deps: update to ICU 67.1

Refs: https://github.com/unicode-org/icu/releases/tag/release-67-1

PR-URL: #33324
Reviewed-By: Ujjwal Sharma <[email protected]>
Reviewed-By: Steven R Loomis <[email protected]>

Author: targos

deps/icu-small/README-FULL-ICU.txt

@@ -1,8 +1,8 @@
 ICU sources - auto generated by shrink-icu-src.py
 
 This directory contains the ICU subset used by --with-intl=full-icu
-It is a strict subset of ICU 66 source files with the following exception(s):
-* deps/icu-small/source/data/in/icudt66l.dat.bz2 : compressed data file
+It is a strict subset of ICU 67 source files with the following exception(s):
+* deps/icu-small/source/data/in/icudt67l.dat.bz2 : compressed data file
 
 
 To rebuild this directory, see ../../tools/icu/README.md

deps/icu-small/source/common/brkiter.cpp

@@ -38,6 +38,7 @@
 #include "uresimp.h"
 #include "uassert.h"
 #include "ubrkimpl.h"
+#include "utracimp.h"
 #include "charstr.h"
 
 // *****************************************************************************
@@ -412,28 +413,41 @@ BreakIterator::makeInstance(const Locale& loc, int32_t kind, UErrorCode& status)
     BreakIterator *result = NULL;
     switch (kind) {
     case UBRK_CHARACTER:
-        result = BreakIterator::buildInstance(loc, "grapheme", status);
+        {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_CHARACTER);
+            result = BreakIterator::buildInstance(loc, "grapheme", status);
+            UTRACE_EXIT_STATUS(status);
+        }
         break;
     case UBRK_WORD:
-        result = BreakIterator::buildInstance(loc, "word", status);
+        {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_WORD);
+            result = BreakIterator::buildInstance(loc, "word", status);
+            UTRACE_EXIT_STATUS(status);
+        }
         break;
     case UBRK_LINE:
-        uprv_strcpy(lbType, "line");
         {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_LINE);
+            uprv_strcpy(lbType, "line");
             char lbKeyValue[kKeyValueLenMax] = {0};
             UErrorCode kvStatus = U_ZERO_ERROR;
             int32_t kLen = loc.getKeywordValue("lb", lbKeyValue, kKeyValueLenMax, kvStatus);
             if (U_SUCCESS(kvStatus) && kLen > 0 && (uprv_strcmp(lbKeyValue,"strict")==0 || uprv_strcmp(lbKeyValue,"normal")==0 || uprv_strcmp(lbKeyValue,"loose")==0)) {
                 uprv_strcat(lbType, "_");
                 uprv_strcat(lbType, lbKeyValue);
             }
+            result = BreakIterator::buildInstance(loc, lbType, status);
+
+            UTRACE_DATA1(UTRACE_INFO, "lb=%s", lbKeyValue);
+            UTRACE_EXIT_STATUS(status);
         }
-        result = BreakIterator::buildInstance(loc, lbType, status);
         break;
     case UBRK_SENTENCE:
-        result = BreakIterator::buildInstance(loc, "sentence", status);
-#if !UCONFIG_NO_FILTERED_BREAK_ITERATION
         {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_SENTENCE);
+            result = BreakIterator::buildInstance(loc, "sentence", status);
+#if !UCONFIG_NO_FILTERED_BREAK_ITERATION
             char ssKeyValue[kKeyValueLenMax] = {0};
             UErrorCode kvStatus = U_ZERO_ERROR;
             int32_t kLen = loc.getKeywordValue("ss", ssKeyValue, kKeyValueLenMax, kvStatus);
@@ -444,11 +458,16 @@ BreakIterator::makeInstance(const Locale& loc, int32_t kind, UErrorCode& status)
                     delete fbiBuilder;
                 }
             }
-        }
 #endif
+            UTRACE_EXIT_STATUS(status);
+        }
         break;
     case UBRK_TITLE:
-        result = BreakIterator::buildInstance(loc, "title", status);
+        {
+            UTRACE_ENTRY(UTRACE_UBRK_CREATE_TITLE);
+            result = BreakIterator::buildInstance(loc, "title", status);
+            UTRACE_EXIT_STATUS(status);
+        }
         break;
     default:
         status = U_ILLEGAL_ARGUMENT_ERROR;

deps/icu-small/source/common/cmemory.h

@@ -274,7 +274,10 @@ inline T *LocalMemory<T>::allocateInsteadAndCopy(int32_t newCapacity, int32_t le
  *
  * WARNING: MaybeStackArray only works with primitive (plain-old data) types.
  * It does NOT know how to call a destructor! If you work with classes with
- * destructors, consider LocalArray in localpointer.h or MemoryPool.
+ * destructors, consider:
+ *
+ * - LocalArray in localpointer.h if you know the length ahead of time
+ * - MaybeStackVector if you know the length at runtime
  */
 template<typename T, int32_t stackCapacity>
 class MaybeStackArray {
@@ -684,26 +687,26 @@ inline H *MaybeStackHeaderAndArray<H, T, stackCapacity>::orphanOrClone(int32_t l
 template<typename T, int32_t stackCapacity = 8>
 class MemoryPool : public UMemory {
 public:
-    MemoryPool() : count(0), pool() {}
+    MemoryPool() : fCount(0), fPool() {}
 
     ~MemoryPool() {
-        for (int32_t i = 0; i < count; ++i) {
-            delete pool[i];
+        for (int32_t i = 0; i < fCount; ++i) {
+            delete fPool[i];
         }
     }
 
     MemoryPool(const MemoryPool&) = delete;
     MemoryPool& operator=(const MemoryPool&) = delete;
 
-    MemoryPool(MemoryPool&& other) U_NOEXCEPT : count(other.count),
-                                                pool(std::move(other.pool)) {
-        other.count = 0;
+    MemoryPool(MemoryPool&& other) U_NOEXCEPT : fCount(other.fCount),
+                                                fPool(std::move(other.fPool)) {
+        other.fCount = 0;
     }
 
     MemoryPool& operator=(MemoryPool&& other) U_NOEXCEPT {
-        count = other.count;
-        pool = std::move(other.pool);
-        other.count = 0;
+        fCount = other.fCount;
+        fPool = std::move(other.fPool);
+        other.fCount = 0;
         return *this;
     }
 
@@ -716,20 +719,101 @@ class MemoryPool : public UMemory {
      */
     template<typename... Args>
     T* create(Args&&... args) {
-        int32_t capacity = pool.getCapacity();
-        if (count == capacity &&
-            pool.resize(capacity == stackCapacity ? 4 * capacity : 2 * capacity,
-                        capacity) == nullptr) {
+        int32_t capacity = fPool.getCapacity();
+        if (fCount == capacity &&
+            fPool.resize(capacity == stackCapacity ? 4 * capacity : 2 * capacity,
+                         capacity) == nullptr) {
             return nullptr;
         }
-        return pool[count++] = new T(std::forward<Args>(args)...);
+        return fPool[fCount++] = new T(std::forward<Args>(args)...);
     }
 
-private:
-    int32_t count;
-    MaybeStackArray<T*, stackCapacity> pool;
+    /**
+     * @return Number of elements that have been allocated.
+     */
+    int32_t count() const {
+        return fCount;
+    }
+
+protected:
+    int32_t fCount;
+    MaybeStackArray<T*, stackCapacity> fPool;
+};
+
+/**
+ * An internal Vector-like implementation based on MemoryPool.
+ *
+ * Heap-allocates each element and stores pointers.
+ *
+ * To append an item to the vector, use emplaceBack.
+ *
+ *     MaybeStackVector<MyType> vector;
+ *     MyType* element = vector.emplaceBack();
+ *     if (!element) {
+ *         status = U_MEMORY_ALLOCATION_ERROR;
+ *     }
+ *     // do stuff with element
+ *
+ * To loop over the vector, use a for loop with indices:
+ *
+ *     for (int32_t i = 0; i < vector.length(); i++) {
+ *         MyType* element = vector[i];
+ *     }
+ */
+template<typename T, int32_t stackCapacity = 8>
+class MaybeStackVector : protected MemoryPool<T, stackCapacity> {
+public:
+    using MemoryPool<T, stackCapacity>::MemoryPool;
+    using MemoryPool<T, stackCapacity>::operator=;
+
+    template<typename... Args>
+    T* emplaceBack(Args&&... args) {
+        return this->create(args...);
+    }
+
+    int32_t length() const {
+        return this->fCount;
+    }
+
+    T** getAlias() {
+        return this->fPool.getAlias();
+    }
+
+    /**
+     * Array item access (read-only).
+     * No index bounds check.
+     * @param i array index
+     * @return reference to the array item
+     */
+    const T* operator[](ptrdiff_t i) const {
+        return this->fPool[i];
+    }
+
+    /**
+     * Array item access (writable).
+     * No index bounds check.
+     * @param i array index
+     * @return reference to the array item
+     */
+    T* operator[](ptrdiff_t i) {
+        return this->fPool[i];
+    }
+
+    /**
+     * Append all the items from another MaybeStackVector to this one.
+     */
+    void appendAll(const MaybeStackVector& other, UErrorCode& status) {
+        for (int32_t i = 0; i < other.fCount; i++) {
+            T* item = emplaceBack(*other[i]);
+            if (!item) {
+                status = U_MEMORY_ALLOCATION_ERROR;
+                return;
+            }
+        }
+    }
 };
 
+
 U_NAMESPACE_END
 
 #endif  /* __cplusplus */

deps/icu-small/source/common/cstring.h

@@ -52,6 +52,8 @@
 U_CAPI UBool U_EXPORT2
 uprv_isASCIILetter(char c);
 
+// NOTE: For u_asciiToUpper that takes a UChar, see ustr_imp.h
+
 U_CAPI char U_EXPORT2
 uprv_toupper(char c);
 

deps/icu-small/source/common/dictbe.cpp

@@ -18,6 +18,7 @@
 #include "unicode/uniset.h"
 #include "unicode/chariter.h"
 #include "unicode/ubrk.h"
+#include "utracimp.h"
 #include "uvectr32.h"
 #include "uvector.h"
 #include "uassert.h"
@@ -194,6 +195,8 @@ ThaiBreakEngine::ThaiBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode
     : DictionaryBreakEngine(),
       fDictionary(adoptDictionary)
 {
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Thai");
     fThaiWordSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Thai:]&[:LineBreak=SA:]]"), status);
     if (U_SUCCESS(status)) {
         setCharacters(fThaiWordSet);
@@ -213,6 +216,7 @@ ThaiBreakEngine::ThaiBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode
     fEndWordSet.compact();
     fBeginWordSet.compact();
     fSuffixSet.compact();
+    UTRACE_EXIT_STATUS(status);
 }
 
 ThaiBreakEngine::~ThaiBreakEngine() {
@@ -436,6 +440,8 @@ LaoBreakEngine::LaoBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &s
     : DictionaryBreakEngine(),
       fDictionary(adoptDictionary)
 {
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Laoo");
     fLaoWordSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Laoo:]&[:LineBreak=SA:]]"), status);
     if (U_SUCCESS(status)) {
         setCharacters(fLaoWordSet);
@@ -452,6 +458,7 @@ LaoBreakEngine::LaoBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCode &s
     fMarkSet.compact();
     fEndWordSet.compact();
     fBeginWordSet.compact();
+    UTRACE_EXIT_STATUS(status);
 }
 
 LaoBreakEngine::~LaoBreakEngine() {
@@ -632,6 +639,8 @@ BurmeseBreakEngine::BurmeseBreakEngine(DictionaryMatcher *adoptDictionary, UErro
     : DictionaryBreakEngine(),
       fDictionary(adoptDictionary)
 {
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Mymr");
     fBurmeseWordSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Mymr:]&[:LineBreak=SA:]]"), status);
     if (U_SUCCESS(status)) {
         setCharacters(fBurmeseWordSet);
@@ -645,6 +654,7 @@ BurmeseBreakEngine::BurmeseBreakEngine(DictionaryMatcher *adoptDictionary, UErro
     fMarkSet.compact();
     fEndWordSet.compact();
     fBeginWordSet.compact();
+    UTRACE_EXIT_STATUS(status);
 }
 
 BurmeseBreakEngine::~BurmeseBreakEngine() {
@@ -825,6 +835,8 @@ KhmerBreakEngine::KhmerBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCod
     : DictionaryBreakEngine(),
       fDictionary(adoptDictionary)
 {
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Khmr");
     fKhmerWordSet.applyPattern(UNICODE_STRING_SIMPLE("[[:Khmr:]&[:LineBreak=SA:]]"), status);
     if (U_SUCCESS(status)) {
         setCharacters(fKhmerWordSet);
@@ -850,6 +862,7 @@ KhmerBreakEngine::KhmerBreakEngine(DictionaryMatcher *adoptDictionary, UErrorCod
     fEndWordSet.compact();
     fBeginWordSet.compact();
 //    fSuffixSet.compact();
+    UTRACE_EXIT_STATUS(status);
 }
 
 KhmerBreakEngine::~KhmerBreakEngine() {
@@ -1045,6 +1058,8 @@ KhmerBreakEngine::divideUpDictionaryRange( UText *text,
 static const uint32_t kuint32max = 0xFFFFFFFF;
 CjkBreakEngine::CjkBreakEngine(DictionaryMatcher *adoptDictionary, LanguageType type, UErrorCode &status)
 : DictionaryBreakEngine(), fDictionary(adoptDictionary) {
+    UTRACE_ENTRY(UTRACE_UBRK_CREATE_BREAK_ENGINE);
+    UTRACE_DATA1(UTRACE_INFO, "dictbe=%s", "Hani");
     // Korean dictionary only includes Hangul syllables
     fHangulWordSet.applyPattern(UNICODE_STRING_SIMPLE("[\\uac00-\\ud7a3]"), status);
     fHanWordSet.applyPattern(UNICODE_STRING_SIMPLE("[:Han:]"), status);
@@ -1066,6 +1081,7 @@ CjkBreakEngine::CjkBreakEngine(DictionaryMatcher *adoptDictionary, LanguageType
             setCharacters(cjSet);
         }
     }
+    UTRACE_EXIT_STATUS(status);
 }
 
 CjkBreakEngine::~CjkBreakEngine(){