charcnv.c

samba-3.0.22.tar.gz 编译smb服务器的源码

				retval += 2;
				if (!lastp)
					break;
			} else {
#ifdef BROKEN_UNICODE_COMPOSE_CHARACTERS
				goto general_case;
#else
				return retval + convert_string_internal(from, to, p, slen, q, dlen, allow_bad_conv);
#endif
			}
		}
		if (!dlen) {
			/* Even if we fast path we should note if we ran out of room. */
			if (((slen != (size_t)-1) && slen) ||
					((slen == (size_t)-1) && lastp)) {
				errno = E2BIG;
			}
		}
		return retval;
	}

#ifdef BROKEN_UNICODE_COMPOSE_CHARACTERS
  general_case:
#endif
	return convert_string_internal(from, to, src, srclen, dest, destlen, allow_bad_conv);
}

/**
 * Convert between character sets, allocating a new buffer for the result.
 *
 * @param ctx TALLOC_CTX to use to allocate with. If NULL use malloc.
 * @param srclen length of source buffer.
 * @param dest always set at least to NULL
 * @note -1 is not accepted for srclen.
 *
 * @returns Size in bytes of the converted string; or -1 in case of error.
 *
 * Ensure the srclen contains the terminating zero.
 * 
 * I hate the goto's in this function. It's embarressing.....
 * There has to be a cleaner way to do this. JRA.
 **/

size_t convert_string_allocate(TALLOC_CTX *ctx, charset_t from, charset_t to,
			       void const *src, size_t srclen, void **dest, BOOL allow_bad_conv)
{
	size_t i_len, o_len, destlen = MAX(srclen, 512);
	size_t retval;
	const char *inbuf = (const char *)src;
	char *outbuf = NULL, *ob = NULL;
	smb_iconv_t descriptor;

	*dest = NULL;

	if (src == NULL || srclen == (size_t)-1)
		return (size_t)-1;
	if (srclen == 0)
		return 0;

	lazy_initialize_conv();

	descriptor = conv_handles[from][to];

	if (descriptor == (smb_iconv_t)-1 || descriptor == (smb_iconv_t)0) {
		if (!conv_silent)
			DEBUG(0,("convert_string_allocate: Conversion not supported.\n"));
		return (size_t)-1;
	}

  convert:

	if ((destlen*2) < destlen) {
		/* wrapped ! abort. */
		if (!conv_silent)
			DEBUG(0, ("convert_string_allocate: destlen wrapped !\n"));
		if (!ctx)
			SAFE_FREE(outbuf);
		return (size_t)-1;
	} else {
		destlen = destlen * 2;
	}

	if (ctx)
		ob = (char *)TALLOC_REALLOC(ctx, ob, destlen);
	else
		ob = (char *)SMB_REALLOC(ob, destlen);

	if (!ob) {
		DEBUG(0, ("convert_string_allocate: realloc failed!\n"));
		if (!ctx)
			SAFE_FREE(outbuf);
		return (size_t)-1;
	} else {
		outbuf = ob;
	}
	i_len = srclen;
	o_len = destlen;

 again:

	retval = smb_iconv(descriptor,
			   &inbuf, &i_len,
			   &outbuf, &o_len);
	if(retval == (size_t)-1) 		{
	    	const char *reason="unknown error";
		switch(errno) {
			case EINVAL:
				reason="Incomplete multibyte sequence";
				if (!conv_silent)
					DEBUG(3,("convert_string_allocate: Conversion error: %s(%s)\n",reason,inbuf));
				if (allow_bad_conv)
					goto use_as_is;
				break;
			case E2BIG:
				goto convert;		
			case EILSEQ:
				reason="Illegal multibyte sequence";
				if (!conv_silent)
					DEBUG(3,("convert_string_allocate: Conversion error: %s(%s)\n",reason,inbuf));
				if (allow_bad_conv)
					goto use_as_is;
				break;
		}
		if (!conv_silent)
			DEBUG(0,("Conversion error: %s(%s)\n",reason,inbuf));
		/* smb_panic(reason); */
		return (size_t)-1;
	}

  out:

	destlen = destlen - o_len;
	if (ctx)
		*dest = (char *)TALLOC_REALLOC(ctx,ob,destlen);
	else
		*dest = (char *)SMB_REALLOC(ob,destlen);
	if (destlen && !*dest) {
		DEBUG(0, ("convert_string_allocate: out of memory!\n"));
		if (!ctx)
			SAFE_FREE(ob);
		return (size_t)-1;
	}

	return destlen;

 use_as_is:

	/* 
	 * Conversion not supported. This is actually an error, but there are so
	 * many misconfigured iconv systems and smb.conf's out there we can't just
	 * fail. Do a very bad conversion instead.... JRA.
	 */

	{
		if (o_len == 0 || i_len == 0)
			goto out;

		if (from == CH_UCS2 && to != CH_UCS2) {
			/* Can't convert from ucs2 to multibyte. Just use the default fail char. */
			if (i_len < 2)
				goto out;

			if (i_len >= 2) {
				*outbuf = lp_failed_convert_char();

				outbuf++;
				o_len--;

				inbuf += 2;
				i_len -= 2;
			}

			if (o_len == 0 || i_len == 0)
				goto out;

			/* Keep trying with the next char... */
			goto again;

		} else if (from != CH_UCS2 && to == CH_UCS2) {
			/* Can't convert to ucs2 - just widen by adding the default fail char then zero. */
			if (o_len < 2)
				goto out;

			outbuf[0] = lp_failed_convert_char();
			outbuf[1] = '\0';

			inbuf++;
			i_len--;

			outbuf += 2;
			o_len -= 2;

			if (o_len == 0 || i_len == 0)
				goto out;

			/* Keep trying with the next char... */
			goto again;

		} else if (from != CH_UCS2 && to != CH_UCS2) {
			/* Failed multibyte to multibyte. Just copy the default fail char and
				try again. */
			outbuf[0] = lp_failed_convert_char();

			inbuf++;
			i_len--;

			outbuf++;
			o_len--;

			if (o_len == 0 || i_len == 0)
				goto out;

			/* Keep trying with the next char... */
			goto again;

		} else {
			/* Keep compiler happy.... */
			goto out;
		}
	}
}

/**
 * Convert between character sets, allocating a new buffer using talloc for the result.
 *
 * @param srclen length of source buffer.
 * @param dest always set at least to NULL 
 * @note -1 is not accepted for srclen.
 *
 * @returns Size in bytes of the converted string; or -1 in case of error.
 **/
static size_t convert_string_talloc(TALLOC_CTX *ctx, charset_t from, charset_t to,
		      		void const *src, size_t srclen, void **dest, BOOL allow_bad_conv)
{
	size_t dest_len;

	*dest = NULL;
	dest_len=convert_string_allocate(ctx, from, to, src, srclen, dest, allow_bad_conv);
	if (dest_len == (size_t)-1)
		return (size_t)-1;
	if (*dest == NULL)
		return (size_t)-1;
	return dest_len;
}

size_t unix_strupper(const char *src, size_t srclen, char *dest, size_t destlen)
{
	size_t size;
	smb_ucs2_t *buffer;
	
	size = push_ucs2_allocate(&buffer, src);
	if (size == (size_t)-1) {
		smb_panic("failed to create UCS2 buffer");
	}
	if (!strupper_w(buffer) && (dest == src)) {
		free(buffer);
		return srclen;
	}
	
	size = convert_string(CH_UCS2, CH_UNIX, buffer, size, dest, destlen, True);
	free(buffer);
	return size;
}

/**
 strdup() a unix string to upper case.
 Max size is pstring.
**/

char *strdup_upper(const char *s)
{
	pstring out_buffer;
	const unsigned char *p = (const unsigned char *)s;
	unsigned char *q = (unsigned char *)out_buffer;

	/* this is quite a common operation, so we want it to be
	   fast. We optimise for the ascii case, knowing that all our
	   supported multi-byte character sets are ascii-compatible
	   (ie. they match for the first 128 chars) */

	while (1) {
		if (*p & 0x80)
			break;
		*q++ = toupper_ascii(*p);
		if (!*p)
			break;
		p++;
		if (p - ( const unsigned char *)s >= sizeof(pstring))
			break;
	}

	if (*p) {
		/* MB case. */
		size_t size;
		wpstring buffer;
		size = convert_string(CH_UNIX, CH_UCS2, s, -1, buffer, sizeof(buffer), True);
		if (size == (size_t)-1) {
			return NULL;
		}

		strupper_w(buffer);
	
		size = convert_string(CH_UCS2, CH_UNIX, buffer, -1, out_buffer, sizeof(out_buffer), True);
		if (size == (size_t)-1) {
			return NULL;
		}
	}

	return SMB_STRDUP(out_buffer);
}

size_t unix_strlower(const char *src, size_t srclen, char *dest, size_t destlen)
{
	size_t size;
	smb_ucs2_t *buffer = NULL;
	
	size = convert_string_allocate(NULL, CH_UNIX, CH_UCS2, src, srclen,
				       (void **)(void *)&buffer, True);
	if (size == (size_t)-1 || !buffer) {
		smb_panic("failed to create UCS2 buffer");
	}
	if (!strlower_w(buffer) && (dest == src)) {
		SAFE_FREE(buffer);
		return srclen;
	}
	size = convert_string(CH_UCS2, CH_UNIX, buffer, size, dest, destlen, True);
	SAFE_FREE(buffer);
	return size;
}

/**
 strdup() a unix string to lower case.
**/

char *strdup_lower(const char *s)
{
	size_t size;
	smb_ucs2_t *buffer = NULL;
	char *out_buffer;
	
	size = push_ucs2_allocate(&buffer, s);
	if (size == -1 || !buffer) {
		return NULL;
	}

	strlower_w(buffer);
	
	size = pull_ucs2_allocate(&out_buffer, buffer);
	SAFE_FREE(buffer);

	if (size == (size_t)-1) {
		return NULL;
	}
	
	return out_buffer;
}

static size_t ucs2_align(const void *base_ptr, const void *p, int flags)
{
	if (flags & (STR_NOALIGN|STR_ASCII))
		return 0;
	return PTR_DIFF(p, base_ptr) & 1;
}


/**
 * Copy a string from a char* unix src to a dos codepage string destination.
 *
 * @return the number of bytes occupied by the string in the destination.
 *
 * @param flags can include
 * <dl>
 * <dt>STR_TERMINATE</dt> <dd>means include the null termination</dd>
 * <dt>STR_UPPER</dt> <dd>means uppercase in the destination</dd>
 * </dl>
 *
 * @param dest_len the maximum length in bytes allowed in the
 * destination.  If @p dest_len is -1 then no maximum is used.
 **/
size_t push_ascii(void *dest, const char *src, size_t dest_len, int flags)
{
	size_t src_len = strlen(src);
	pstring tmpbuf;

	/* treat a pstring as "unlimited" length */
	if (dest_len == (size_t)-1)
		dest_len = sizeof(pstring);

	if (flags & STR_UPPER) {
		pstrcpy(tmpbuf, src);
		strupper_m(tmpbuf);
		src = tmpbuf;
	}

	if (flags & (STR_TERMINATE | STR_TERMINATE_ASCII))
		src_len++;

	return convert_string(CH_UNIX, CH_DOS, src, src_len, dest, dest_len, True);
}

size_t push_ascii_fstring(void *dest, const char *src)
{
	return push_ascii(dest, src, sizeof(fstring), STR_TERMINATE);
}

size_t push_ascii_pstring(void *dest, const char *src)
{
	return push_ascii(dest, src, sizeof(pstring), STR_TERMINATE);
}

/********************************************************************
 Push an nstring - ensure null terminated. Written by
 [e-mail] (MORIYAMA Masayuki).
********************************************************************/

size_t push_ascii_nstring(void *dest, const char *src)
{
	size_t i, buffer_len, dest_len;
	smb_ucs2_t *buffer;

	conv_silent = True;
	buffer_len = push_ucs2_allocate(&buffer, src);
	if (buffer_len == (size_t)-1) {
		smb_panic("failed to create UCS2 buffer");
	}

	/* We're using buffer_len below to count ucs2 characters, not bytes. */
	buffer_len /= sizeof(smb_ucs2_t);

	dest_len = 0;
	for (i = 0; buffer[i] != 0 && (i < buffer_len); i++) {
		unsigned char mb[10];
		/* Convert one smb_ucs2_t character at a time. */
		size_t mb_len = convert_string(CH_UCS2, CH_DOS, buffer+i, sizeof(smb_ucs2_t), mb, sizeof(mb), False);
		if ((mb_len != (size_t)-1) && (dest_len + mb_len <= MAX_NETBIOSNAME_LEN - 1)) {
			memcpy((char *)dest + dest_len, mb, mb_len);
			dest_len += mb_len;
		} else {
			errno = E2BIG;
			break;
		}
	}
	((char *)dest)[dest_len] = '\0';

	SAFE_FREE(buffer);
	conv_silent = False;
	return dest_len;
}

/**
 * Copy a string from a dos codepage source to a unix char* destination.
 *
 * The resulting string in "dest" is always null terminated.
 *
 * @param flags can have:
 * <dl>
 * <dt>STR_TERMINATE</dt>
 * <dd>STR_TERMINATE means the string in @p src
 * is null terminated, and src_len is ignored.</dd>
 * </dl>
 *
 * @param src_len is the length of the source area in bytes.
 * @returns the number of bytes occupied by the string in @p src.
 **/
size_t pull_ascii(char *dest, const void *src, size_t dest_len, size_t src_len, int flags)
{