipping 84 matching lines @@
     // This is a bit tricky. We look at a number of factors.
     // The number of scripts in the text.
     // Plus 1 if there is some commonality among the alternates (eg [Arab Thaa]; [Arab Syrc])
     // Plus number of alternates otherwise (this only works because we only test cardinality up to 2.)
 
     // Note: the requiredScripts set omits COMMON and INHERITED; they are taken out at the
     //       time it is created, in setIdentifier().
     int32_t cardinalityPlus = fRequiredScripts->countMembers() + 
             (fCommonAmongAlternates->countMembers() == 0 ? uhash_count(fScriptSetSet) : 1);
     if (cardinalityPlus < 2) {
-        return USPOOF_HIGHLY_RESTRICTIVE;
+        return USPOOF_SINGLE_SCRIPT_RESTRICTIVE;
     }
     if (containsWithAlternates(*JAPANESE, *fRequiredScripts) || containsWithAlternates(*CHINESE, *fRequiredScripts)
             || containsWithAlternates(*KOREAN, *fRequiredScripts)) {
         return USPOOF_HIGHLY_RESTRICTIVE;
     }
     if (cardinalityPlus == 2 && 
             fRequiredScripts->test(USCRIPT_LATIN, status) && 
             !fRequiredScripts->intersects(*CONFUSABLE_WITH_LATIN)) {
         return USPOOF_MODERATELY_RESTRICTIVE;
     }
@@ skipping 48 matching lines @@
             dest.append(separator);
         }
         ScriptSet *ss = static_cast<ScriptSet *>(sorted.elementAt(i));
         ss->displayScripts(dest);
     }
     return dest;
 }
 
 U_NAMESPACE_END