去除相邻重复字符函数

字符去重函数
功能：去重字符串相邻重复的字符，不相邻的不用去重
参数：

1. arg1 -- 输入字符串
2. arg2 -- 字符串开始位置
3. arg3 -- 字符串结束位置

要求：

1. 输入参数为arg1时， 对这个字符串去重
2. 输入参数为arg1，arg2时， 从arg2位置到字符串结束，去重
3. 输入参数为arg1，arg2，arg3时，从arg2到arg3位置，去重

src/include/catalog/pg_proc.h 

DATA(insert OID = 6669 (  remove_dup_char PGNSP PGUID 12 1 0 0 0 f f f f t f i 1 0 25 "25" _null_ _null_ _null_ _null_ _null_ remove_dup_char_arg1 _null_ _null_ _null_ ));
DESCR("Remove duplicate characters.");
DATA(insert OID = 6670 (  remove_dup_char PGNSP PGUID 12 1 0 0 0 f f f f t f i 2 0 25 "25 23" _null_ _null_ _null_ _null_ _null_ remove_dup_char_arg2 _null_ _null_ _null_ ));
DESCR("Remove duplicate characters.");
DATA(insert OID = 6671 (  remove_dup_char PGNSP PGUID 12 1 0 0 0 f f f f t f i 3 0 25 "25 23 23" _null_ _null_ _null_ _null_ _null_ remove_dup_char_arg3 _null_ _null_ _null_ ));
DESCR("Remove duplicate characters.");

 src/backend/utils/adt/myfuncs.c

/* 
 * Remove duplicate characters 
 * author:young
 */
Datum 
remove_dup_char_arg1 (PG_FUNCTION_ARGS)
{
	int n = 0;
	text *arg0 = PG_GETARG_TEXT_P(0);

	char *str = text_to_cstring(arg0);
	n = strlen(str);

	remove_dup(str, 0, n);

	PG_RETURN_TEXT_P(cstring_to_text(str));
}

Datum 
remove_dup_char_arg2 (PG_FUNCTION_ARGS)
{
	int n = 0;
	text *arg0 = PG_GETARG_TEXT_P(0);
	int32 arg1 = PG_GETARG_INT32(1);

	char *str = text_to_cstring(arg0);
	n = strlen(str);

	if (!(1 <= arg1 && arg1 <= n))
	{
		ereport(ERROR,
			(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
			errmsg("out of range")));
	}

	remove_dup(str, arg1 - 1, n);

	PG_RETURN_TEXT_P(cstring_to_text(str));
}

Datum 
remove_dup_char_arg3 (PG_FUNCTION_ARGS)
{
	int n = 0;
	text *arg0 = PG_GETARG_TEXT_P(0);
	int32 arg1 = PG_GETARG_INT32(1);
	int32 arg2 = PG_GETARG_INT32(2);

	char *str = text_to_cstring(arg0);
	n = strlen(str);

	if (!(1 <= arg1 && arg1 <= arg2 && arg2 <= n))
	{
		ereport(ERROR,
			(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
			errmsg("out of range")));
	}

	remove_dup(str, arg1 - 1, arg2 - 1);

	PG_RETURN_TEXT_P(cstring_to_text(str));
}

void 
remove_dup(char *str, int start, int end)
{
	int i = start, k = start;

	for (i = start; i <= end; i++)  
	{
		if (str[i + 1] && str[i + 1] == str[i] && i + 1 <= end)
		{
			k++;
		} 
		else  
		{
			str[i-k] = str[i];
		}			  
	}
	str[i-k] = '\0';
}

比较繁琐，再做一下修改，三个函数放到一个中

src/include/catalog/pg_proc.h 

DATA(insert OID = 6669 (  remove_dup_char PGNSP PGUID 12 1 0 0 0 f f f f t f i 1 0 25 "25" _null_ _null_ _null_ _null_ _null_ remove_dup_char _null_ _null_ _null_ ));
DESCR("Remove duplicate characters.");
DATA(insert OID = 6670 (  remove_dup_char PGNSP PGUID 12 1 0 0 0 f f f f t f i 2 0 25 "25 23" _null_ _null_ _null_ _null_ _null_ remove_dup_char _null_ _null_ _null_ ));
DESCR("Remove duplicate characters.");
DATA(insert OID = 6671 (  remove_dup_char PGNSP PGUID 12 1 0 0 0 f f f f t f i 3 0 25 "25 23 23" _null_ _null_ _null_ _null_ _null_ remove_dup_char _null_ _null_ _null_ ));
DESCR("Remove duplicate characters.");

src/backend/utils/adt/myfuncs.c

添加定义：

#define PG_GETARG_IF_EXISTS(n, type, defval) \
	((PG_NARGS() > (n) && !PG_ARGISNULL(n)) ? PG_GETARG_##type(n) : (defval)) 

 修改方法：

/* 
 * Remove duplicate characters 
 * author:yangjie
 */
Datum 
remove_dup_char (PG_FUNCTION_ARGS)
{
	text *arg0 = PG_GETARG_IF_EXISTS(0, TEXT_P, NULL);
	int32 arg1 = PG_GETARG_IF_EXISTS(1, INT32, 0);
	int32 arg2 = PG_GETARG_IF_EXISTS(2, INT32, 0);
	int n = 0;

	char *str = text_to_cstring(arg0);
	n = strlen(str);

	if(PG_NARGS() == 1)
	{
		remove_dup(str, 0, n);
	}

	if(PG_NARGS() == 2)
	{
		if (!(1 <= arg1 && arg1 <= n))
		{
			ereport(ERROR,
				(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
				errmsg("out of range")));
		}
		remove_dup(str, arg1 - 1, n);
	}

	if(PG_NARGS() == 3)
	{
		if (!(1 <= arg1 && arg1 <= arg2 && arg2 <= n))
		{
			ereport(ERROR,
				(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
				errmsg("out of range")));
		}
		remove_dup(str, arg1 - 1, arg2 - 1);
	}

	PG_RETURN_TEXT_P(cstring_to_text(str));
}

 再修改一下，如果有输入参数就用 没有就用默认值  最后再去重处理减少代码重用

/* 
 * Remove duplicate characters 
 * author:yangjie
 */
Datum 
remove_dup_char (PG_FUNCTION_ARGS)
{
	text *arg0 = PG_GETARG_IF_EXISTS(0, TEXT_P, NULL);
	int n = 0;
	char *str = text_to_cstring(arg0);
	n = strlen(str);
	int32 arg1 = PG_GETARG_IF_EXISTS(1, INT32, 0);
	int32 arg2 = PG_GETARG_IF_EXISTS(2, INT32, n);
	
	if (!(1 <= arg1 && arg1 <= n))
	{
		ereport(ERROR,
			(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
			errmsg("1 <= arg1 && arg1 <= n")));
	}

	if (!(1 <= arg1 && arg1 <= arg2 && arg2 <= n))
	{
		ereport(ERROR,
			(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
			errmsg("1 <= arg1 && arg1 <= arg2 && arg2 <= n")));
	}

	remove_dup(str, arg1, arg2);
	
	PG_RETURN_TEXT_P(cstring_to_text(str));
}

void 
remove_dup(char *str, int start, int end)
{
	int i = start -1, k = start - 1;

	for (i = start - 1; i <= end - 1; i++)  
	{
		if (str[i + 1] && str[i + 1] == str[i] && i + 1 <= end - 1)
		{
			k++;
		} 
		else  
		{
			str[i-k] = str[i];
		}			  
	}
	str[i-k] = '\0';
}

 

转载于:https://my.oschina.net/yonj1e/blog/873559