[unix-history] / usr / src / old / awk / lib.c

#ifndef lint
static char sccsid[] = "@(#)lib.c	4.8 (Berkeley) %G%";
#endif

#include <stdarg.h>
#include "stdio.h"
#include "awk.def"
#include "awk.h"
#include "ctype.h"

extern FILE	*yyin;	/* lex input file */
extern char	*lexprog;	/* points to program argument if it exists */
FILE	*infile	= NULL;
char	*file;
#define	RECSIZE	(5 * 512)
char	record[RECSIZE];
char	fields[RECSIZE];
char	EMPTY[] = "";

#define	MAXFLD	100
int	donefld;	/* 1 = implies rec broken into fields */
int	donerec;	/* 1 = record is valid (no flds have changed) */
int	mustfld;	/* 1 = NF seen, so always break*/

#define	FINIT	{EMPTY, EMPTY, 0.0, FLD|STR}
cell fldtab[MAXFLD] = {	/*room for fields */
	{ "$record", record, 0.0, STR|FLD},
	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
};
int	maxfld	= 0;	/* last used field */


getrec()
{
	register char *rr;
	extern int svargc;
	extern char **svargv;
	register c, sep;

	dprintf("**RS=%o, **FS=%o\n", **RS, **FS, NULL);
	donefld = 0;
	donerec = 1;
	record[0] = 0;
	while (svargc > 0) {
		dprintf("svargc=%d, *svargv=%s\n", svargc, *svargv, NULL);
		if (infile == NULL) {	/* have to open a new file */
			if (member('=', *svargv)) {	/* it's a var=value argument */
				setclvar(*svargv);
				svargv++;
				svargc--;
				continue;
			}
			*FILENAME = file = *svargv;
			dprintf("opening file %s\n", file, NULL, NULL);
			if (*file == '-') {
				if (yyin == stdin && ! lexprog)
					error(FATAL, "standard input already used for reading commands");
				else
					infile = stdin;
			}
			else if ((infile = fopen(file, "r")) == NULL)
				error(FATAL, "can't open %s", file);
		}
		if ((sep = **RS) == 0)
			sep = '\n';
		for (rr = record; ; ) {
			for (; (c=getc(infile)) != sep && c != EOF; *rr++ = c)
				;
			if (**RS == sep || c == EOF)
				break;
			if ((c = getc(infile)) == '\n' || c == EOF)	/* 2 in a row */
				break;
			*rr++ = '\n';
			*rr++ = c;
		}
		if (rr > record+RECSIZE)
			error(FATAL, "record `%.20s...' too long", record);
		*rr = 0;
		if (mustfld)
			fldbld();
		if (c != EOF || rr > record) {	/* normal record */
			recloc->tval &= ~NUM;
			recloc->tval |= STR;
			++nrloc->fval;
			nrloc->tval &= ~STR;
			nrloc->tval |= NUM;
			return(1);
		}
		/* EOF arrived on this file; set up next */
		if (infile != stdin)
			fclose(infile);
		infile = NULL;
		svargc--;
		svargv++;
	}
	return(0);	/* true end of file */
}

setclvar(s)	/* set var=value from s */
char *s;
{
	char *p;
	cell *q;

	for (p=s; *p != '='; p++)
		;
	*p++ = 0;
	q = setsymtab(s, tostring(p), 0.0, STR, symtab);
	setsval(q, p);
	dprintf("command line set %s to |%s|\n", s, p, NULL);
}

fldbld()
{
	register char *r, *fr, sep;
	int i, j;

	r = record;
	fr = fields;
	i = 0;	/* number of fields accumulated here */
	if ((sep = **FS) == ' ')
		for (i = 0; ; ) {
			while (*r == ' ' || *r == '\t' || *r == '\n')
				r++;
			if (*r == 0)
				break;
			i++;
			if (i >= MAXFLD)
				error(FATAL, "record `%.20s...' has too many fields", record);
			if (!(fldtab[i].tval&FLD))
				strfree(fldtab[i].sval);
			fldtab[i].sval = fr;
			fldtab[i].tval = FLD | STR;
			do
				*fr++ = *r++;
			while (*r != ' ' && *r != '\t' && *r != '\n' && *r != '\0');
			*fr++ = 0;
		}
	else if (*r != 0)	/* if 0, it's a null field */
		for (;;) {
			i++;
			if (i >= MAXFLD)
				error(FATAL, "record `%.20s...' has too many fields", record);
			if (!(fldtab[i].tval&FLD))
				strfree(fldtab[i].sval);
			fldtab[i].sval = fr;
			fldtab[i].tval = FLD | STR;
			while (*r != sep && *r != '\n' && *r != '\0')	/* \n always a separator */
				*fr++ = *r++;
			*fr++ = 0;
			if (*r++ == 0)
				break;
		}
	*fr = 0;
	for (j=MAXFLD-1; j>i; j--) {	/* clean out junk from previous record */
		if (!(fldtab[j].tval&FLD))
			strfree(fldtab[j].sval);
		fldtab[j].tval = STR | FLD;
		fldtab[j].sval = EMPTY;
	}
	maxfld = i;
	donefld = 1;
	for(i=1; i<=maxfld; i++)
		if(isnumber(fldtab[i].sval)) {
			fldtab[i].fval = atof(fldtab[i].sval);
			fldtab[i].tval |= NUM;
		}
	setfval(lookup("NF", symtab, 0), (awkfloat) maxfld);
	if (dbg)
		for (i = 0; i <= maxfld; i++)
			printf("field %d: |%s|\n", i, fldtab[i].sval);
}

recbld()
{
	int i;
	register char *r, *p;

	if (donefld == 0 || donerec == 1)
		return;
	r = record;
	for (i = 1; i <= *NF; i++) {
		p = getsval(&fldtab[i]);
		while (*r++ = *p++)
			;
		*(r-1) = **OFS;
	}
	*(r-1) = '\0';
	dprintf("in recbld FS=%o, recloc=%o\n", **FS, recloc, NULL);
	recloc->tval = STR | FLD;
	dprintf("in recbld FS=%o, recloc=%o\n", **FS, recloc, NULL);
	if (r > record+RECSIZE)
		error(FATAL, "built giant record `%.20s...'", record);
	dprintf("recbld = |%s|\n", record, NULL, NULL);
}

cell *fieldadr(n)
{
	if (n >= MAXFLD)
		error(FATAL, "trying to access field %d", n);
	return(&fldtab[n]);
}

int	errorflag	= 0;

yyerror(s) char *s; {
	fprintf(stderr, "awk: %s near line %d\n", s, lineno);
	errorflag = 2;
}

error(isfatal, fmt)
	int isfatal;
	char *fmt;
{
	va_list ap;

	va_start(ap, fmt);
	(void)fprintf(stderr, "awk: ");
	(void)vfprintf(stderr, fmt, ap);
	va_end(ap);
	(void)fprintf(stderr, "\n");
	if (NR && *NR > 0)
		(void)fprintf(stderr, " record number %g\n", *NR);
	if (isfatal)
		exit(2);
}

PUTS(s) char *s; {
	dprintf("%s\n", s, NULL, NULL);
}

#define	MAXEXPON	38	/* maximum exponenet for fp number */

isnumber(s)
register char *s;
{
	register d1, d2;
	int point;
	char *es;

	if (s == NULL) 
		return (0);
	d1 = d2 = point = 0;
	while (*s == ' ' || *s == '\t' || *s == '\n')
		s++;
	if (*s == '\0')
		return(0);	/* empty stuff isn't number */
	if (*s == '+' || *s == '-')
		s++;
	if (!isdigit(*s) && *s != '.')
		return(0);
	if (isdigit(*s)) {
		do {
			d1++;
			s++;
		} while (isdigit(*s));
	}
	if(d1 >= MAXEXPON)
		return(0);	/* too many digits to convert */
	if (*s == '.') {
		point++;
		s++;
	}
	if (isdigit(*s)) {
		d2++;
		do {
			s++;
		} while (isdigit(*s));
	}
	if (!(d1 || point && d2))
		return(0);
	if (*s == 'e' || *s == 'E') {
		s++;
		if (*s == '+' || *s == '-')
			s++;
		if (!isdigit(*s))
			return(0);
		es = s;
		do {
			s++;
		} while (isdigit(*s));
		if (s - es > 2)
			return(0);
		else if (s - es == 2 && 10 * (*es-'0') + *(es+1)-'0' >= MAXEXPON)
			return(0);
	}
	while (*s == ' ' || *s == '\t' || *s == '\n')
		s++;
	if (*s == '\0')
		return(1);
	else
		return(0);
}
/*
isnumber(s) char *s; {return(0);}
*/
Commit	Line	Data
b803c326	1	#ifndef lint
2404822a	2	static char sccsid[] = "@(#)lib.c 4.8 (Berkeley) %G%";
b803c326	3	#endif
b9bbf539	4
2404822a	5	#include <stdarg.h>
b9bbf539 KM	6	#include "stdio.h"
	7	#include "awk.def"
	8	#include "awk.h"
	9	#include "ctype.h"
	10
26758489	11	extern FILE yyin; / lex input file */
e068ed21	12	extern char lexprog; / points to program argument if it exists */
b9bbf539 KM	13	FILE *infile = NULL;
	14	char *file;
	15	#define RECSIZE (5 * 512)
	16	char record[RECSIZE];
	17	char fields[RECSIZE];
c48090b0	18	char EMPTY[] = "";
b9bbf539 KM	19
	20	#define MAXFLD 100
	21	int donefld; /* 1 = implies rec broken into fields */
	22	int donerec; /* 1 = record is valid (no flds have changed) */
	23	int mustfld; /* 1 = NF seen, so always break*/
	24
c48090b0	25	#define FINIT {EMPTY, EMPTY, 0.0, FLD\|STR}
b9bbf539 KM	26	cell fldtab[MAXFLD] = { /room for fields /
b9bbf539 KM	27	{ "$record", record, 0.0, STR\|FLD},
83b07d63 KB	28	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	29	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	30	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	31	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	32	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	33	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	34	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	35	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	36	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
b9bbf539 KM	37	};
	38	int maxfld = 0; /* last used field */
	39
	40
	41	getrec()
	42	{
	43	register char *rr;
	44	extern int svargc;
	45	extern char **svargv;
	46	register c, sep;
	47
	48	dprintf("RS=%o, FS=%o\n", RS, FS, NULL);
	49	donefld = 0;
	50	donerec = 1;
	51	record[0] = 0;
	52	while (svargc > 0) {
	53	dprintf("svargc=%d, svargv=%s\n", svargc, svargv, NULL);
	54	if (infile == NULL) { /* have to open a new file */
	55	if (member('=', svargv)) { / it's a var=value argument */
	56	setclvar(*svargv);
	57	svargv++;
	58	svargc--;
	59	continue;
	60	}
	61	FILENAME = file = svargv;
	62	dprintf("opening file %s\n", file, NULL, NULL);
26758489	63	if (*file == '-') {
e068ed21	64	if (yyin == stdin && ! lexprog)
26758489 KB	65	error(FATAL, "standard input already used for reading commands");
	66	else
	67	infile = stdin;
	68	}
b9bbf539 KM	69	else if ((infile = fopen(file, "r")) == NULL)
	70	error(FATAL, "can't open %s", file);
	71	}
	72	if ((sep = **RS) == 0)
	73	sep = '\n';
	74	for (rr = record; ; ) {
	75	for (; (c=getc(infile)) != sep && c != EOF; *rr++ = c)
	76	;
	77	if (**RS == sep \|\| c == EOF)
	78	break;
	79	if ((c = getc(infile)) == '\n' \|\| c == EOF) /* 2 in a row */
	80	break;
	81	*rr++ = '\n';
	82	*rr++ = c;
	83	}
	84	if (rr > record+RECSIZE)
	85	error(FATAL, "record `%.20s...' too long", record);
	86	*rr = 0;
	87	if (mustfld)
	88	fldbld();
	89	if (c != EOF \|\| rr > record) { /* normal record */
	90	recloc->tval &= ~NUM;
	91	recloc->tval \|= STR;
	92	++nrloc->fval;
	93	nrloc->tval &= ~STR;
	94	nrloc->tval \|= NUM;
	95	return(1);
	96	}
	97	/* EOF arrived on this file; set up next */
	98	if (infile != stdin)
	99	fclose(infile);
	100	infile = NULL;
	101	svargc--;
	102	svargv++;
	103	}
	104	return(0); /* true end of file */
	105	}
	106
	107	setclvar(s) /* set var=value from s */
	108	char *s;
	109	{
	110	char *p;
	111	cell *q;
	112
	113	for (p=s; *p != '='; p++)
	114	;
	115	*p++ = 0;
	116	q = setsymtab(s, tostring(p), 0.0, STR, symtab);
	117	setsval(q, p);
	118	dprintf("command line set %s to \|%s\|\n", s, p, NULL);
	119	}
	120
	121	fldbld()
	122	{
	123	register char r, fr, sep;
	124	int i, j;
	125
	126	r = record;
	127	fr = fields;
	128	i = 0; /* number of fields accumulated here */
	129	if ((sep = **FS) == ' ')
	130	for (i = 0; ; ) {
	131	while (r == ' ' \|\| r == '\t' \|\| *r == '\n')
	132	r++;
133	if (*r == 0)
134	break;
135	i++;
136	if (i >= MAXFLD)
137	error(FATAL, "record `%.20s...' has too many fields", record);
138	if (!(fldtab[i].tval&FLD))
c48090b0	139	strfree(fldtab[i].sval);
b9bbf539 KM	140	fldtab[i].sval = fr;
	141	fldtab[i].tval = FLD \| STR;
	142	do
	143	fr++ = r++;
	144	while (r != ' ' && r != '\t' && r != '\n' && r != '\0');
	145	*fr++ = 0;
	146	}
	147	else if (r != 0) / if 0, it's a null field */
	148	for (;;) {
	149	i++;
	150	if (i >= MAXFLD)
	151	error(FATAL, "record `%.20s...' has too many fields", record);
	152	if (!(fldtab[i].tval&FLD))
c48090b0	153	strfree(fldtab[i].sval);
b9bbf539 KM	154	fldtab[i].sval = fr;
	155	fldtab[i].tval = FLD \| STR;
	156	while (r != sep && r != '\n' && r != '\0') / \n always a separator */
	157	fr++ = r++;
	158	*fr++ = 0;
	159	if (*r++ == 0)
	160	break;
	161	}
	162	*fr = 0;
	163	for (j=MAXFLD-1; j>i; j--) { /* clean out junk from previous record */
	164	if (!(fldtab[j].tval&FLD))
c48090b0	165	strfree(fldtab[j].sval);
b9bbf539	166	fldtab[j].tval = STR \| FLD;
c48090b0	167	fldtab[j].sval = EMPTY;
b9bbf539 KM	168	}
	169	maxfld = i;
	170	donefld = 1;
	171	for(i=1; i<=maxfld; i++)
	172	if(isnumber(fldtab[i].sval)) {
	173	fldtab[i].fval = atof(fldtab[i].sval);
	174	fldtab[i].tval \|= NUM;
	175	}
	176	setfval(lookup("NF", symtab, 0), (awkfloat) maxfld);
	177	if (dbg)
	178	for (i = 0; i <= maxfld; i++)
	179	printf("field %d: \|%s\|\n", i, fldtab[i].sval);
	180	}
	181
	182	recbld()
	183	{
	184	int i;
	185	register char r, p;
	186
	187	if (donefld == 0 \|\| donerec == 1)
	188	return;
	189	r = record;
	190	for (i = 1; i <= *NF; i++) {
	191	p = getsval(&fldtab[i]);
	192	while (r++ = p++)
	193	;
	194	(r-1) = *OFS;
	195	}
	196	*(r-1) = '\0';
	197	dprintf("in recbld FS=%o, recloc=%o\n", **FS, recloc, NULL);
	198	recloc->tval = STR \| FLD;
	199	dprintf("in recbld FS=%o, recloc=%o\n", **FS, recloc, NULL);
	200	if (r > record+RECSIZE)
	201	error(FATAL, "built giant record `%.20s...'", record);
	202	dprintf("recbld = \|%s\|\n", record, NULL, NULL);
	203	}
	204
	205	cell *fieldadr(n)
	206	{
	207	if (n >= MAXFLD)
	208	error(FATAL, "trying to access field %d", n);
	209	return(&fldtab[n]);
	210	}
	211
	212	int errorflag = 0;
	213
	214	yyerror(s) char *s; {
	215	fprintf(stderr, "awk: %s near line %d\n", s, lineno);
	216	errorflag = 2;
	217	}
	218
2404822a KB	219	error(isfatal, fmt)
	220	int isfatal;
	221	char *fmt;
	222	{
	223	va_list ap;
	224
	225	va_start(ap, fmt);
	226	(void)fprintf(stderr, "awk: ");
	227	(void)vfprintf(stderr, fmt, ap);
	228	va_end(ap);
	229	(void)fprintf(stderr, "\n");
8820cba9	230	if (NR && *NR > 0)
2404822a KB	231	(void)fprintf(stderr, " record number %g\n", *NR);
2404822a KB	232	if (isfatal)
b9bbf539 KM	233	exit(2);
	234	}
	235
	236	PUTS(s) char *s; {
	237	dprintf("%s\n", s, NULL, NULL);
	238	}
	239
	240	#define MAXEXPON 38 /* maximum exponenet for fp number */
	241
	242	isnumber(s)
	243	register char *s;
	244	{
	245	register d1, d2;
	246	int point;
	247	char *es;
	248
c48090b0 SL	249	if (s == NULL)
c48090b0 SL	250	return (0);
b9bbf539 KM	251	d1 = d2 = point = 0;
	252	while (s == ' ' \|\| s == '\t' \|\| *s == '\n')
	253	s++;
	254	if (*s == '\0')
	255	return(0); /* empty stuff isn't number */
	256	if (s == '+' \|\| s == '-')
	257	s++;
	258	if (!isdigit(s) && s != '.')
	259	return(0);
	260	if (isdigit(*s)) {
	261	do {
	262	d1++;
	263	s++;
	264	} while (isdigit(*s));
	265	}
	266	if(d1 >= MAXEXPON)
	267	return(0); /* too many digits to convert */
	268	if (*s == '.') {
	269	point++;
	270	s++;
	271	}
	272	if (isdigit(*s)) {
	273	d2++;
	274	do {
	275	s++;
	276	} while (isdigit(*s));
	277	}
	278	if (!(d1 \|\| point && d2))
	279	return(0);
	280	if (s == 'e' \|\| s == 'E') {
	281	s++;
	282	if (s == '+' \|\| s == '-')
	283	s++;
	284	if (!isdigit(*s))
	285	return(0);
	286	es = s;
	287	do {
	288	s++;
	289	} while (isdigit(*s));
	290	if (s - es > 2)
	291	return(0);
	292	else if (s - es == 2 && 10 * (es-'0') + (es+1)-'0' >= MAXEXPON)
	293	return(0);
	294	}
	295	while (s == ' ' \|\| s == '\t' \|\| *s == '\n')
	296	s++;
	297	if (*s == '\0')
	298	return(1);
	299	else
	300	return(0);
	301	}
	302	/*
	303	isnumber(s) char *s; {return(0);}
	304	*/