C++ U16_LEAD函数代码示例

ConversionResult convertUTF8ToUTF16(
    const char** sourceStart, const char* sourceEnd, 
    UChar** targetStart, UChar* targetEnd, bool strict)
{
    ConversionResult result = conversionOK;
    const char* source = *sourceStart;
    UChar* target = *targetStart;
    while (source < sourceEnd) {
        int utf8SequenceLength = inlineUTF8SequenceLength(*source);
        if (sourceEnd - source < utf8SequenceLength)  {
            result = sourceExhausted;
            break;
        }
        // Do this check whether lenient or strict
        if (!isLegalUTF8(reinterpret_cast<const unsigned char*>(source), utf8SequenceLength)) {
            result = sourceIllegal;
            break;
        }

        UChar32 character = readUTF8Sequence(source, utf8SequenceLength);

        if (target >= targetEnd) {
            source -= utf8SequenceLength; // Back up source pointer!
            result = targetExhausted;
            break;
        }

        if (U_IS_BMP(character)) {
            // UTF-16 surrogate values are illegal in UTF-32
            if (U_IS_SURROGATE(character)) {
                if (strict) {
                    source -= utf8SequenceLength; // return to the illegal value itself
                    result = sourceIllegal;
                    break;
                } else
                    *target++ = replacementCharacter;
            } else
                *target++ = character; // normal case
        } else if (U_IS_SUPPLEMENTARY(character)) {
            // target is a character in range 0xFFFF - 0x10FFFF
            if (target + 1 >= targetEnd) {
                source -= utf8SequenceLength; // Back up source pointer!
                result = targetExhausted;
                break;
            }
            *target++ = U16_LEAD(character);
            *target++ = U16_TRAIL(character);
        } else {
            if (strict) {
                source -= utf8SequenceLength; // return to the start
                result = sourceIllegal;
                break; // Bail out; shouldn't continue
            } else
                *target++ = replacementCharacter;
        }
    }
    *sourceStart = source;
    *targetStart = target;
    return result;
}

U_CAPI UChar32 U_EXPORT2 /* U_CAPI ... U_EXPORT2 added by Peter Kirk 17 Nov 2001 */
u_fungetc(UChar32        ch,
          UFILE        *f)
{
    u_localized_string *str;

    str = &f->str;

    /* if we're at the beginning of the buffer, sorry! */
    if (str->fPos == str->fBuffer
            || (U_IS_LEAD(ch) && (str->fPos - 1) == str->fBuffer))
    {
        ch = U_EOF;
    }
    else {
        /* otherwise, put the character back */
        /* Remember, read them back on in the reverse order. */
        if (U_IS_LEAD(ch)) {
            if (*--(str->fPos) != U16_TRAIL(ch)
                    || *--(str->fPos) != U16_LEAD(ch))
            {
                ch = U_EOF;
            }
        }
        else if (*--(str->fPos) != ch) {
            ch = U_EOF;
        }
    }
    return ch;
}

void XMLTreeBuilder::processHTMLEntity(const AtomicXMLToken& token)
{
    HTMLEntitySearch search;
    const AtomicString& name = token.name();
    for (size_t i = 0; i < name.length(); ++i) {
        search.advance(name[i]);
        if (!search.isEntityPrefix()) {
            m_parser->stopParsing();
            return;
        }
    }
    search.advance(';');
    if (!search.isEntityPrefix()) {
        m_parser->stopParsing();
        return;
    }
    UChar32 entityValue = search.mostRecentMatch()->firstValue;
    // FIXME: We need to account for secondValue if any XML entities are longer
    // than one unicode character.
    ASSERT_NOT_REACHED();
    // Darin Adler writes:
    //   You can see given the code above that this else is dead code. This code is in a strange state.
    //   And the reinterpret_cast to UChar* makes the code little-endian-specific. That is not good!
    if (entityValue <= 0xFFFF)
        appendToText(reinterpret_cast<UChar*>(&entityValue), 1);
    else {
        UChar utf16Pair[2] = { U16_LEAD(entityValue), U16_TRAIL(entityValue) };
        appendToText(utf16Pair, 2);
    }
}

static UChar32 U_CALLCONV
lenient8IteratorPrevious(UCharIterator *iter) {
    int32_t index;

    if(iter->reservedField!=0) {
        UChar lead=U16_LEAD(iter->reservedField);
        iter->reservedField=0;
        iter->start-=4; /* we stayed behind the supplementary code point; go before it now */
        if((index=iter->index)>0) {
            iter->index=index-1;
        }
        return lead;
    } else if(iter->start>0) {
        const uint8_t *s=(const uint8_t *)iter->context;
        UChar32 c;

        L8_PREV(s, 0, iter->start, c);
        if((index=iter->index)>0) {
            iter->index=index-1;
        } else if(iter->start<=1) {
            iter->index= c<=0xffff ? iter->start : iter->start+1;
        }
        if(c<0) {
            return 0xfffd;
        } else if(c<=0xffff) {
            return c;
        } else {
            iter->start+=4; /* back to behind this supplementary code point for consistent state */
            iter->reservedField=c;
            return U16_TRAIL(c);
        }
    } else {
        return U_SENTINEL;
    }
}

RefPtr<Font> Font::systemFallbackFontForCharacter(UChar32 character, const FontDescription& description, bool isForPlatformFont) const
{
    auto fontAddResult = systemFallbackCache().add(this, CharacterFallbackMap());

    if (!character) {
        UChar codeUnit = 0;
        return FontCache::singleton().systemFallbackForCharacters(description, this, isForPlatformFont, &codeUnit, 1);
    }

    auto key = CharacterFallbackMapKey(description.locale(), character, isForPlatformFont);
    auto characterAddResult = fontAddResult.iterator->value.add(WTF::move(key), nullptr);

    Font*& fallbackFont = characterAddResult.iterator->value;

    if (!fallbackFont) {
        UChar codeUnits[2];
        unsigned codeUnitsLength;
        if (U_IS_BMP(character)) {
            codeUnits[0] = FontCascade::normalizeSpaces(character);
            codeUnitsLength = 1;
        } else {
            codeUnits[0] = U16_LEAD(character);
            codeUnits[1] = U16_TRAIL(character);
            codeUnitsLength = 2;
        }

        fallbackFont = FontCache::singleton().systemFallbackForCharacters(description, this, isForPlatformFont, codeUnits, codeUnitsLength).get();
        if (fallbackFont)
            fallbackFont->m_isUsedInSystemFallbackCache = true;
    }

    return fallbackFont;
}

/* Explain <xxxxx> tag to a native value
 *
 * Since <xxxxx> is always larger than the native value,
 * the operation will replace the tag directly in the buffer,
 * and, of course, will shift tail elements.
 */
void IdnaConfTest::ExplainCodePointTag(UnicodeString& buf){
    buf.append((UChar)0);    // add a terminal NULL
    UChar* bufBase = buf.getBuffer(buf.length());
    UChar* p = bufBase;
    while (*p != 0){
        if ( *p != 0x3C){    // <
            *bufBase++ = *p++;
        } else {
            p++;    // skip <
            UChar32 cp = 0;
            for ( ;*p != 0x3E; p++){   // >
                if (0x30 <= *p && *p <= 0x39){        // 0-9
                    cp = (cp * 16) + (*p - 0x30);
                } else if (0x61 <= *p && *p <= 0x66){ // a-f
                    cp = (cp * 16) + (*p - 0x61) + 10;
                } else if (0x41 <= *p && *p <= 0x46) {// A-F
                    cp = (cp * 16) + (*p - 0x41) + 10;
                }
                // no else. hope everything is good.
            }
            p++;    // skip >
            if (U_IS_BMP(cp)){
                *bufBase++ = cp;
            } else {
                *bufBase++ = U16_LEAD(cp);
                *bufBase++ = U16_TRAIL(cp);
            }
        }
    }
    *bufBase = 0;  // close our buffer
    buf.releaseBuffer();
}

UStringTrieResult
UCharsTrie::nextForCodePoint(UChar32 cp) {
    return cp<=0xffff ?
        next(cp) :
        (USTRINGTRIE_HAS_NEXT(next(U16_LEAD(cp))) ?
            next(U16_TRAIL(cp)) :
            USTRINGTRIE_NO_MATCH);
}

UBool
Appendable::appendCodePoint(UChar32 c) {
    if(c<=0xffff) {
        return appendCodeUnit((UChar)c);
    } else {
        return appendCodeUnit(U16_LEAD(c)) && appendCodeUnit(U16_TRAIL(c));
    }
}

void git__utf8_to_16(wchar_t *dest, size_t length, const char *src)
{
	wchar_t *pDest = dest;
	uint32_t ch;
	const uint8_t* pSrc = (uint8_t*) src;

	assert(dest && src && length);

	length--;

	while(*pSrc && length > 0) {
		ch = *pSrc++;
		length--;

		if(ch < 0xc0) {
			/*
			 * ASCII, or a trail byte in lead position which is treated like
			 * a single-byte sequence for better character boundary
			 * resynchronization after illegal sequences.
			 */
			*pDest++ = (wchar_t)ch;
			continue;
		} else if(ch < 0xe0) { /* U+0080..U+07FF */
			if (pSrc[0]) {
				/* 0x3080 = (0xc0 << 6) + 0x80 */
				*pDest++ = (wchar_t)((ch << 6) + *pSrc++ - 0x3080);
				continue;
			}
		} else if(ch < 0xf0) { /* U+0800..U+FFFF */
			if (pSrc[0] && pSrc[1]) {
				/* no need for (ch & 0xf) because the upper bits are truncated after <<12 in the cast to (UChar) */
				/* 0x2080 = (0x80 << 6) + 0x80 */
				ch = (ch << 12) + (*pSrc++ << 6);
				*pDest++ = (wchar_t)(ch + *pSrc++ - 0x2080);
				continue;
			}
		} else /* f0..f4 */ { /* U+10000..U+10FFFF */
			if (length >= 1 && pSrc[0] && pSrc[1] && pSrc[2]) {
				/* 0x3c82080 = (0xf0 << 18) + (0x80 << 12) + (0x80 << 6) + 0x80 */
				ch = (ch << 18) + (*pSrc++ << 12);
				ch += *pSrc++ << 6;
				ch += *pSrc++ - 0x3c82080;
				*(pDest++) = U16_LEAD(ch);
				*(pDest++) = U16_TRAIL(ch);
				length--; /* two bytes for this character */
				continue;
			}
		}

		/* truncated character at the end */
		*pDest++ = 0xfffd;
		break;
	}

	*pDest++ = 0x0;
}

UBool
FCDUTF8CollationIterator::previousHasTccc() const {
    U_ASSERT(state == CHECK_BWD && pos != 0);
    UChar32 c = u8[pos - 1];
    if(c < 0x80) { return FALSE; }
    int32_t i = pos;
    U8_PREV_OR_FFFD(u8, 0, i, c);
    if(c > 0xffff) { c = U16_LEAD(c); }
    return CollationFCD::hasTccc(c);
}

static inline UChar* appendCharacter(UChar* destination, int character)
{
    ASSERT(character != nonCharacter);
    ASSERT(!U_IS_SURROGATE(character));
    if (U_IS_BMP(character))
        *destination++ = character;
    else {
        *destination++ = U16_LEAD(character);
        *destination++ = U16_TRAIL(character);
    }
    return destination;
}

static size_t appendUChar32ToUCharArray(UChar32 value, UChar* result)
{
    if (U_IS_BMP(value)) {
        UChar character = static_cast<UChar>(value);
        ASSERT(character == value);
        result[0] = character;
        return 1;
    }

    result[0] = U16_LEAD(value);
    result[1] = U16_TRAIL(value);
    return 2;
}

U_CFUNC void 
ustr_u32cat(struct UString *dst, UChar32 c, UErrorCode *status){
    if(c > 0x10FFFF){
        *status = U_ILLEGAL_CHAR_FOUND;
        return;
    }
    if(c >0xFFFF){
        ustr_ucat(dst, U16_LEAD(c), status);
        ustr_ucat(dst, U16_TRAIL(c), status);
    }else{
        ustr_ucat(dst, (UChar) c, status);
    }
}

static inline String singleCharacterString(UChar32 c) 
{
    if (!c)
        return String();
    if (c > 0xffff) {
        UChar lead = U16_LEAD(c);
        UChar trail = U16_TRAIL(c);
        UChar utf16[2] = {lead, trail};
        return String(utf16, 2);
    }
    UChar n = (UChar)c;
    return String(&n, 1);
}

unsigned calculateStringHashAndLengthFromUTF8MaskingTop8Bits(const char* data, const char* dataEnd, unsigned& dataLength, unsigned& utf16Length)
{
    if (!data)
        return 0;

    StringHasher stringHasher;
    dataLength = 0;
    utf16Length = 0;

    while (data < dataEnd || (!dataEnd && *data)) {
        if (isASCII(*data)) {
            stringHasher.addCharacter(*data++);
            dataLength++;
            utf16Length++;
            continue;
        }

        int utf8SequenceLength = inlineUTF8SequenceLengthNonASCII(*data);
        dataLength += utf8SequenceLength;

        if (!dataEnd) {
            for (int i = 1; i < utf8SequenceLength; ++i) {
                if (!data[i])
                    return 0;
            }
        } else if (dataEnd - data < utf8SequenceLength) {
            return 0;
        }

        if (!isLegalUTF8(reinterpret_cast<const unsigned char*>(data), utf8SequenceLength))
            return 0;

        UChar32 character = readUTF8Sequence(data, utf8SequenceLength);
        ASSERT(!isASCII(character));

        if (U_IS_BMP(character)) {
            // UTF-16 surrogate values are illegal in UTF-32
            if (U_IS_SURROGATE(character))
                return 0;
            stringHasher.addCharacter(static_cast<UChar>(character)); // normal case
            utf16Length++;
        } else if (U_IS_SUPPLEMENTARY(character)) {
            stringHasher.addCharacters(static_cast<UChar>(U16_LEAD(character)), static_cast<UChar>(U16_TRAIL(character)));
            utf16Length += 2;
        } else {
            return 0;
        }
    }

    return stringHasher.hashWithTop8BitsMasked();
}