1/*-
2 * Copyright (c) 1992, 1993
3 *	The Regents of the University of California.  All rights reserved.
4 *
5 * This code is derived from software contributed to Berkeley by
6 * Christos Zoulas of Cornell University.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 *    notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 *    notice, this list of conditions and the following disclaimer in the
15 *    documentation and/or other materials provided with the distribution.
16 * 3. Neither the name of the University nor the names of its contributors
17 *    may be used to endorse or promote products derived from this software
18 *    without specific prior written permission.
19 *
20 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
30 * SUCH DAMAGE.
31 *
32 *	$NetBSD: tokenizer.c,v 1.15 2009/02/15 21:55:23 christos Exp $
33 */
34
35#if !defined(lint) && !defined(SCCSID)
36static char sccsid[] = "@(#)tokenizer.c	8.1 (Berkeley) 6/4/93";
37#endif /* not lint && not SCCSID */
38#include <sys/cdefs.h>
39__FBSDID("$FreeBSD$");
40
41/*
42 * tokenize.c: Bourne shell like tokenizer
43 */
44#include "sys.h"
45#include <string.h>
46#include <stdlib.h>
47#include "histedit.h"
48
49typedef enum {
50	Q_none, Q_single, Q_double, Q_one, Q_doubleone
51} quote_t;
52
53#define	IFS		"\t \n"
54
55#define	TOK_KEEP	1
56#define	TOK_EAT		2
57
58#define	WINCR		20
59#define	AINCR		10
60
61#define	tok_strdup(a)		strdup(a)
62#define	tok_malloc(a)		malloc(a)
63#define	tok_free(a)		free(a)
64#define	tok_realloc(a, b)	realloc(a, b)
65
66
67struct tokenizer {
68	char	*ifs;		/* In field separator			 */
69	int	 argc, amax;	/* Current and maximum number of args	 */
70	char   **argv;		/* Argument list			 */
71	char	*wptr, *wmax;	/* Space and limit on the word buffer	 */
72	char	*wstart;	/* Beginning of next word		 */
73	char	*wspace;	/* Space of word buffer			 */
74	quote_t	 quote;		/* Quoting state			 */
75	int	 flags;		/* flags;				 */
76};
77
78
79private void tok_finish(Tokenizer *);
80
81
82/* tok_finish():
83 *	Finish a word in the tokenizer.
84 */
85private void
86tok_finish(Tokenizer *tok)
87{
88
89	*tok->wptr = '\0';
90	if ((tok->flags & TOK_KEEP) || tok->wptr != tok->wstart) {
91		tok->argv[tok->argc++] = tok->wstart;
92		tok->argv[tok->argc] = NULL;
93		tok->wstart = ++tok->wptr;
94	}
95	tok->flags &= ~TOK_KEEP;
96}
97
98
99/* tok_init():
100 *	Initialize the tokenizer
101 */
102public Tokenizer *
103tok_init(const char *ifs)
104{
105	Tokenizer *tok = (Tokenizer *) tok_malloc(sizeof(Tokenizer));
106
107	if (tok == NULL)
108		return NULL;
109	tok->ifs = tok_strdup(ifs ? ifs : IFS);
110	if (tok->ifs == NULL) {
111		tok_free((ptr_t)tok);
112		return NULL;
113	}
114	tok->argc = 0;
115	tok->amax = AINCR;
116	tok->argv = (char **) tok_malloc(sizeof(char *) * tok->amax);
117	if (tok->argv == NULL) {
118		tok_free((ptr_t)tok->ifs);
119		tok_free((ptr_t)tok);
120		return NULL;
121	}
122	tok->argv[0] = NULL;
123	tok->wspace = (char *) tok_malloc(WINCR);
124	if (tok->wspace == NULL) {
125		tok_free((ptr_t)tok->argv);
126		tok_free((ptr_t)tok->ifs);
127		tok_free((ptr_t)tok);
128		return NULL;
129	}
130	tok->wmax = tok->wspace + WINCR;
131	tok->wstart = tok->wspace;
132	tok->wptr = tok->wspace;
133	tok->flags = 0;
134	tok->quote = Q_none;
135
136	return (tok);
137}
138
139
140/* tok_reset():
141 *	Reset the tokenizer
142 */
143public void
144tok_reset(Tokenizer *tok)
145{
146
147	tok->argc = 0;
148	tok->wstart = tok->wspace;
149	tok->wptr = tok->wspace;
150	tok->flags = 0;
151	tok->quote = Q_none;
152}
153
154
155/* tok_end():
156 *	Clean up
157 */
158public void
159tok_end(Tokenizer *tok)
160{
161
162	tok_free((ptr_t) tok->ifs);
163	tok_free((ptr_t) tok->wspace);
164	tok_free((ptr_t) tok->argv);
165	tok_free((ptr_t) tok);
166}
167
168
169
170/* tok_line():
171 *	Bourne shell (sh(1)) like tokenizing
172 *	Arguments:
173 *		tok	current tokenizer state (setup with tok_init())
174 *		line	line to parse
175 *	Returns:
176 *		-1	Internal error
177 *		 3	Quoted return
178 *		 2	Unmatched double quote
179 *		 1	Unmatched single quote
180 *		 0	Ok
181 *	Modifies (if return value is 0):
182 *		argc	number of arguments
183 *		argv	argument array
184 *		cursorc	if !NULL, argv element containing cursor
185 *		cursorv	if !NULL, offset in argv[cursorc] of cursor
186 */
187public int
188tok_line(Tokenizer *tok, const LineInfo *line,
189    int *argc, const char ***argv, int *cursorc, int *cursoro)
190{
191	const char *ptr;
192	int cc, co;
193
194	cc = co = -1;
195	ptr = line->buffer;
196	for (ptr = line->buffer; ;ptr++) {
197		if (ptr >= line->lastchar)
198			ptr = "";
199		if (ptr == line->cursor) {
200			cc = tok->argc;
201			co = (int)(tok->wptr - tok->wstart);
202		}
203		switch (*ptr) {
204		case '\'':
205			tok->flags |= TOK_KEEP;
206			tok->flags &= ~TOK_EAT;
207			switch (tok->quote) {
208			case Q_none:
209				tok->quote = Q_single;	/* Enter single quote
210							 * mode */
211				break;
212
213			case Q_single:	/* Exit single quote mode */
214				tok->quote = Q_none;
215				break;
216
217			case Q_one:	/* Quote this ' */
218				tok->quote = Q_none;
219				*tok->wptr++ = *ptr;
220				break;
221
222			case Q_double:	/* Stay in double quote mode */
223				*tok->wptr++ = *ptr;
224				break;
225
226			case Q_doubleone:	/* Quote this ' */
227				tok->quote = Q_double;
228				*tok->wptr++ = *ptr;
229				break;
230
231			default:
232				return (-1);
233			}
234			break;
235
236		case '"':
237			tok->flags &= ~TOK_EAT;
238			tok->flags |= TOK_KEEP;
239			switch (tok->quote) {
240			case Q_none:	/* Enter double quote mode */
241				tok->quote = Q_double;
242				break;
243
244			case Q_double:	/* Exit double quote mode */
245				tok->quote = Q_none;
246				break;
247
248			case Q_one:	/* Quote this " */
249				tok->quote = Q_none;
250				*tok->wptr++ = *ptr;
251				break;
252
253			case Q_single:	/* Stay in single quote mode */
254				*tok->wptr++ = *ptr;
255				break;
256
257			case Q_doubleone:	/* Quote this " */
258				tok->quote = Q_double;
259				*tok->wptr++ = *ptr;
260				break;
261
262			default:
263				return (-1);
264			}
265			break;
266
267		case '\\':
268			tok->flags |= TOK_KEEP;
269			tok->flags &= ~TOK_EAT;
270			switch (tok->quote) {
271			case Q_none:	/* Quote next character */
272				tok->quote = Q_one;
273				break;
274
275			case Q_double:	/* Quote next character */
276				tok->quote = Q_doubleone;
277				break;
278
279			case Q_one:	/* Quote this, restore state */
280				*tok->wptr++ = *ptr;
281				tok->quote = Q_none;
282				break;
283
284			case Q_single:	/* Stay in single quote mode */
285				*tok->wptr++ = *ptr;
286				break;
287
288			case Q_doubleone:	/* Quote this \ */
289				tok->quote = Q_double;
290				*tok->wptr++ = *ptr;
291				break;
292
293			default:
294				return (-1);
295			}
296			break;
297
298		case '\n':
299			tok->flags &= ~TOK_EAT;
300			switch (tok->quote) {
301			case Q_none:
302				goto tok_line_outok;
303
304			case Q_single:
305			case Q_double:
306				*tok->wptr++ = *ptr;	/* Add the return */
307				break;
308
309			case Q_doubleone:   /* Back to double, eat the '\n' */
310				tok->flags |= TOK_EAT;
311				tok->quote = Q_double;
312				break;
313
314			case Q_one:	/* No quote, more eat the '\n' */
315				tok->flags |= TOK_EAT;
316				tok->quote = Q_none;
317				break;
318
319			default:
320				return (0);
321			}
322			break;
323
324		case '\0':
325			switch (tok->quote) {
326			case Q_none:
327				/* Finish word and return */
328				if (tok->flags & TOK_EAT) {
329					tok->flags &= ~TOK_EAT;
330					return (3);
331				}
332				goto tok_line_outok;
333
334			case Q_single:
335				return (1);
336
337			case Q_double:
338				return (2);
339
340			case Q_doubleone:
341				tok->quote = Q_double;
342				*tok->wptr++ = *ptr;
343				break;
344
345			case Q_one:
346				tok->quote = Q_none;
347				*tok->wptr++ = *ptr;
348				break;
349
350			default:
351				return (-1);
352			}
353			break;
354
355		default:
356			tok->flags &= ~TOK_EAT;
357			switch (tok->quote) {
358			case Q_none:
359				if (strchr(tok->ifs, *ptr) != NULL)
360					tok_finish(tok);
361				else
362					*tok->wptr++ = *ptr;
363				break;
364
365			case Q_single:
366			case Q_double:
367				*tok->wptr++ = *ptr;
368				break;
369
370
371			case Q_doubleone:
372				*tok->wptr++ = '\\';
373				tok->quote = Q_double;
374				*tok->wptr++ = *ptr;
375				break;
376
377			case Q_one:
378				tok->quote = Q_none;
379				*tok->wptr++ = *ptr;
380				break;
381
382			default:
383				return (-1);
384
385			}
386			break;
387		}
388
389		if (tok->wptr >= tok->wmax - 4) {
390			size_t size = tok->wmax - tok->wspace + WINCR;
391			char *s = (char *) tok_realloc(tok->wspace, size);
392			if (s == NULL)
393				return (-1);
394
395			if (s != tok->wspace) {
396				int i;
397				for (i = 0; i < tok->argc; i++) {
398				    tok->argv[i] =
399					(tok->argv[i] - tok->wspace) + s;
400				}
401				tok->wptr = (tok->wptr - tok->wspace) + s;
402				tok->wstart = (tok->wstart - tok->wspace) + s;
403				tok->wspace = s;
404			}
405			tok->wmax = s + size;
406		}
407		if (tok->argc >= tok->amax - 4) {
408			char **p;
409			tok->amax += AINCR;
410			p = (char **) tok_realloc(tok->argv,
411			    tok->amax * sizeof(char *));
412			if (p == NULL)
413				return (-1);
414			tok->argv = p;
415		}
416	}
417 tok_line_outok:
418	if (cc == -1 && co == -1) {
419		cc = tok->argc;
420		co = (int)(tok->wptr - tok->wstart);
421	}
422	if (cursorc != NULL)
423		*cursorc = cc;
424	if (cursoro != NULL)
425		*cursoro = co;
426	tok_finish(tok);
427	*argv = (const char **)tok->argv;
428	*argc = tok->argc;
429	return (0);
430}
431
432/* tok_str():
433 *	Simpler version of tok_line, taking a NUL terminated line
434 *	and splitting into words, ignoring cursor state.
435 */
436public int
437tok_str(Tokenizer *tok, const char *line, int *argc, const char ***argv)
438{
439	LineInfo li;
440
441	memset(&li, 0, sizeof(li));
442	li.buffer = line;
443	li.cursor = li.lastchar = strchr(line, '\0');
444	return (tok_line(tok, &li, argc, argv, NULL, NULL));
445}
446