fontinfo.cpp

linux下的一款播放器

    }

    if (CHARSET__iso_8859_5 == ulCharset  ||
	    CHARSET__koi8_r == ulCharset)
    {
	HX_ASSERT(0  &&  "handle_MacCyrillic(Ukranian?)_charset");
    }
#else
    // /What do you see when an infant eats food for the first time?:
    const UINT8 ucFirstHighChar = FIRST_UPPER_128_CHAR_INDEX;
    if(CHARSET__x_mac_roman == ulCharset)
    {
	//WINDOWS and UNIX need conversion, if mac-roman charset, to the
	// iso-8859-1 charset:
	while (*pBuf)
	{
	    if ((UINT8)(*pBuf) >= ucFirstHighChar)
	    {
		_CHAR tmpCh = macRoman_to_iso88591[
			(UINT8)(*pBuf) - ucFirstHighChar];
		//Translate only if there is a matching character:
		if (UNKNOWN_CHAR_CONVERSION != tmpCh  &&
			UNDEFINED_CHAR_CONVERSION != tmpCh)
		{
		    *pBuf = tmpCh;
		}
	    }
	    pBuf++;
	};
	goto cleanup;
    }

    if (uiMaxSupportedLevel < 1) // /RealText 1.0 is level 0 only.
    {
	goto cleanup;
    }

    if  (CHARSET__us_ascii == ulCharset)
    {
	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if ((UINT8)(*pBuf) >= FIRST_UPPER_128_CHAR_INDEX)
	    {
		// /None of these chars are defined, so do what browsers do
		// and ignore the high-order bit:
		*pBuf &= 0x7F;
	    }
	    pBuf++;
	};
	// /Caller should now be sure to handle the change in actual charset!
    }
    else if  (CHARSET__iso_8859_1 == ulCharset) // /Western European
    {
	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if (uCh >= FIRST_UPPER_ISO_8859_1_INDEX)
	    {
		; // /They're all OK; WinLatin1 maps to iso-8859-1
	    }
	    else if ((UINT8)(*pBuf) >= FIRST_UPPER_128_CHAR_INDEX)
	    {
		// /Convert 0x80-0x9F chars that are not in iso-8859-9 to nbsp:
		*pBuf = (_CHAR)0xA0;
	    }
	    pBuf++;
	};
	// /Caller should now be sure to handle the change in actual charset!
    }
    else if (CHARSET__iso_8859_2 == ulCharset) // /Eastern European
    {
	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if ((UINT8)(*pBuf) >= FIRST_UPPER_ISO_8859_2_INDEX)
	    {
		// /Note: I manually made these translations by comparing the
		// codepages  -eh:
		if (0xA0 == uCh  ||  (uCh >= 0xA2  &&  uCh <= 0xA4)  ||
			0xA7 == uCh  ||  0xA8 == uCh  ||  0xAA == uCh  ||
			0xAD == uCh  ||  (uCh >= 0xAF  &&  uCh <= 0xB0)  ||
			(uCh >= 0xB2  &&  uCh <= 0xB4)  ||  0xB8 == uCh  ||
			0xBA == uCh  ||  (uCh >= 0xC0  &&  uCh <= 0xFF) )
		{
		    ; // /It's OK as is.
		}
		else if (0xA1 == uCh)
		{
		    *pBuf = (_CHAR)0xA5;
		}
		else if (0xA5 == uCh)
		{
		    *pBuf = (_CHAR)0xBC;
		}
		else if (0xA6 == uCh  ||  0xB6 == uCh)
		{
		    *pBuf -= 0x1A; // /A6->8C, B6->9C
		}
		else if (0xA9 == uCh  ||  0xAE == uCh  ||  0xB9 == uCh)
		{
		    *pBuf -= 0x1F; // /A9->8A, AE->8E, B9->9A
		}
		else if (0xAB == uCh  ||  0xBB == uCh)
		{
		    *pBuf -= 0x1E; // /AB->8D, BB->9D
		}
		else if (0xAC == uCh  ||  0xBC == uCh)
		{
		    *pBuf -= 0x1D; // /AC->8F, BC->9F
		}
		else if (0xB1 == uCh)
		{
		    *pBuf = (_CHAR)0xB9;
		}
		else if (0xB5 == uCh)
		{
		    *pBuf = (_CHAR)0xBE;
		}
		else if (0xB7 == uCh)
		{
		    *pBuf = (_CHAR)0xA1;
		}
		else if (0xBD == uCh)
		{
		    *pBuf = (_CHAR)0x94;
		}
		else if (0xBE == uCh  ||  0xBF == uCh)
		{
		    *pBuf -= 0x20; // /BE->9E, BF->9F
		}
	    }
	    else if ((UINT8)(*pBuf) >= FIRST_UPPER_128_CHAR_INDEX)
	    {
		// /Convert 0x80-0x9F chars that are not in iso-8859-5 to nbsp:
		*pBuf = (_CHAR)0xA0;
	    }
	    pBuf++;
	};
	// /Caller should now be sure to handle the change in actual charset!
    }
    else if (CHARSET__iso_8859_5 == ulCharset) // /Cyrillic
    {
	while (*pBuf)
	{
	    if ((UINT8)(*pBuf) >= FIRST_UPPER_ISO_8859_5_INDEX)
	    {
		// /Note: I manually made these translations by comparing the
		// codepages  -eh:
		if ((UINT8)(*pBuf) < 0xB0)
		{
		    *pBuf = (_CHAR)iso88595_to_windows1251_A0_to_AF[
			    (UINT8)(*pBuf) - 0xA0];
		}
		else if ((UINT8)(*pBuf) >= 0xF0)
		{
		    *pBuf = (_CHAR)iso88595_to_windows1251_F0_to_FF[
			    (UINT8)(*pBuf) - 0xF0];
		}
		else // /add 0x10 to convert the middle 4 rows:
		{
		    *pBuf += 0x10;
		}
	    }
	    else if ((UINT8)(*pBuf) >= FIRST_UPPER_128_CHAR_INDEX)
	    {
		// /Convert 0x80-0x9F chars that are not in iso-8859-5 to nbsp:
		*pBuf = (_CHAR)0xA0;
	    }
	    pBuf++;
	};
	// /Caller should now be sure to handle the change in actual charset!
    }
    else if (CHARSET__koi8_r == ulCharset) // /Cyrillic
    {
	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if ((UINT8)(*pBuf) >= ucFirstHighChar)
	    {
		// /Note: I manually made these translations by comparing the
		// codepages  -eh:
		if (uCh >= 0xE0)
		{
		    *pBuf = (_CHAR)koi8r_to_windows1251_E0_to_FF[
			    uCh - 0xE0];
		}
		else if (uCh >= 0xC0) // /and we know it is < 0xE0
		{
		    *pBuf = (_CHAR)koi8r_to_windows1251_E0_to_FF[
			    uCh - 0xC0 /*-0xC0==(0x20-0xE0) */] +
			    KOI8_R_LOWER_CASE_CHAR_OFFSET;
		}
		else if (0x9A == uCh)
		{
		    *pBuf = (_CHAR)0xA0;// /XXXEH- should be non-breaking space
		}
		else if (0x9C == uCh)
		{
		    *pBuf = (_CHAR)0xB0;
		}
		else if (0x9E == uCh)
		{
		    *pBuf = (_CHAR)0xB7;
		}
		else if (0xA3 == uCh)
		{
		    *pBuf = (_CHAR)0xB8;
		}
		else if (0xB3 == uCh)
		{
		    *pBuf = (_CHAR)0xA8;
		}
		else if (0xBF == uCh)
		{
		    *pBuf = (_CHAR)0xA9;
		}
		else if (0x95 == uCh)
		{
		    ; // /It's stays the same.
		}
		else // /All the others are from the original IBM Personal
		    // Computer with the code page CP437's box-drawing
		    // characters.  We *could* bother to convert them (and
		    // draw each in US-ASCII charset), but won't go to that
		    // trouble unless we get complaints which I *strongly*
		    // doubt will happen.  Instead, just draw them as blanks:
		{
		    *pBuf = (_CHAR)0x98; // /0x98 in Windows-1251 is a box
		}
	    }
	    pBuf++;
	};
	// /Caller should now be sure to handle the change in actual charset!
    }
    else if (CHARSET__iso_8859_6 == ulCharset) // /Arabic
    {
#if defined(RTL_PERFORM_REVERSAL_OF_RTL_CHARS)
	// /First, this is Arabic, so we need to reverse the text of each line:
	_CHAR* pTmpBuf = pBuf;
	_CHAR* pNewline = pTmpBuf;
	while (*pTmpBuf  &&  *pNewline)
	{
	    while (*pNewline  &&  '\r' != *pNewline  &&  '\n' != *pNewline)
	    {
		pNewline++;
	    };
	    _CHAR* pFromLastChar = pNewline-1;
	    while (*pTmpBuf  &&  pTmpBuf < pFromLastChar)
	    {
		_CHAR chTmp = *pFromLastChar;
		*pFromLastChar = *pTmpBuf;
		*pTmpBuf = chTmp;
		pTmpBuf++;
		pFromLastChar--;
	    }
	    if ('\r' == *pNewline  &&  '\n' == pNewline[1])
	    {
		pNewline++; // /Treat \r\n as single char.
	    }
	    if (*pNewline)
	    {
		pNewline++; // /Go look for the next one.
	    }
	    pTmpBuf = pNewline;
	}
#endif

	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if (uCh >= FIRST_UPPER_ISO_8859_6_INDEX)
	    {
		// /Note: I manually made these translations by comparing the
		// codepages  -eh:
		if (0xA0 == uCh  ||  0xA4 == uCh  ||  0xAD == uCh  ||
			0xBF == uCh  ||  (uCh >= 0xC1  &&  uCh <= 0xD6))
		{
		    ; // /It's OK as is.
		}
		else if (0xAC == uCh)  // /[Could also translate to 0x60 (?)]
		{
		    *pBuf = (_CHAR)0x91;
		}
		else if (0xBB == uCh)
		{
		    *pBuf = (_CHAR)0xBA;
		}
		else if (uCh >= 0xD7  &&  uCh <= 0xDA)
		{
		    *pBuf += 1;
		}
		else if (uCh >= 0xE0  &&  uCh <= 0xE3)
		{
		    *pBuf -= 4;
		}
		else if (0xE4 == uCh)
		{
		    *pBuf = (_CHAR)0xE1;
		}
		else if (uCh >= 0xE5  &&  uCh <= 0xE8)
		{
		    *pBuf -= 2;
		}
		else if (uCh >= 0xE9  &&  uCh <= 0xEA)
		{
		    *pBuf += 3;
		}
		else if (0xEC == uCh)  // /[Could also translate to 0xF1 (?)]
		{
		    *pBuf = (_CHAR)0x94;
		}
		else if (0xEB == uCh  ||  0xED == uCh  ||  0xEE == uCh)
		{
		    *pBuf +=5;
		}
		else if (0xEF == uCh)  // /[Could also translate to 0xF5 (?)]
		{
		    *pBuf = (_CHAR)0x92;
		}
		else if (0xF0 == uCh)
		{
		    *pBuf +=6;
		}
		else if (0xF1 == uCh)
		{
		    *pBuf +=7;
		}
		else if (0xF2 == uCh)
		{
		    *pBuf += 8;
		}
		else
		{
		    // /Convert chars that are not in iso-8859-6 to nbsp char:
		    *pBuf = (_CHAR)0xA0;
		}
	    }
	    else if ((UINT8)(*pBuf) >= FIRST_UPPER_128_CHAR_INDEX)
	    {
		// /Convert 0x80-0x9F chars that are not in iso-8859-5 to nbsp:
		*pBuf = (_CHAR)0xA0;
	    }
	    pBuf++;
	};
	// /Caller should now be sure to handle the change in actual charset!
    }
    else if (CHARSET__iso_8859_7 == ulCharset) // /Greek
    {
	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if (uCh >= FIRST_UPPER_ISO_8859_7_INDEX)
	    {
		// /Note: I manually made these translations by comparing the
		// codepages  -eh:
		if (0xA0 == uCh  ||  0xA3 == uCh  ||
			(uCh >= 0xA6  &&  uCh <= 0xA9)  ||
			(uCh >= 0xAB  &&  uCh <= 0xAD)  ||
			(uCh >= 0xAF  &&  uCh <= 0xB4)  ||
			(uCh >= 0xB7  &&  uCh <= 0xD1)  ||
			(uCh >= 0xD3  &&  uCh <= 0xFE) )
		{
		    ; // /It's OK as is.
		}
		else if (0xA1 == uCh  ||  0xA2 == uCh)
		{
		    *pBuf -= 0x10; // /A1->91, A2->92
		}
		else if (0xB5 == uCh  ||  0xB6 == uCh)
		{
		    *pBuf -= 0x14;  // /B5->A1, B6->A2.
		}
		else
		{
		    // /Convert chars that are not in iso-8859-7 to nbsp char:
		    *pBuf = (_CHAR)0xA0;
		}
	    }
	    else if ((UINT8)(*pBuf) >= FIRST_UPPER_128_CHAR_INDEX)
	    {
		// /Convert 0x80-0x9F chars that are not in iso-8859-7 to nbsp:
		*pBuf = (_CHAR)0xA0;
	    }
	    pBuf++;
	};
	// /Caller should now be sure to handle the change in actual charset!
    }
    else if  (CHARSET__iso_8859_8 == ulCharset) // /Hebrew
    {
	// /XXXEH- NOTE: iso-8859-8's 0xDF doesn't seem to map to anything on
	// Windows; it appears to be a double underline char (low '=').
	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if (uCh >= FIRST_UPPER_ISO_8859_8_INDEX)
	    {
		if (0xA1 == uCh  ||  (uCh >= 0xBF  &&  uCh <= 0xDE)  ||
			(uCh >= 0xFB  &&  uCh <= 0xFF) )
		{
		    // /Convert chars that are not in iso-8859-8 to nbsp char:
		    *pBuf = (_CHAR)0xA0;
		}
	    }
	    else if ((UINT8)(*pBuf) >= FIRST_UPPER_128_CHAR_INDEX)
	    {
		// /Convert 0x80-0x9F chars that are not in iso-8859-8 to nbsp:
		*pBuf = (_CHAR)0xA0;
	    }
	    pBuf++;
	};
	// /Caller should now be sure to handle the change in actual charset!
    }
    else if  (CHARSET__iso_8859_9 == ulCharset) // /Turkish
    {
	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if (uCh >= FIRST_UPPER_ISO_8859_9_INDEX)
	    {
		; // /All are OK as-is; WinTurkish maps exactly to iso-8859-9
	    }
	    else if ((UINT8)(*pBuf) >= FIRST_UPPER_128_CHAR_INDEX)
	    {
		// /Convert 0x80-0x9F chars that are not in iso-8859-9 to nbsp:
		*pBuf = (_CHAR)0xA0;
	    }
	    pBuf++;
	};
    }
    else if  (CHARSET__iso_ir_166 == ulCharset  ||  // /Thai
	    CHARSET__iso_8859_11 == ulCharset)
    {
	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if (uCh >= FIRST_UPPER_ISO_8859_11_INDEX)
	    {
		if ((uCh >= 0xDB  &&  uCh <= 0xDE)  ||
			(uCh >= 0xFC  &&  uCh <= 0xFF) )
		{
		    // /Convert chars that are not in iso-8859-11 to nbsp char:
		    *pBuf = (_CHAR)0xA0;
		}
	    }
	    else if ((UINT8)(*pBuf) >= FIRST_UPPER_128_CHAR_INDEX)
	    {
		// /Convert 0x80-0x9F chars that are not in iso-8859-5 to nbsp:
		*pBuf = (_CHAR)0xA0;
	    }
	    pBuf++;
	};
	// /Caller should now be sure to handle the change in actual charset!
    }
    else if  (CHARSET__iso_8859_13 == ulCharset) // /Baltic (Latin 7)
    {
	while (*pBuf)
	{
	    UINT8 uCh = (UINT8)(*pBuf);
	    if (uCh >= FIRST_UPPER_ISO_8859_13_INDEX)
	    {
		if (0xA0 == uCh  ||
			(uCh >= 0xA2  &&  uCh <= 0xA4)  ||
			(uCh >= 0xA6  &&  uCh <= 0xB3)  ||
			(uCh >= 0xB5  &&  uCh <= 0xFF) )
		{
		    ; // /It's OK as is.
		}
		else if (0xA1 == uCh)
		{
		    *pBuf = (_CHAR)0x94;
		}
		else if (0xA5 == uCh)
		{