ustring.cpp

来自「konqueror3 embedded版本, KDE环境下的当家浏览器的嵌入式版」· C++ 代码 · 共 978 行 · 第 1/2 页

    delete [] statBuffer;
    statBuffer = new char [neededSize];
    statBufferSize = neededSize;
  }

  const UChar *p = data();
  char *q = statBuffer;
  const UChar *limit = p + length;
  while (p != limit) {
    *q = p->uc;
    ++p;
    ++q;
  }
  *q = '\0';

  return statBuffer;
}

#ifdef KJS_DEBUG_MEM
void UString::globalClear()
{
  delete [] statBuffer;
  statBuffer = 0;
  statBufferSize = 0;
}
#endif

UString &UString::operator=(const char *c)
{
  int l = c ? strlen(c) : 0;
  UChar *d;
  if (rep->rc == 1 && l <= rep->capacity) {
    d = rep->dat;
    rep->_hash = 0;
  } else {
    release();
    d = allocateChars(l);
    rep = Rep::create(d, l);
  }
  for (int i = 0; i < l; i++)
    d[i].uc = c[i];

  return *this;
}

UString &UString::operator=(const UString &str)
{
  str.rep->ref();
  release();
  rep = str.rep;

  return *this;
}

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

  return true;
}

UChar UString::operator[](int pos) const
{
  if (pos >= size())
    return UChar::null;

  return ((UChar *)data())[pos];
}

UCharReference UString::operator[](int pos)
{
  /* TODO: boundary check */
  return UCharReference(this, pos);
}

static int skipInfString(const char *start)
{
  const char *c = start;
  if (*c == '+' || *c == '-')
    c++;
  if (!strncmp(c,"Infinity",8))
    return c+8-start;

  while (*c >= '0' && *c <= '9')
    c++;
  const char * const at_dot = c;
  if (*c == '.')
    c++;
  while (*c >= '0' && *c <= '9')
    c++;

  // don't accept a single dot as a number
  if (c - at_dot == 1 && *at_dot == '.')
    return at_dot-start;

  if (*c != 'e')
    return c-start;

  c++;
  if (*c == '+' || *c == '-')
    c++;
  while (*c >= '0' && *c <= '9')
    c++;
  return c-start;
}

double UString::toDouble(bool tolerateTrailingJunk, bool tolerateEmptyString) const
{
  double d;
  double sign = 1;

  // FIXME: If tolerateTrailingJunk is true, then we want to tolerate non-8-bit junk
  // after the number, so is8Bit is too strict a check.
  if (!is8Bit())
    return NaN;

  const char *c = ascii();

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

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

  if (*c == '-') {
    sign = -1;
    c++;
  }
  else if (*c == '+') {
    sign = 1;
    c++;
  }

  // hex number ?
  if (*c == '0' && (*(c+1) == 'x' || *(c+1) == 'X')) {
    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;
    }
  } else {
    // regular number ?
    char *end;
    d = kjs_strtod(c, &end);
    if ((d != 0.0 || end != c) && d != HUGE_VAL && d != -HUGE_VAL) {
      c = end;
    } else {
      // infinity ?

      int count = skipInfString(c);
      if (count == 0)
	return NaN;
      d = Inf;
      c += count;
    }
  }

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

  return d*sign;
}

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

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

unsigned long UString::toULong(bool *ok, bool tolerateEmptyString) const
{
  double d = toDouble(false, tolerateEmptyString);
  bool b = true;

  if (isNaN(d) || d != static_cast<unsigned long>(d)) {
    b = false;
    d = 0;
  }

  if (ok)
    *ok = b;

  return static_cast<unsigned long>(d);
}

unsigned long UString::toULong(bool *ok) const
{
  return toULong(ok, true);
}

UString UString::toLower() const
{
  UString u = *this;
  for (int i = 0; i < size(); i++)
    u[i] = u[i].toLower();
  return u;
}

UString UString::toUpper() const
{
  UString u = *this;
  for (int i = 0; i < size(); i++)
    u[i] = u[i].toUpper();
  return u;
}

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

  if (isNaN(d) || d != static_cast<unsigned>(d)) {
    b = false;
    d = 0;
  }

  if (ok)
    *ok = b;

  return static_cast<unsigned>(d);
}

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

  // Empty string is not OK.
  int len = rep->len;
  if (len == 0)
    return 0;
  const UChar *p = rep->dat;
  unsigned short c = p->unicode();

  // 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.
  unsigned int 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)->unicode();
  }
}

// Rule from ECMA 15.2 about what an array index is.
// Must exactly match string form of an unsigned integer, and be less than 2^32 - 1.
unsigned UString::toArrayIndex(bool *ok) const
{
  unsigned i = toStrictUInt32(ok);
  if (i >= 0xFFFFFFFFU && ok)
    *ok = false;
  return i;
}

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

  return -1;
}

UString UString::substr(int pos, int len) const
{
  if (pos < 0)
    pos = 0;
  else if (pos >= (int) size())
    pos = size();
  if (len < 0)
    len = size();
  if (pos + len >= (int) size())
    len = size() - pos;

  UChar *tmp = allocateChars(len);
  memcpy(tmp, data()+pos, len * sizeof(UChar));
  UString result(tmp, len);
  delete [] tmp;

  return result;
}

void UString::attach(Rep *r)
{
  rep = r;
  rep->ref();
}

void UString::detach()
{
  if (rep->rc > 1) {
    int l = size();
    UChar *n = allocateChars(l);
    memcpy(n, data(), l * sizeof(UChar));
    release();
    rep = Rep::create(n, l);
  }
}

void UString::release()
{
  rep->deref();
}

bool KJS::operator==(const UString& s1, const UString& s2)
{
  if (s1.rep->len != s2.rep->len)
    return false;

#ifndef NDEBUG
  if ((s1.isNull() && s2.isEmpty() && !s2.isNull()) ||
      (s2.isNull() && s1.isEmpty() && !s1.isNull()))
    fprintf(stderr,
            "KJS warning: comparison between empty and null string\n");
#endif

  return (memcmp(s1.rep->dat, s2.rep->dat,
		 s1.rep->len * sizeof(UChar)) == 0);
}

bool KJS::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->uc != (unsigned char)*s2)
      return false;
    s2++;
    u++;
  }

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

bool KJS::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->uc < c2->uc);

  return (l1 < l2);
}

int KJS::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->uc > c2->uc) ? 1 : -1;

  if (l1 == l2) {
    return 0;
  }
  return (l1 < l2) ? 1 : -1;
}