Magic-Mirror/gecko-dev
1
0
Fork 0
mirror of https://github.com/mozilla/gecko-dev.git synced 2024-10-13 05:15:45 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

183354 patch by alexey@optus.net (Alexey Chernyak) r=ftang sr=blizzard Make Universal Charset Autodetector recognise UTF by BOM

Browse Source
This commit is contained in:
cbiesinger%web.de 2003-04-08 19:54:42 +00:00
parent 3bc4a0884a
commit 63b8f071bc
1 changed files with 34 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
extensions/universalchardet/src/nsUniversalDetector.cpp
View File

@ -118,15 +118,43 @@ void nsUniversalDetector::HandleData(const char* aBuf, PRUint32 aLen)
if (aLen > 0)
mGotData = PR_TRUE;
//If the data start with BOM, we know it is UCS2
//If the data starts with BOM, we know it is UTF
if (mStart)
{
mStart = PR_FALSE;
if (aLen > 1)
if (aBuf[0] == '\376' && aBuf[1] == '\377')
mDetectedCharset = "UTF-16BE";
else if (aBuf[0] == '\377' && aBuf[1] == '\376')
mDetectedCharset = "UTF-16LE";
if (aLen > 3)
switch (aBuf[0])
{
case 0xEF:
if ((0xBB == aBuf[1]) && (0xBF == aBuf[2]))
// EF BB BF UTF-8 encoded BOM
mDetectedCharset = "UTF-8";
break;
case 0xFE:
if ((0xFF == aBuf[1]) && (0x00 == aBuf[2]) && (0x00 == aBuf[3]))
// FE FF 00 00 UCS-4, unusual octet order BOM (3412)
mDetectedCharset = "X-ISO-10646-UCS-4-3412";
else if (0xFF == aBuf[1])
// FE FF UTF-16, big endian BOM
mDetectedCharset = "UTF-16BE";
break;
case 0x00:
if ((0x00 == aBuf[1]) && (0xFE == aBuf[2]) && (0xFF == aBuf[3]))
// 00 00 FE FF UTF-32, big-endian BOM
mDetectedCharset = "UTF-32BE";
else if ((0x00 == aBuf[1]) && (0xFF == aBuf[2]) && (0xFE == aBuf[3]))
// 00 00 FF FE UCS-4, unusual octet order BOM (2143)
mDetectedCharset = "X-ISO-10646-UCS-4-2143";
break;
case 0xFF:
if ((0xFE == aBuf[1]) && (0x00 == aBuf[2]) && (0x00 == aBuf[3]))
// FF FE 00 00 UTF-32, little-endian BOM
mDetectedCharset = "UTF-32LE";
else if (0xFE == aBuf[1])
// FF FE UTF-16, little endian BOM
mDetectedCharset = "UTF-16LE";
break;
} // switch
if (mDetectedCharset)
{