xref: /openbsd/lib/libedit/tokenizer.c (revision 73471bf0)
1 /*	$OpenBSD: tokenizer.c,v 1.21 2016/04/11 21:17:29 schwarze Exp $	*/
2 /*	$NetBSD: tokenizer.c,v 1.28 2016/04/11 18:56:31 christos Exp $	*/
3 
4 /*-
5  * Copyright (c) 1992, 1993
6  *	The Regents of the University of California.  All rights reserved.
7  *
8  * This code is derived from software contributed to Berkeley by
9  * Christos Zoulas of Cornell University.
10  *
11  * Redistribution and use in source and binary forms, with or without
12  * modification, are permitted provided that the following conditions
13  * are met:
14  * 1. Redistributions of source code must retain the above copyright
15  *    notice, this list of conditions and the following disclaimer.
16  * 2. Redistributions in binary form must reproduce the above copyright
17  *    notice, this list of conditions and the following disclaimer in the
18  *    documentation and/or other materials provided with the distribution.
19  * 3. Neither the name of the University nor the names of its contributors
20  *    may be used to endorse or promote products derived from this software
21  *    without specific prior written permission.
22  *
23  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
24  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
26  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
27  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
29  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
30  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33  * SUCH DAMAGE.
34  */
35 
36 #include "config.h"
37 
38 /* We build this file twice, once as NARROW, once as WIDE. */
39 /*
40  * tokenize.c: Bourne shell like tokenizer
41  */
42 #include <stdlib.h>
43 #include <string.h>
44 
45 #include "histedit.h"
46 
47 typedef enum {
48 	Q_none, Q_single, Q_double, Q_one, Q_doubleone
49 } quote_t;
50 
51 #define	TOK_KEEP	1
52 #define	TOK_EAT		2
53 
54 #define	WINCR		20
55 #define	AINCR		10
56 
57 #define	IFS		STR("\t \n")
58 
59 #ifdef NARROWCHAR
60 #define	Char			char
61 #define	FUN(prefix, rest)	prefix ## _ ## rest
62 #define	TYPE(type)		type
63 #define	STR(x)			x
64 #define	Strchr(s, c)		strchr(s, c)
65 #define	tok_strdup(s)		strdup(s)
66 #else
67 #define	Char			wchar_t
68 #define	FUN(prefix, rest)	prefix ## _w ## rest
69 #define	TYPE(type)		type ## W
70 #define	STR(x)			L ## x
71 #define	Strchr(s, c)		wcschr(s, c)
72 #define	tok_strdup(s)		wcsdup(s)
73 #endif
74 
75 struct TYPE(tokenizer) {
76 	Char	*ifs;		/* In field separator			 */
77 	int	 argc, amax;	/* Current and maximum number of args	 */
78 	Char   **argv;		/* Argument list			 */
79 	Char	*wptr, *wmax;	/* Space and limit on the word buffer	 */
80 	Char	*wstart;	/* Beginning of next word		 */
81 	Char	*wspace;	/* Space of word buffer			 */
82 	quote_t	 quote;		/* Quoting state			 */
83 	int	 flags;		/* flags;				 */
84 };
85 
86 
87 static void FUN(tok,finish)(TYPE(Tokenizer) *);
88 
89 
90 /* FUN(tok,finish)():
91  *	Finish a word in the tokenizer.
92  */
93 static void
94 FUN(tok,finish)(TYPE(Tokenizer) *tok)
95 {
96 
97 	*tok->wptr = '\0';
98 	if ((tok->flags & TOK_KEEP) || tok->wptr != tok->wstart) {
99 		tok->argv[tok->argc++] = tok->wstart;
100 		tok->argv[tok->argc] = NULL;
101 		tok->wstart = ++tok->wptr;
102 	}
103 	tok->flags &= ~TOK_KEEP;
104 }
105 
106 
107 /* FUN(tok,init)():
108  *	Initialize the tokenizer
109  */
110 TYPE(Tokenizer) *
111 FUN(tok,init)(const Char *ifs)
112 {
113 	TYPE(Tokenizer) *tok = malloc(sizeof(TYPE(Tokenizer)));
114 
115 	if (tok == NULL)
116 		return NULL;
117 	tok->ifs = tok_strdup(ifs ? ifs : IFS);
118 	if (tok->ifs == NULL) {
119 		free(tok);
120 		return NULL;
121 	}
122 	tok->argc = 0;
123 	tok->amax = AINCR;
124 	tok->argv = reallocarray(NULL, tok->amax, sizeof(*tok->argv));
125 	if (tok->argv == NULL) {
126 		free(tok->ifs);
127 		free(tok);
128 		return NULL;
129 	}
130 	tok->argv[0] = NULL;
131 	tok->wspace = reallocarray(NULL, WINCR, sizeof(*tok->wspace));
132 	if (tok->wspace == NULL) {
133 		free(tok->argv);
134 		free(tok->ifs);
135 		free(tok);
136 		return NULL;
137 	}
138 	tok->wmax = tok->wspace + WINCR;
139 	tok->wstart = tok->wspace;
140 	tok->wptr = tok->wspace;
141 	tok->flags = 0;
142 	tok->quote = Q_none;
143 
144 	return tok;
145 }
146 
147 
148 /* FUN(tok,reset)():
149  *	Reset the tokenizer
150  */
151 void
152 FUN(tok,reset)(TYPE(Tokenizer) *tok)
153 {
154 
155 	tok->argc = 0;
156 	tok->wstart = tok->wspace;
157 	tok->wptr = tok->wspace;
158 	tok->flags = 0;
159 	tok->quote = Q_none;
160 }
161 
162 
163 /* FUN(tok,end)():
164  *	Clean up
165  */
166 void
167 FUN(tok,end)(TYPE(Tokenizer) *tok)
168 {
169 
170 	free(tok->ifs);
171 	free(tok->wspace);
172 	free(tok->argv);
173 	free(tok);
174 }
175 
176 
177 
178 /* FUN(tok,line)():
179  *	Bourne shell (sh(1)) like tokenizing
180  *	Arguments:
181  *		tok	current tokenizer state (setup with FUN(tok,init)())
182  *		line	line to parse
183  *	Returns:
184  *		-1	Internal error
185  *		 3	Quoted return
186  *		 2	Unmatched double quote
187  *		 1	Unmatched single quote
188  *		 0	Ok
189  *	Modifies (if return value is 0):
190  *		argc	number of arguments
191  *		argv	argument array
192  *		cursorc	if !NULL, argv element containing cursor
193  *		cursorv	if !NULL, offset in argv[cursorc] of cursor
194  */
195 int
196 FUN(tok,line)(TYPE(Tokenizer) *tok, const TYPE(LineInfo) *line,
197     int *argc, const Char ***argv, int *cursorc, int *cursoro)
198 {
199 	const Char *ptr;
200 	int cc, co;
201 
202 	cc = co = -1;
203 	ptr = line->buffer;
204 	for (ptr = line->buffer; ;ptr++) {
205 		if (ptr >= line->lastchar)
206 			ptr = STR("");
207 		if (ptr == line->cursor) {
208 			cc = tok->argc;
209 			co = (int)(tok->wptr - tok->wstart);
210 		}
211 		switch (*ptr) {
212 		case '\'':
213 			tok->flags |= TOK_KEEP;
214 			tok->flags &= ~TOK_EAT;
215 			switch (tok->quote) {
216 			case Q_none:
217 				tok->quote = Q_single;	/* Enter single quote
218 							 * mode */
219 				break;
220 
221 			case Q_single:	/* Exit single quote mode */
222 				tok->quote = Q_none;
223 				break;
224 
225 			case Q_one:	/* Quote this ' */
226 				tok->quote = Q_none;
227 				*tok->wptr++ = *ptr;
228 				break;
229 
230 			case Q_double:	/* Stay in double quote mode */
231 				*tok->wptr++ = *ptr;
232 				break;
233 
234 			case Q_doubleone:	/* Quote this ' */
235 				tok->quote = Q_double;
236 				*tok->wptr++ = *ptr;
237 				break;
238 
239 			default:
240 				return -1;
241 			}
242 			break;
243 
244 		case '"':
245 			tok->flags &= ~TOK_EAT;
246 			tok->flags |= TOK_KEEP;
247 			switch (tok->quote) {
248 			case Q_none:	/* Enter double quote mode */
249 				tok->quote = Q_double;
250 				break;
251 
252 			case Q_double:	/* Exit double quote mode */
253 				tok->quote = Q_none;
254 				break;
255 
256 			case Q_one:	/* Quote this " */
257 				tok->quote = Q_none;
258 				*tok->wptr++ = *ptr;
259 				break;
260 
261 			case Q_single:	/* Stay in single quote mode */
262 				*tok->wptr++ = *ptr;
263 				break;
264 
265 			case Q_doubleone:	/* Quote this " */
266 				tok->quote = Q_double;
267 				*tok->wptr++ = *ptr;
268 				break;
269 
270 			default:
271 				return -1;
272 			}
273 			break;
274 
275 		case '\\':
276 			tok->flags |= TOK_KEEP;
277 			tok->flags &= ~TOK_EAT;
278 			switch (tok->quote) {
279 			case Q_none:	/* Quote next character */
280 				tok->quote = Q_one;
281 				break;
282 
283 			case Q_double:	/* Quote next character */
284 				tok->quote = Q_doubleone;
285 				break;
286 
287 			case Q_one:	/* Quote this, restore state */
288 				*tok->wptr++ = *ptr;
289 				tok->quote = Q_none;
290 				break;
291 
292 			case Q_single:	/* Stay in single quote mode */
293 				*tok->wptr++ = *ptr;
294 				break;
295 
296 			case Q_doubleone:	/* Quote this \ */
297 				tok->quote = Q_double;
298 				*tok->wptr++ = *ptr;
299 				break;
300 
301 			default:
302 				return -1;
303 			}
304 			break;
305 
306 		case '\n':
307 			tok->flags &= ~TOK_EAT;
308 			switch (tok->quote) {
309 			case Q_none:
310 				goto tok_line_outok;
311 
312 			case Q_single:
313 			case Q_double:
314 				*tok->wptr++ = *ptr;	/* Add the return */
315 				break;
316 
317 			case Q_doubleone:   /* Back to double, eat the '\n' */
318 				tok->flags |= TOK_EAT;
319 				tok->quote = Q_double;
320 				break;
321 
322 			case Q_one:	/* No quote, more eat the '\n' */
323 				tok->flags |= TOK_EAT;
324 				tok->quote = Q_none;
325 				break;
326 
327 			default:
328 				return 0;
329 			}
330 			break;
331 
332 		case '\0':
333 			switch (tok->quote) {
334 			case Q_none:
335 				/* Finish word and return */
336 				if (tok->flags & TOK_EAT) {
337 					tok->flags &= ~TOK_EAT;
338 					return 3;
339 				}
340 				goto tok_line_outok;
341 
342 			case Q_single:
343 				return 1;
344 
345 			case Q_double:
346 				return 2;
347 
348 			case Q_doubleone:
349 				tok->quote = Q_double;
350 				*tok->wptr++ = *ptr;
351 				break;
352 
353 			case Q_one:
354 				tok->quote = Q_none;
355 				*tok->wptr++ = *ptr;
356 				break;
357 
358 			default:
359 				return -1;
360 			}
361 			break;
362 
363 		default:
364 			tok->flags &= ~TOK_EAT;
365 			switch (tok->quote) {
366 			case Q_none:
367 				if (Strchr(tok->ifs, *ptr) != NULL)
368 					FUN(tok,finish)(tok);
369 				else
370 					*tok->wptr++ = *ptr;
371 				break;
372 
373 			case Q_single:
374 			case Q_double:
375 				*tok->wptr++ = *ptr;
376 				break;
377 
378 
379 			case Q_doubleone:
380 				*tok->wptr++ = '\\';
381 				tok->quote = Q_double;
382 				*tok->wptr++ = *ptr;
383 				break;
384 
385 			case Q_one:
386 				tok->quote = Q_none;
387 				*tok->wptr++ = *ptr;
388 				break;
389 
390 			default:
391 				return -1;
392 
393 			}
394 			break;
395 		}
396 
397 		if (tok->wptr >= tok->wmax - 4) {
398 			size_t size = tok->wmax - tok->wspace + WINCR;
399 			Char *s = reallocarray(tok->wspace, size, sizeof(*s));
400 			if (s == NULL)
401 				return -1;
402 
403 			if (s != tok->wspace) {
404 				int i;
405 				for (i = 0; i < tok->argc; i++) {
406 				    tok->argv[i] =
407 					(tok->argv[i] - tok->wspace) + s;
408 				}
409 				tok->wptr = (tok->wptr - tok->wspace) + s;
410 				tok->wstart = (tok->wstart - tok->wspace) + s;
411 				tok->wspace = s;
412 			}
413 			tok->wmax = s + size;
414 		}
415 		if (tok->argc >= tok->amax - 4) {
416 			Char **p;
417 			tok->amax += AINCR;
418 			p = reallocarray(tok->argv, tok->amax, sizeof(*p));
419 			if (p == NULL) {
420 				tok->amax -= AINCR;
421 				return -1;
422 			}
423 			tok->argv = p;
424 		}
425 	}
426  tok_line_outok:
427 	if (cc == -1 && co == -1) {
428 		cc = tok->argc;
429 		co = (int)(tok->wptr - tok->wstart);
430 	}
431 	if (cursorc != NULL)
432 		*cursorc = cc;
433 	if (cursoro != NULL)
434 		*cursoro = co;
435 	FUN(tok,finish)(tok);
436 	*argv = (const Char **)tok->argv;
437 	*argc = tok->argc;
438 	return 0;
439 }
440 
441 /* FUN(tok,str)():
442  *	Simpler version of tok_line, taking a NUL terminated line
443  *	and splitting into words, ignoring cursor state.
444  */
445 int
446 FUN(tok,str)(TYPE(Tokenizer) *tok, const Char *line, int *argc,
447     const Char ***argv)
448 {
449 	TYPE(LineInfo) li;
450 
451 	memset(&li, 0, sizeof(li));
452 	li.buffer = line;
453 	li.cursor = li.lastchar = Strchr(line, '\0');
454 	return FUN(tok,line)(tok, &li, argc, argv, NULL, NULL);
455 }
456