utf.c

在x86平台上运行不可信任代码的sandbox。

	0x1e1c, 501,	/* Ḝ ḝ */
	0x1e1e, 501,	/* Ḟ ḟ */
	0x1e20, 501,	/* Ḡ ḡ */
	0x1e22, 501,	/* Ḣ ḣ */
	0x1e24, 501,	/* Ḥ ḥ */
	0x1e26, 501,	/* Ḧ ḧ */
	0x1e28, 501,	/* Ḩ ḩ */
	0x1e2a, 501,	/* Ḫ ḫ */
	0x1e2c, 501,	/* Ḭ ḭ */
	0x1e2e, 501,	/* Ḯ ḯ */
	0x1e30, 501,	/* Ḱ ḱ */
	0x1e32, 501,	/* Ḳ ḳ */
	0x1e34, 501,	/* Ḵ ḵ */
	0x1e36, 501,	/* Ḷ ḷ */
	0x1e38, 501,	/* Ḹ ḹ */
	0x1e3a, 501,	/* Ḻ ḻ */
	0x1e3c, 501,	/* Ḽ ḽ */
	0x1e3e, 501,	/* Ḿ ḿ */
	0x1e40, 501,	/* Ṁ ṁ */
	0x1e42, 501,	/* Ṃ ṃ */
	0x1e44, 501,	/* Ṅ ṅ */
	0x1e46, 501,	/* Ṇ ṇ */
	0x1e48, 501,	/* Ṉ ṉ */
	0x1e4a, 501,	/* Ṋ ṋ */
	0x1e4c, 501,	/* Ṍ ṍ */
	0x1e4e, 501,	/* Ṏ ṏ */
	0x1e50, 501,	/* Ṑ ṑ */
	0x1e52, 501,	/* Ṓ ṓ */
	0x1e54, 501,	/* Ṕ ṕ */
	0x1e56, 501,	/* Ṗ ṗ */
	0x1e58, 501,	/* Ṙ ṙ */
	0x1e5a, 501,	/* Ṛ ṛ */
	0x1e5c, 501,	/* Ṝ ṝ */
	0x1e5e, 501,	/* Ṟ ṟ */
	0x1e60, 501,	/* Ṡ ṡ */
	0x1e62, 501,	/* Ṣ ṣ */
	0x1e64, 501,	/* Ṥ ṥ */
	0x1e66, 501,	/* Ṧ ṧ */
	0x1e68, 501,	/* Ṩ ṩ */
	0x1e6a, 501,	/* Ṫ ṫ */
	0x1e6c, 501,	/* Ṭ ṭ */
	0x1e6e, 501,	/* Ṯ ṯ */
	0x1e70, 501,	/* Ṱ ṱ */
	0x1e72, 501,	/* Ṳ ṳ */
	0x1e74, 501,	/* Ṵ ṵ */
	0x1e76, 501,	/* Ṷ ṷ */
	0x1e78, 501,	/* Ṹ ṹ */
	0x1e7a, 501,	/* Ṻ ṻ */
	0x1e7c, 501,	/* Ṽ ṽ */
	0x1e7e, 501,	/* Ṿ ṿ */
	0x1e80, 501,	/* Ẁ ẁ */
	0x1e82, 501,	/* Ẃ ẃ */
	0x1e84, 501,	/* Ẅ ẅ */
	0x1e86, 501,	/* Ẇ ẇ */
	0x1e88, 501,	/* Ẉ ẉ */
	0x1e8a, 501,	/* Ẋ ẋ */
	0x1e8c, 501,	/* Ẍ ẍ */
	0x1e8e, 501,	/* Ẏ ẏ */
	0x1e90, 501,	/* Ẑ ẑ */
	0x1e92, 501,	/* Ẓ ẓ */
	0x1e94, 501,	/* Ẕ ẕ */
	0x1ea0, 501,	/* Ạ ạ */
	0x1ea2, 501,	/* Ả ả */
	0x1ea4, 501,	/* Ấ ấ */
	0x1ea6, 501,	/* Ầ ầ */
	0x1ea8, 501,	/* Ẩ ẩ */
	0x1eaa, 501,	/* Ẫ ẫ */
	0x1eac, 501,	/* Ậ ậ */
	0x1eae, 501,	/* Ắ ắ */
	0x1eb0, 501,	/* Ằ ằ */
	0x1eb2, 501,	/* Ẳ ẳ */
	0x1eb4, 501,	/* Ẵ ẵ */
	0x1eb6, 501,	/* Ặ ặ */
	0x1eb8, 501,	/* Ẹ ẹ */
	0x1eba, 501,	/* Ẻ ẻ */
	0x1ebc, 501,	/* Ẽ ẽ */
	0x1ebe, 501,	/* Ế ế */
	0x1ec0, 501,	/* Ề ề */
	0x1ec2, 501,	/* Ể ể */
	0x1ec4, 501,	/* Ễ ễ */
	0x1ec6, 501,	/* Ệ ệ */
	0x1ec8, 501,	/* Ỉ ỉ */
	0x1eca, 501,	/* Ị ị */
	0x1ecc, 501,	/* Ọ ọ */
	0x1ece, 501,	/* Ỏ ỏ */
	0x1ed0, 501,	/* Ố ố */
	0x1ed2, 501,	/* Ồ ồ */
	0x1ed4, 501,	/* Ổ ổ */
	0x1ed6, 501,	/* Ỗ ỗ */
	0x1ed8, 501,	/* Ộ ộ */
	0x1eda, 501,	/* Ớ ớ */
	0x1edc, 501,	/* Ờ ờ */
	0x1ede, 501,	/* Ở ở */
	0x1ee0, 501,	/* Ỡ ỡ */
	0x1ee2, 501,	/* Ợ ợ */
	0x1ee4, 501,	/* Ụ ụ */
	0x1ee6, 501,	/* Ủ ủ */
	0x1ee8, 501,	/* Ứ ứ */
	0x1eea, 501,	/* Ừ ừ */
	0x1eec, 501,	/* Ử ử */
	0x1eee, 501,	/* Ữ ữ */
	0x1ef0, 501,	/* Ự ự */
	0x1ef2, 501,	/* Ỳ ỳ */
	0x1ef4, 501,	/* Ỵ ỵ */
	0x1ef6, 501,	/* Ỷ ỷ */
	0x1ef8, 501,	/* Ỹ ỹ */
	0x1f59, 492,	/* Ὑ ὑ */
	0x1f5b, 492,	/* Ὓ ὓ */
	0x1f5d, 492,	/* Ὕ ὕ */
	0x1f5f, 492,	/* Ὗ ὗ */
	0x1fbc, 491,	/* ᾼ ᾳ */
	0x1fcc, 491,	/* ῌ ῃ */
	0x1fec, 493,	/* Ῥ ῥ */
	0x1ffc, 491,	/* ῼ ῳ */
};

/*
 * title characters are those between
 * upper and lower case. ie DZ Dz dz
 */
static
Rune	__totitle1[] =
{
	0x01c4, 501,	/* DŽ Dž */
	0x01c6, 499,	/* dž Dž */
	0x01c7, 501,	/* LJ Lj */
	0x01c9, 499,	/* lj Lj */
	0x01ca, 501,	/* NJ Nj */
	0x01cc, 499,	/* nj Nj */
	0x01f1, 501,	/* DZ Dz */
	0x01f3, 499,	/* dz Dz */
};

static Rune*
bsearch(Rune c, Rune *t, int n, int ne)
{
	Rune *p;
	int m;

	while(n > 1) {
		m = n/2;
		p = t + m*ne;
		if(c >= p[0]) {
			t = p;
			n = n-m;
		} else
			n = m;
	}
	if(n && c >= t[0])
		return t;
	return 0;
}

Rune
tolowerrune(Rune c)
{
	Rune *p;

	p = bsearch(c, __tolower2, nelem(__tolower2)/3, 3);
	if(p && c >= p[0] && c <= p[1])
		return c + p[2] - 500;
	p = bsearch(c, __tolower1, nelem(__tolower1)/2, 2);
	if(p && c == p[0])
		return c + p[1] - 500;
	return c;
}

Rune
toupperrune(Rune c)
{
	Rune *p;

	p = bsearch(c, __toupper2, nelem(__toupper2)/3, 3);
	if(p && c >= p[0] && c <= p[1])
		return c + p[2] - 500;
	p = bsearch(c, __toupper1, nelem(__toupper1)/2, 2);
	if(p && c == p[0])
		return c + p[1] - 500;
	return c;
}

Rune
totitlerune(Rune c)
{
	Rune *p;

	p = bsearch(c, __totitle1, nelem(__totitle1)/2, 2);
	if(p && c == p[0])
		return c + p[1] - 500;
	return c;
}

int
islowerrune(Rune c)
{
	Rune *p;

	p = bsearch(c, __toupper2, nelem(__toupper2)/3, 3);
	if(p && c >= p[0] && c <= p[1])
		return 1;
	p = bsearch(c, __toupper1, nelem(__toupper1)/2, 2);
	if(p && c == p[0])
		return 1;
	return 0;
}

int
isupperrune(Rune c)
{
	Rune *p;

	p = bsearch(c, __tolower2, nelem(__tolower2)/3, 3);
	if(p && c >= p[0] && c <= p[1])
		return 1;
	p = bsearch(c, __tolower1, nelem(__tolower1)/2, 2);
	if(p && c == p[0])
		return 1;
	return 0;
}

int
isalpharune(Rune c)
{
	Rune *p;

	if(isupperrune(c) || islowerrune(c))
		return 1;
	p = bsearch(c, __alpha2, nelem(__alpha2)/2, 2);
	if(p && c >= p[0] && c <= p[1])
		return 1;
	p = bsearch(c, __alpha1, nelem(__alpha1), 1);
	if(p && c == p[0])
		return 1;
	return 0;
}

int
istitlerune(Rune c)
{
	return isupperrune(c) && islowerrune(c);
}

int
isspacerune(Rune c)
{
	Rune *p;

	p = bsearch(c, __space2, nelem(__space2)/2, 2);
	if(p && c >= p[0] && c <= p[1])
		return 1;
	return 0;
}
/* -------------- utfecpy.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.
 */
#define _BSD_SOURCE 1	/* memccpy */
// #include <stdarg.h>
// #include <string.h>
// #include "plan9.h"
// #include "utf.h"

char*
utfecpy(char *to, char *e, char *from)
{
	char *end;

	if(to >= e)
		return to;
	end = memccpy(to, from, '\0', e - to);
	if(end == nil){
		end = e-1;
		while(end>to && (*--end&0xC0)==0x80)
			;
		*end = '\0';
	}else{
		end--;
	}
	return end;
}
/* -------------- utflen.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
utflen(char *s)
{
	int c;
	long n;
	Rune rune;

	n = 0;
	for(;;) {
		c = *(uchar*)s;
		if(c < Runeself) {
			if(c == 0)
				return n;
			s++;
		} else
			s += chartorune(&rune, s);
		n++;
	}
}
/* -------------- utfnlen.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
utfnlen(char *s, long m)
{
	int c;
	long n;
	Rune rune;
	char *es;

	es = s + m;
	for(n = 0; s < es; n++) {
		c = *(uchar*)s;
		if(c < Runeself){
			if(c == '\0')
				break;
			s++;
			continue;
		}
		if(!fullrune(s, es-s))
			break;
		s += chartorune(&rune, s);
	}
	return n;
}
/* -------------- utfrrune.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"

char*
utfrrune(char *s, long c)
{
	long c1;
	Rune r;
	char *s1;

	if(c < Runesync)		/* not part of utf sequence */
		return strrchr(s, c);

	s1 = 0;
	for(;;) {
		c1 = *(uchar*)s;
		if(c1 < Runeself) {	/* one byte rune */
			if(c1 == 0)
				return s1;
			if(c1 == c)
				s1 = s;
			s++;
			continue;
		}
		c1 = chartorune(&r, s);
		if(r == c)
			s1 = s;
		s += c1;
	}
}
/* -------------- utfrune.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"

char*
utfrune(char *s, long c)
{
	long c1;
	Rune r;
	int n;

	if(c < Runesync)		/* not part of utf sequence */
		return strchr(s, c);

	for(;;) {
		c1 = *(uchar*)s;
		if(c1 < Runeself) {	/* one byte rune */
			if(c1 == 0)
				return 0;
			if(c1 == c)
				return s;
			s++;
			continue;
		}
		n = chartorune(&r, s);
		if(r == c)
			return s;
		s += n;
	}
}
/* -------------- utfutf.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"


/*
 * Return pointer to first occurrence of s2 in s1,
 * 0 if none
 */
char*
utfutf(char *s1, char *s2)
{
	char *p;
	long f, n1, n2;
	Rune r;

	n1 = chartorune(&r, s2);
	f = r;
	if(f <= Runesync)		/* represents self */
		return strstr(s1, s2);

	n2 = strlen(s2);
	for(p=s1; (p=utfrune(p, f)); p+=n1)
		if(strncmp(p, s2, n2) == 0)
			return p;
	return 0;
}