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

#ifndef lint
static char sccsid[] = "@(#)lib.c	4.3 %G%";
#endif

#include "stdio.h"
#include "awk.def"
#include "awk.h"
#include "ctype.h"

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
};
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 == '-')
				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(f, s, a1, a2, a3, a4, a5, a6, a7) {
	fprintf(stderr, "awk: ");
	fprintf(stderr, s, a1, a2, a3, a4, a5, a6, a7);
	fprintf(stderr, "\n");
	if (*NR > 0)
		fprintf(stderr, " record number %g\n", *NR);
	if (f)
		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 SL	1	#ifndef lint
	2	static char sccsid[] = "@(#)lib.c 4.3 %G%";
	3	#endif
b9bbf539 KM	4
	5	#include "stdio.h"
	6	#include "awk.def"
	7	#include "awk.h"
	8	#include "ctype.h"
	9
	10	FILE *infile = NULL;
	11	char *file;
	12	#define RECSIZE (5 * 512)
	13	char record[RECSIZE];
	14	char fields[RECSIZE];
c48090b0	15	char EMPTY[] = "";
b9bbf539 KM	16
	17	#define MAXFLD 100
	18	int donefld; /* 1 = implies rec broken into fields */
	19	int donerec; /* 1 = record is valid (no flds have changed) */
	20	int mustfld; /* 1 = NF seen, so always break*/
	21
c48090b0	22	#define FINIT {EMPTY, EMPTY, 0.0, FLD\|STR}
b9bbf539 KM	23	cell fldtab[MAXFLD] = { /room for fields /
	24	{ "$record", record, 0.0, STR\|FLD},
	25	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	26	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	27	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	28	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	29	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	30	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	31	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT
	32	};
	33	int maxfld = 0; /* last used field */
	34
	35
	36	getrec()
	37	{
	38	register char *rr;
	39	extern int svargc;
	40	extern char **svargv;
	41	register c, sep;
	42
	43	dprintf("RS=%o, FS=%o\n", RS, FS, NULL);
	44	donefld = 0;
	45	donerec = 1;
	46	record[0] = 0;
	47	while (svargc > 0) {
	48	dprintf("svargc=%d, svargv=%s\n", svargc, svargv, NULL);
	49	if (infile == NULL) { /* have to open a new file */
	50	if (member('=', svargv)) { / it's a var=value argument */
	51	setclvar(*svargv);
	52	svargv++;
	53	svargc--;
	54	continue;
	55	}
	56	FILENAME = file = svargv;
	57	dprintf("opening file %s\n", file, NULL, NULL);
	58	if (*file == '-')
	59	infile = stdin;
	60	else if ((infile = fopen(file, "r")) == NULL)
	61	error(FATAL, "can't open %s", file);
	62	}
	63	if ((sep = **RS) == 0)
	64	sep = '\n';
	65	for (rr = record; ; ) {
	66	for (; (c=getc(infile)) != sep && c != EOF; *rr++ = c)
	67	;
	68	if (**RS == sep \|\| c == EOF)
	69	break;
	70	if ((c = getc(infile)) == '\n' \|\| c == EOF) /* 2 in a row */
	71	break;
	72	*rr++ = '\n';
	73	*rr++ = c;
	74	}
	75	if (rr > record+RECSIZE)
	76	error(FATAL, "record `%.20s...' too long", record);
	77	*rr = 0;
	78	if (mustfld)
	79	fldbld();
	80	if (c != EOF \|\| rr > record) { /* normal record */
	81	recloc->tval &= ~NUM;
	82	recloc->tval \|= STR;
	83	++nrloc->fval;
	84	nrloc->tval &= ~STR;
	85	nrloc->tval \|= NUM;
	86	return(1);
87	}
88	/* EOF arrived on this file; set up next */
89	if (infile != stdin)
90	fclose(infile);
91	infile = NULL;
92	svargc--;
93	svargv++;
94	}
95	return(0); /* true end of file */
96	}
97
98	setclvar(s) /* set var=value from s */
99	char *s;
100	{
101	char *p;
102	cell *q;
103
104	for (p=s; *p != '='; p++)
105	;
106	*p++ = 0;
107	q = setsymtab(s, tostring(p), 0.0, STR, symtab);
108	setsval(q, p);
109	dprintf("command line set %s to \|%s\|\n", s, p, NULL);
110	}
111
112	fldbld()
113	{
114	register char r, fr, sep;
115	int i, j;
116
117	r = record;
118	fr = fields;
119	i = 0; /* number of fields accumulated here */
120	if ((sep = **FS) == ' ')
121	for (i = 0; ; ) {
122	while (r == ' ' \|\| r == '\t' \|\| *r == '\n')
123	r++;
124	if (*r == 0)
125	break;
126	i++;
127	if (i >= MAXFLD)
128	error(FATAL, "record `%.20s...' has too many fields", record);
129	if (!(fldtab[i].tval&FLD))
c48090b0	130	strfree(fldtab[i].sval);
b9bbf539 KM	131	fldtab[i].sval = fr;
	132	fldtab[i].tval = FLD \| STR;
	133	do
	134	fr++ = r++;
	135	while (r != ' ' && r != '\t' && r != '\n' && r != '\0');
	136	*fr++ = 0;
	137	}
	138	else if (r != 0) / if 0, it's a null field */
	139	for (;;) {
	140	i++;
	141	if (i >= MAXFLD)
	142	error(FATAL, "record `%.20s...' has too many fields", record);
	143	if (!(fldtab[i].tval&FLD))
c48090b0	144	strfree(fldtab[i].sval);
b9bbf539 KM	145	fldtab[i].sval = fr;
	146	fldtab[i].tval = FLD \| STR;
	147	while (r != sep && r != '\n' && r != '\0') / \n always a separator */
	148	fr++ = r++;
	149	*fr++ = 0;
	150	if (*r++ == 0)
	151	break;
	152	}
	153	*fr = 0;
	154	for (j=MAXFLD-1; j>i; j--) { /* clean out junk from previous record */
	155	if (!(fldtab[j].tval&FLD))
c48090b0	156	strfree(fldtab[j].sval);
b9bbf539	157	fldtab[j].tval = STR \| FLD;
c48090b0	158	fldtab[j].sval = EMPTY;
b9bbf539 KM	159	}
	160	maxfld = i;
	161	donefld = 1;
	162	for(i=1; i<=maxfld; i++)
	163	if(isnumber(fldtab[i].sval)) {
	164	fldtab[i].fval = atof(fldtab[i].sval);
	165	fldtab[i].tval \|= NUM;
	166	}
	167	setfval(lookup("NF", symtab, 0), (awkfloat) maxfld);
	168	if (dbg)
	169	for (i = 0; i <= maxfld; i++)
	170	printf("field %d: \|%s\|\n", i, fldtab[i].sval);
	171	}
	172
	173	recbld()
	174	{
	175	int i;
	176	register char r, p;
	177
	178	if (donefld == 0 \|\| donerec == 1)
	179	return;
	180	r = record;
	181	for (i = 1; i <= *NF; i++) {
	182	p = getsval(&fldtab[i]);
	183	while (r++ = p++)
	184	;
	185	(r-1) = *OFS;
	186	}
	187	*(r-1) = '\0';
	188	dprintf("in recbld FS=%o, recloc=%o\n", **FS, recloc, NULL);
	189	recloc->tval = STR \| FLD;
	190	dprintf("in recbld FS=%o, recloc=%o\n", **FS, recloc, NULL);
	191	if (r > record+RECSIZE)
	192	error(FATAL, "built giant record `%.20s...'", record);
	193	dprintf("recbld = \|%s\|\n", record, NULL, NULL);
	194	}
	195
	196	cell *fieldadr(n)
	197	{
	198	if (n >= MAXFLD)
	199	error(FATAL, "trying to access field %d", n);
	200	return(&fldtab[n]);
	201	}
	202
	203	int errorflag = 0;
	204
	205	yyerror(s) char *s; {
	206	fprintf(stderr, "awk: %s near line %d\n", s, lineno);
	207	errorflag = 2;
	208	}
	209
	210	error(f, s, a1, a2, a3, a4, a5, a6, a7) {
	211	fprintf(stderr, "awk: ");
	212	fprintf(stderr, s, a1, a2, a3, a4, a5, a6, a7);
	213	fprintf(stderr, "\n");
	214	if (*NR > 0)
	215	fprintf(stderr, " record number %g\n", *NR);
	216	if (f)
	217	exit(2);
	218	}
	219
	220	PUTS(s) char *s; {
	221	dprintf("%s\n", s, NULL, NULL);
	222	}
223
224	#define MAXEXPON 38 /* maximum exponenet for fp number */
225
226	isnumber(s)
227	register char *s;
228	{
229	register d1, d2;
230	int point;
231	char *es;
232
c48090b0 SL	233	if (s == NULL)
c48090b0 SL	234	return (0);
b9bbf539 KM	235	d1 = d2 = point = 0;
	236	while (s == ' ' \|\| s == '\t' \|\| *s == '\n')
	237	s++;
	238	if (*s == '\0')
	239	return(0); /* empty stuff isn't number */
	240	if (s == '+' \|\| s == '-')
	241	s++;
	242	if (!isdigit(s) && s != '.')
	243	return(0);
	244	if (isdigit(*s)) {
	245	do {
	246	d1++;
	247	s++;
	248	} while (isdigit(*s));
	249	}
	250	if(d1 >= MAXEXPON)
	251	return(0); /* too many digits to convert */
	252	if (*s == '.') {
	253	point++;
	254	s++;
	255	}
	256	if (isdigit(*s)) {
	257	d2++;
	258	do {
	259	s++;
	260	} while (isdigit(*s));
	261	}
	262	if (!(d1 \|\| point && d2))
	263	return(0);
	264	if (s == 'e' \|\| s == 'E') {
	265	s++;
	266	if (s == '+' \|\| s == '-')
	267	s++;
	268	if (!isdigit(*s))
	269	return(0);
	270	es = s;
	271	do {
	272	s++;
	273	} while (isdigit(*s));
	274	if (s - es > 2)
	275	return(0);
	276	else if (s - es == 2 && 10 * (es-'0') + (es+1)-'0' >= MAXEXPON)
	277	return(0);
	278	}
	279	while (s == ' ' \|\| s == '\t' \|\| *s == '\n')
	280	s++;
	281	if (*s == '\0')
	282	return(1);
	283	else
	284	return(0);
	285	}
	286	/*
	287	isnumber(s) char *s; {return(0);}
	288	*/