oracle_compat.c

来自「postgresql8.3.4源码,开源数据库」· C语言 代码 · 共 1,498 行 · 第 1/3 页

	{
		ptr2 = ptr2start;
		while (ptr2 <= end2)
		{
			if (*ptr == *ptr2)
				break;
			++ptr2;
		}
		if (ptr2 > end2)
			break;
		ptr++;
		m--;
	}

	while (m > 0)
	{
		ptr2 = ptr2start;
		while (ptr2 <= end2)
		{
			if (*end == *ptr2)
				break;
			++ptr2;
		}
		if (ptr2 > end2)
			break;
		end--;
		m--;
	}

	ret = (bytea *) palloc(VARHDRSZ + m);
	SET_VARSIZE(ret, VARHDRSZ + m);
	memcpy(VARDATA(ret), ptr, m);

	PG_RETURN_BYTEA_P(ret);
}

/********************************************************************
 *
 * ltrim
 *
 * Syntax:
 *
 *	 text ltrim(text string, text set)
 *
 * Purpose:
 *
 *	 Returns string with initial characters removed up to the first
 *	 character not in set.
 *
 ********************************************************************/

Datum
ltrim(PG_FUNCTION_ARGS)
{
	text	   *string = PG_GETARG_TEXT_PP(0);
	text	   *set = PG_GETARG_TEXT_PP(1);
	text	   *ret;

	ret = dotrim(VARDATA_ANY(string), VARSIZE_ANY_EXHDR(string),
				 VARDATA_ANY(set), VARSIZE_ANY_EXHDR(set),
				 true, false);

	PG_RETURN_TEXT_P(ret);
}

/********************************************************************
 *
 * ltrim1 --- ltrim with set fixed as ' '
 *
 ********************************************************************/

Datum
ltrim1(PG_FUNCTION_ARGS)
{
	text	   *string = PG_GETARG_TEXT_PP(0);
	text	   *ret;

	ret = dotrim(VARDATA_ANY(string), VARSIZE_ANY_EXHDR(string),
				 " ", 1,
				 true, false);

	PG_RETURN_TEXT_P(ret);
}

/********************************************************************
 *
 * rtrim
 *
 * Syntax:
 *
 *	 text rtrim(text string, text set)
 *
 * Purpose:
 *
 *	 Returns string with final characters removed after the last
 *	 character not in set.
 *
 ********************************************************************/

Datum
rtrim(PG_FUNCTION_ARGS)
{
	text	   *string = PG_GETARG_TEXT_PP(0);
	text	   *set = PG_GETARG_TEXT_PP(1);
	text	   *ret;

	ret = dotrim(VARDATA_ANY(string), VARSIZE_ANY_EXHDR(string),
				 VARDATA_ANY(set), VARSIZE_ANY_EXHDR(set),
				 false, true);

	PG_RETURN_TEXT_P(ret);
}

/********************************************************************
 *
 * rtrim1 --- rtrim with set fixed as ' '
 *
 ********************************************************************/

Datum
rtrim1(PG_FUNCTION_ARGS)
{
	text	   *string = PG_GETARG_TEXT_PP(0);
	text	   *ret;

	ret = dotrim(VARDATA_ANY(string), VARSIZE_ANY_EXHDR(string),
				 " ", 1,
				 false, true);

	PG_RETURN_TEXT_P(ret);
}


/********************************************************************
 *
 * translate
 *
 * Syntax:
 *
 *	 text translate(text string, text from, text to)
 *
 * Purpose:
 *
 *	 Returns string after replacing all occurrences of characters in from
 *	 with the corresponding character in to.  If from is longer than to,
 *	 occurrences of the extra characters in from are deleted.
 *	 Improved by Edwin Ramirez <[e-mail]>.
 *
 ********************************************************************/

Datum
translate(PG_FUNCTION_ARGS)
{
	text	   *string = PG_GETARG_TEXT_PP(0);
	text	   *from = PG_GETARG_TEXT_PP(1);
	text	   *to = PG_GETARG_TEXT_PP(2);
	text	   *result;
	char	   *from_ptr,
			   *to_ptr;
	char	   *source,
			   *target;
	int			m,
				fromlen,
				tolen,
				retlen,
				i;
	int			worst_len;
	int			len;
	int			source_len;
	int			from_index;

	m = VARSIZE_ANY_EXHDR(string);
	if (m <= 0)
		PG_RETURN_TEXT_P(string);
	source = VARDATA_ANY(string);

	fromlen = VARSIZE_ANY_EXHDR(from);
	from_ptr = VARDATA_ANY(from);
	tolen = VARSIZE_ANY_EXHDR(to);
	to_ptr = VARDATA_ANY(to);

	/*
	 * The worst-case expansion is to substitute a max-length character for a
	 * single-byte character at each position of the string.
	 */
	worst_len = pg_database_encoding_max_length() * m;

	/* check for integer overflow */
	if (worst_len / pg_database_encoding_max_length() != m)
		ereport(ERROR,
				(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
				 errmsg("requested length too large")));

	result = (text *) palloc(worst_len + VARHDRSZ);
	target = VARDATA(result);
	retlen = 0;

	while (m > 0)
	{
		source_len = pg_mblen(source);
		from_index = 0;

		for (i = 0; i < fromlen; i += len)
		{
			len = pg_mblen(&from_ptr[i]);
			if (len == source_len &&
				memcmp(source, &from_ptr[i], len) == 0)
				break;

			from_index++;
		}
		if (i < fromlen)
		{
			/* substitute */
			char	   *p = to_ptr;

			for (i = 0; i < from_index; i++)
			{
				p += pg_mblen(p);
				if (p >= (to_ptr + tolen))
					break;
			}
			if (p < (to_ptr + tolen))
			{
				len = pg_mblen(p);
				memcpy(target, p, len);
				target += len;
				retlen += len;
			}

		}
		else
		{
			/* no match, so copy */
			memcpy(target, source, source_len);
			target += source_len;
			retlen += source_len;
		}

		source += source_len;
		m -= source_len;
	}

	SET_VARSIZE(result, retlen + VARHDRSZ);

	/*
	 * The function result is probably much bigger than needed, if we're using
	 * a multibyte encoding, but it's not worth reallocating it; the result
	 * probably won't live long anyway.
	 */

	PG_RETURN_TEXT_P(result);
}

/********************************************************************
 *
 * ascii
 *
 * Syntax:
 *
 *	 int ascii(text string)
 *
 * Purpose:
 *
 *	 Returns the decimal representation of the first character from
 *	 string.
 *	 If the string is empty we return 0.
 *	 If the database encoding is UTF8, we return the Unicode codepoint.
 *	 If the database encoding is any other multi-byte encoding, we
 *	 return the value of the first byte if it is an ASCII character
 *	 (range 1 .. 127), or raise an error.
 *	 For all other encodings we return the value of the first byte,
 *	 (range 1..255).
 *
 ********************************************************************/

Datum
ascii(PG_FUNCTION_ARGS)
{
	text	   *string = PG_GETARG_TEXT_PP(0);
	int			encoding = GetDatabaseEncoding();
	unsigned char *data;

	if (VARSIZE_ANY_EXHDR(string) <= 0)
		PG_RETURN_INT32(0);

	data = (unsigned char *) VARDATA_ANY(string);

	if (encoding == PG_UTF8 && *data > 127)
	{
		/* return the code point for Unicode */

		int			result = 0,
					tbytes = 0,
					i;

		if (*data >= 0xF0)
		{
			result = *data & 0x07;
			tbytes = 3;
		}
		else if (*data >= 0xE0)
		{
			result = *data & 0x0F;
			tbytes = 2;
		}
		else
		{
			Assert(*data > 0xC0);
			result = *data & 0x1f;
			tbytes = 1;
		}

		Assert(tbytes > 0);

		for (i = 1; i <= tbytes; i++)
		{
			Assert((data[i] & 0xC0) == 0x80);
			result = (result << 6) + (data[i] & 0x3f);
		}

		PG_RETURN_INT32(result);
	}
	else
	{
		if (pg_encoding_max_length(encoding) > 1 && *data > 127)
			ereport(ERROR,
					(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
					 errmsg("requested character too large")));


		PG_RETURN_INT32((int32) *data);
	}
}

/********************************************************************
 *
 * chr
 *
 * Syntax:
 *
 *	 text chr(int val)
 *
 * Purpose:
 *
 *	Returns the character having the binary equivalent to val.
 *
 * For UTF8 we treat the argumwent as a Unicode code point.
 * For other multi-byte encodings we raise an error for arguments
 * outside the strict ASCII range (1..127).
 *
 * It's important that we don't ever return a value that is not valid
 * in the database encoding, so that this doesn't become a way for
 * invalid data to enter the database.
 *
 ********************************************************************/

Datum
chr			(PG_FUNCTION_ARGS)
{
	uint32		cvalue = PG_GETARG_UINT32(0);
	text	   *result;
	int			encoding = GetDatabaseEncoding();

	if (encoding == PG_UTF8 && cvalue > 127)
	{
		/* for Unicode we treat the argument as a code point */
		int			bytes;
		char	   *wch;

		/* We only allow valid Unicode code points */
		if (cvalue > 0x001fffff)
			ereport(ERROR,
					(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
					 errmsg("requested character too large for encoding: %d",
							cvalue)));

		if (cvalue > 0xffff)
			bytes = 4;
		else if (cvalue > 0x07ff)
			bytes = 3;
		else
			bytes = 2;

		result = (text *) palloc(VARHDRSZ + bytes);
		SET_VARSIZE(result, VARHDRSZ + bytes);
		wch = VARDATA(result);

		if (bytes == 2)
		{
			wch[0] = 0xC0 | ((cvalue >> 6) & 0x1F);
			wch[1] = 0x80 | (cvalue & 0x3F);;
		}
		else if (bytes == 3)
		{
			wch[0] = 0xE0 | ((cvalue >> 12) & 0x0F);
			wch[1] = 0x80 | ((cvalue >> 6) & 0x3F);
			wch[2] = 0x80 | (cvalue & 0x3F);
		}
		else
		{
			wch[0] = 0xF0 | ((cvalue >> 18) & 0x07);
			wch[1] = 0x80 | ((cvalue >> 12) & 0x3F);
			wch[2] = 0x80 | ((cvalue >> 6) & 0x3F);
			wch[3] = 0x80 | (cvalue & 0x3F);
		}

	}

	else
	{
		bool		is_mb;

		/*
		 * Error out on arguments that make no sense or that we can't validly
		 * represent in the encoding.
		 */

		if (cvalue == 0)
			ereport(ERROR,
					(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
					 errmsg("null character not permitted")));

		is_mb = pg_encoding_max_length(encoding) > 1;

		if ((is_mb && (cvalue > 127)) || (!is_mb && (cvalue > 255)))
			ereport(ERROR,
					(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
					 errmsg("requested character too large for encoding: %d",
							cvalue)));


		result = (text *) palloc(VARHDRSZ + 1);
		SET_VARSIZE(result, VARHDRSZ + 1);
		*VARDATA(result) = (char) cvalue;
	}

	PG_RETURN_TEXT_P(result);
}

/********************************************************************
 *
 * repeat
 *
 * Syntax:
 *
 *	 text repeat(text string, int val)
 *
 * Purpose:
 *
 *	Repeat string by val.
 *
 ********************************************************************/

Datum
repeat(PG_FUNCTION_ARGS)
{
	text	   *string = PG_GETARG_TEXT_PP(0);
	int32		count = PG_GETARG_INT32(1);
	text	   *result;
	int			slen,
				tlen;
	int			i;
	char	   *cp,
			   *sp;

	if (count < 0)
		count = 0;

	slen = VARSIZE_ANY_EXHDR(string);
	tlen = VARHDRSZ + (count * slen);

	/* Check for integer overflow */
	if (slen != 0 && count != 0)
	{
		int			check = count * slen;
		int			check2 = check + VARHDRSZ;

		if ((check / slen) != count || check2 <= check)
			ereport(ERROR,
					(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
					 errmsg("requested length too large")));
	}

	result = (text *) palloc(tlen);

	SET_VARSIZE(result, tlen);
	cp = VARDATA(result);
	sp = VARDATA_ANY(string);
	for (i = 0; i < count; i++)
	{
		memcpy(cp, sp, slen);
		cp += slen;
	}

	PG_RETURN_TEXT_P(result);
}