xref: /illumos-gate/usr/src/cmd/awk_xpg4/awk.h (revision b2be350e)
1 /*
2  * CDDL HEADER START
3  *
4  * The contents of this file are subject to the terms of the
5  * Common Development and Distribution License (the "License").
6  * You may not use this file except in compliance with the License.
7  *
8  * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
9  * or http://www.opensolaris.org/os/licensing.
10  * See the License for the specific language governing permissions
11  * and limitations under the License.
12  *
13  * When distributing Covered Code, include this CDDL HEADER in each
14  * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
15  * If applicable, add the following below this CDDL HEADER, with the
16  * fields enclosed by brackets "[]" replaced with your own identifying
17  * information: Portions Copyright [yyyy] [name of copyright owner]
18  *
19  * CDDL HEADER END
20  */
21 /*
22  * Copyright 2007 Sun Microsystems, Inc.  All rights reserved.
23  * Use is subject to license terms.
24  */
25 
26 /*
27  * awk -- common header file.
28  *
29  * Copyright 1986, 1994 by Mortice Kern Systems Inc.  All rights reserved.
30  *
31  * This version uses the POSIX.2 compatible <regex.h> routines.
32  *
33  * Based on MKS awk(1) ported to be /usr/xpg4/bin/awk with POSIX/XCU4 changes
34  *
35  */
36 
37 #include <stdio.h>
38 #include <ctype.h>
39 #include <string.h>
40 #include <math.h>
41 #include <limits.h>
42 #include <stdlib.h>
43 #include <regex.h>
44 #include <errno.h>
45 #include <sys/types.h>
46 #include <locale.h>
47 #include <wchar.h>
48 #include <widec.h>
49 
50 #define	YYMAXDEPTH	300	/* Max # of productions (used by yacc) */
51 #define	YYSSIZE		300	/* Size of State/Value stacks (MKS YACC) */
52 #define	MAXDIGINT	19	/* Number of digits in an INT */
53 #define	FNULL		((FILE *)0)
54 #define	NNULL		((NODE *)0)
55 #define	SNULL		((STRING)0)
56 #define	LARGE		INT_MAX	/* Large integer */
57 #define	NPFILE		32	/* Number of -[fl] options allowed */
58 #define	NRECUR		3000	/* Maximum recursion depth */
59 
60 #define	M_LDATA	1
61 #ifdef M_LDATA
62 #define	NLINE	20000	/* Longest input record */
63 #define	NFIELD	4000	/* Number of fields allowed */
64 #define	NBUCKET	1024	/* # of symtab buckets (power of 2) */
65 #else
66 #define	NLINE	2048	/* Longest input record */
67 #define	NFIELD	1024	/* Number of fields allowed */
68 #define	NBUCKET	256	/* # of symtab buckets (power of 2) */
69 #endif
70 
71 #define	NSNODE		40	/* Number of cached nodes */
72 #define	NCONTEXT	50	/* Amount of context for error msgs */
73 #define	hashbuck(n)	((n)&(NBUCKET-1))
74 #if	BSD
75 /*
76  * A speedup for BSD.  Use their routines which are
77  * already optimised.  Note that BSD bcopy does not
78  * return a value.
79  */
80 int	bcmp();
81 #define	memcmp(b1, b2, n)	bcmp(b1, b2, n)
82 void	bcopy();
83 #define	memcpy(b1, b2, n)	bcopy(b2, b1, (int)n)
84 #endif	/* BSD */
85 #define	vlook(n)	vlookup(n, 0)
86 
87 /*
88  * Basic AWK internal types.
89  */
90 typedef	double		REAL;
91 typedef	long long	INT;
92 typedef	wchar_t		*STRING;
93 typedef	struct NODE	*(*FUNCTION)(struct NODE *np);
94 typedef	void		*REGEXP;
95 
96 /*
97  * Node in the AWK interpreter expression tree.
98  */
99 typedef	struct	NODE	{
100 	ushort_t	n_type;
101 	struct NODE	*n_next;		/* Symbol table/PARM link */
102 	ushort_t	n_flags;		/* Node flags, type */
103 
104 
105 
106 
107 	union	{
108 		struct	{
109 			ushort_t	N_hash;		/* Full hash value */
110 			struct NODE	*N_alink;	/* Array link */
111 			union	{
112 				struct	{
113 					STRING	N_string;
114 					size_t	N_strlen;
115 				}	n_str;
116 				INT	N_int;
117 				REAL	N_real;
118 				FUNCTION	N_function;
119 				struct	NODE	*N_ufunc;
120 			}	n_tun;
121 			wchar_t	N_name[1];
122 		}	n_term;
123 		struct	{
124 			struct	NODE	*N_left;
125 			struct	NODE	*N_right;
126 			ushort_t	N_lineno;
127 		}	n_op;
128 		struct {
129 			struct	NODE	*N_left;	/* Used for fliplist */
130 			struct	NODE	*N_right;
131 			REGEXP		N_regexp;	/* Regular expression */
132 		}	n_re;
133 	}	n_un;
134 }	NODE;
135 
136 /*
137  * Definitions to make the node access much easier.
138  */
139 #define	n_hash		n_un.n_term.N_hash	/* full hash value is sym tbl */
140 #define	n_scope		n_un.n_term.N_hash	/* local variable scope level */
141 #define	n_alink		n_un.n_term.N_alink	/* link to array list */
142 #define	n_string	n_un.n_term.n_tun.n_str.N_string
143 #define	n_strlen	n_un.n_term.n_tun.n_str.N_strlen
144 #define	n_int		n_un.n_term.n_tun.N_int
145 #define	n_real		n_un.n_term.n_tun.N_real
146 #define	n_function	n_un.n_term.n_tun.N_function
147 #define	n_ufunc		n_un.n_term.n_tun.N_ufunc
148 #define	n_name		n_un.n_term.N_name
149 #define	n_left		n_un.n_op.N_left
150 #define	n_right		n_un.n_op.N_right
151 #define	n_lineno	n_un.n_op.N_lineno
152 #define	n_keywtype	n_un.n_op.N_lineno
153 #define	n_regexp	n_un.n_re.N_regexp
154 /*
155  * Compress the types that are actually used in the final tree
156  * to save space in the intermediate file. Allows 1 byte to
157  * represent all types
158  */
159 
160 
161 
162 
163 
164 
165 
166 /*
167  * n_flags bit assignments.
168  */
169 #define	FALLOC		0x01	/* Allocated node */
170 #define	FSTATIC		0x00	/* Not allocated */
171 #define	FMATCH		0x02	/* pattern,pattern (first part matches) */
172 #define	FSPECIAL	0x04	/* Special pre-computed variable */
173 #define	FINARRAY	0x08	/* NODE installed in N_alink array list */
174 #define	FNOALLOC	0x10	/* mark node FALLOC, but don't malloc */
175 #define	FSENSE		0x20	/* Sense if string looks like INT/REAL */
176 #define	FSAVE		(FSPECIAL|FINARRAY)	/* assign leaves on */
177 
178 #define	FINT		0x40	/* Node has integer type */
179 #define	FREAL		0x80	/* Node has real type */
180 #define	FSTRING		0x100	/* Node has string type */
181 #define	FNONTOK		0x200	/* Node has non-token type */
182 #define	FVINT		0x400	/* Node looks like an integer */
183 #define	FVREAL		0x800	/* Node looks like a real number */
184 #define	FLARRAY		0x1000	/* Local array node */
185 
186 /*
187  * n_flags macros
188  * These work when given an argument of np->n_flags
189  */
190 #define	isleaf(f)	(!((f)&FNONTOK))
191 #define	isstring(f)	((f)&FSTRING)
192 #define	isastring(f)	(((f)&(FSTRING|FALLOC)) == (FSTRING|FALLOC))
193 #define	isnumber(f)	((f)&(FINT|FVINT|FREAL|FVREAL))
194 #define	isreal(f)	((f)&(FREAL|FVREAL))
195 #define	isint(f)	((f)&(FINT|FVINT))
196 
197 /*
198  * Prototype file size is defined in awksize.h
199  */
200 
201 
202 
203 
204 
205 /*
206  * Awkrun prototype default name
207  */
208 #if defined(DOS)
209 #if defined(__386__)
210 #define	AWK_PROTOTYPE  M_ETCDIR(awkrunf.dos)
211 #define	AWK_LPROTOTYPE M_ETCDIR(awkrunf.dos)
212 #else
213 #define	AWK_PROTOTYPE  M_ETCDIR(awkrun.dos)
214 #define	AWK_LPROTOTYPE M_ETCDIR(awkrunl.dos)
215 #endif
216 #elif defined(OS2)
217 #define	AWK_PROTOTYPE M_ETCDIR(awkrun.os2)
218 #elif defined(NT)
219 #define	AWK_PROTOTYPE M_ETCDIR(awkrun.nt)
220 #else
221 #define	AWK_PROTOTYPE M_ETCDIR(awkrun.mod)
222 #endif
223 
224 /*
225  * This is a kludge that gets around a bug in compact & large
226  * models under DOS.  It also makes the generated
227  * code faster even if there wasn't a bug.  UNIX people: try
228  * to ignore these noisy "near" declarations.
229  */
230 #ifndef	DOS
231 #define	near
232 #endif
233 
234 typedef	wchar_t	near	*LOCCHARP;	/* pointer to local strings */
235 /*
236  * Form of builtin symbols
237  * This should be a union because only one of r_ivalue
238  * and r_svalue is needed, but (alas) unions cannot be
239  * initialised.
240  */
241 typedef	struct	RESERVED {
242 	LOCCHARP	r_name;
243 	int		r_type;		/* Type of node */
244 	INT		r_ivalue;	/* Integer value or wcslen(r_svalue) */
245 	STRING		r_svalue;	/* String value */
246 }	RESERVED;
247 
248 /*
249  * Table of builtin functions.
250  */
251 typedef	struct	RESFUNC {
252 	LOCCHARP	rf_name;
253 	int		rf_type;	/* FUNC || GETLINE */
254 	FUNCTION	rf_func;	/* Function pointer */
255 }	RESFUNC;
256 
257 /*
258  * Structure holding list of open files.
259  */
260 typedef	struct	OFILE	{
261 	ushort_t f_mode;		/* Open mode: WRITE, APPEND, PIPE */
262 	FILE	*f_fp;			/* File pointer if open */
263 	char	*f_name;		/* Remembered file name */
264 }	OFILE;
265 
266 /* Global functions -- awk.y */
267 int	yyparse(void);
268 
269 /* Global functions -- awk1.c */
270 #ifdef __WATCOMC__
271 #pragma aux yyerror aborts;
272 #pragma aux awkerr aborts;
273 #pragma aux awkperr aborts;
274 #endif
275 void	yyerror(char *msg, ...);
276 void	awkerr(char *fmt, ...) __NORETURN;
277 void	awkperr(char *fmt, ...);
278 void	uexit(NODE *);
279 int	yylex(void);
280 NODE	*renode(wchar_t *restr);
281 wchar_t	*emalloc(unsigned);
282 wchar_t	*erealloc(wchar_t *, unsigned);
283 
284 /* Global functions -- awk2.c */
285 void	awk(void);
286 void	dobegin(void);
287 void	doend(int status) __NORETURN;
288 int	nextrecord(wchar_t *buf, FILE *fp);
289 wchar_t	*defrecord(wchar_t *bp, int lim, FILE *fp);
290 wchar_t	*charrecord(wchar_t *bp, int lim, FILE *fp);
291 wchar_t	*multirecord(wchar_t *bp, int lim, FILE *fp);
292 wchar_t	*whitefield(wchar_t **endp);
293 wchar_t	*blackfield(wchar_t **endp);
294 wchar_t	*refield(wchar_t **endp);
295 void	s_print(NODE *np);
296 void	s_prf(NODE *np);
297 size_t	xprintf(NODE *np, FILE *fp, wchar_t **cp);
298 void	awkclose(OFILE *op);
299 
300 /* Global functions -- awk3.c */
301 void	strassign(NODE *np, STRING string, int flags, size_t length);
302 NODE	*nassign(NODE *np, NODE *value);
303 NODE	*assign(NODE *np, NODE *value);
304 void	delarray(NODE *np);
305 NODE	*node(int type, NODE *left, NODE *right);
306 NODE	*intnode(INT i);
307 NODE	*realnode(REAL r);
308 NODE	*stringnode(STRING str, int aflag, size_t wcslen);
309 NODE	*vlookup(wchar_t *name, int nocreate);
310 NODE	*emptynode(int type, size_t nlength);
311 void	freenode(NODE *np);
312 void	execute(NODE *np);
313 INT	exprint(NODE *np);
314 REAL	exprreal(NODE *np);
315 STRING	exprstring(NODE *np);
316 STRING	strsave(wchar_t *string);
317 NODE	*exprreduce(NODE *np);
318 NODE	*getlist(NODE **npp);
319 NODE	*symwalk(int *buckp, NODE **npp);
320 REGEXP	getregexp(NODE *np);
321 void	addsymtab(NODE *np);
322 void	delsymtab(NODE *np, int fflag);
323 NODE	* finstall(LOCCHARP name, FUNCTION f, int type);
324 void	kinstall(LOCCHARP name, int type);
325 void	fieldsplit(void);
326 void	promote(NODE *);
327 
328 
329 
330 
331 
332 
333 
334 /* Global functions -- awk4.c */
335 NODE	*f_exp(NODE *np);
336 NODE	*f_int(NODE *np);
337 NODE	*f_log(NODE *np);
338 NODE	*f_sqrt(NODE *np);
339 NODE	*f_getline(NODE *np);
340 NODE	*f_index(NODE *np);
341 NODE	*f_length(NODE *np);
342 NODE	*f_split(NODE *np);
343 NODE	*f_sprintf(NODE *np);
344 NODE	*f_substr(NODE *np);
345 NODE	*f_rand(NODE *np);
346 NODE	*f_srand(NODE *np);
347 NODE	*f_sin(NODE *np);
348 NODE	*f_cos(NODE *np);
349 NODE	*f_atan2(NODE *np);
350 NODE	*f_sub(NODE *np);
351 NODE	*f_gsub(NODE *np);
352 NODE	*f_match(NODE *np);
353 NODE	*f_system(NODE *np);
354 NODE	*f_ord(NODE *np);
355 NODE	*f_tolower(NODE *np);
356 NODE	*f_toupper(NODE *np);
357 NODE	*f_close(NODE *np);
358 NODE	*f_asort(NODE *np);
359 
360 /* In awk0.c */
361 
362 
363 
364 extern	wchar_t	_null[];
365 extern	char	r[];
366 extern	char	w[];
367 extern	wchar_t	s_OFMT[];
368 extern	wchar_t	s_CONVFMT[];
369 extern	wchar_t	s_NR[];
370 extern	wchar_t	s_NF[];
371 extern	wchar_t	s_OFS[];
372 extern	wchar_t	s_ORS[];
373 extern	wchar_t	s_RS[];
374 extern	wchar_t	s_FS[];
375 extern	wchar_t	s_FNR[];
376 extern	wchar_t	s_SUBSEP[];
377 extern	wchar_t	s_ARGC[], s_ARGV[], s_ENVIRON[];
378 extern	wchar_t	s_FILENAME[], s_SYMTAB[];
379 extern	wchar_t	s_BEGIN[], s_END[], s_next[];
380 extern	wchar_t	_begin[], _end[];
381 extern	wchar_t	s_exp[], s_getline[], s_index[], s_int[], s_length[], s_log[];
382 extern	wchar_t	s_split[], s_sprintf[], s_sqrt[], s_substr[];
383 extern	wchar_t	s_rand[], s_srand[], s_sin[], s_cos[], s_atan2[];
384 extern	wchar_t	s_sub[], s_gsub[], s_match[], s_system[], s_ord[];
385 extern	wchar_t	s_toupper[], s_tolower[], s_asort[];
386 extern	wchar_t	s_close[];
387 extern	wchar_t	redelim;
388 extern	unsigned char	inprint;
389 extern	unsigned char	funparm;
390 extern	unsigned char	splitdone;
391 extern	uint_t	npattern;
392 extern	uint_t	nfield;
393 extern	uint_t	fcount;
394 extern	uint_t	phase;
395 extern	uint_t	running;
396 extern	uchar_t	catterm;
397 extern	uint_t	lexlast;
398 extern	uint_t	lineno;
399 extern	uchar_t	needsplit, needenviron, doing_begin, begin_getline;
400 extern	ushort_t	slevel;
401 extern	ushort_t	loopexit;
402 extern	wchar_t	radixpoint;
403 extern	REGEXP	resep;
404 extern	RESERVED	reserved[];
405 extern	RESFUNC		resfuncs[];
406 extern	long	NIOSTREAM;	/* Maximum open I/O streams */
407 extern	OFILE	*ofiles;
408 extern	wchar_t	*linebuf;
409 extern	size_t	lbuflen;
410 extern	char	interr[];
411 extern	char	nomem[];
412 extern	NODE	*symtab[NBUCKET];
413 extern	NODE	*yytree;
414 extern	NODE	*freelist;
415 extern	wchar_t	*(*awkrecord)(wchar_t *, int, FILE *);
416 extern	wchar_t	*(*awkfield)(wchar_t **);
417 
418 extern	NODE	*constant;
419 extern	NODE	*const0;
420 extern	NODE	*const1;
421 extern	NODE	*constundef;
422 extern	NODE	*field0;
423 extern	NODE	*incNR;
424 extern	NODE	*incFNR;
425 extern	NODE	*clrFNR;
426 extern	NODE	*ARGVsubi;
427 extern	NODE	*varNR;
428 extern	NODE	*varFNR;
429 extern	NODE	*varNF;
430 extern	NODE	*varOFMT;
431 extern	NODE	*varCONVFMT;
432 extern	NODE	*varOFS;
433 extern	NODE	*varORS;
434 extern	NODE	*varFS;
435 extern	NODE	*varRS;
436 extern	NODE	*varARGC;
437 extern	NODE	*varSUBSEP;
438 extern	NODE	*varENVIRON;
439 extern	NODE	*varSYMTAB;
440 extern	NODE	*varFILENAME;
441 extern	NODE	*curnode;
442 extern	NODE    *inc_oper;
443 extern	NODE	*asn_oper;
444 
445 extern char *mbunconvert(wchar_t *);
446 extern	wchar_t 	*mbstowcsdup(char *);
447 extern	char		*wcstombsdup(wchar_t *);
448 extern	void		awkerr(char *, ...);
449 /*
450  * The following defines the expected max length in chars of a printed number.
451  * This should be the longest expected size for any type of number
452  * ie. float, long etc. This number is used to calculate the approximate
453  * number of chars needed to hold the number.
454  */
455 #ifdef M_NUMSIZE
456 #define	NUMSIZE M_NUMSIZE
457 #else
458 #define	NUMSIZE 30
459 #endif
460 
461 #define	M_MB_L(s)	L##s
462 #ifdef  __STDC__
463 #define	ANSI(x) x
464 #else
465 #define	const
466 #define	signed
467 #define	volatile
468 #define	ANSI(x) ()
469 #endif
470 
471 #define	isWblank(x) (((x) == ' ' || (x) == '\t') ? 1 : 0)
472 
473 
474 /*
475  * Wide character version of regular expression functions.
476  */
477 #define	REGWMATCH_T	int_regwmatch_t
478 #define	REGWCOMP	int_regwcomp
479 #define	REGWEXEC	int_regwexec
480 #define	REGWFREE	int_regwfree
481 #define	REGWERROR	int_regwerror
482 #define	REGWDOSUBA	int_regwdosuba
483 
484 typedef struct {
485 	const wchar_t	*rm_sp, *rm_ep;
486 	regoff_t	rm_so, rm_eo;
487 } int_regwmatch_t;
488 
489 extern int int_regwcomp(REGEXP *, const wchar_t *);
490 extern int int_regwexec(REGEXP, const wchar_t *, size_t,
491 			int_regwmatch_t *, int);
492 extern void int_regwfree(REGEXP);
493 extern size_t int_regwerror(int, REGEXP, char *, size_t);
494 extern int int_regwdosuba(REGEXP, const wchar_t *,
495 			const wchar_t *, wchar_t **, int, int *);
496