Home | History | Annotate | Line # | Download | only in libedit
tokenizer.c revision 1.7
      1 /*	$NetBSD: tokenizer.c,v 1.7 2001/01/04 15:56:32 christos Exp $	*/
      2 
      3 /*-
      4  * Copyright (c) 1992, 1993
      5  *	The Regents of the University of California.  All rights reserved.
      6  *
      7  * This code is derived from software contributed to Berkeley by
      8  * Christos Zoulas of Cornell University.
      9  *
     10  * Redistribution and use in source and binary forms, with or without
     11  * modification, are permitted provided that the following conditions
     12  * are met:
     13  * 1. Redistributions of source code must retain the above copyright
     14  *    notice, this list of conditions and the following disclaimer.
     15  * 2. Redistributions in binary form must reproduce the above copyright
     16  *    notice, this list of conditions and the following disclaimer in the
     17  *    documentation and/or other materials provided with the distribution.
     18  * 3. All advertising materials mentioning features or use of this software
     19  *    must display the following acknowledgement:
     20  *	This product includes software developed by the University of
     21  *	California, Berkeley and its contributors.
     22  * 4. Neither the name of the University nor the names of its contributors
     23  *    may be used to endorse or promote products derived from this software
     24  *    without specific prior written permission.
     25  *
     26  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
     27  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     28  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     29  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
     30  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     31  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     32  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     33  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     34  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     35  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     36  * SUCH DAMAGE.
     37  */
     38 
     39 #include <sys/cdefs.h>
     40 #if !defined(lint) && !defined(SCCSID)
     41 #if 0
     42 static char sccsid[] = "@(#)tokenizer.c	8.1 (Berkeley) 6/4/93";
     43 #else
     44 __RCSID("$NetBSD: tokenizer.c,v 1.7 2001/01/04 15:56:32 christos Exp $");
     45 #endif
     46 #endif /* not lint && not SCCSID */
     47 
     48 /*
     49  * tokenize.c: Bourne shell like tokenizer
     50  */
     51 #include "sys.h"
     52 #include <string.h>
     53 #include <stdlib.h>
     54 #include "tokenizer.h"
     55 
     56 typedef enum {
     57 	Q_none, Q_single, Q_double, Q_one, Q_doubleone
     58 } quote_t;
     59 
     60 #define	IFS		"\t \n"
     61 
     62 #define	TOK_KEEP	1
     63 #define	TOK_EAT		2
     64 
     65 #define	WINCR		20
     66 #define	AINCR		10
     67 
     68 #define	tok_malloc(a)		malloc(a)
     69 #define	tok_free(a)		free(a)
     70 #define	tok_realloc(a, b)	realloc(a, b)
     71 
     72 
     73 struct tokenizer {
     74 	char	*ifs;		/* In field separator			 */
     75 	int	 argc, amax;	/* Current and maximum number of args	 */
     76 	char   **argv;		/* Argument list			 */
     77 	char	*wptr, *wmax;	/* Space and limit on the word buffer	 */
     78 	char	*wstart;	/* Beginning of next word		 */
     79 	char	*wspace;	/* Space of word buffer			 */
     80 	quote_t	 quote;		/* Quoting state			 */
     81 	int	 flags;		/* flags;				 */
     82 };
     83 
     84 
     85 private void tok_finish(Tokenizer *);
     86 
     87 
     88 /* tok_finish():
     89  *	Finish a word in the tokenizer.
     90  */
     91 private void
     92 tok_finish(Tokenizer *tok)
     93 {
     94 
     95 	*tok->wptr = '\0';
     96 	if ((tok->flags & TOK_KEEP) || tok->wptr != tok->wstart) {
     97 		tok->argv[tok->argc++] = tok->wstart;
     98 		tok->argv[tok->argc] = NULL;
     99 		tok->wstart = ++tok->wptr;
    100 	}
    101 	tok->flags &= ~TOK_KEEP;
    102 }
    103 
    104 
    105 /* tok_init():
    106  *	Initialize the tokenizer
    107  */
    108 public Tokenizer *
    109 tok_init(const char *ifs)
    110 {
    111 	Tokenizer *tok = (Tokenizer *) tok_malloc(sizeof(Tokenizer));
    112 
    113 	tok->ifs = strdup(ifs ? ifs : IFS);
    114 	tok->argc = 0;
    115 	tok->amax = AINCR;
    116 	tok->argv = (char **) tok_malloc(sizeof(char *) * tok->amax);
    117 	if (tok->argv == NULL)
    118 		return (NULL);
    119 	tok->argv[0] = NULL;
    120 	tok->wspace = (char *) tok_malloc(WINCR);
    121 	if (tok->wspace == NULL)
    122 		return (NULL);
    123 	tok->wmax = tok->wspace + WINCR;
    124 	tok->wstart = tok->wspace;
    125 	tok->wptr = tok->wspace;
    126 	tok->flags = 0;
    127 	tok->quote = Q_none;
    128 
    129 	return (tok);
    130 }
    131 
    132 
    133 /* tok_reset():
    134  *	Reset the tokenizer
    135  */
    136 public void
    137 tok_reset(Tokenizer *tok)
    138 {
    139 
    140 	tok->argc = 0;
    141 	tok->wstart = tok->wspace;
    142 	tok->wptr = tok->wspace;
    143 	tok->flags = 0;
    144 	tok->quote = Q_none;
    145 }
    146 
    147 
    148 /* tok_end():
    149  *	Clean up
    150  */
    151 public void
    152 tok_end(Tokenizer *tok)
    153 {
    154 
    155 	tok_free((ptr_t) tok->ifs);
    156 	tok_free((ptr_t) tok->wspace);
    157 	tok_free((ptr_t) tok->argv);
    158 	tok_free((ptr_t) tok);
    159 }
    160 
    161 
    162 
    163 /* tok_line():
    164  *	Bourne shell like tokenizing
    165  *	Return:
    166  *		-1: Internal error
    167  *		 3: Quoted return
    168  *		 2: Unmatched double quote
    169  *		 1: Unmatched single quote
    170  *		 0: Ok
    171  */
    172 public int
    173 tok_line(Tokenizer *tok, const char *line, int *argc, char ***argv)
    174 {
    175 	const char *ptr;
    176 
    177 	for (;;) {
    178 		switch (*(ptr = line++)) {
    179 		case '\'':
    180 			tok->flags |= TOK_KEEP;
    181 			tok->flags &= ~TOK_EAT;
    182 			switch (tok->quote) {
    183 			case Q_none:
    184 				tok->quote = Q_single;	/* Enter single quote
    185 							 * mode */
    186 				break;
    187 
    188 			case Q_single:	/* Exit single quote mode */
    189 				tok->quote = Q_none;
    190 				break;
    191 
    192 			case Q_one:	/* Quote this ' */
    193 				tok->quote = Q_none;
    194 				*tok->wptr++ = *ptr;
    195 				break;
    196 
    197 			case Q_double:	/* Stay in double quote mode */
    198 				*tok->wptr++ = *ptr;
    199 				break;
    200 
    201 			case Q_doubleone:	/* Quote this ' */
    202 				tok->quote = Q_double;
    203 				*tok->wptr++ = *ptr;
    204 				break;
    205 
    206 			default:
    207 				return (-1);
    208 			}
    209 			break;
    210 
    211 		case '"':
    212 			tok->flags &= ~TOK_EAT;
    213 			tok->flags |= TOK_KEEP;
    214 			switch (tok->quote) {
    215 			case Q_none:	/* Enter double quote mode */
    216 				tok->quote = Q_double;
    217 				break;
    218 
    219 			case Q_double:	/* Exit double quote mode */
    220 				tok->quote = Q_none;
    221 				break;
    222 
    223 			case Q_one:	/* Quote this " */
    224 				tok->quote = Q_none;
    225 				*tok->wptr++ = *ptr;
    226 				break;
    227 
    228 			case Q_single:	/* Stay in single quote mode */
    229 				*tok->wptr++ = *ptr;
    230 				break;
    231 
    232 			case Q_doubleone:	/* Quote this " */
    233 				tok->quote = Q_double;
    234 				*tok->wptr++ = *ptr;
    235 				break;
    236 
    237 			default:
    238 				return (-1);
    239 			}
    240 			break;
    241 
    242 		case '\\':
    243 			tok->flags |= TOK_KEEP;
    244 			tok->flags &= ~TOK_EAT;
    245 			switch (tok->quote) {
    246 			case Q_none:	/* Quote next character */
    247 				tok->quote = Q_one;
    248 				break;
    249 
    250 			case Q_double:	/* Quote next character */
    251 				tok->quote = Q_doubleone;
    252 				break;
    253 
    254 			case Q_one:	/* Quote this, restore state */
    255 				*tok->wptr++ = *ptr;
    256 				tok->quote = Q_none;
    257 				break;
    258 
    259 			case Q_single:	/* Stay in single quote mode */
    260 				*tok->wptr++ = *ptr;
    261 				break;
    262 
    263 			case Q_doubleone:	/* Quote this \ */
    264 				tok->quote = Q_double;
    265 				*tok->wptr++ = *ptr;
    266 				break;
    267 
    268 			default:
    269 				return (-1);
    270 			}
    271 			break;
    272 
    273 		case '\n':
    274 			tok->flags &= ~TOK_EAT;
    275 			switch (tok->quote) {
    276 			case Q_none:
    277 				tok_finish(tok);
    278 				*argv = tok->argv;
    279 				*argc = tok->argc;
    280 				return (0);
    281 
    282 			case Q_single:
    283 			case Q_double:
    284 				*tok->wptr++ = *ptr;	/* Add the return */
    285 				break;
    286 
    287 			case Q_doubleone:   /* Back to double, eat the '\n' */
    288 				tok->flags |= TOK_EAT;
    289 				tok->quote = Q_double;
    290 				break;
    291 
    292 			case Q_one:	/* No quote, more eat the '\n' */
    293 				tok->flags |= TOK_EAT;
    294 				tok->quote = Q_none;
    295 				break;
    296 
    297 			default:
    298 				return (0);
    299 			}
    300 			break;
    301 
    302 		case '\0':
    303 			switch (tok->quote) {
    304 			case Q_none:
    305 				/* Finish word and return */
    306 				if (tok->flags & TOK_EAT) {
    307 					tok->flags &= ~TOK_EAT;
    308 					return (3);
    309 				}
    310 				tok_finish(tok);
    311 				*argv = tok->argv;
    312 				*argc = tok->argc;
    313 				return (0);
    314 
    315 			case Q_single:
    316 				return (1);
    317 
    318 			case Q_double:
    319 				return (2);
    320 
    321 			case Q_doubleone:
    322 				tok->quote = Q_double;
    323 				*tok->wptr++ = *ptr;
    324 				break;
    325 
    326 			case Q_one:
    327 				tok->quote = Q_none;
    328 				*tok->wptr++ = *ptr;
    329 				break;
    330 
    331 			default:
    332 				return (-1);
    333 			}
    334 			break;
    335 
    336 		default:
    337 			tok->flags &= ~TOK_EAT;
    338 			switch (tok->quote) {
    339 			case Q_none:
    340 				if (strchr(tok->ifs, *ptr) != NULL)
    341 					tok_finish(tok);
    342 				else
    343 					*tok->wptr++ = *ptr;
    344 				break;
    345 
    346 			case Q_single:
    347 			case Q_double:
    348 				*tok->wptr++ = *ptr;
    349 				break;
    350 
    351 
    352 			case Q_doubleone:
    353 				*tok->wptr++ = '\\';
    354 				tok->quote = Q_double;
    355 				*tok->wptr++ = *ptr;
    356 				break;
    357 
    358 			case Q_one:
    359 				tok->quote = Q_none;
    360 				*tok->wptr++ = *ptr;
    361 				break;
    362 
    363 			default:
    364 				return (-1);
    365 
    366 			}
    367 			break;
    368 		}
    369 
    370 		if (tok->wptr >= tok->wmax - 4) {
    371 			size_t size = tok->wmax - tok->wspace + WINCR;
    372 			char *s = (char *) tok_realloc(tok->wspace, size);
    373 			/* SUPPRESS 22 */
    374 			int offs = s - tok->wspace;
    375 			if (s == NULL)
    376 				return (-1);
    377 
    378 			if (offs != 0) {
    379 				int i;
    380 				for (i = 0; i < tok->argc; i++)
    381 					tok->argv[i] = tok->argv[i] + offs;
    382 				tok->wptr = tok->wptr + offs;
    383 				tok->wstart = tok->wstart + offs;
    384 				tok->wmax = s + size;
    385 				tok->wspace = s;
    386 			}
    387 		}
    388 		if (tok->argc >= tok->amax - 4) {
    389 			char **p;
    390 			tok->amax += AINCR;
    391 			p = (char **) tok_realloc(tok->argv,
    392 			    tok->amax * sizeof(char *));
    393 			if (p == NULL)
    394 				return (-1);
    395 			tok->argv = p;
    396 		}
    397 	}
    398 }
    399