示例#1
0
/*
 * Extract the next non-wildcard part of a search string, ie, a word bounded
 * by '_' or '%' meta-characters, non-word characters or string end.
 *
 * str: source string, of length lenstr bytes (need not be null-terminated)
 * buf: where to return the substring (must be long enough)
 * *bytelen: receives byte length of the found substring
 * *charlen: receives character length of the found substring
 *
 * Returns pointer to end+1 of the found substring in the source string.
 * Returns NULL if no word found (in which case buf, bytelen, charlen not set)
 *
 * If the found word is bounded by non-word characters or string boundaries
 * then this function will include corresponding padding spaces into buf.
 */
static const char *
get_wildcard_part(const char *str, int lenstr,
				  char *buf, int *bytelen, int *charlen)
{
	const char *beginword = str;
	const char *endword;
	char	   *s = buf;
	bool		in_leading_wildcard_meta = false;
	bool		in_trailing_wildcard_meta = false;
	bool		in_escape = false;
	int			clen;

	/*
	 * Find the first word character, remembering whether preceding character
	 * was wildcard meta-character.  Note that the in_escape state persists
	 * from this loop to the next one, since we may exit at a word character
	 * that is in_escape.
	 */
	while (beginword - str < lenstr)
	{
		if (in_escape)
		{
			if (iswordchr(beginword))
				break;
			in_escape = false;
			in_leading_wildcard_meta = false;
		}
		else
		{
			if (ISESCAPECHAR(beginword))
				in_escape = true;
			else if (ISWILDCARDCHAR(beginword))
				in_leading_wildcard_meta = true;
			else if (iswordchr(beginword))
				break;
			else
				in_leading_wildcard_meta = false;
		}
		beginword += pg_mblen(beginword);
	}

	/*
	 * Handle string end.
	 */
	if (beginword - str >= lenstr)
		return NULL;

	/*
	 * Add left padding spaces if preceding character wasn't wildcard
	 * meta-character.
	 */
	*charlen = 0;
	if (!in_leading_wildcard_meta)
	{
		if (LPADDING > 0)
		{
			*s++ = ' ';
			(*charlen)++;
			if (LPADDING > 1)
			{
				*s++ = ' ';
				(*charlen)++;
			}
		}
	}

	/*
	 * Copy data into buf until wildcard meta-character, non-word character or
	 * string boundary.  Strip escapes during copy.
	 */
	endword = beginword;
	while (endword - str < lenstr)
	{
		clen = pg_mblen(endword);
		if (in_escape)
		{
			if (iswordchr(endword))
			{
				memcpy(s, endword, clen);
				(*charlen)++;
				s += clen;
			}
			else
			{
				/*
				 * Back up endword to the escape character when stopping at
				 * an escaped char, so that subsequent get_wildcard_part will
				 * restart from the escape character.  We assume here that
				 * escape chars are single-byte.
				 */
				endword--;
				break;
			}
			in_escape = false;
		}
		else
		{
			if (ISESCAPECHAR(endword))
				in_escape = true;
			else if (ISWILDCARDCHAR(endword))
			{
				in_trailing_wildcard_meta = true;
				break;
			}
			else if (iswordchr(endword))
			{
				memcpy(s, endword, clen);
				(*charlen)++;
				s += clen;
			}
			else
				break;
		}
		endword += clen;
	}

	/*
	 * Add right padding spaces if next character isn't wildcard
	 * meta-character.
	 */
	if (!in_trailing_wildcard_meta)
	{
		if (RPADDING > 0)
		{
			*s++ = ' ';
			(*charlen)++;
			if (RPADDING > 1)
			{
				*s++ = ' ';
				(*charlen)++;
			}
		}
	}

	*bytelen = s - buf;
	return endword;
}
示例#2
0
文件: trgm_op.c 项目: GisKook/Gis
/*
 * Extract the next non-wildcard part of a search string, ie, a word bounded
 * by '_' or '%' meta-characters, non-word characters or string end.
 *
 * str: source string, of length lenstr bytes (need not be null-terminated)
 * buf: where to return the substring (must be long enough)
 * *bytelen: receives byte length of the found substring
 * *charlen: receives character length of the found substring
 *
 * Returns pointer to end+1 of the found substring in the source string.
 * Returns NULL if no word found (in which case buf, bytelen, charlen not set)
 *
 * If the found word is bounded by non-word characters or string boundaries
 * then this function will include corresponding padding spaces into buf.
 */
static const char *
get_wildcard_part(const char *str, int lenstr,
				  char *buf, int *bytelen, int *charlen)
{
	const char *beginword = str;
	const char *endword;
	char	   *s = buf;
	bool		in_wildcard_meta = false;
	bool		in_escape = false;
	int			clen;

	/*
	 * Find the first word character remembering whether last character was
	 * wildcard meta-character.
	 */
	while (beginword - str < lenstr)
	{
		if (in_escape)
		{
			in_escape = false;
			in_wildcard_meta = false;
			if (iswordchr(beginword))
				break;
		}
		else
		{
			if (ISESCAPECHAR(beginword))
				in_escape = true;
			else if (ISWILDCARDCHAR(beginword))
				in_wildcard_meta = true;
			else if (iswordchr(beginword))
				break;
			else
				in_wildcard_meta = false;
		}
		beginword += pg_mblen(beginword);
	}

	/*
	 * Handle string end.
	 */
	if (beginword - str >= lenstr)
		return NULL;

	/*
	 * Add left padding spaces if last character wasn't wildcard
	 * meta-character.
	 */
	*charlen = 0;
	if (!in_wildcard_meta)
	{
		if (LPADDING > 0)
		{
			*s++ = ' ';
			(*charlen)++;
			if (LPADDING > 1)
			{
				*s++ = ' ';
				(*charlen)++;
			}
		}
	}

	/*
	 * Copy data into buf until wildcard meta-character, non-word character or
	 * string boundary.  Strip escapes during copy.
	 */
	endword = beginword;
	in_wildcard_meta = false;
	in_escape = false;
	while (endword - str < lenstr)
	{
		clen = pg_mblen(endword);
		if (in_escape)
		{
			in_escape = false;
			in_wildcard_meta = false;
			if (iswordchr(endword))
			{
				memcpy(s, endword, clen);
				(*charlen)++;
				s += clen;
			}
			else
				break;
		}
		else
		{
			if (ISESCAPECHAR(endword))
				in_escape = true;
			else if (ISWILDCARDCHAR(endword))
			{
				in_wildcard_meta = true;
				break;
			}
			else if (iswordchr(endword))
			{
				memcpy(s, endword, clen);
				(*charlen)++;
				s += clen;
			}
			else
			{
				in_wildcard_meta = false;
				break;
			}
		}
		endword += clen;
	}

	/*
	 * Add right padding spaces if last character wasn't wildcard
	 * meta-character.
	 */
	if (!in_wildcard_meta)
	{
		if (RPADDING > 0)
		{
			*s++ = ' ';
			(*charlen)++;
			if (RPADDING > 1)
			{
				*s++ = ' ';
				(*charlen)++;
			}
		}
	}

	*bytelen = s - buf;
	return endword;
}