varlena.c

关系型数据库 Postgresql 6.5.2

/*-------------------------------------------------------------------------
 *
 * varlena.c
 *	  Functions for the variable-length built-in types.
 *
 * Copyright (c) 1994, Regents of the University of California
 *
 *
 * IDENTIFICATION
 *	  $Header: /usr/local/cvsroot/pgsql/src/backend/utils/adt/varlena.c,v 1.48.2.1 1999/08/02 05:24:58 scrappy Exp $
 *
 *-------------------------------------------------------------------------
 */
#include <ctype.h>

#include "postgres.h"

#include "mb/pg_wchar.h"
#include "utils/builtins.h"

static int	text_cmp(text *arg1, text *arg2);


/*****************************************************************************
 *	 USER I/O ROUTINES														 *
 *****************************************************************************/


#define VAL(CH)			((CH) - '0')
#define DIG(VAL)		((VAL) + '0')

/*
 *		byteain			- converts from printable representation of byte array
 *
 *		Non-printable characters must be passed as '\nnn' (octal) and are
 *		converted to internal form.  '\' must be passed as '\\'.
 *		elog(ERROR, ...) if bad form.
 *
 *		BUGS:
 *				The input is scaned twice.
 *				The error checking of input is minimal.
 */
text *
byteain(char *inputText)
{
	char	   *tp;
	char	   *rp;
	int			byte;
	text	   *result;

	if (inputText == NULL)
		elog(ERROR, "Bad input string for type bytea");

	for (byte = 0, tp = inputText; *tp != '\0'; byte++)
		if (*tp++ == '\\')
		{
			if (*tp == '\\')
				tp++;
			else if (!isdigit(*tp++) ||
					 !isdigit(*tp++) ||
					 !isdigit(*tp++))
				elog(ERROR, "Bad input string for type bytea");
		}
	tp = inputText;
	byte += VARHDRSZ;
	result = (text *) palloc(byte);
	result->vl_len = byte;		/* varlena? */
	rp = result->vl_dat;
	while (*tp != '\0')
		if (*tp != '\\' || *++tp == '\\')
			*rp++ = *tp++;
		else
		{
			byte = VAL(*tp++);
			byte <<= 3;
			byte += VAL(*tp++);
			byte <<= 3;
			*rp++ = byte + VAL(*tp++);
		}
	return result;
}

/*
 *		byteaout		- converts to printable representation of byte array
 *
 *		Non-printable characters are inserted as '\nnn' (octal) and '\' as
 *		'\\'.
 *
 *		NULL vlena should be an error--returning string with NULL for now.
 */
char *
byteaout(text *vlena)
{
	char	   *result;

	char	   *vp;
	char	   *rp;
	int			val;			/* holds unprintable chars */
	int			i;
	int			len;

	if (vlena == NULL)
	{
		result = (char *) palloc(2);
		result[0] = '-';
		result[1] = '\0';
		return result;
	}
	vp = vlena->vl_dat;
	len = 1;					/* empty string has 1 char */
	for (i = vlena->vl_len - VARHDRSZ; i != 0; i--, vp++)
		if (*vp == '\\')
			len += 2;
		else if (isascii(*vp) && isprint(*vp))
			len++;
		else
			len += VARHDRSZ;
	rp = result = (char *) palloc(len);
	vp = vlena->vl_dat;
	for (i = vlena->vl_len - VARHDRSZ; i != 0; i--)
		if (*vp == '\\')
		{
			vp++;
			*rp++ = '\\';
			*rp++ = '\\';
		}
		else if (isascii(*vp) && isprint(*vp))
			*rp++ = *vp++;
		else
		{
			val = *vp++;
			*rp = '\\';
			rp += 3;
			*rp-- = DIG(val & 07);
			val >>= 3;
			*rp-- = DIG(val & 07);
			val >>= 3;
			*rp = DIG(val & 03);
			rp += 3;
		}
	*rp = '\0';
	return result;
}


/*
 *		textin			- converts "..." to internal representation
 */
text *
textin(char *inputText)
{
	text	   *result;
	int			len;

	if (inputText == NULL)
		return NULL;

	len = strlen(inputText) + VARHDRSZ;
	result = (text *) palloc(len);
	VARSIZE(result) = len;

	memmove(VARDATA(result), inputText, len - VARHDRSZ);

#ifdef CYR_RECODE
	convertstr(VARDATA(result), len - VARHDRSZ, 0);
#endif

	return result;
}

/*
 *		textout			- converts internal representation to "..."
 */
char *
textout(text *vlena)
{
	int			len;
	char	   *result;

	if (vlena == NULL)
	{
		result = (char *) palloc(2);
		result[0] = '-';
		result[1] = '\0';
		return result;
	}
	len = VARSIZE(vlena) - VARHDRSZ;
	result = (char *) palloc(len + 1);
	memmove(result, VARDATA(vlena), len);
	result[len] = '\0';

#ifdef CYR_RECODE
	convertstr(result, len, 1);
#endif

	return result;
}


/* ========== PUBLIC ROUTINES ========== */

/*
 * textlen -
 *	  returns the logical length of a text*
 *	   (which is less than the VARSIZE of the text*)
 */
int32
textlen(text *t)
{
#ifdef MULTIBYTE
	unsigned char *s;
	int			len,
				l,
				wl;

#endif

	if (!PointerIsValid(t))
		elog(ERROR, "Null input to textlen");

#ifdef MULTIBYTE
	len = 0;
	s = VARDATA(t);
	l = VARSIZE(t) - VARHDRSZ;
	while (l > 0)
	{
		wl = pg_mblen(s);
		l -= wl;
		s += wl;
		len++;
	}
	return (len);
#else
	return VARSIZE(t) - VARHDRSZ;
#endif

}	/* textlen() */

/*
 * textoctetlen -
 *	  returns the physical length of a text*
 *	   (which is less than the VARSIZE of the text*)
 */
int32
textoctetlen(text *t)
{
	if (!PointerIsValid(t))
		elog(ERROR, "Null input to textoctetlen");

	return VARSIZE(t) - VARHDRSZ;

}	/* textoctetlen() */

/*
 * textcat -
 *	  takes two text* and returns a text* that is the concatentation of
 *	  the two.
 *
 * Rewritten by Sapa, [e-mail]. 8-Jul-96.
 * Updated by Thomas, [e-mail] 1997-07-10.
 * Allocate space for output in all cases.
 * XXX - thomas 1997-07-10
 */
text *
textcat(text *t1, text *t2)
{
	int			len1,
				len2,
				len;
	char	   *ptr;
	text	   *result;

	if (!PointerIsValid(t1) || !PointerIsValid(t2))
		return NULL;

	len1 = (VARSIZE(t1) - VARHDRSZ);
	if (len1 < 0)
		len1 = 0;
	while (len1 > 0 && VARDATA(t1)[len1 - 1] == '\0')
		len1--;

	len2 = (VARSIZE(t2) - VARHDRSZ);
	if (len2 < 0)
		len2 = 0;
	while (len2 > 0 && VARDATA(t2)[len2 - 1] == '\0')
		len2--;

	len = len1 + len2 + VARHDRSZ;
	result = palloc(len);

	/* Set size of result string... */
	VARSIZE(result) = len;

	/* Fill data field of result string... */
	ptr = VARDATA(result);
	if (len1 > 0)
		memcpy(ptr, VARDATA(t1), len1);
	if (len2 > 0)
		memcpy(ptr + len1, VARDATA(t2), len2);

	return result;
}	/* textcat() */

/*
 * text_substr()
 * Return a substring starting at the specified position.
 * - thomas 1997-12-31
 *
 * Input:
 *	- string
 *	- starting position (is one-based)
 *	- string length
 *
 * If the starting position is zero or less, then return from the start of the string
 *	adjusting the length to be consistant with the "negative start" per SQL92.
 * If the length is less than zero, return the remaining string.
 *
 * Note that the arguments operate on octet length,
 *	so not aware of multi-byte character sets.
 *
 * Added multi-byte support.
 * - Tatsuo Ishii 1998-4-21
 * Changed behavior if starting position is less than one to conform to SQL92 behavior.
 * Formerly returned the entire string; now returns a portion.
 * - Thomas Lockhart 1998-12-10
 */
text *
text_substr(text *string, int32 m, int32 n)
{
	text	   *ret;
	int			len;

#ifdef MULTIBYTE
	int			i;
	char	   *p;

#endif

	if (string == (text *) NULL)
		return string;

	len = VARSIZE(string) - VARHDRSZ;
#ifdef MULTIBYTE
	len = pg_mbstrlen_with_len(VARDATA(string), len);
#endif

	/* starting position after the end of the string? */
	if (m > len)
	{
		m = 1;
		n = 0;
	}

	/*
	 * starting position before the start of the string? then offset into
	 * the string per SQL92 spec...
	 */
	else if (m < 1)
	{
		n += (m - 1);
		m = 1;
	}

	/* m will now become a zero-based starting position */
	m--;
	if (((m + n) > len) || (n < 0))
		n = (len - m);

#ifdef MULTIBYTE
	p = VARDATA(string);
	for (i = 0; i < m; i++)
		p += pg_mblen(p);
	m = p - VARDATA(string);
	for (i = 0; i < n; i++)
		p += pg_mblen(p);
	n = p - (VARDATA(string) + m);
#endif
	ret = (text *) palloc(VARHDRSZ + n);
	VARSIZE(ret) = VARHDRSZ + n;

	memcpy(VARDATA(ret), VARDATA(string) + m, n);

	return ret;
}	/* text_substr() */

/*
 * textpos -
 *	  Return the position of the specified substring.
 *	  Implements the SQL92 POSITION() function.
 *	  Ref: A Guide To The SQL Standard, Date & Darwen, 1997
 * - thomas 1997-07-27
 *
 * Added multi-byte support.
 * - Tatsuo Ishii 1998-4-21
 */
int32
textpos(text *t1, text *t2)
{
	int			pos;
	int			px,
				p;
	int			len1,
				len2;
	pg_wchar   *p1,
			   *p2;

#ifdef MULTIBYTE
	pg_wchar   *ps1,
			   *ps2;

#endif

	if (!PointerIsValid(t1) || !PointerIsValid(t2))
		return 0;

	if (VARSIZE(t2) <= 0)
		return 1;

	len1 = (VARSIZE(t1) - VARHDRSZ);
	len2 = (VARSIZE(t2) - VARHDRSZ);
#ifdef MULTIBYTE