xref: /freebsd/contrib/one-true-awk/main.c (revision d7d962ead0b6e5e8a39202d0590022082bf5bfb6)
1 /****************************************************************
2 Copyright (C) Lucent Technologies 1997
3 All Rights Reserved
4 
5 Permission to use, copy, modify, and distribute this software and
6 its documentation for any purpose and without fee is hereby
7 granted, provided that the above copyright notice appear in all
8 copies and that both that the copyright notice and this
9 permission notice and warranty disclaimer appear in supporting
10 documentation, and that the name Lucent Technologies or any of
11 its entities not be used in advertising or publicity pertaining
12 to distribution of the software without specific, written prior
13 permission.
14 
15 LUCENT DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
16 INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS.
17 IN NO EVENT SHALL LUCENT OR ANY OF ITS ENTITIES BE LIABLE FOR ANY
18 SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
19 WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER
20 IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION,
21 ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF
22 THIS SOFTWARE.
23 ****************************************************************/
24 
25 const char	*version = "version 20210724";
26 
27 #define DEBUG
28 #include <stdio.h>
29 #include <ctype.h>
30 #include <locale.h>
31 #include <stdlib.h>
32 #include <string.h>
33 #include <signal.h>
34 #include "awk.h"
35 
36 extern	char	**environ;
37 extern	int	nfields;
38 
39 int	dbg	= 0;
40 Awkfloat	srand_seed = 1;
41 char	*cmdname;	/* gets argv[0] for error messages */
42 extern	FILE	*yyin;	/* lex input file */
43 char	*lexprog;	/* points to program argument if it exists */
44 extern	int errorflag;	/* non-zero if any syntax errors; set by yyerror */
45 enum compile_states	compile_time = ERROR_PRINTING;
46 
47 static char	**pfile;	/* program filenames from -f's */
48 static size_t	maxpfile;	/* max program filename */
49 static size_t	npfile;		/* number of filenames */
50 static size_t	curpfile;	/* current filename */
51 
52 bool	safe = false;	/* true => "safe" mode */
53 
54 static noreturn void fpecatch(int n
55 #ifdef SA_SIGINFO
56 	, siginfo_t *si, void *uc
57 #endif
58 )
59 {
60 #ifdef SA_SIGINFO
61 	static const char *emsg[] = {
62 		[0] = "Unknown error",
63 		[FPE_INTDIV] = "Integer divide by zero",
64 		[FPE_INTOVF] = "Integer overflow",
65 		[FPE_FLTDIV] = "Floating point divide by zero",
66 		[FPE_FLTOVF] = "Floating point overflow",
67 		[FPE_FLTUND] = "Floating point underflow",
68 		[FPE_FLTRES] = "Floating point inexact result",
69 		[FPE_FLTINV] = "Invalid Floating point operation",
70 		[FPE_FLTSUB] = "Subscript out of range",
71 	};
72 #endif
73 	FATAL("floating point exception"
74 #ifdef SA_SIGINFO
75 		": %s", (size_t)si->si_code < sizeof(emsg) / sizeof(emsg[0]) &&
76 		emsg[si->si_code] ? emsg[si->si_code] : emsg[0]
77 #endif
78 	    );
79 }
80 
81 /* Can this work with recursive calls?  I don't think so.
82 void segvcatch(int n)
83 {
84 	FATAL("segfault.  Do you have an unbounded recursive call?", n);
85 }
86 */
87 
88 static const char *
89 setfs(char *p)
90 {
91 	/* wart: t=>\t */
92 	if (p[0] == 't' && p[1] == '\0') {
93 		WARNING("-Ft to imply tab separator is deprecated behavior.");
94 		return "\t";
95 	}
96 	return p;
97 }
98 
99 static char *
100 getarg(int *argc, char ***argv, const char *msg)
101 {
102 	if ((*argv)[1][2] != '\0') {	/* arg is -fsomething */
103 		return &(*argv)[1][2];
104 	} else {			/* arg is -f something */
105 		(*argc)--; (*argv)++;
106 		if (*argc <= 1)
107 			FATAL("%s", msg);
108 		return (*argv)[1];
109 	}
110 }
111 
112 int main(int argc, char *argv[])
113 {
114 	const char *fs = NULL;
115 	char *fn, *vn;
116 
117 	setlocale(LC_CTYPE, "");
118 	setlocale(LC_NUMERIC, "C"); /* for parsing cmdline & prog */
119 	cmdname = argv[0];
120 	if (argc == 1) {
121 		fprintf(stderr,
122 		  "usage: %s [-F fs] [-v var=value] [-f progfile | 'prog'] [file ...]\n",
123 		  cmdname);
124 		exit(1);
125 	}
126 #ifdef SA_SIGINFO
127 	{
128 		struct sigaction sa;
129 		sa.sa_sigaction = fpecatch;
130 		sa.sa_flags = SA_SIGINFO;
131 		sigemptyset(&sa.sa_mask);
132 		(void)sigaction(SIGFPE, &sa, NULL);
133 	}
134 #else
135 	(void)signal(SIGFPE, fpecatch);
136 #endif
137 	/*signal(SIGSEGV, segvcatch); experiment */
138 
139 	/* Set and keep track of the random seed */
140 	srand_seed = 1;
141 	srandom((unsigned long) srand_seed);
142 
143 	yyin = NULL;
144 	symtab = makesymtab(NSYMTAB/NSYMTAB);
145 	while (argc > 1 && argv[1][0] == '-' && argv[1][1] != '\0') {
146 		if (strcmp(argv[1], "-version") == 0 || strcmp(argv[1], "--version") == 0) {
147 			printf("awk %s\n", version);
148 			return 0;
149 		}
150 		if (strcmp(argv[1], "--") == 0) {	/* explicit end of args */
151 			argc--;
152 			argv++;
153 			break;
154 		}
155 		switch (argv[1][1]) {
156 		case 's':
157 			if (strcmp(argv[1], "-safe") == 0)
158 				safe = true;
159 			break;
160 		case 'f':	/* next argument is program filename */
161 			fn = getarg(&argc, &argv, "no program filename");
162 			if (npfile >= maxpfile) {
163 				maxpfile += 20;
164 				pfile = (char **) realloc(pfile, maxpfile * sizeof(*pfile));
165 				if (pfile == NULL)
166 					FATAL("error allocating space for -f options");
167  			}
168 			pfile[npfile++] = fn;
169  			break;
170 		case 'F':	/* set field separator */
171 			fs = setfs(getarg(&argc, &argv, "no field separator"));
172 			break;
173 		case 'v':	/* -v a=1 to be done NOW.  one -v for each */
174 			vn = getarg(&argc, &argv, "no variable name");
175 			if (isclvar(vn))
176 				setclvar(vn);
177 			else
178 				FATAL("invalid -v option argument: %s", vn);
179 			break;
180 		case 'd':
181 			dbg = atoi(&argv[1][2]);
182 			if (dbg == 0)
183 				dbg = 1;
184 			printf("awk %s\n", version);
185 			break;
186 		default:
187 			WARNING("unknown option %s ignored", argv[1]);
188 			break;
189 		}
190 		argc--;
191 		argv++;
192 	}
193 	/* argv[1] is now the first argument */
194 	if (npfile == 0) {	/* no -f; first argument is program */
195 		if (argc <= 1) {
196 			if (dbg)
197 				exit(0);
198 			FATAL("no program given");
199 		}
200 		DPRINTF("program = |%s|\n", argv[1]);
201 		lexprog = argv[1];
202 		argc--;
203 		argv++;
204 	}
205 	recinit(recsize);
206 	syminit();
207 	compile_time = COMPILING;
208 	argv[0] = cmdname;	/* put prog name at front of arglist */
209 	DPRINTF("argc=%d, argv[0]=%s\n", argc, argv[0]);
210 	arginit(argc, argv);
211 	if (!safe)
212 		envinit(environ);
213 	yyparse();
214 #if 0
215 	// Doing this would comply with POSIX, but is not compatible with
216 	// other awks and with what most users expect. So comment it out.
217 	setlocale(LC_NUMERIC, ""); /* back to whatever it is locally */
218 #endif
219 	if (fs)
220 		*FS = qstring(fs, '\0');
221 	DPRINTF("errorflag=%d\n", errorflag);
222 	if (errorflag == 0) {
223 		compile_time = RUNNING;
224 		run(winner);
225 	} else
226 		bracecheck();
227 	return(errorflag);
228 }
229 
230 int pgetc(void)		/* get 1 character from awk program */
231 {
232 	int c;
233 
234 	for (;;) {
235 		if (yyin == NULL) {
236 			if (curpfile >= npfile)
237 				return EOF;
238 			if (strcmp(pfile[curpfile], "-") == 0)
239 				yyin = stdin;
240 			else if ((yyin = fopen(pfile[curpfile], "r")) == NULL)
241 				FATAL("can't open file %s", pfile[curpfile]);
242 			lineno = 1;
243 		}
244 		if ((c = getc(yyin)) != EOF)
245 			return c;
246 		if (yyin != stdin)
247 			fclose(yyin);
248 		yyin = NULL;
249 		curpfile++;
250 	}
251 }
252 
253 char *cursource(void)	/* current source file name */
254 {
255 	if (npfile > 0)
256 		return pfile[curpfile < npfile ? curpfile : curpfile - 1];
257 	else
258 		return NULL;
259 }
260