encoding.c.svn-base

这是一个用于解析xml文件的类库。使用这个类库

		*outlen = out - outstart;
		*inlenb = processed - inb;
	        return(-2);
	    }
        }

	/* assertion: c is a single UTF-4 value */
        if (out >= outend)
	    break;
        if      (c <    0x80) {  *out++=  c;                bits= -6; }
        else if (c <   0x800) {  *out++= ((c >>  6) & 0x1F) | 0xC0;  bits=  0; }
        else if (c < 0x10000) {  *out++= ((c >> 12) & 0x0F) | 0xE0;  bits=  6; }
        else                  {  *out++= ((c >> 18) & 0x07) | 0xF0;  bits= 12; }
 
        for ( ; bits >= 0; bits-= 6) {
            if (out >= outend)
	        break;
            *out++= ((c >> bits) & 0x3F) | 0x80;
        }
	processed = (const unsigned char*) in;
    }
    *outlen = out - outstart;
    *inlenb = processed - inb;
    return(0);
}

#ifdef LIBXML_OUTPUT_ENABLED
/**
 * UTF8ToUTF16LE:
 * @outb:  a pointer to an array of bytes to store the result
 * @outlen:  the length of @outb
 * @in:  a pointer to an array of UTF-8 chars
 * @inlen:  the length of @in
 *
 * Take a block of UTF-8 chars in and try to convert it to an UTF-16LE
 * block of chars out.
 *
 * Returns the number of bytes written, or -1 if lack of space, or -2
 *     if the transcoding failed. 
 */
static int
UTF8ToUTF16LE(unsigned char* outb, int *outlen,
            const unsigned char* in, int *inlen)
{
    unsigned short* out = (unsigned short*) outb;
    const unsigned char* processed = in;
    const unsigned char *const instart = in;
    unsigned short* outstart= out;
    unsigned short* outend;
    const unsigned char* inend= in+*inlen;
    unsigned int c, d;
    int trailing;
    unsigned char *tmp;
    unsigned short tmp1, tmp2;

    /* UTF16LE encoding has no BOM */
    if (in == NULL) {
	*outlen = 0;
	*inlen = 0;
	return(0);
    }
    outend = out + (*outlen / 2);
    while (in < inend) {
      d= *in++;
      if      (d < 0x80)  { c= d; trailing= 0; }
      else if (d < 0xC0) {
          /* trailing byte in leading position */
	  *outlen = (out - outstart) * 2;
	  *inlen = processed - instart;
	  return(-2);
      } else if (d < 0xE0)  { c= d & 0x1F; trailing= 1; }
      else if (d < 0xF0)  { c= d & 0x0F; trailing= 2; }
      else if (d < 0xF8)  { c= d & 0x07; trailing= 3; }
      else {
	/* no chance for this in UTF-16 */
	*outlen = (out - outstart) * 2;
	*inlen = processed - instart;
	return(-2);
      }

      if (inend - in < trailing) {
          break;
      } 

      for ( ; trailing; trailing--) {
          if ((in >= inend) || (((d= *in++) & 0xC0) != 0x80))
	      break;
          c <<= 6;
          c |= d & 0x3F;
      }

      /* assertion: c is a single UTF-4 value */
        if (c < 0x10000) {
            if (out >= outend)
	        break;
	    if (xmlLittleEndian) {
		*out++ = c;
	    } else {
		tmp = (unsigned char *) out;
		*tmp = c ;
		*(tmp + 1) = c >> 8 ;
		out++;
	    }
        }
        else if (c < 0x110000) {
            if (out+1 >= outend)
	        break;
            c -= 0x10000;
	    if (xmlLittleEndian) {
		*out++ = 0xD800 | (c >> 10);
		*out++ = 0xDC00 | (c & 0x03FF);
	    } else {
		tmp1 = 0xD800 | (c >> 10);
		tmp = (unsigned char *) out;
		*tmp = (unsigned char) tmp1;
		*(tmp + 1) = tmp1 >> 8;
		out++;

		tmp2 = 0xDC00 | (c & 0x03FF);
		tmp = (unsigned char *) out;
		*tmp  = (unsigned char) tmp2;
		*(tmp + 1) = tmp2 >> 8;
		out++;
	    }
        }
        else
	    break;
	processed = in;
    }
    *outlen = (out - outstart) * 2;
    *inlen = processed - instart;
    return(0);
}

/**
 * UTF8ToUTF16:
 * @outb:  a pointer to an array of bytes to store the result
 * @outlen:  the length of @outb
 * @in:  a pointer to an array of UTF-8 chars
 * @inlen:  the length of @in
 *
 * Take a block of UTF-8 chars in and try to convert it to an UTF-16
 * block of chars out.
 *
 * Returns the number of bytes written, or -1 if lack of space, or -2
 *     if the transcoding failed. 
 */
static int
UTF8ToUTF16(unsigned char* outb, int *outlen,
            const unsigned char* in, int *inlen)
{
    if (in == NULL) {
	/*
	 * initialization, add the Byte Order Mark for UTF-16LE
	 */
        if (*outlen >= 2) {
	    outb[0] = 0xFF;
	    outb[1] = 0xFE;
	    *outlen = 2;
	    *inlen = 0;
#ifdef DEBUG_ENCODING
            xmlGenericError(xmlGenericErrorContext,
		    "Added FFFE Byte Order Mark\n");
#endif
	    return(2);
	}
	*outlen = 0;
	*inlen = 0;
	return(0);
    }
    return (UTF8ToUTF16LE(outb, outlen, in, inlen));
}
#endif /* LIBXML_OUTPUT_ENABLED */

/**
 * UTF16BEToUTF8:
 * @out:  a pointer to an array of bytes to store the result
 * @outlen:  the length of @out
 * @inb:  a pointer to an array of UTF-16 passed as a byte array
 * @inlenb:  the length of @in in UTF-16 chars
 *
 * Take a block of UTF-16 ushorts in and try to convert it to an UTF-8
 * block of chars out. This function assumes the endian property
 * is the same between the native type of this machine and the
 * inputed one.
 *
 * Returns the number of bytes written, or -1 if lack of space, or -2
 *     if the transcoding fails (if *in is not a valid utf16 string)
 * The value of *inlen after return is the number of octets consumed
 *     if the return value is positive, else unpredictable.
 */
static int
UTF16BEToUTF8(unsigned char* out, int *outlen,
            const unsigned char* inb, int *inlenb)
{
    unsigned char* outstart = out;
    const unsigned char* processed = inb;
    unsigned char* outend = out + *outlen;
    unsigned short* in = (unsigned short*) inb;
    unsigned short* inend;
    unsigned int c, d, inlen;
    unsigned char *tmp;
    int bits;

    if ((*inlenb % 2) == 1)
        (*inlenb)--;
    inlen = *inlenb / 2;
    inend= in + inlen;
    while (in < inend) {
	if (xmlLittleEndian) {
	    tmp = (unsigned char *) in;
	    c = *tmp++;
	    c = c << 8;
	    c = c | (unsigned int) *tmp;
	    in++;
	} else {
	    c= *in++;
	} 
        if ((c & 0xFC00) == 0xD800) {    /* surrogates */
	    if (in >= inend) {           /* (in > inend) shouldn't happens */
		*outlen = out - outstart;
		*inlenb = processed - inb;
	        return(-2);
	    }
	    if (xmlLittleEndian) {
		tmp = (unsigned char *) in;
		d = *tmp++;
		d = d << 8;
		d = d | (unsigned int) *tmp;
		in++;
	    } else {
		d= *in++;
	    }
            if ((d & 0xFC00) == 0xDC00) {
                c &= 0x03FF;
                c <<= 10;
                c |= d & 0x03FF;
                c += 0x10000;
            }
            else {
		*outlen = out - outstart;
		*inlenb = processed - inb;
	        return(-2);
	    }
        }

	/* assertion: c is a single UTF-4 value */
        if (out >= outend) 
	    break;
        if      (c <    0x80) {  *out++=  c;                bits= -6; }
        else if (c <   0x800) {  *out++= ((c >>  6) & 0x1F) | 0xC0;  bits=  0; }
        else if (c < 0x10000) {  *out++= ((c >> 12) & 0x0F) | 0xE0;  bits=  6; }
        else                  {  *out++= ((c >> 18) & 0x07) | 0xF0;  bits= 12; }
 
        for ( ; bits >= 0; bits-= 6) {
            if (out >= outend) 
	        break;
            *out++= ((c >> bits) & 0x3F) | 0x80;
        }
	processed = (const unsigned char*) in;
    }
    *outlen = out - outstart;
    *inlenb = processed - inb;
    return(0);
}

#ifdef LIBXML_OUTPUT_ENABLED
/**
 * UTF8ToUTF16BE:
 * @outb:  a pointer to an array of bytes to store the result
 * @outlen:  the length of @outb
 * @in:  a pointer to an array of UTF-8 chars
 * @inlen:  the length of @in
 *
 * Take a block of UTF-8 chars in and try to convert it to an UTF-16BE
 * block of chars out.
 *
 * Returns the number of byte written, or -1 by lack of space, or -2
 *     if the transcoding failed. 
 */
static int
UTF8ToUTF16BE(unsigned char* outb, int *outlen,
            const unsigned char* in, int *inlen)
{
    unsigned short* out = (unsigned short*) outb;
    const unsigned char* processed = in;
    const unsigned char *const instart = in;
    unsigned short* outstart= out;
    unsigned short* outend;
    const unsigned char* inend= in+*inlen;
    unsigned int c, d;
    int trailing;
    unsigned char *tmp;
    unsigned short tmp1, tmp2;

    /* UTF-16BE has no BOM */
    if (in == NULL) {
	*outlen = 0;
	*inlen = 0;
	return(0);
    }
    outend = out + (*outlen / 2);
    while (in < inend) {
      d= *in++;
      if      (d < 0x80)  { c= d; trailing= 0; }
      else if (d < 0xC0)  {
          /* trailing byte in leading position */
	  *outlen = out - outstart;
	  *inlen = processed - instart;
	  return(-2);
      } else if (d < 0xE0)  { c= d & 0x1F; trailing= 1; }
      else if (d < 0xF0)  { c= d & 0x0F; trailing= 2; }
      else if (d < 0xF8)  { c= d & 0x07; trailing= 3; }
      else {
          /* no chance for this in UTF-16 */
	  *outlen = out - outstart;
	  *inlen = processed - instart;
	  return(-2);
      }

      if (inend - in < trailing) {
          break;
      } 

      for ( ; trailing; trailing--) {
          if ((in >= inend) || (((d= *in++) & 0xC0) != 0x80))  break;
          c <<= 6;
          c |= d & 0x3F;
      }

      /* assertion: c is a single UTF-4 value */
        if (c < 0x10000) {
            if (out >= outend)  break;
	    if (xmlLittleEndian) {
		tmp = (unsigned char *) out;
		*tmp = c >> 8;
		*(tmp + 1) = c;
		out++;
	    } else {
		*out++ = c;
	    }
        }
        else if (c < 0x110000) {
            if (out+1 >= outend)  break;
            c -= 0x10000;
	    if (xmlLittleEndian) {
		tmp1 = 0xD800 | (c >> 10);
		tmp = (unsigned char *) out;
		*tmp = tmp1 >> 8;
		*(tmp + 1) = (unsigned char) tmp1;
		out++;

		tmp2 = 0xDC00 | (c & 0x03FF);
		tmp = (unsigned char *) out;
		*tmp = tmp2 >> 8;
		*(tmp + 1) = (unsigned char) tmp2;
		out++;
	    } else {
		*out++ = 0xD800 | (c >> 10);
		*out++ = 0xDC00 | (c & 0x03FF);
	    }
        }
        else
	    break;
	processed = in;
    }
    *outlen = (out - outstart) * 2;
    *inlen = processed - instart;
    return(0);
}
#endif /* LIBXML_OUTPUT_ENABLED */

/************************************************************************
 *									*
 *		Generic encoding handling routines			*
 *									*
 ************************************************************************/

/**
 * xmlDetectCharEncoding:
 * @in:  a pointer to the first bytes of the XML entity, must be at least
 *       2 bytes long (at least 4 if encoding is UTF4 variant).
 * @len:  pointer to the length of the buffer
 *
 * Guess the encoding of the entity using the first bytes of the entity content
 * according to the non-normative appendix F of the XML-1.0 recommendation.
 * 
 * Returns one of the XML_CHAR_ENCODING_... values.
 */
xmlCharEncoding
xmlDetectCharEncoding(const unsigned char* in, int len)
{
    if (len >= 4) {
	if ((in[0] == 0x00) && (in[1] == 0x00) &&
	    (in[2] == 0x00) && (in[3] == 0x3C))
	    return(XML_CHAR_ENCODING_UCS4BE);
	if ((in[0] == 0x3C) && (in[1] == 0x00) &&
	    (in[2] == 0x00) && (in[3] == 0x00))
	    return(XML_CHAR_ENCODING_UCS4LE);
	if ((in[0] == 0x00) && (in[1] == 0x00) &&
	    (in[2] == 0x3C) && (in[3] == 0x00))
	    return(XML_CHAR_ENCODING_UCS4_2143);
	if ((in[0] == 0x00) && (in[1] == 0x3C) &&
	    (in[2] == 0x00) && (in[3] == 0x00))
	    return(XML_CHAR_ENCODING_UCS4_3412);
	if ((in[0] == 0x4C) && (in[1] == 0x6F) &&
	    (in[2] == 0xA7) && (in[3] == 0x94))
	    return(XML_CHAR_ENCODING_EBCDIC);
	if ((in[0] == 0x3C) && (in[1] == 0x3F) &&
	    (in[2] == 0x78) && (in[3] == 0x6D))
	    return(XML_CHAR_ENCODING_UTF8);
	/*
	 * Although not part of the recommendation, we also
	 * attempt an "auto-recognition" of UTF-16LE and
	 * UTF-16BE encodings.
	 */
	if ((in[0] == 0x3C) && (in[1] == 0x00) &&
	    (in[2] == 0x3F) && (in[3] == 0x00))
	    return(XML_CHAR_ENCODING_UTF16LE);
	if ((in[0] == 0x00) && (in[1] == 0x3C) &&
	    (in[2] == 0x00) && (in[3] == 0x3F))
	    return(XML_CHAR_ENCODING_UTF16BE);
    }
    if (len >= 3) {
	/*
	 * Errata on XML-1.0 June 20 2001
	 * We now allow an UTF8 encoded BOM
	 */
	if ((in[0] == 0xEF) && (in[1] == 0xBB) &&
	    (in[2] == 0xBF))
	    return(XML_CHAR_ENCODING_UTF8);
    }
    /* For UTF-16 we can recognize by the BOM */
    if (len >= 2) {
	if ((in[0] == 0xFE) && (in[1] == 0xFF))
	    return(XML_CHAR_ENCODING_UTF16BE);
	if ((in[0] == 0xFF) && (in[1] == 0xFE))
	    return(XML_CHAR_ENCODING_UTF16LE);
    }
    return(XML_CHAR_ENCODING_NONE);
}

/**
 * xmlCleanupEncodingAliases:
 *
 * Unregisters all aliases
 */
void
xmlCleanupEncodingAliases(void) {