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

/*	lib.c	4.2	83/02/09	*/

#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
c48090b0	1	/* lib.c 4.2 83/02/09 */
b9bbf539 KM	2
	3	#include "stdio.h"
	4	#include "awk.def"
	5	#include "awk.h"
	6	#include "ctype.h"
	7
	8	FILE *infile = NULL;
	9	char *file;
	10	#define RECSIZE (5 * 512)
	11	char record[RECSIZE];
	12	char fields[RECSIZE];
c48090b0	13	char EMPTY[] = "";
b9bbf539 KM	14
	15	#define MAXFLD 100
	16	int donefld; /* 1 = implies rec broken into fields */
	17	int donerec; /* 1 = record is valid (no flds have changed) */
	18	int mustfld; /* 1 = NF seen, so always break*/
	19
c48090b0	20	#define FINIT {EMPTY, EMPTY, 0.0, FLD\|STR}
b9bbf539 KM	21	cell fldtab[MAXFLD] = { /room for fields /
	22	{ "$record", record, 0.0, STR\|FLD},
	23	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	24	FINIT, FINIT, FINIT, FINIT, FINIT, FINIT, FINIT,
	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	};
	31	int maxfld = 0; /* last used field */
	32
	33
	34	getrec()
	35	{
	36	register char *rr;
	37	extern int svargc;
	38	extern char **svargv;
	39	register c, sep;
	40
	41	dprintf("RS=%o, FS=%o\n", RS, FS, NULL);
	42	donefld = 0;
	43	donerec = 1;
	44	record[0] = 0;
	45	while (svargc > 0) {
	46	dprintf("svargc=%d, svargv=%s\n", svargc, svargv, NULL);
	47	if (infile == NULL) { /* have to open a new file */
	48	if (member('=', svargv)) { / it's a var=value argument */
	49	setclvar(*svargv);
	50	svargv++;
	51	svargc--;
	52	continue;
	53	}
	54	FILENAME = file = svargv;
	55	dprintf("opening file %s\n", file, NULL, NULL);
	56	if (*file == '-')
	57	infile = stdin;
	58	else if ((infile = fopen(file, "r")) == NULL)
	59	error(FATAL, "can't open %s", file);
	60	}
	61	if ((sep = **RS) == 0)
	62	sep = '\n';
	63	for (rr = record; ; ) {
	64	for (; (c=getc(infile)) != sep && c != EOF; *rr++ = c)
	65	;
	66	if (**RS == sep \|\| c == EOF)
	67	break;
	68	if ((c = getc(infile)) == '\n' \|\| c == EOF) /* 2 in a row */
	69	break;
	70	*rr++ = '\n';
	71	*rr++ = c;
	72	}
	73	if (rr > record+RECSIZE)
	74	error(FATAL, "record `%.20s...' too long", record);
	75	*rr = 0;
	76	if (mustfld)
	77	fldbld();
	78	if (c != EOF \|\| rr > record) { /* normal record */
	79	recloc->tval &= ~NUM;
	80	recloc->tval \|= STR;
	81	++nrloc->fval;
	82	nrloc->tval &= ~STR;
	83	nrloc->tval \|= NUM;
	84	return(1);
85	}
86	/* EOF arrived on this file; set up next */
87	if (infile != stdin)
88	fclose(infile);
89	infile = NULL;
90	svargc--;
91	svargv++;
92	}
93	return(0); /* true end of file */
94	}
95
96	setclvar(s) /* set var=value from s */
97	char *s;
98	{
99	char *p;
100	cell *q;
101
102	for (p=s; *p != '='; p++)
103	;
104	*p++ = 0;
105	q = setsymtab(s, tostring(p), 0.0, STR, symtab);
106	setsval(q, p);
107	dprintf("command line set %s to \|%s\|\n", s, p, NULL);
108	}
109
110	fldbld()
111	{
112	register char r, fr, sep;
113	int i, j;
114
115	r = record;
116	fr = fields;
117	i = 0; /* number of fields accumulated here */
118	if ((sep = **FS) == ' ')
119	for (i = 0; ; ) {
120	while (r == ' ' \|\| r == '\t' \|\| *r == '\n')
121	r++;
122	if (*r == 0)
123	break;
124	i++;
125	if (i >= MAXFLD)
126	error(FATAL, "record `%.20s...' has too many fields", record);
127	if (!(fldtab[i].tval&FLD))
c48090b0	128	strfree(fldtab[i].sval);
b9bbf539 KM	129	fldtab[i].sval = fr;
	130	fldtab[i].tval = FLD \| STR;
	131	do
	132	fr++ = r++;
	133	while (r != ' ' && r != '\t' && r != '\n' && r != '\0');
	134	*fr++ = 0;
	135	}
	136	else if (r != 0) / if 0, it's a null field */
	137	for (;;) {
	138	i++;
	139	if (i >= MAXFLD)
	140	error(FATAL, "record `%.20s...' has too many fields", record);
	141	if (!(fldtab[i].tval&FLD))
c48090b0	142	strfree(fldtab[i].sval);
b9bbf539 KM	143	fldtab[i].sval = fr;
	144	fldtab[i].tval = FLD \| STR;
	145	while (r != sep && r != '\n' && r != '\0') / \n always a separator */
	146	fr++ = r++;
	147	*fr++ = 0;
	148	if (*r++ == 0)
	149	break;
	150	}
	151	*fr = 0;
	152	for (j=MAXFLD-1; j>i; j--) { /* clean out junk from previous record */
	153	if (!(fldtab[j].tval&FLD))
c48090b0	154	strfree(fldtab[j].sval);
b9bbf539	155	fldtab[j].tval = STR \| FLD;
c48090b0	156	fldtab[j].sval = EMPTY;
b9bbf539 KM	157	}
	158	maxfld = i;
	159	donefld = 1;
	160	for(i=1; i<=maxfld; i++)
	161	if(isnumber(fldtab[i].sval)) {
	162	fldtab[i].fval = atof(fldtab[i].sval);
	163	fldtab[i].tval \|= NUM;
	164	}
	165	setfval(lookup("NF", symtab, 0), (awkfloat) maxfld);
	166	if (dbg)
	167	for (i = 0; i <= maxfld; i++)
	168	printf("field %d: \|%s\|\n", i, fldtab[i].sval);
	169	}
	170
	171	recbld()
	172	{
	173	int i;
	174	register char r, p;
	175
	176	if (donefld == 0 \|\| donerec == 1)
	177	return;
	178	r = record;
	179	for (i = 1; i <= *NF; i++) {
	180	p = getsval(&fldtab[i]);
	181	while (r++ = p++)
	182	;
	183	(r-1) = *OFS;
	184	}
	185	*(r-1) = '\0';
	186	dprintf("in recbld FS=%o, recloc=%o\n", **FS, recloc, NULL);
	187	recloc->tval = STR \| FLD;
	188	dprintf("in recbld FS=%o, recloc=%o\n", **FS, recloc, NULL);
	189	if (r > record+RECSIZE)
	190	error(FATAL, "built giant record `%.20s...'", record);
	191	dprintf("recbld = \|%s\|\n", record, NULL, NULL);
	192	}
	193
	194	cell *fieldadr(n)
	195	{
	196	if (n >= MAXFLD)
	197	error(FATAL, "trying to access field %d", n);
	198	return(&fldtab[n]);
	199	}
	200
	201	int errorflag = 0;
	202
	203	yyerror(s) char *s; {
	204	fprintf(stderr, "awk: %s near line %d\n", s, lineno);
	205	errorflag = 2;
	206	}
	207
	208	error(f, s, a1, a2, a3, a4, a5, a6, a7) {
	209	fprintf(stderr, "awk: ");
	210	fprintf(stderr, s, a1, a2, a3, a4, a5, a6, a7);
	211	fprintf(stderr, "\n");
	212	if (*NR > 0)
	213	fprintf(stderr, " record number %g\n", *NR);
	214	if (f)
	215	exit(2);
	216	}
	217
	218	PUTS(s) char *s; {
	219	dprintf("%s\n", s, NULL, NULL);
	220	}
221
222	#define MAXEXPON 38 /* maximum exponenet for fp number */
223
224	isnumber(s)
225	register char *s;
226	{
227	register d1, d2;
228	int point;
229	char *es;
230
c48090b0 SL	231	if (s == NULL)
c48090b0 SL	232	return (0);
b9bbf539 KM	233	d1 = d2 = point = 0;
	234	while (s == ' ' \|\| s == '\t' \|\| *s == '\n')
	235	s++;
	236	if (*s == '\0')
	237	return(0); /* empty stuff isn't number */
	238	if (s == '+' \|\| s == '-')
	239	s++;
	240	if (!isdigit(s) && s != '.')
	241	return(0);
	242	if (isdigit(*s)) {
	243	do {
	244	d1++;
	245	s++;
	246	} while (isdigit(*s));
	247	}
	248	if(d1 >= MAXEXPON)
	249	return(0); /* too many digits to convert */
	250	if (*s == '.') {
	251	point++;
	252	s++;
	253	}
	254	if (isdigit(*s)) {
	255	d2++;
	256	do {
	257	s++;
	258	} while (isdigit(*s));
	259	}
	260	if (!(d1 \|\| point && d2))
	261	return(0);
	262	if (s == 'e' \|\| s == 'E') {
	263	s++;
	264	if (s == '+' \|\| s == '-')
	265	s++;
	266	if (!isdigit(*s))
	267	return(0);
	268	es = s;
	269	do {
	270	s++;
	271	} while (isdigit(*s));
	272	if (s - es > 2)
	273	return(0);
	274	else if (s - es == 2 && 10 * (es-'0') + (es+1)-'0' >= MAXEXPON)
	275	return(0);
	276	}
	277	while (s == ' ' \|\| s == '\t' \|\| *s == '\n')
	278	s++;
	279	if (*s == '\0')
	280	return(1);
	281	else
	282	return(0);
	283	}
	284	/*
	285	isnumber(s) char *s; {return(0);}
	286	*/