deroff.c

<B>Digital的Unix操作系统VAX 4.2源码</B>

#ifndef lint
static char *sccsid = "@(#)deroff.c	4.1	ULTRIX	7/17/90";
#endif lint

/************************************************************************
 *									*
 *			Copyright (c) 1985 by				*
 *		Digital Equipment Corporation, Maynard, MA		*
 *			All rights reserved.				*
 *									*
 *   This software is furnished under a license and may be used and	*
 *   copied  only  in accordance with the terms of such license and	*
 *   with the  inclusion  of  the  above  copyright  notice.   This	*
 *   software  or  any  other copies thereof may not be provided or	*
 *   otherwise made available to any other person.  No title to and	*
 *   ownership of the software is hereby transferred.			*
 *									*
 *   This software is  derived  from  software  received  from  the	*
 *   University    of   California,   Berkeley,   and   from   Bell	*
 *   Laboratories.  Use, duplication, or disclosure is  subject  to	*
 *   restrictions  under  license  agreements  with  University  of	*
 *   California and with AT&T.						*
 *									*
 *   The information in this software is subject to change  without	*
 *   notice  and should not be construed as a commitment by Digital	*
 *   Equipment Corporation.						*
 *									*
 *   Digital assumes no responsibility for the use  or  reliability	*
 *   of its software on equipment which is not supplied by Digital.	*
 *									*
 ************************************************************************/
/************************************************************************
 *			Modification History
 *
 * 001	David L Ballenger, 14-Mar-1986
 *	Change use of "const" as an identifier, since it is now a keyword.
 *
 * Based on:  deroff.c	4.4	(Berkeley)	83/04/29
 *
 ************************************************************************/

#include <stdio.h>

/*
 *	Deroff command -- strip troff, eqn, and Tbl sequences from
 *	a file.  Has two flags argument, -w, to cause output one word per line
 *	rather than in the original format.
 *	-mm (or -ms) causes the corresponding macro's to be interpreted
 *	so that just sentences are output
 *	-ml  also gets rid of lists.
 *	Deroff follows .so and .nx commands, removes contents of macro
 *	definitions, equations (both .EQ ... .EN and $...$),
 *	Tbl command sequences, and Troff backslash constructions.
 *
 *	All input is through the Cget macro;
 *	the most recently read character is in c.
 *
 *	Modified by Robert Henry to process -me and -man macros.
 */

#define Cget ( (c=getc(infile)) == EOF ? eof() : ((c==ldelim)&&(filesp==files) ? skeqn() : c) )
#define C1get ( (c=getc(infile)) == EOF ? eof() :  c)

#ifdef DEBUG
#  define C	_C()
#  define C1	_C1()
#else not DEBUG
#  define C	Cget
#  define C1	C1get
#endif not DEBUG

#define SKIP while(C != '\n') 
#define SKIP_TO_COM SKIP; SKIP; pc=c; while(C != '.' || pc != '\n' || C > 'Z')pc=c

#define	YES 1
#define	NO 0
#define	MS 0	/* -ms */
#define	MM 1	/* -mm */
#define	ME 2	/* -me */
#define	MA 3	/* -man */

#ifdef DEBUG
char *mactab[] = {"-ms", "-mm", "-me", "-ma"};
#endif DEBUG

#define	ONE 1
#define	TWO 2

#define NOCHAR -2
#define SPECIAL 0
#define APOS 1
#define PUNCT 2
#define DIGIT 3
#define LETTER 4

int	wordflag;
int	msflag;		/* processing a source written using a mac package */
int	mac;		/* which package */
int	disp;
int	parag;
int	inmacro;
int	intable;
int	keepblock;	/* keep blocks of text; normally false when msflag */

char chars[128];  /* SPECIAL, PUNCT, APOS, DIGIT, or LETTER */

char line[512];
char *lp;

int c;
int pc;
int ldelim;
int rdelim;


int argc;
char **argv;

char fname[50];
FILE *files[15];
FILE **filesp;
FILE *infile;
FILE	*opn();
/*
 *	Flags for matching conditions other than
 *	the macro name
 */
#define	NONE		0
#define	FNEST		1		/* no nested files */
#define	NOMAC		2		/* no macro */
#define	MAC		3		/* macro */
#define	PARAG		4		/* in a paragraph */
#define	MSF		5		/* msflag is on */
#define	NBLK		6		/* set if no blocks to be kept */

/*
 *	Return codes from macro minions, determine where to jump,
 *	how to repeat/reprocess text
 */
#define	COMX		1		/* goto comx */
#define	COM		2		/* goto com */

main(ac, av)
int ac;
char **av;
{
	register int i;
	int	errflg = 0;
	register	optchar;
	FILE *opn();
	int	kflag = NO;
	char	*p;

	wordflag = NO;
	msflag = NO;
	mac = ME;
	disp = NO;
	parag = NO;
	inmacro = NO;
	intable = NO;
	ldelim	= NOCHAR;
	rdelim	= NOCHAR;
	keepblock = YES;

	for(argc = ac - 1, argv = av + 1;
	    (   (argc > 0)
	     && (argv[0][0] == '-')
	     && (argv[0][1] != '\0') );
	    --argc, ++argv
	){
		for(p = argv[0]+1; *p; ++p) {
			switch(*p) {
			case 'p':
				parag=YES;
				break;
			case 'k':
				kflag = YES;
				break;
			case 'w':
				wordflag = YES;
				kflag = YES;
				break;
			case 'm':
				msflag = YES;
				keepblock = NO;
				switch(p[1]){
				case 'm':	mac = MM; p++; break;
				case 's':	mac = MS; p++; break;
				case 'e':	mac = ME; p++; break;
				case 'a':	mac = MA; p++; break;
				case 'l':	disp = YES; p++; break;
				default:	errflg++; break;
				}
				break;
			default:
				errflg++;
			}
		}
	}

	if (kflag)
		keepblock = YES;
	if (errflg)
		fatal("usage: deroff [ -w ] [ -k] [ -m (a e m s l) ] [ file ] ... \n",
			(char *) NULL);

#ifdef DEBUG
	printf("msflag = %d, mac = %s, keepblock = %d, disp = %d\n",
		msflag, mactab[mac], keepblock, disp);
#endif DEBUG
	if (argc == 0){
		infile = stdin;
	} else {
		infile = opn(argv[0]);
		--argc;
		++argv;
	}


	files[0] = infile;
	filesp = &files[0];

	for(i='a'; i<='z' ; ++i)
		chars[i] = LETTER;
	for(i='A'; i<='Z'; ++i)
		chars[i] = LETTER;
	for(i='0'; i<='9'; ++i)
		chars[i] = DIGIT;
	chars['\''] = APOS;
	chars['&'] = APOS;
	chars['.'] = PUNCT;
	chars[','] = PUNCT;
	chars[';'] = PUNCT;
	chars['?'] = PUNCT;
	chars[':'] = PUNCT;
	work();
}
char *calloc();

skeqn()
{
	while((c = getc(infile)) != rdelim)
		if(c == EOF)
			c = eof();
		else if(c == '"')
			while( (c = getc(infile)) != '"')
				if(c == EOF)
					c = eof();
				else if(c == '\\')
					if((c = getc(infile)) == EOF)
						c = eof();
	if(msflag)return(c='x');
	return(c = ' ');
}

FILE *opn(p)
register char *p;
{
	FILE *fd;

	if( (fd = fopen(p, "r")) == NULL) {
		fprintf(stderr, "Deroff: ");
		perror(p);
		exit(1);
	}

	return(fd);
}

eof()
{
	if(infile != stdin)
		fclose(infile);
	if(filesp > files)
		infile = *--filesp;
	else if (argc > 0) {
		infile = opn(argv[0]);
		--argc;
		++argv;
	} else
		exit(0);
	return(C);
}

getfname()
{
	register char *p;
	struct chain { 
		struct chain *nextp; 
		char *datap; 
	} *chainblock;
	register struct chain *q;
	static struct chain *namechain	= NULL;
	char *copys();

	while(C == ' ') ;

	for(p = fname ; (*p=c)!= '\n' && c!=' ' && c!='\t' && c!='\\' ; ++p)
		C;
	*p = '\0';
	while(c != '\n')
		C;

	/* see if this name has already been used */

	for(q = namechain ; q; q = q->nextp)
		if( ! strcmp(fname, q->datap))
		{
			fname[0] = '\0';
			return;
		}

	q = (struct chain *) calloc(1, sizeof(*chainblock));
	q->nextp = namechain;
	q->datap = copys(fname);
	namechain = q;
}

fatal(s,p)
char *s, *p;
{
	fprintf(stderr, "Deroff: ");
	fprintf(stderr, s, p);
	exit(1);
}

/*ARGSUSED*/
textline(str, const_val)
	char	*str;
	int	const_val;
{
	if (wordflag) {
		msputwords(0);
		return;
	}
	puts(str);
}

work()
{
	for( ;; )
	{
		C;
#ifdef FULLDEBUG
		printf("Starting work with `%c'\n", c);
#endif FULLDEBUG
		if(c == '.'  ||  c == '\'')
			comline();
		else
			regline(textline, TWO);
	}
}

regline(pfunc, const_val)
	int	(*pfunc)();
	int	const_val;
{
	line[0] = c;
	lp = line;
	for( ; ; )
	{
		if(c == '\\') {
			*lp = ' ';
			backsl();
		}
		if(c == '\n')
			break;
		if(intable && c=='T') {
			*++lp = C;
			if(c=='{' || c=='}') {
				lp[-1] = ' ';
				*lp = C;
			}
		} else {
			*++lp = C;
		}
	}

	*lp = '\0';

	if(line[0] != '\0')
		(*pfunc)(line, const_val);
}

macro()
{
	if(msflag){
		do { 
			SKIP; 
		}		while(C!='.' || C!='.' || C=='.');	/* look for  .. */
		if(c != '\n')SKIP;
		return;
	}
	SKIP;
	inmacro = YES;
}

tbl()
{
	while(C != '.');
	SKIP;
	intable = YES;
}
stbl()
{
	while(C != '.');
	SKIP_TO_COM;
	if(c != 'T' || C != 'E'){
		SKIP;
		pc=c;
		while(C != '.' || pc != '\n' || C != 'T' || C != 'E')pc=c;
	}
}

eqn()
{
	register int c1, c2;
	register int dflg;
	char last;

	last=0;
	dflg = 1;
	SKIP;

	for( ;;)
	{
		if(C1 == '.'  || c == '\'')
		{
			while(C1==' ' || c=='\t')
				;
			if(c=='E' && C1=='N')
			{
				SKIP;
				if(msflag && dflg){
					putchar('x');
					putchar(' ');
					if(last){
						putchar(last); 
						putchar('\n'); 
					}
				}
				return;
			}
		}
		else if(c == 'd')	/* look for delim */
		{
			if(C1=='e' && C1=='l')
				if( C1=='i' && C1=='m')
				{
					while(C1 == ' ');
					if((c1=c)=='\n' || (c2=C1)=='\n'
					    || (c1=='o' && c2=='f' && C1=='f') )
					{
						ldelim = NOCHAR;
						rdelim = NOCHAR;
					}
					else	{
						ldelim = c1;
						rdelim = c2;
					}
				}
			dflg = 0;
		}

		if(c != '\n') while(C1 != '\n'){ 
			if(chars[c] == PUNCT)last = c;
			else if(c != ' ')last = 0;
		}
	}
}

backsl()	/* skip over a complete backslash construction */
{
	int bdelim;

sw:  
	switch(C)
	{
	case '"':
		SKIP;
		return;
	case 's':
		if(C == '\\') backsl();
		else	{
			while(C>='0' && c<='9') ;
			ungetc(c,infile);
			c = '0';
		}
		--lp;
		return;

	case 'f':
	case 'n':
	case '*':
		if(C != '(')
			return;

	case '(':
		if(msflag){
			if(C == 'e'){
				if(C == 'm'){
					*lp = '-';
					return;
				}
			}
			else if(c != '\n')C;
			return;
		}
		if(C != '\n') C;
		return;

	case '$':
		C;	/* discard argument number */
		return;

	case 'b':
	case 'x':
	case 'v':
	case 'h':
	case 'w':
	case 'o':
	case 'l':
	case 'L':
		if( (bdelim=C) == '\n')
			return;
		while(C!='\n' && c!=bdelim)
			if(c == '\\') backsl();
		return;

	case '\\':
		if(inmacro)
			goto sw;
	default:
		return;
	}
}

char *copys(s)
register char *s;
{
	register char *t, *t0;

	if( (t0 = t = calloc( (unsigned)(strlen(s)+1), sizeof(*t) ) ) == NULL)
		fatal("Cannot allocate memory", (char *) NULL);

	while( *t++ = *s++ )
		;
	return(t0);
}

sce()
{
	register char *ap;
	register int n, i;
	char a[10];
	for(ap=a;C != '\n';ap++){
		*ap = c;
		if(ap == &a[9]){
			SKIP;
			ap=a;
			break;
		}
	}
	if(ap != a)n = atoi(a);
	else n = 1;
	for(i=0;i<n;){
		if(C == '.'){
			if(C == 'c'){
				if(C == 'e'){
					while(C == ' ');
					if(c == '0'){
						SKIP;
						break;
					}
					else SKIP;
				}
				else SKIP;
			}
			else if(c == 'P' || C == 'P'){
				if(c != '\n')SKIP;
				break;
			}
			else if(c != '\n')SKIP;
		}
		else {
			SKIP;
			i++;
		}
	}
}

refer(c1)
{
	register int c2;
	if(c1 != '\n')
		SKIP;
	while(1){
		if(C != '.')
			SKIP;
		else {
			if(C != ']')
				SKIP;
			else {
				while(C != '\n')
					c2=c;
				if(chars[c2] == PUNCT)putchar(c2);
				return;
			}
		}
	}
}

inpic()
{
	register int c1;
	register char *p1;
	SKIP;
	p1 = line;
	c = '\n';
	while(1){
		c1 = c;
		if(C == '.' && c1 == '\n'){
			if(C != 'P'){
				if(c == '\n')continue;
				else { SKIP; c='\n'; continue;}
			}
			if(C != 'E'){
				if(c == '\n')continue;
				else { SKIP; c='\n';continue; }
			}
			SKIP;
			return;
		}
		else if(c == '\"'){
			while(C != '\"'){
				if(c == '\\'){
					if(C == '\"')continue;
					ungetc(c,infile);
					backsl();
				}
				else *p1++ = c;
			}
			*p1++ = ' ';
		}
		else if(c == '\n' && p1 != line){
			*p1 = '\0';
			if(wordflag)msputwords(NO);
			else {
				puts(line);
				putchar('\n');
			}
			p1 = line;
		}
	}
}

#ifdef DEBUG
_C1()
{
	return(C1get);
}
_C()
{
	return(Cget);
}
#endif DEBUG

/*
 *	Macro processing
 *
 *	Macro table definitions
 */
#define	reg	register
typedef	int pacmac;		/* compressed macro name */
int	argconcat = 0;		/* concat arguments together (-me only) */

#define	tomac(c1, c2)		((((c1) & 0xFF) << 8) | ((c2) & 0xFF))
#define	frommac(src, c1, c2)	(((c1)=((src)>>8)&0xFF),((c2) =(src)&0xFF))

struct	mactab{
	int	condition;
	pacmac	macname;
	int	(*func)();
};
extern struct	mactab	troffmactab[];
extern struct	mactab	ppmactab[];
extern struct	mactab	msmactab[];
extern struct	mactab	mmmactab[];
extern struct	mactab	memactab[];
extern struct	mactab	manmactab[];
/*
 *	macro table initialization
 */
#define	M(cond, c1, c2, func) {cond, tomac(c1, c2), func}

/*
 *	Put out a macro line, using ms and mm conventions.
 */
msputmac(s, const_val)
	register char *s;
	int	const_val;
{
	register char *t;
	register found;
	int last;
	found = 0;

	if (wordflag) {
		msputwords(YES);
		return;
	}
	while(*s)
	{
		while(*s==' ' || *s=='\t')
			putchar(*s++);
		for(t = s ; *t!=' ' && *t!='\t' && *t!='\0' ; ++t)
			;
		if(*s == '\"')s++;
		if(t>s+const_val && chars[ s[0] ]==LETTER && chars[ s[1] ]==LETTER){
			while(s < t)
				if(*s == '\"')s++;
				else
					putchar(*s++);
			last = *(t-1);
			found++;
		}
		else if(found && chars[ s[0] ] == PUNCT && s[1] == '\0')
			putchar(*s++);
		else{
			last = *(t-1);
			s = t;
		}
	}
	putchar('\n');
	if(msflag && chars[last] == PUNCT){
		putchar(last);
		putchar('\n');
	}
}