blob: b70d505ce60cb193d3540451a1718d0db8ce31e3 [file] [log] [blame]
/*
* Copyright (C) 1999-2000 Harri Porten (porten@kde.org)
* Copyright (C) 2004, 2005, 2006, 2007, 2008, 2009 Apple Inc. All rights reserved.
* Copyright (C) 2007 Cameron Zwarich (cwzwarich@uwaterloo.ca)
* Copyright (C) 2009 Google Inc. All rights reserved.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Library General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Library General Public License for more details.
*
* You should have received a copy of the GNU Library General Public License
* along with this library; see the file COPYING.LIB. If not, write to
* the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
* Boston, MA 02110-1301, USA.
*
*/
#include "config.h"
#include "UString.h"
#include "JSGlobalObjectFunctions.h"
#include "Heap.h"
#include "Identifier.h"
#include "Operations.h"
#include <ctype.h>
#include <limits.h>
#include <limits>
#include <stdio.h>
#include <stdlib.h>
#include <wtf/ASCIICType.h>
#include <wtf/Assertions.h>
#include <wtf/DecimalNumber.h>
#include <wtf/MathExtras.h>
#include <wtf/StringExtras.h>
#include <wtf/Vector.h>
#include <wtf/unicode/UTF8.h>
#if HAVE(STRINGS_H)
#include <strings.h>
#endif
using namespace WTF;
using namespace WTF::Unicode;
using namespace std;
namespace JSC {
extern const double NaN;
extern const double Inf;
COMPILE_ASSERT(sizeof(UString) == sizeof(void*), UString_should_stay_small);
// Construct a string with UTF-16 data.
UString::UString(const UChar* characters, unsigned length)
: m_impl(characters ? StringImpl::create(characters, length) : 0)
{
}
// Construct a string with UTF-16 data, from a null-terminated source.
UString::UString(const UChar* characters)
{
if (!characters)
return;
int length = 0;
while (characters[length] != UChar(0))
++length;
m_impl = StringImpl::create(characters, length);
}
// Construct a string with latin1 data.
UString::UString(const char* characters, unsigned length)
: m_impl(characters ? StringImpl::create(characters, length) : 0)
{
}
// Construct a string with latin1 data, from a null-terminated source.
UString::UString(const char* characters)
: m_impl(characters ? StringImpl::create(characters) : 0)
{
}
UString UString::number(int i)
{
UChar buf[1 + sizeof(i) * 3];
UChar* end = buf + WTF_ARRAY_LENGTH(buf);
UChar* p = end;
if (i == 0)
*--p = '0';
else if (i == INT_MIN) {
char minBuf[1 + sizeof(i) * 3];
snprintf(minBuf, sizeof(minBuf), "%d", INT_MIN);
return UString(minBuf);
} else {
bool negative = false;
if (i < 0) {
negative = true;
i = -i;
}
while (i) {
*--p = static_cast<unsigned short>((i % 10) + '0');
i /= 10;
}
if (negative)
*--p = '-';
}
return UString(p, static_cast<unsigned>(end - p));
}
UString UString::number(long long i)
{
UChar buf[1 + sizeof(i) * 3];
UChar* end = buf + WTF_ARRAY_LENGTH(buf);
UChar* p = end;
if (i == 0)
*--p = '0';
else if (i == std::numeric_limits<long long>::min()) {
char minBuf[1 + sizeof(i) * 3];
#if OS(WINDOWS)
snprintf(minBuf, sizeof(minBuf), "%I64d", std::numeric_limits<long long>::min());
#else
snprintf(minBuf, sizeof(minBuf), "%lld", std::numeric_limits<long long>::min());
#endif
return UString(minBuf);
} else {
bool negative = false;
if (i < 0) {
negative = true;
i = -i;
}
while (i) {
*--p = static_cast<unsigned short>((i % 10) + '0');
i /= 10;
}
if (negative)
*--p = '-';
}
return UString(p, static_cast<unsigned>(end - p));
}
UString UString::number(unsigned u)
{
UChar buf[sizeof(u) * 3];
UChar* end = buf + WTF_ARRAY_LENGTH(buf);
UChar* p = end;
if (u == 0)
*--p = '0';
else {
while (u) {
*--p = static_cast<unsigned short>((u % 10) + '0');
u /= 10;
}
}
return UString(p, static_cast<unsigned>(end - p));
}
UString UString::number(long l)
{
UChar buf[1 + sizeof(l) * 3];
UChar* end = buf + WTF_ARRAY_LENGTH(buf);
UChar* p = end;
if (l == 0)
*--p = '0';
else if (l == LONG_MIN) {
char minBuf[1 + sizeof(l) * 3];
snprintf(minBuf, sizeof(minBuf), "%ld", LONG_MIN);
return UString(minBuf);
} else {
bool negative = false;
if (l < 0) {
negative = true;
l = -l;
}
while (l) {
*--p = static_cast<unsigned short>((l % 10) + '0');
l /= 10;
}
if (negative)
*--p = '-';
}
return UString(p, end - p);
}
UString UString::number(double d)
{
NumberToStringBuffer buffer;
unsigned length = numberToString(d, buffer);
return UString(buffer, length);
}
UString UString::substringSharingImpl(unsigned offset, unsigned length) const
{
// FIXME: We used to check against a limit of Heap::minExtraCost / sizeof(UChar).
unsigned stringLength = this->length();
offset = min(offset, stringLength);
length = min(length, stringLength - offset);
if (!offset && length == stringLength)
return *this;
return UString(StringImpl::create(m_impl, offset, length));
}
bool operator==(const UString& s1, const char *s2)
{
if (s2 == 0)
return s1.isEmpty();
const UChar* u = s1.characters();
const UChar* uend = u + s1.length();
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 unsigned l1 = s1.length();
const unsigned l2 = s2.length();
const unsigned lmin = l1 < l2 ? l1 : l2;
const UChar* c1 = s1.characters();
const UChar* c2 = s2.characters();
unsigned 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 unsigned l1 = s1.length();
const unsigned l2 = s2.length();
const unsigned lmin = l1 < l2 ? l1 : l2;
const UChar* c1 = s1.characters();
const UChar* c2 = s2.characters();
unsigned l = 0;
while (l < lmin && *c1 == *c2) {
c1++;
c2++;
l++;
}
if (l < lmin)
return (c1[0] > c2[0]);
return (l1 > l2);
}
CString UString::ascii() const
{
// Basic Latin1 (ISO) encoding - Unicode characters 0..255 are
// preserved, characters outside of this range are converted to '?'.
unsigned length = this->length();
const UChar* characters = this->characters();
char* characterBuffer;
CString result = CString::newUninitialized(length, characterBuffer);
for (unsigned i = 0; i < length; ++i) {
UChar ch = characters[i];
characterBuffer[i] = ch && (ch < 0x20 || ch >= 0x7f) ? '?' : ch;
}
return result;
}
CString UString::latin1() const
{
// Basic Latin1 (ISO) encoding - Unicode characters 0..255 are
// preserved, characters outside of this range are converted to '?'.
unsigned length = this->length();
const UChar* characters = this->characters();
char* characterBuffer;
CString result = CString::newUninitialized(length, characterBuffer);
for (unsigned i = 0; i < length; ++i) {
UChar ch = characters[i];
characterBuffer[i] = ch > 0xff ? '?' : ch;
}
return result;
}
// Helper to write a three-byte UTF-8 code point to the buffer, caller must check room is available.
static inline void putUTF8Triple(char*& buffer, UChar ch)
{
ASSERT(ch >= 0x0800);
*buffer++ = static_cast<char>(((ch >> 12) & 0x0F) | 0xE0);
*buffer++ = static_cast<char>(((ch >> 6) & 0x3F) | 0x80);
*buffer++ = static_cast<char>((ch & 0x3F) | 0x80);
}
CString UString::utf8(bool strict) const
{
unsigned length = this->length();
const UChar* characters = this->characters();
// Allocate a buffer big enough to hold all the characters
// (an individual UTF-16 UChar can only expand to 3 UTF-8 bytes).
// Optimization ideas, if we find this function is hot:
// * We could speculatively create a CStringBuffer to contain 'length'
// characters, and resize if necessary (i.e. if the buffer contains
// non-ascii characters). (Alternatively, scan the buffer first for
// ascii characters, so we know this will be sufficient).
// * We could allocate a CStringBuffer with an appropriate size to
// have a good chance of being able to write the string into the
// buffer without reallocing (say, 1.5 x length).
if (length > numeric_limits<unsigned>::max() / 3)
return CString();
Vector<char, 1024> bufferVector(length * 3);
char* buffer = bufferVector.data();
ConversionResult result = convertUTF16ToUTF8(&characters, characters + length, &buffer, buffer + bufferVector.size(), strict);
ASSERT(result != targetExhausted); // (length * 3) should be sufficient for any conversion
// Only produced from strict conversion.
if (result == sourceIllegal)
return CString();
// Check for an unconverted high surrogate.
if (result == sourceExhausted) {
if (strict)
return CString();
// This should be one unpaired high surrogate. Treat it the same
// was as an unpaired high surrogate would have been handled in
// the middle of a string with non-strict conversion - which is
// to say, simply encode it to UTF-8.
ASSERT((characters + 1) == (this->characters() + length));
ASSERT((*characters >= 0xD800) && (*characters <= 0xDBFF));
// There should be room left, since one UChar hasn't been converted.
ASSERT((buffer + 3) <= (buffer + bufferVector.size()));
putUTF8Triple(buffer, *characters);
}
return CString(bufferVector.data(), buffer - bufferVector.data());
}
} // namespace JSC