| @@ 3009-3015 (lines=7) @@ | ||
| 3006 | if (0 == (0x80 & $in)) { |
|
| 3007 | // US-ASCII, pass straight through. |
|
| 3008 | $mBytes = 1; |
|
| 3009 | } elseif (0xC0 == (0xE0 & $in)) { |
|
| 3010 | // First octet of 2 octet sequence. |
|
| 3011 | $mUcs4 = $in; |
|
| 3012 | $mUcs4 = ($mUcs4 & 0x1F) << 6; |
|
| 3013 | $mState = 1; |
|
| 3014 | $mBytes = 2; |
|
| 3015 | } elseif (0xE0 == (0xF0 & $in)) { |
|
| 3016 | // First octet of 3 octet sequence. |
|
| 3017 | $mUcs4 = $in; |
|
| 3018 | $mUcs4 = ($mUcs4 & 0x0F) << 12; |
|
| @@ 3021-3027 (lines=7) @@ | ||
| 3018 | $mUcs4 = ($mUcs4 & 0x0F) << 12; |
|
| 3019 | $mState = 2; |
|
| 3020 | $mBytes = 3; |
|
| 3021 | } elseif (0xF0 == (0xF8 & $in)) { |
|
| 3022 | // First octet of 4 octet sequence. |
|
| 3023 | $mUcs4 = $in; |
|
| 3024 | $mUcs4 = ($mUcs4 & 0x07) << 18; |
|
| 3025 | $mState = 3; |
|
| 3026 | $mBytes = 4; |
|
| 3027 | } elseif (0xF8 == (0xFC & $in)) { |
|
| 3028 | /* First octet of 5 octet sequence. |
|
| 3029 | * |
|
| 3030 | * This is illegal because the encoded codepoint must be either |
|
| @@ 3040-3046 (lines=7) @@ | ||
| 3037 | $mUcs4 = ($mUcs4 & 0x03) << 24; |
|
| 3038 | $mState = 4; |
|
| 3039 | $mBytes = 5; |
|
| 3040 | } elseif (0xFC == (0xFE & $in)) { |
|
| 3041 | // First octet of 6 octet sequence, see comments for 5 octet sequence. |
|
| 3042 | $mUcs4 = $in; |
|
| 3043 | $mUcs4 = ($mUcs4 & 1) << 30; |
|
| 3044 | $mState = 5; |
|
| 3045 | $mBytes = 6; |
|
| 3046 | } else { |
|
| 3047 | /* Current octet is neither in the US-ASCII range nor a legal first |
|
| 3048 | * octet of a multi-octet sequence. |
|
| 3049 | */ |
|