StringView CachedScript::script()
{
if (!m_data)
return { };
if (m_decodingState == NeverDecoded
&& TextEncoding(encoding()).isByteBasedEncoding()
&& m_data->size()
&& charactersAreAllASCII(reinterpret_cast<const LChar*>(m_data->data()), m_data->size())) {
m_decodingState = DataAndDecodedStringHaveSameBytes;
// If the encoded and decoded data are the same, there is no decoded data cost!
setDecodedSize(0);
m_decodedDataDeletionTimer.stop();
m_scriptHash = StringHasher::computeHashAndMaskTop8Bits(reinterpret_cast<const LChar*>(m_data->data()), m_data->size());
}
if (m_decodingState == DataAndDecodedStringHaveSameBytes)
return { reinterpret_cast<const LChar*>(m_data->data()), m_data->size() };
if (!m_script) {
m_script = m_decoder->decodeAndFlush(m_data->data(), encodedSize());
m_scriptHash = m_script.impl()->hash();
m_decodingState = DataAndDecodedStringHaveDifferentBytes;
setDecodedSize(m_script.sizeInBytes());
}
m_decodedDataDeletionTimer.restart();
return m_script;
}
String String::fromUTF8(const LChar* stringStart, size_t length)
{
RELEASE_ASSERT(length <= std::numeric_limits<unsigned>::max());
if (!stringStart)
return String();
if (!length)
return emptyString();
if (charactersAreAllASCII(stringStart, length))
return StringImpl::create(stringStart, length);
Vector<UChar, 1024> buffer(length);
UChar* bufferStart = buffer.data();
UChar* bufferCurrent = bufferStart;
const char* stringCurrent = reinterpret_cast<const char*>(stringStart);
if (convertUTF8ToUTF16(&stringCurrent, reinterpret_cast<const char *>(stringStart + length), &bufferCurrent, bufferCurrent + buffer.size()) != conversionOK)
return String();
unsigned utf16Length = bufferCurrent - bufferStart;
ASSERT(utf16Length < length);
return StringImpl::create(bufferStart, utf16Length);
}
ALWAYS_INLINE static void makeLower(const CharacterType* input, CharacterType* output, unsigned length)
{
// FIXME: If we need Unicode lowercasing here, then we probably want the real kind
// that can potentially change the length of the string rather than the character
// by character kind. If we don't need Unicode lowercasing, it would be good to
// simplify this function.
if (charactersAreAllASCII(input, length)) {
// Fast case for all-ASCII.
for (unsigned i = 0; i < length; i++)
output[i] = toASCIILower(input[i]);
} else {
for (unsigned i = 0; i < length; i++)
output[i] = Unicode::toLower(input[i]);
}
}
static bool isValidYouTubeVideo(const String& path)
{
if (!charactersAreAllASCII(path.characters(), path.length()))
return false;
unsigned int len = path.length();
if (len <= sizeof(slash_v_slash)) // check for more than just /v/
return false;
CString str = path.lower().utf8();
const char* data = str.data();
if (memcmp(data, slash_v_slash, sizeof(slash_v_slash)) != 0)
return false;
// Start after /v/
for (unsigned int i = sizeof(slash_v_slash); i < len; i++) {
char c = data[i];
// Check for alpha-numeric characters only.
if (WTF::isASCIIAlphanumeric(c) || c == '_' || c == '-')
continue;
// The url can have more parameters such as &hl=en after the video id.
// Once we start seeing extra parameters we can return true.
return c == '&' && i > sizeof(slash_v_slash);
}
return true;
}
StringView CachedScript::script()
{
if (!m_script && m_data) {
if (m_ASCIIOptimizationState == Unknown
&& encodingMayBeAllASCII(encoding())
&& m_data->size()
&& charactersAreAllASCII(reinterpret_cast<const LChar*>(m_data->data()), m_data->size())) {
m_script = StringImpl::createWithoutCopying(reinterpret_cast<const LChar*>(m_data->data()), m_data->size());
m_ASCIIOptimizationState = DataAndDecodedStringHaveSameBytes;
// If the encoded and decoded data are the same, there is no decoded data cost!
setDecodedSize(0);
m_decodedDataDeletionTimer.stop();
return m_script;
}
m_script = m_decoder->decodeAndFlush(m_data->data(), encodedSize());
m_ASCIIOptimizationState = DataAndDecodedStringHaveDifferentBytes;
setDecodedSize(m_script.sizeInBytes());
}
if (m_ASCIIOptimizationState == DataAndDecodedStringHaveDifferentBytes)
m_decodedDataDeletionTimer.restart();
return m_script;
}
// When asked for a CJK font with a native name under a non-CJK locale or
// asked for a CJK font with a Romanized name under a CJK locale,
// |GetTextFace| (after |CreateFont*|) returns a 'bogus' value (e.g. Arial).
// This is not consistent with what MSDN says !!
// Therefore, before we call |CreateFont*|, we have to map a Romanized name to
// the corresponding native name under a CJK locale and vice versa
// under a non-CJK locale.
// See the corresponding gecko bugs at
// https://bugzilla.mozilla.org/show_bug.cgi?id=373952
// https://bugzilla.mozilla.org/show_bug.cgi?id=231426
static bool LookupAltName(const String& name, String& altName)
{
struct FontCodepage {
WCHAR* name;
int codePage;
};
struct NamePair {
WCHAR* name;
FontCodepage altNameCodepage;
};
const int japaneseCodepage = 932;
const int simplifiedChineseCodepage = 936;
const int koreanCodepage = 949;
const int traditionalChineseCodepage = 950;
// FIXME(jungshik) : This list probably covers 99% of cases.
// To cover the remaining 1% and cut down the file size,
// consider accessing 'NAME' table of a truetype font
// using |GetFontData| and caching the mapping.
// In the table below, the ASCII keys are all lower-cased for
// case-insensitive matching.
static const NamePair namePairs[] = {
// ＭＳ Ｐゴシック, MS PGothic
{L"\xFF2D\xFF33 \xFF30\x30B4\x30B7\x30C3\x30AF", {L"MS PGothic", japaneseCodepage}},
{L"ms pgothic", {L"\xFF2D\xFF33 \xFF30\x30B4\x30B7\x30C3\x30AF", japaneseCodepage}},
// ＭＳ Ｐ明朝, MS PMincho
{L"\xFF2D\xFF33 \xFF30\x660E\x671D", {L"MS PMincho", japaneseCodepage}},
{L"ms pmincho", {L"\xFF2D\xFF33 \xFF30\x660E\x671D", japaneseCodepage}},
// ＭＳゴシック, MS Gothic
{L"\xFF2D\xFF33 \x30B4\x30B7\x30C3\x30AF", {L"MS Gothic", japaneseCodepage}},
{L"ms gothic", {L"\xFF2D\xFF33 \x30B4\x30B7\x30C3\x30AF", japaneseCodepage}},
// ＭＳ 明朝, MS Mincho
{L"\xFF2D\xFF33 \x660E\x671D", {L"MS Mincho", japaneseCodepage}},
{L"ms mincho", {L"\xFF2D\xFF33 \x660E\x671D", japaneseCodepage}},
// メイリオ, Meiryo
{L"\x30E1\x30A4\x30EA\x30AA", {L"Meiryo", japaneseCodepage}},
{L"meiryo", {L"\x30E1\x30A4\x30EA\x30AA", japaneseCodepage}},
// 바탕, Batang
{L"\xBC14\xD0D5", {L"Batang", koreanCodepage}},
{L"batang", {L"\xBC14\xD0D5", koreanCodepage}},
// 바탕체, Batangche
{L"\xBC14\xD0D5\xCCB4", {L"Batangche", koreanCodepage}},
{L"batangche", {L"\xBC14\xD0D5\xCCB4", koreanCodepage}},
// 굴림, Gulim
{L"\xAD74\xB9BC", {L"Gulim", koreanCodepage}},
{L"gulim", {L"\xAD74\xB9BC", koreanCodepage}},
// 굴림체, Gulimche
{L"\xAD74\xB9BC\xCCB4", {L"Gulimche", koreanCodepage}},
{L"gulimche", {L"\xAD74\xB9BC\xCCB4", koreanCodepage}},
// 돋움, Dotum
{L"\xB3CB\xC6C0", {L"Dotum", koreanCodepage}},
{L"dotum", {L"\xB3CB\xC6C0", koreanCodepage}},
// 돋움체, Dotumche
{L"\xB3CB\xC6C0\xCCB4", {L"Dotumche", koreanCodepage}},
{L"dotumche", {L"\xB3CB\xC6C0\xCCB4", koreanCodepage}},
// 궁서, Gungsuh
{L"\xAD81\xC11C", {L"Gungsuh", koreanCodepage}},
{L"gungsuh", {L"\xAD81\xC11C", koreanCodepage}},
// 궁서체, Gungsuhche
{L"\xAD81\xC11C\xCCB4", {L"Gungsuhche", koreanCodepage}},
{L"gungsuhche", {L"\xAD81\xC11C\xCCB4", koreanCodepage}},
// 맑은 고딕, Malgun Gothic
{L"\xB9D1\xC740 \xACE0\xB515", {L"Malgun Gothic", koreanCodepage}},
{L"malgun gothic", {L"\xB9D1\xC740 \xACE0\xB515", koreanCodepage}},
// 宋体, SimSun
{L"\x5B8B\x4F53", {L"SimSun", simplifiedChineseCodepage}},
{L"simsun", {L"\x5B8B\x4F53", simplifiedChineseCodepage}},
// 宋体-ExtB, SimSun-ExtB
{L"\x5B8B\x4F53-ExtB", {L"SimSun-ExtB", simplifiedChineseCodepage}},
{L"simsun-extb", {L"\x5B8B\x4F53-extb", simplifiedChineseCodepage}},
// 黑体, SimHei
{L"\x9ED1\x4F53", {L"SimHei", simplifiedChineseCodepage}},
{L"simhei", {L"\x9ED1\x4F53", simplifiedChineseCodepage}},
// 新宋体, NSimSun
{L"\x65B0\x5B8B\x4F53", {L"NSimSun", simplifiedChineseCodepage}},
{L"nsimsun", {L"\x65B0\x5B8B\x4F53", simplifiedChineseCodepage}},
// 微软雅黑, Microsoft Yahei
{L"\x5FAE\x8F6F\x96C5\x9ED1", {L"Microsoft Yahei", simplifiedChineseCodepage}},
{L"microsoft yahei", {L"\x5FAE\x8F6F\x96C5\x9ED1", simplifiedChineseCodepage}},
// 仿宋, FangSong
{L"\x4EFF\x5B8B", {L"FangSong", simplifiedChineseCodepage}},
{L"fangsong", {L"\x4EFF\x5B8B", simplifiedChineseCodepage}},
// 楷体, KaiTi
{L"\x6977\x4F53", {L"KaiTi", simplifiedChineseCodepage}},
{L"kaiti", {L"\x6977\x4F53", simplifiedChineseCodepage}},
// 仿宋_GB2312, FangSong_GB2312
{L"\x4EFF\x5B8B_GB2312", {L"FangSong_GB2312", simplifiedChineseCodepage}},
{L"fangsong_gb2312", {L"\x4EFF\x5B8B_gb2312", simplifiedChineseCodepage}},
// 楷体_GB2312, KaiTi_GB2312
{L"\x6977\x4F53", {L"KaiTi_GB2312", simplifiedChineseCodepage}},
{L"kaiti_gb2312", {L"\x6977\x4F53_gb2312", simplifiedChineseCodepage}},
// 新細明體, PMingLiu
{L"\x65B0\x7D30\x660E\x9AD4", {L"PMingLiu", traditionalChineseCodepage}},
{L"pmingliu", {L"\x65B0\x7D30\x660E\x9AD4", traditionalChineseCodepage}},
// 新細明體-ExtB, PMingLiu-ExtB
{L"\x65B0\x7D30\x660E\x9AD4-ExtB", {L"PMingLiu-ExtB", traditionalChineseCodepage}},
{L"pmingliu-extb", {L"\x65B0\x7D30\x660E\x9AD4-extb", traditionalChineseCodepage}},
// 細明體, MingLiu
{L"\x7D30\x660E\x9AD4", {L"MingLiu", traditionalChineseCodepage}},
{L"mingliu", {L"\x7D30\x660E\x9AD4", traditionalChineseCodepage}},
// 細明體-ExtB, MingLiu-ExtB
{L"\x7D30\x660E\x9AD4-ExtB", {L"MingLiu-ExtB", traditionalChineseCodepage}},
{L"mingliu-extb", {L"x65B0\x7D30\x660E\x9AD4-extb", traditionalChineseCodepage}},
// 微軟正黑體, Microsoft JhengHei
{L"\x5FAE\x8EDF\x6B63\x9ED1\x9AD4", {L"Microsoft JhengHei", traditionalChineseCodepage}},
{L"microsoft jhengHei", {L"\x5FAE\x8EDF\x6B63\x9ED1\x9AD4", traditionalChineseCodepage}},
// 標楷體, DFKai-SB
{L"\x6A19\x6977\x9AD4", {L"DFKai-SB", traditionalChineseCodepage}},
{L"dfkai-sb", {L"\x6A19\x6977\x9AD4", traditionalChineseCodepage}},
// WenQuanYi Zen Hei
{L"\x6587\x6cc9\x9a5b\x6b63\x9ed1", {L"WenQuanYi Zen Hei", traditionalChineseCodepage}},
{L"wenquanyi zen hei", {L"\x6587\x6cc9\x9a5b\x6b63\x9ed1", traditionalChineseCodepage}},
// WenQuanYi Zen Hei
{L"\x6587\x6cc9\x9a7f\x6b63\x9ed1", {L"WenQuanYi Zen Hei", simplifiedChineseCodepage}},
{L"wenquanyi zen hei", {L"\x6587\x6cc9\x9a7f\x6b63\x9ed1", simplifiedChineseCodepage}},
// AR PL ShanHeiSun Uni,
{L"\x6587\x9f0e\x0050\x004c\x7d30\x4e0a\x6d77\x5b8b\x0055\x006e\x0069",
{L"AR PL ShanHeiSun Uni", traditionalChineseCodepage}},
{L"ar pl shanheisun uni",
{L"\x6587\x9f0e\x0050\x004c\x7d30\x4e0a\x6d77\x5b8b\x0055\x006e\x0069", traditionalChineseCodepage}},
// AR PL ShanHeiSun Uni,
{L"\x6587\x9f0e\x0050\x004c\x7ec6\x4e0a\x6d77\x5b8b\x0055\x006e\x0069",
{L"AR PL ShanHeiSun Uni", simplifiedChineseCodepage}},
{L"ar pl shanheisun uni",
{L"\x6587\x9f0e\x0050\x004c\x7ec6\x4e0a\x6d77\x5b8b\x0055\x006e\x0069", simplifiedChineseCodepage}},
// AR PL ZenKai Uni
// Traditional Chinese and Simplified Chinese names are
// identical.
{L"\x6587\x0050\x004C\x4E2D\x6977\x0055\x006E\x0069", {L"AR PL ZenKai Uni", traditionalChineseCodepage}},
{L"ar pl zenkai uni", {L"\x6587\x0050\x004C\x4E2D\x6977\x0055\x006E\x0069", traditionalChineseCodepage}},
{L"\x6587\x0050\x004C\x4E2D\x6977\x0055\x006E\x0069", {L"AR PL ZenKai Uni", simplifiedChineseCodepage}},
{L"ar pl zenkai uni", {L"\x6587\x0050\x004C\x4E2D\x6977\x0055\x006E\x0069", simplifiedChineseCodepage}},
};
typedef HashMap<String, const FontCodepage*> NameMap;
static NameMap* fontNameMap = 0;
if (!fontNameMap) {
fontNameMap = new NameMap;
for (size_t i = 0; i < WTF_ARRAY_LENGTH(namePairs); ++i)
fontNameMap->set(String(namePairs[i].name), &(namePairs[i].altNameCodepage));
}
bool isAscii = false;
String n;
// use |lower| only for ASCII names
// For non-ASCII names, we don't want to invoke an expensive
// and unnecessary |lower|.
if (charactersAreAllASCII(name)) {
isAscii = true;
n = name.lower();
} else
n = name;
NameMap::iterator iter = fontNameMap->find(n);
if (iter == fontNameMap->end())
return false;
static int systemCp = ::GetACP();
int fontCp = iter->second->codePage;
if ((isAscii && systemCp == fontCp) || (!isAscii && systemCp != fontCp)) {
altName = String(iter->second->name);
return true;
}
return false;
}
// FIXME: consider adding to WebKit String class
static bool charactersAreAllASCII(const String& s)
{
return charactersAreAllASCII(s.characters(), s.length());
}