| OLD | NEW |
| 1 /* | 1 /* |
| 2 * Copyright (C) 2004, 2006, 2008, 2010 Apple Inc. All rights reserved. | 2 * Copyright (C) 2004, 2006, 2008, 2010 Apple Inc. All rights reserved. |
| 3 * | 3 * |
| 4 * Redistribution and use in source and binary forms, with or without | 4 * Redistribution and use in source and binary forms, with or without |
| 5 * modification, are permitted provided that the following conditions | 5 * modification, are permitted provided that the following conditions |
| 6 * are met: | 6 * are met: |
| 7 * 1. Redistributions of source code must retain the above copyright | 7 * 1. Redistributions of source code must retain the above copyright |
| 8 * notice, this list of conditions and the following disclaimer. | 8 * notice, this list of conditions and the following disclaimer. |
| 9 * 2. Redistributions in binary form must reproduce the above copyright | 9 * 2. Redistributions in binary form must reproduce the above copyright |
| 10 * notice, this list of conditions and the following disclaimer in the | 10 * notice, this list of conditions and the following disclaimer in the |
| (...skipping 18 matching lines...) Expand all Loading... |
| 29 #include "wtf/PassOwnPtr.h" | 29 #include "wtf/PassOwnPtr.h" |
| 30 #include "wtf/text/CString.h" | 30 #include "wtf/text/CString.h" |
| 31 #include "wtf/text/CharacterNames.h" | 31 #include "wtf/text/CharacterNames.h" |
| 32 #include "wtf/text/StringBuffer.h" | 32 #include "wtf/text/StringBuffer.h" |
| 33 #include "wtf/text/WTFString.h" | 33 #include "wtf/text/WTFString.h" |
| 34 | 34 |
| 35 using namespace std; | 35 using namespace std; |
| 36 | 36 |
| 37 namespace WTF { | 37 namespace WTF { |
| 38 | 38 |
| 39 void TextCodecUTF16::registerEncodingNames(EncodingNameRegistrar registrar) | 39 void TextCodecUTF16::registerEncodingNames(EncodingNameRegistrar registrar) { |
| 40 { | 40 registrar("UTF-16LE", "UTF-16LE"); |
| 41 registrar("UTF-16LE", "UTF-16LE"); | 41 registrar("UTF-16BE", "UTF-16BE"); |
| 42 registrar("UTF-16BE", "UTF-16BE"); | |
| 43 | 42 |
| 44 registrar("ISO-10646-UCS-2", "UTF-16LE"); | 43 registrar("ISO-10646-UCS-2", "UTF-16LE"); |
| 45 registrar("UCS-2", "UTF-16LE"); | 44 registrar("UCS-2", "UTF-16LE"); |
| 46 registrar("UTF-16", "UTF-16LE"); | 45 registrar("UTF-16", "UTF-16LE"); |
| 47 registrar("Unicode", "UTF-16LE"); | 46 registrar("Unicode", "UTF-16LE"); |
| 48 registrar("csUnicode", "UTF-16LE"); | 47 registrar("csUnicode", "UTF-16LE"); |
| 49 registrar("unicodeFEFF", "UTF-16LE"); | 48 registrar("unicodeFEFF", "UTF-16LE"); |
| 50 | 49 |
| 51 registrar("unicodeFFFE", "UTF-16BE"); | 50 registrar("unicodeFFFE", "UTF-16BE"); |
| 52 } | 51 } |
| 53 | 52 |
| 54 static PassOwnPtr<TextCodec> newStreamingTextDecoderUTF16LE(const TextEncoding&,
const void*) | 53 static PassOwnPtr<TextCodec> newStreamingTextDecoderUTF16LE(const TextEncoding&,
const void*) { |
| 55 { | 54 return adoptPtr(new TextCodecUTF16(true)); |
| 56 return adoptPtr(new TextCodecUTF16(true)); | |
| 57 } | 55 } |
| 58 | 56 |
| 59 static PassOwnPtr<TextCodec> newStreamingTextDecoderUTF16BE(const TextEncoding&,
const void*) | 57 static PassOwnPtr<TextCodec> newStreamingTextDecoderUTF16BE(const TextEncoding&,
const void*) { |
| 60 { | 58 return adoptPtr(new TextCodecUTF16(false)); |
| 61 return adoptPtr(new TextCodecUTF16(false)); | |
| 62 } | 59 } |
| 63 | 60 |
| 64 void TextCodecUTF16::registerCodecs(TextCodecRegistrar registrar) | 61 void TextCodecUTF16::registerCodecs(TextCodecRegistrar registrar) { |
| 65 { | 62 registrar("UTF-16LE", newStreamingTextDecoderUTF16LE, 0); |
| 66 registrar("UTF-16LE", newStreamingTextDecoderUTF16LE, 0); | 63 registrar("UTF-16BE", newStreamingTextDecoderUTF16BE, 0); |
| 67 registrar("UTF-16BE", newStreamingTextDecoderUTF16BE, 0); | |
| 68 } | 64 } |
| 69 | 65 |
| 70 String TextCodecUTF16::decode(const char* bytes, size_t length, FlushBehavior fl
ush, bool, bool& sawError) | 66 String TextCodecUTF16::decode(const char* bytes, size_t length, FlushBehavior fl
ush, bool, bool& sawError) { |
| 71 { | 67 // For compatibility reasons, ignore flush from fetch EOF. |
| 72 // For compatibility reasons, ignore flush from fetch EOF. | 68 const bool reallyFlush = flush != DoNotFlush && flush != FetchEOF; |
| 73 const bool reallyFlush = flush != DoNotFlush && flush != FetchEOF; | |
| 74 | 69 |
| 75 if (!length) { | 70 if (!length) { |
| 76 if (!reallyFlush || !m_haveBufferedByte) | 71 if (!reallyFlush || !m_haveBufferedByte) |
| 77 return String(); | 72 return String(); |
| 78 sawError = true; | 73 sawError = true; |
| 79 return String(&replacementCharacter, 1); | 74 return String(&replacementCharacter, 1); |
| 75 } |
| 76 |
| 77 // FIXME: This should generate an error if there is an unpaired surrogate. |
| 78 |
| 79 const unsigned char* p = reinterpret_cast<const unsigned char*>(bytes); |
| 80 size_t numBytes = length + m_haveBufferedByte; |
| 81 size_t numCharsIn = numBytes / 2; |
| 82 size_t numCharsOut = ((numBytes & 1) && reallyFlush) ? numCharsIn + 1 : numCha
rsIn; |
| 83 |
| 84 StringBuffer<UChar> buffer(numCharsOut); |
| 85 UChar* q = buffer.characters(); |
| 86 |
| 87 if (m_haveBufferedByte) { |
| 88 UChar c; |
| 89 if (m_littleEndian) |
| 90 c = m_bufferedByte | (p[0] << 8); |
| 91 else |
| 92 c = (m_bufferedByte << 8) | p[0]; |
| 93 *q++ = c; |
| 94 m_haveBufferedByte = false; |
| 95 p += 1; |
| 96 numCharsIn -= 1; |
| 97 } |
| 98 |
| 99 if (m_littleEndian) { |
| 100 for (size_t i = 0; i < numCharsIn; ++i) { |
| 101 UChar c = p[0] | (p[1] << 8); |
| 102 p += 2; |
| 103 *q++ = c; |
| 80 } | 104 } |
| 105 } else { |
| 106 for (size_t i = 0; i < numCharsIn; ++i) { |
| 107 UChar c = (p[0] << 8) | p[1]; |
| 108 p += 2; |
| 109 *q++ = c; |
| 110 } |
| 111 } |
| 81 | 112 |
| 82 // FIXME: This should generate an error if there is an unpaired surrogate. | 113 if (numBytes & 1) { |
| 114 ASSERT(!m_haveBufferedByte); |
| 83 | 115 |
| 84 const unsigned char* p = reinterpret_cast<const unsigned char*>(bytes); | 116 if (reallyFlush) { |
| 85 size_t numBytes = length + m_haveBufferedByte; | 117 sawError = true; |
| 86 size_t numCharsIn = numBytes / 2; | 118 *q++ = replacementCharacter; |
| 87 size_t numCharsOut = ((numBytes & 1) && reallyFlush) ? numCharsIn + 1 : numC
harsIn; | 119 } else { |
| 120 m_haveBufferedByte = true; |
| 121 m_bufferedByte = p[0]; |
| 122 } |
| 123 } |
| 88 | 124 |
| 89 StringBuffer<UChar> buffer(numCharsOut); | 125 buffer.shrink(q - buffer.characters()); |
| 90 UChar* q = buffer.characters(); | |
| 91 | 126 |
| 92 if (m_haveBufferedByte) { | 127 return String::adopt(buffer); |
| 93 UChar c; | |
| 94 if (m_littleEndian) | |
| 95 c = m_bufferedByte | (p[0] << 8); | |
| 96 else | |
| 97 c = (m_bufferedByte << 8) | p[0]; | |
| 98 *q++ = c; | |
| 99 m_haveBufferedByte = false; | |
| 100 p += 1; | |
| 101 numCharsIn -= 1; | |
| 102 } | |
| 103 | |
| 104 if (m_littleEndian) { | |
| 105 for (size_t i = 0; i < numCharsIn; ++i) { | |
| 106 UChar c = p[0] | (p[1] << 8); | |
| 107 p += 2; | |
| 108 *q++ = c; | |
| 109 } | |
| 110 } else { | |
| 111 for (size_t i = 0; i < numCharsIn; ++i) { | |
| 112 UChar c = (p[0] << 8) | p[1]; | |
| 113 p += 2; | |
| 114 *q++ = c; | |
| 115 } | |
| 116 } | |
| 117 | |
| 118 if (numBytes & 1) { | |
| 119 ASSERT(!m_haveBufferedByte); | |
| 120 | |
| 121 if (reallyFlush) { | |
| 122 sawError = true; | |
| 123 *q++ = replacementCharacter; | |
| 124 } else { | |
| 125 m_haveBufferedByte = true; | |
| 126 m_bufferedByte = p[0]; | |
| 127 } | |
| 128 } | |
| 129 | |
| 130 buffer.shrink(q - buffer.characters()); | |
| 131 | |
| 132 return String::adopt(buffer); | |
| 133 } | 128 } |
| 134 | 129 |
| 135 CString TextCodecUTF16::encode(const UChar* characters, size_t length, Unencodab
leHandling) | 130 CString TextCodecUTF16::encode(const UChar* characters, size_t length, Unencodab
leHandling) { |
| 136 { | 131 // We need to be sure we can double the length without overflowing. |
| 137 // We need to be sure we can double the length without overflowing. | 132 // Since the passed-in length is the length of an actual existing |
| 138 // Since the passed-in length is the length of an actual existing | 133 // character buffer, each character is two bytes, and we know |
| 139 // character buffer, each character is two bytes, and we know | 134 // the buffer doesn't occupy the entire address space, we can |
| 140 // the buffer doesn't occupy the entire address space, we can | 135 // assert here that doubling the length does not overflow size_t |
| 141 // assert here that doubling the length does not overflow size_t | 136 // and there's no need for a runtime check. |
| 142 // and there's no need for a runtime check. | 137 ASSERT(length <= numeric_limits<size_t>::max() / 2); |
| 143 ASSERT(length <= numeric_limits<size_t>::max() / 2); | |
| 144 | 138 |
| 145 char* bytes; | 139 char* bytes; |
| 146 CString result = CString::newUninitialized(length * 2, bytes); | 140 CString result = CString::newUninitialized(length * 2, bytes); |
| 147 | 141 |
| 148 // FIXME: CString is not a reasonable data structure for encoded UTF-16, whi
ch will have | 142 // FIXME: CString is not a reasonable data structure for encoded UTF-16, which
will have |
| 149 // null characters inside it. Perhaps the result of encode should not be a C
String. | 143 // null characters inside it. Perhaps the result of encode should not be a CSt
ring. |
| 150 if (m_littleEndian) { | 144 if (m_littleEndian) { |
| 151 for (size_t i = 0; i < length; ++i) { | 145 for (size_t i = 0; i < length; ++i) { |
| 152 UChar c = characters[i]; | 146 UChar c = characters[i]; |
| 153 bytes[i * 2] = static_cast<char>(c); | 147 bytes[i * 2] = static_cast<char>(c); |
| 154 bytes[i * 2 + 1] = c >> 8; | 148 bytes[i * 2 + 1] = c >> 8; |
| 155 } | |
| 156 } else { | |
| 157 for (size_t i = 0; i < length; ++i) { | |
| 158 UChar c = characters[i]; | |
| 159 bytes[i * 2] = c >> 8; | |
| 160 bytes[i * 2 + 1] = static_cast<char>(c); | |
| 161 } | |
| 162 } | 149 } |
| 150 } else { |
| 151 for (size_t i = 0; i < length; ++i) { |
| 152 UChar c = characters[i]; |
| 153 bytes[i * 2] = c >> 8; |
| 154 bytes[i * 2 + 1] = static_cast<char>(c); |
| 155 } |
| 156 } |
| 163 | 157 |
| 164 return result; | 158 return result; |
| 165 } | 159 } |
| 166 | 160 |
| 167 CString TextCodecUTF16::encode(const LChar* characters, size_t length, Unencodab
leHandling) | 161 CString TextCodecUTF16::encode(const LChar* characters, size_t length, Unencodab
leHandling) { |
| 168 { | 162 // In the LChar case, we do actually need to perform this check in release. :
) |
| 169 // In the LChar case, we do actually need to perform this check in release.
:) | 163 RELEASE_ASSERT(length <= numeric_limits<size_t>::max() / 2); |
| 170 RELEASE_ASSERT(length <= numeric_limits<size_t>::max() / 2); | |
| 171 | 164 |
| 172 char* bytes; | 165 char* bytes; |
| 173 CString result = CString::newUninitialized(length * 2, bytes); | 166 CString result = CString::newUninitialized(length * 2, bytes); |
| 174 | 167 |
| 175 if (m_littleEndian) { | 168 if (m_littleEndian) { |
| 176 for (size_t i = 0; i < length; ++i) { | 169 for (size_t i = 0; i < length; ++i) { |
| 177 bytes[i * 2] = characters[i]; | 170 bytes[i * 2] = characters[i]; |
| 178 bytes[i * 2 + 1] = 0; | 171 bytes[i * 2 + 1] = 0; |
| 179 } | |
| 180 } else { | |
| 181 for (size_t i = 0; i < length; ++i) { | |
| 182 bytes[i * 2] = 0; | |
| 183 bytes[i * 2 + 1] = characters[i]; | |
| 184 } | |
| 185 } | 172 } |
| 173 } else { |
| 174 for (size_t i = 0; i < length; ++i) { |
| 175 bytes[i * 2] = 0; |
| 176 bytes[i * 2 + 1] = characters[i]; |
| 177 } |
| 178 } |
| 186 | 179 |
| 187 return result; | 180 return result; |
| 188 } | 181 } |
| 189 | 182 |
| 190 } // namespace WTF | 183 } // namespace WTF |
| OLD | NEW |