| @@ 1125-1131 (lines=7) @@ | ||
| 1122 | if (0 == (0x80 & $in)) { |
|
| 1123 | // US-ASCII, pass straight through. |
|
| 1124 | $mBytes = 1; |
|
| 1125 | } elseif (0xC0 == (0xE0 & $in)) { |
|
| 1126 | // First octet of 2 octet sequence. |
|
| 1127 | $mUcs4 = $in; |
|
| 1128 | $mUcs4 = ($mUcs4 & 0x1F) << 6; |
|
| 1129 | $mState = 1; |
|
| 1130 | $mBytes = 2; |
|
| 1131 | } elseif (0xE0 == (0xF0 & $in)) { |
|
| 1132 | // First octet of 3 octet sequence. |
|
| 1133 | $mUcs4 = $in; |
|
| 1134 | $mUcs4 = ($mUcs4 & 0x0F) << 12; |
|
| @@ 1137-1143 (lines=7) @@ | ||
| 1134 | $mUcs4 = ($mUcs4 & 0x0F) << 12; |
|
| 1135 | $mState = 2; |
|
| 1136 | $mBytes = 3; |
|
| 1137 | } elseif (0xF0 == (0xF8 & $in)) { |
|
| 1138 | // First octet of 4 octet sequence. |
|
| 1139 | $mUcs4 = $in; |
|
| 1140 | $mUcs4 = ($mUcs4 & 0x07) << 18; |
|
| 1141 | $mState = 3; |
|
| 1142 | $mBytes = 4; |
|
| 1143 | } elseif (0xF8 == (0xFC & $in)) { |
|
| 1144 | /* First octet of 5 octet sequence. |
|
| 1145 | * |
|
| 1146 | * This is illegal because the encoded codepoint must be either |
|
| @@ 1156-1162 (lines=7) @@ | ||
| 1153 | $mUcs4 = ($mUcs4 & 0x03) << 24; |
|
| 1154 | $mState = 4; |
|
| 1155 | $mBytes = 5; |
|
| 1156 | } elseif (0xFC == (0xFE & $in)) { |
|
| 1157 | // First octet of 6 octet sequence, see comments for 5 octet sequence. |
|
| 1158 | $mUcs4 = $in; |
|
| 1159 | $mUcs4 = ($mUcs4 & 1) << 30; |
|
| 1160 | $mState = 5; |
|
| 1161 | $mBytes = 6; |
|
| 1162 | } else { |
|
| 1163 | /* Current octet is neither in the US-ASCII range nor a legal first |
|
| 1164 | * octet of a multi-octet sequence. |
|
| 1165 | */ |
|