ustring.cpp

linux下开源浏览器WebKit的源码,市面上的很多商用浏览器都是移植自WebKit

            d[0] = c;
            m_rep = Rep::create(d, 1);
            m_rep->baseString()->capacity = newCapacity;
        }
    } else if (m_rep == base && m_rep->rc == 1) {
        // this is direct and has refcount of 1 (so we can just alter it directly)
        expandCapacity(thisOffset + length + 1);
        UChar* d = m_rep->data();
        if (d) {
            d[length] = c;
            m_rep->len = length + 1;
            m_rep->_hash = 0;
        }
    } else if (thisOffset + length == base->usedCapacity && length >= minShareSize) {
        // this reaches the end of the string - extend it and share
        expandCapacity(thisOffset + length + 1);
        UChar* d = m_rep->data();
        if (d) {
            d[length] = c;
            m_rep = Rep::create(m_rep, 0, length + 1);
        }
    } else {
        // this is shared with someone using more capacity, gotta make a whole new string
        size_t newCapacity = expandedSize(length + 1, 0);
        UChar* d = allocChars(newCapacity);
        if (!d)
            makeNull();
        else {
            copyChars(d, data(), length);
            d[length] = c;
            m_rep = Rep::create(d, length + 1);
            m_rep->baseString()->capacity = newCapacity;
        }
    }

    m_rep->checkConsistency();

    return *this;
}

bool UString::getCString(CStringBuffer& buffer) const
{
    int length = size();
    int neededSize = length + 1;
    buffer.resize(neededSize);
    char* buf = buffer.data();

    UChar ored = 0;
    const UChar* p = data();
    char* q = buf;
    const UChar* limit = p + length;
    while (p != limit) {
        UChar c = p[0];
        ored |= c;
        *q = static_cast<char>(c);
        ++p;
        ++q;
    }
    *q = '\0';

    return !(ored & 0xFF00);
}

char* UString::ascii() const
{
    int length = size();
    int neededSize = length + 1;
    delete[] statBuffer;
    statBuffer = new char[neededSize];

    const UChar* p = data();
    char* q = statBuffer;
    const UChar* limit = p + length;
    while (p != limit) {
        *q = static_cast<char>(p[0]);
        ++p;
        ++q;
    }
    *q = '\0';

    return statBuffer;
}

UString& UString::operator=(const char* c)
{
    if (!c) {
        m_rep = &Rep::null();
        return *this;
    }

    if (!c[0]) {
        m_rep = &Rep::empty();
        return *this;
    }

    int l = static_cast<int>(strlen(c));
    UChar* d;
    BaseString* base = m_rep->baseString();
    if (m_rep->rc == 1 && l <= base->capacity && m_rep == base && m_rep->offset == 0 && base->preCapacity == 0) {
        d = base->buf;
        m_rep->_hash = 0;
        m_rep->len = l;
    } else {
        d = allocChars(l);
        if (!d) {
            makeNull();
            return *this;
        }
        m_rep = Rep::create(d, l);
    }
    for (int i = 0; i < l; i++)
        d[i] = static_cast<unsigned char>(c[i]); // use unsigned char to zero-extend instead of sign-extend

    return *this;
}

bool UString::is8Bit() const
{
    const UChar* u = data();
    const UChar* limit = u + size();
    while (u < limit) {
        if (u[0] > 0xFF)
            return false;
        ++u;
    }

    return true;
}

UChar UString::operator[](int pos) const
{
    if (pos >= size())
        return '\0';
    return data()[pos];
}

double UString::toDouble(bool tolerateTrailingJunk, bool tolerateEmptyString) const
{
    if (size() == 1) {
        UChar c = data()[0];
        if (isASCIIDigit(c))
            return c - '0';
        if (isASCIISpace(c) && tolerateEmptyString)
            return 0;
        return NaN;
    }

    // FIXME: If tolerateTrailingJunk is true, then we want to tolerate non-8-bit junk
    // after the number, so this is too strict a check.
    CStringBuffer s;
    if (!getCString(s))
        return NaN;
    const char* c = s.data();

    // skip leading white space
    while (isASCIISpace(*c))
        c++;

    // empty string ?
    if (*c == '\0')
        return tolerateEmptyString ? 0.0 : NaN;

    double d;

    // hex number ?
    if (*c == '0' && (*(c + 1) == 'x' || *(c + 1) == 'X')) {
        const char* firstDigitPosition = c + 2;
        c++;
        d = 0.0;
        while (*(++c)) {
            if (*c >= '0' && *c <= '9')
                d = d * 16.0 + *c - '0';
            else if ((*c >= 'A' && *c <= 'F') || (*c >= 'a' && *c <= 'f'))
                d = d * 16.0 + (*c & 0xdf) - 'A' + 10.0;
            else
                break;
        }

        if (d >= mantissaOverflowLowerBound)
            d = parseIntOverflow(firstDigitPosition, c - firstDigitPosition, 16);
    } else {
        // regular number ?
        char* end;
        d = WTF::strtod(c, &end);
        if ((d != 0.0 || end != c) && d != Inf && d != -Inf) {
            c = end;
        } else {
            double sign = 1.0;

            if (*c == '+')
                c++;
            else if (*c == '-') {
                sign = -1.0;
                c++;
            }

            // We used strtod() to do the conversion. However, strtod() handles
            // infinite values slightly differently than JavaScript in that it
            // converts the string "inf" with any capitalization to infinity,
            // whereas the ECMA spec requires that it be converted to NaN.

            if (c[0] == 'I' && c[1] == 'n' && c[2] == 'f' && c[3] == 'i' && c[4] == 'n' && c[5] == 'i' && c[6] == 't' && c[7] == 'y') {
                d = sign * Inf;
                c += 8;
            } else if ((d == Inf || d == -Inf) && *c != 'I' && *c != 'i')
                c = end;
            else
                return NaN;
        }
    }

    // allow trailing white space
    while (isASCIISpace(*c))
        c++;
    // don't allow anything after - unless tolerant=true
    if (!tolerateTrailingJunk && *c != '\0')
        d = NaN;

    return d;
}

double UString::toDouble(bool tolerateTrailingJunk) const
{
    return toDouble(tolerateTrailingJunk, true);
}

double UString::toDouble() const
{
    return toDouble(false, true);
}

uint32_t UString::toUInt32(bool* ok) const
{
    double d = toDouble();
    bool b = true;

    if (d != static_cast<uint32_t>(d)) {
        b = false;
        d = 0;
    }

    if (ok)
        *ok = b;

    return static_cast<uint32_t>(d);
}

uint32_t UString::toUInt32(bool* ok, bool tolerateEmptyString) const
{
    double d = toDouble(false, tolerateEmptyString);
    bool b = true;

    if (d != static_cast<uint32_t>(d)) {
        b = false;
        d = 0;
    }

    if (ok)
        *ok = b;

    return static_cast<uint32_t>(d);
}

uint32_t UString::toStrictUInt32(bool* ok) const
{
    if (ok)
        *ok = false;

    // Empty string is not OK.
    int len = m_rep->len;
    if (len == 0)
        return 0;
    const UChar* p = m_rep->data();
    unsigned short c = p[0];

    // If the first digit is 0, only 0 itself is OK.
    if (c == '0') {
        if (len == 1 && ok)
            *ok = true;
        return 0;
    }

    // Convert to UInt32, checking for overflow.
    uint32_t i = 0;
    while (1) {
        // Process character, turning it into a digit.
        if (c < '0' || c > '9')
            return 0;
        const unsigned d = c - '0';

        // Multiply by 10, checking for overflow out of 32 bits.
        if (i > 0xFFFFFFFFU / 10)
            return 0;
        i *= 10;

        // Add in the digit, checking for overflow out of 32 bits.
        const unsigned max = 0xFFFFFFFFU - d;
        if (i > max)
            return 0;
        i += d;

        // Handle end of string.
        if (--len == 0) {
            if (ok)
                *ok = true;
            return i;
        }

        // Get next character.
        c = *(++p);
    }
}

int UString::find(const UString& f, int pos) const
{
    int fsz = f.size();

    if (pos < 0)
        pos = 0;

    if (fsz == 1) {
        UChar ch = f[0];
        const UChar* end = data() + size();
        for (const UChar* c = data() + pos; c < end; c++) {
            if (*c == ch)
                return static_cast<int>(c - data());
        }
        return -1;
    }

    int sz = size();
    if (sz < fsz)
        return -1;
    if (fsz == 0)
        return pos;
    const UChar* end = data() + sz - fsz;
    int fsizeminusone = (fsz - 1) * sizeof(UChar);
    const UChar* fdata = f.data();
    unsigned short fchar = fdata[0];
    ++fdata;
    for (const UChar* c = data() + pos; c <= end; c++) {
        if (c[0] == fchar && !memcmp(c + 1, fdata, fsizeminusone))
            return static_cast<int>(c - data());
    }

    return -1;
}

int UString::find(UChar ch, int pos) const
{
    if (pos < 0)
        pos = 0;
    const UChar* end = data() + size();
    for (const UChar* c = data() + pos; c < end; c++) {
        if (*c == ch)
            return static_cast<int>(c - data());
    }
    
    return -1;
}

int UString::rfind(const UString& f, int pos) const
{
    int sz = size();
    int fsz = f.size();
    if (sz < fsz)
        return -1;
    if (pos < 0)
        pos = 0;
    if (pos > sz - fsz)
        pos = sz - fsz;
    if (fsz == 0)
        return pos;
    int fsizeminusone = (fsz - 1) * sizeof(UChar);
    const UChar* fdata = f.data();
    for (const UChar* c = data() + pos; c >= data(); c--) {
        if (*c == *fdata && !memcmp(c + 1, fdata + 1, fsizeminusone))
            return static_cast<int>(c - data());
    }

    return -1;
}

int UString::rfind(UChar ch, int pos) const
{
    if (isEmpty())
        return -1;
    if (pos + 1 >= size())
        pos = size() - 1;
    for (const UChar* c = data() + pos; c >= data(); c--) {
        if (*c == ch)
            return static_cast<int>(c - data());
    }

    return -1;
}

UString UString::substr(int pos, int len) const
{
    int s = size();

    if (pos < 0)
        pos = 0;
    else if (pos >= s)
        pos = s;
    if (len < 0)
        len = s;
    if (pos + len >= s)
        len = s - pos;

    if (pos == 0 && len == s)
        return *this;

    return UString(Rep::create(m_rep, pos, len));
}

bool operator==(const UString& s1, const UString& s2)
{
    int size = s1.size();
    switch (size) {
        case 0:
            return !s2.size();
        case 1:
            return s2.size() == 1 && s1.data()[0] == s2.data()[0];
        default:
            return s2.size() == size && memcmp(s1.data(), s2.data(), size * sizeof(UChar)) == 0;
    }
}

bool operator==(const UString& s1, const char *s2)
{
    if (s2 == 0)
        return s1.isEmpty();

    const UChar* u = s1.data();
    const UChar* uend = u + s1.size();
    while (u != uend && *s2) {
        if (u[0] != (unsigned char)*s2)
            return false;
        s2++;
        u++;
    }

    return u == uend && *s2 == 0;
}

bool operator<(const UString& s1, const UString& s2)
{
    const int l1 = s1.size();
    const int l2 = s2.size();
    const int lmin = l1 < l2 ? l1 : l2;
    const UChar* c1 = s1.data();
    const UChar* c2 = s2.data();
    int l = 0;
    while (l < lmin && *c1 == *c2) {
        c1++;
        c2++;
        l++;
    }
    if (l < lmin)
        return (c1[0] < c2[0]);

    return (l1 < l2);
}

bool operator>(const UString& s1, const UString& s2)
{
    const int l1 = s1.size();
    const int l2 = s2.size();
    const int lmin = l1 < l2 ? l1 : l2;
    const UChar* c1 = s1.data();
    const UChar* c2 = s2.data();
    int l = 0;
    while (l < lmin && *c1 == *c2) {
        c1++;
        c2++;
        l++;
    }
    if (l < lmin)
        return (c1[0] > c2[0]);

    return (l1 > l2);
}

int compare(const UString& s1, const UString& s2)
{
    const int l1 = s1.size();
    const int l2 = s2.size();
    const int lmin = l1 < l2 ? l1 : l2;
    const UChar* c1 = s1.data();
    const UChar* c2 = s2.data();
    int l = 0;
    while (l < lmin && *c1 == *c2) {
        c1++;
        c2++;
        l++;
    }

    if (l < lmin)
        return (c1[0] > c2[0]) ? 1 : -1;

    if (l1 == l2)
        return 0;

    return (l1 > l2) ? 1 : -1;
}

bool equal(const UString::Rep* r, const UString::Rep* b)
{
    int length = r->len;
    if (length != b->len)
        return false;
    const UChar* d = r->data();
    const UChar* s = b->data();
    for (int i = 0; i != length; ++i) {
        if (d[i] != s[i])
            return false;
    }
    return true;
}

CString UString::UTF8String(bool strict) const
{
    // Allocate a buffer big enough to hold all the characters.
    const int length = size();
    Vector<char, 1024> buffer(length * 3);

    // Convert to runs of 8-bit characters.
    char* p = buffer.data();
    const UChar* d = reinterpret_cast<const UChar*>(&data()[0]);
    ConversionResult result = convertUTF16ToUTF8(&d, d + length, &p, p + buffer.size(), strict);
    if (result != conversionOK)
        return CString();

    return CString(buffer.data(), p - buffer.data());
}

// For use in error handling code paths -- having this not be inlined helps avoid PIC branches to fetch the global on Mac OS X.
NEVER_INLINE void UString::makeNull()
{
    m_rep = &Rep::null();
}

// For use in error handling code paths -- having this not be inlined helps avoid PIC branches to fetch the global on Mac OS X.
NEVER_INLINE UString::Rep* UString::nullRep()
{
    return &Rep::null();
}

} // namespace JSC