src/cmd/dict/world.c

#include <u.h>
#include <libc.h>
#include <bio.h>
#include "dict.h"
#include "kuten.h"

/*
 * Routines for handling dictionaries in the "Languages of the World"
 * format.  worldnextoff *must* be called with <address of valid entry>+1.
 */

#define	GSHORT(p)	(((p)[0]<<8)|(p)[1])

#define putchar dictputchar

static void	putchar(int, int*);

#define	NONE	0xffff

/* adapted from jhelling@cs.ruu.nl (Jeroen Hellingman) */

static Rune chartab[] = {

/*00*/	NONE,	NONE,	NONE,	NONE,	NONE,	NONE,	NONE,	NONE,
	NONE,	NONE,'\n',	0xe6,	0xf8,	0xe5,	0xe4,	0xf6,
/*10*/	NONE,	NONE,	NONE,	NONE,	NONE,	NONE,	NONE,	NONE,
	NONE,	NONE,	NONE,	0xc6,	0xd8,	0xc5,	0xc4,	0xd6,

/*20*/	0x20,	0x21,	0x22,	0x23,	0x24,	0x25,	0x26,	'\'',
	0x28,	0x29,	0x2a,	0x2b,	0x2c,	0x2d,	0x2e,	0x2f,
/*30*/  0x30,	0x31,	0x32,	0x33,	0x34,	0x35,	0x36,	0x37,
	0x38,	0x39,	0x3a,	0x3b,	0x3c,	0x3d,	0x3e,	0x3f,
/*40*/  0x40,	0x41,	0x42,	0x43,	0x44,	0x45,	0x46,	0x47,
	0x48,	0x49,	0x4a,	0x4b,'L',	0x4d,	0x4e,	0x4f,
/*50*/	0x50,	0x51,	0x52,	0x53,	0x54,	0x55,	0x56,	0x57,
	0x58,	0x59,	0x5a,	0x5b,'\\',	0x5d,	0x5e,	0x5f,
/*60*/	0x60,	0x61,	0x62,	0x63,	0x64,	0x65,	0x66,	0x67,
	0x68,	0x69,	0x6a,	0x6b,	0x6c,	0x6d,	0x6e,	0x6f,
/*70*/	0x70,	0x71,	0x72,	0x73,	0x74,	0x75,	0x76,	0x77,
	0x78,	0x79,	0x7a,	0x7b,	0x7c,	0x7d,	0x7e,	NONE,

/*80*/	0xc7,	0xfc,	0xe9,	0xe2,	0xe4,	0xe0,	0xe5,	0xe7,
	0xea,	0xeb,	0xe8,	0xef,	0xee,	0xec,	0xc4,	0xc5,
/*90*/	0xc9,	0xe6,	0xc6,	0xf4,	0xf6,	0xf2,	0xfb,	0xf9,
	0xff,	0xd6,	0xdc,	0xa2,	0xa3,	0xa5,	0x20a7,	0x283,
/*a0*/	0xe1,	0xed,	0xf3,	0xfa,	0xf1,	0xd1,	0xaa,	0xba,
	0xbf,	0x2310,	0xac,	0xbd,	0xbc,	0xa1,	0xab,	0xbb,

/*b0*/	0x254,	0x259,	0xf0,	0x283,	0x292,	0x14b,	0x251,	0x7a,
	0x26a,	0xf0,	0x292,	0xe3,	0x153,	0x169,	0x28c,	0x265,
/*c0*/	0x280,	0xeb,	0x6c,	0x28c,	0xf5,	0xf1,	0x152,	NONE,
	NONE,	0x53,	0x73,	0x5a,	0x7a,	NONE,	NONE,	NONE,
/*d0*/	0xdf,	NONE,	NONE,	0x101,	0x12b,	0x16b,	0x113,	0x14d,	
	NONE,	NONE,	NONE,	0x20,	NONE,	NONE,	NONE,	NONE,

/*e0*/	0x3b1,	0x3b2,	0x3b3,	0x3c0,	0x3a3,	0x3c3,	0xb5,	0x3c4,
	0x3a6,	0x398,	0x3a9,	0x3b4,	0x221e,	0xd8,	0x3b5,	0x2229,
/*f0*/	0x2261,	0xb1,	0x2265,	0x2264,	0x2320,	0x2321,	0xf7,	0x2248,
	0xb0,	0x2219,	0xb7,	NONE,	NONE,	NONE,	NONE,	NONE
};

enum{ Utf, Kanahi, Kanalo=Kanahi+1, GBhi, GBlo=GBhi+1 };

void
worldprintentry(Entry e, int cmd)
{
	int nh, state[3];
	uchar *p, *pe;

	p = (uchar *)e.start;
	pe = (uchar *)e.end;
	nh = GSHORT(p);
	p += 6;
	if(cmd == 'h')
		pe = p+nh;
	state[0] = Utf;
	state[1] = 0;
	state[2] = 0;
	while(p < pe){
		if(cmd == 'r')
			outchar(*p++);
		else
			putchar(*p++, state);
	}
	outnl(0);
}

long
worldnextoff(long fromoff)
{
	int nh, np, nd;
	uchar buf[6];

	if(Bseek(bdict, fromoff-1, 0) < 0)
		return -1;
	if(Bread(bdict, buf, 6) != 6)
		return -1;
	nh = GSHORT(buf);
	np = GSHORT(buf+2);
	nd = GSHORT(buf+4);
	return fromoff-1 + 6 + nh + np + nd;
}

static void
putchar(int c, int *state)
{
	int xflag = 0;
	Rune r;
	int hi, lo;

	switch(state[0]){
	case Kanahi:
	case GBhi:
		if(CANS2JH(c) || c == 0xff){
			state[0]++;
			state[1] = c;
			break;
		}
		/* fall through */
	case Utf:
		if(c == 0xfe){
			state[0] = Kanahi;
			break;
		}else if(c == 0xff){
			state[0] = GBhi;
			break;
		}
		r = chartab[c];
		if(r < 0x80 && state[2] == 0)
			outchar(r);
		else if(r == NONE){
			switch(c){
			case 0xfb:
				if(!xflag){
					state[2] = 1;
					break;
				}
			case 0xfc:
				if(!xflag){
					state[2] = 0;
					break;
				}
			case 0x10:
			case 0xc7: case 0xc8:
			case 0xd8: case 0xd9: case 0xda:
			case 0xdc: case 0xdd: case 0xde: case 0xdf:
			case 0xfd:
				if(!xflag)
					break;
				/* fall through */
			default:
				outprint("\\%.2ux", c);
			}
		}else if(state[2] == 0)
			outrune(r);
		break;
	case Kanalo:
	case GBlo:
		if(state[1] == 0xff && c == 0xff){
			state[0] = Utf;
			break;
		}
		state[0]--;
		hi = state[1];
		lo = c;
		S2J(hi, lo);		/* convert to JIS */
		r = hi*100 + lo - 3232;	/* convert to jis208 */
		if(state[0] == Kanahi && r < JIS208MAX)
			r = tabjis208[r];
		else if(state[0] == GBhi && r < GB2312MAX)
			r = tabgb2312[r];
		else
			r = NONE;
		if(r == NONE)
			outprint("\\%.2ux\\%.2ux", state[1], c);
		else
			outrune(r);
		break;
	}
}

void
worldprintkey(void)
{
	Bprint(bout, "No pronunciation key.\n");
}