utf.c

在x86平台上运行不可信任代码的sandbox。

/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 1998-2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE ANY
 * REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
#define _BSD_SOURCE 1 /* for memcccpy */
#include <stdarg.h>
#include <string.h>
#include <stdlib.h>
#include "u.h"
#include "utf.h"
#define bsearch utfbsearch
/*
 * compiler directive on Plan 9
 */
#ifndef USED
#define USED(x) if(x);else
#endif

/*
 * nil cannot be ((void*)0) on ANSI C,
 * because it is used for function pointers
 */
#undef	nil
#define	nil	0

#undef	nelem
#define	nelem(x)	(sizeof (x)/sizeof (x)[0])


/* ---------- end preamble -------- */

/* -------------- rune.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

enum
{
	Bit1	= 7,
	Bitx	= 6,
	Bit2	= 5,
	Bit3	= 4,
	Bit4	= 3,

	T1	= ((1<<(Bit1+1))-1) ^ 0xFF,	/* 0000 0000 */
	Tx	= ((1<<(Bitx+1))-1) ^ 0xFF,	/* 1000 0000 */
	T2	= ((1<<(Bit2+1))-1) ^ 0xFF,	/* 1100 0000 */
	T3	= ((1<<(Bit3+1))-1) ^ 0xFF,	/* 1110 0000 */
	T4	= ((1<<(Bit4+1))-1) ^ 0xFF,	/* 1111 0000 */

	Rune1	= (1<<(Bit1+0*Bitx))-1,		/* 0000 0000 0111 1111 */
	Rune2	= (1<<(Bit2+1*Bitx))-1,		/* 0000 0111 1111 1111 */
	Rune3	= (1<<(Bit3+2*Bitx))-1,		/* 1111 1111 1111 1111 */

	Maskx	= (1<<Bitx)-1,			/* 0011 1111 */
	Testx	= Maskx ^ 0xFF,			/* 1100 0000 */

	Bad	= Runeerror
};

int
chartorune(Rune *rune, char *str)
{
	int c, c1, c2;
	long l;

	/*
	 * one character sequence
	 *	00000-0007F => T1
	 */
	c = *(uchar*)str;
	if(c < Tx) {
		*rune = c;
		return 1;
	}

	/*
	 * two character sequence
	 *	0080-07FF => T2 Tx
	 */
	c1 = *(uchar*)(str+1) ^ Tx;
	if(c1 & Testx)
		goto bad;
	if(c < T3) {
		if(c < T2)
			goto bad;
		l = ((c << Bitx) | c1) & Rune2;
		if(l <= Rune1)
			goto bad;
		*rune = l;
		return 2;
	}

	/*
	 * three character sequence
	 *	0800-FFFF => T3 Tx Tx
	 */
	c2 = *(uchar*)(str+2) ^ Tx;
	if(c2 & Testx)
		goto bad;
	if(c < T4) {
		l = ((((c << Bitx) | c1) << Bitx) | c2) & Rune3;
		if(l <= Rune2)
			goto bad;
		*rune = l;
		return 3;
	}

	/*
	 * bad decoding
	 */
bad:
	*rune = Bad;
	return 1;
}

int
runetochar(char *str, Rune *rune)
{
	long c;

	/*
	 * one character sequence
	 *	00000-0007F => 00-7F
	 */
	c = *rune;
	if(c <= Rune1) {
		str[0] = c;
		return 1;
	}

	/*
	 * two character sequence
	 *	0080-07FF => T2 Tx
	 */
	if(c <= Rune2) {
		str[0] = T2 | (c >> 1*Bitx);
		str[1] = Tx | (c & Maskx);
		return 2;
	}

	/*
	 * three character sequence
	 *	0800-FFFF => T3 Tx Tx
	 */
	str[0] = T3 |  (c >> 2*Bitx);
	str[1] = Tx | ((c >> 1*Bitx) & Maskx);
	str[2] = Tx |  (c & Maskx);
	return 3;
}

int
runelen(long c)
{
	Rune rune;
	char str[10];

	rune = c;
	return runetochar(str, &rune);
}

int
runenlen(Rune *r, int nrune)
{
	int nb, c;

	nb = 0;
	while(nrune--) {
		c = *r++;
		if(c <= Rune1)
			nb++;
		else
		if(c <= Rune2)
			nb += 2;
		else
			nb += 3;
	}
	return nb;
}

int
fullrune(char *str, int n)
{
	int c;

	if(n > 0) {
		c = *(uchar*)str;
		if(c < Tx)
			return 1;
		if(n > 1)
			if(c < T3 || n > 2)
				return 1;
	}
	return 0;
}
/* -------------- runestrcat.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

Rune*
runestrcat(Rune *s1, Rune *s2)
{

	runestrcpy(runestrchr(s1, 0), s2);
	return s1;
}
/* -------------- runestrchr.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

Rune*
runestrchr(Rune *s, Rune c)
{
	Rune c0 = c;
	Rune c1;

	if(c == 0) {
		while(*s++)
			;
		return s-1;
	}

	while((c1 = *s++))
		if(c1 == c0)
			return s-1;
	return 0;
}
/* -------------- runestrcmp.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

int
runestrcmp(Rune *s1, Rune *s2)
{
	Rune c1, c2;

	for(;;) {
		c1 = *s1++;
		c2 = *s2++;
		if(c1 != c2) {
			if(c1 > c2)
				return 1;
			return -1;
		}
		if(c1 == 0)
			return 0;
	}
}
/* -------------- runestrcpy.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

Rune*
runestrcpy(Rune *s1, Rune *s2)
{
	Rune *os1;

	os1 = s1;
	while((*s1++ = *s2++))
		;
	return os1;
}
/* -------------- runestrdup.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include <stdlib.h>
// #include "plan9.h"
// #include "utf.h"

Rune*
runestrdup(Rune *s) 
{  
	Rune *ns;

	ns = malloc(sizeof(Rune)*(runestrlen(s) + 1));
	if(ns == 0)
		return 0;

	return runestrcpy(ns, s);
}
/* -------------- runestrlen.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

long
runestrlen(Rune *s)
{

	return runestrchr(s, 0) - s;
}
/* -------------- runestrecpy.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

Rune*
runestrecpy(Rune *s1, Rune *es1, Rune *s2)
{
	if(s1 >= es1)
		return s1;

	while((*s1++ = *s2++)){
		if(s1 == es1){
			*--s1 = '\0';
			break;
		}
	}
	return s1;
}
/* -------------- runestrncat.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

Rune*
runestrncat(Rune *s1, Rune *s2, long n)
{
	Rune *os1;

	os1 = s1;
	s1 = runestrchr(s1, 0);
	while((*s1++ = *s2++))
		if(--n < 0) {
			s1[-1] = 0;
			break;
		}
	return os1;
}
/* -------------- runestrncmp.c --------------- */
/*
 * The authors of this software are Rob Pike and Ken Thompson.
 *              Copyright (c) 2002 by Lucent Technologies.
 * Permission to use, copy, modify, and distribute this software for any
 * purpose without fee is hereby granted, provided that this entire notice
 * is included in all copies of any software which is or includes a copy
 * or modification of this software and in all copies of the supporting
 * documentation for such software.
 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
 * WARRANTY.  IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
 */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

int
runestrncmp(Rune *s1, Rune *s2, long n)
{
	Rune c1, c2;

	while(n > 0) {
		c1 = *s1++;
		c2 = *s2++;
		n--;
		if(c1 != c2) {
			if(c1 > c2)
				return 1;
			return -1;
		}
		if(c1 == 0)
			break;
	}
	return 0;
}