1/*
2 * Copyright (c) Ian F. Darwin 1986-1995.
3 * Software written by Ian F. Darwin and others;
4 * maintained 1995-present by Christos Zoulas and others.
5 *
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
8 * are met:
9 * 1. Redistributions of source code must retain the above copyright
10 *    notice immediately at the beginning of the file, without modification,
11 *    this list of conditions, and the following disclaimer.
12 * 2. Redistributions in binary form must reproduce the above copyright
13 *    notice, this list of conditions and the following disclaimer in the
14 *    documentation and/or other materials provided with the distribution.
15 *
16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
20 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26 * SUCH DAMAGE.
27 */
28/*
29 * ASCII magic -- file types that we know based on keywords
30 * that can appear anywhere in the file.
31 *
32 * Extensively modified by Eric Fischer <enf@pobox.com> in July, 2000,
33 * to handle character codes other than ASCII on a unified basis.
34 */
35
36#include "file.h"
37
38#ifndef	lint
39FILE_RCSID("@(#)$File: ascmagic.c,v 1.75 2009/02/03 20:27:51 christos Exp $")
40#endif	/* lint */
41
42#include "magic.h"
43#include <string.h>
44#include <memory.h>
45#include <ctype.h>
46#include <stdlib.h>
47#ifdef HAVE_UNISTD_H
48#include <unistd.h>
49#endif
50#include "names.h"
51
52#define MAXLINELEN 300	/* longest sane line length */
53#define ISSPC(x) ((x) == ' ' || (x) == '\t' || (x) == '\r' || (x) == '\n' \
54		  || (x) == 0x85 || (x) == '\f')
55
56private int ascmatch(const unsigned char *, const unichar *, size_t);
57private unsigned char *encode_utf8(unsigned char *, size_t, unichar *, size_t);
58private size_t trim_nuls(const unsigned char *, size_t);
59
60/*
61 * Undo the NUL-termination kindly provided by process()
62 * but leave at least one byte to look at
63 */
64private size_t
65trim_nuls(const unsigned char *buf, size_t nbytes)
66{
67	while (nbytes > 1 && buf[nbytes - 1] == '\0')
68		nbytes--;
69
70	return nbytes;
71}
72
73protected int
74file_ascmagic(struct magic_set *ms, const unsigned char *buf, size_t nbytes)
75{
76	unichar *ubuf = NULL;
77	size_t ulen;
78	int rv = 1;
79
80	const char *code = NULL;
81	const char *code_mime = NULL;
82	const char *type = NULL;
83
84	if (ms->flags & MAGIC_APPLE)
85		return 0;
86
87	nbytes = trim_nuls(buf, nbytes);
88
89	/* If file doesn't look like any sort of text, give up. */
90	if (file_encoding(ms, buf, nbytes, &ubuf, &ulen, &code, &code_mime,
91	    &type) == 0) {
92		rv = 0;
93		goto done;
94	}
95
96	rv = file_ascmagic_with_encoding(ms, buf, nbytes, ubuf, ulen, code,
97	    type);
98
99 done:
100	if (ubuf)
101		free(ubuf);
102
103	return rv;
104}
105
106protected int
107file_ascmagic_with_encoding(struct magic_set *ms, const unsigned char *buf,
108    size_t nbytes, unichar *ubuf, size_t ulen, const char *code,
109    const char *type)
110{
111	unsigned char *utf8_buf = NULL, *utf8_end;
112	size_t mlen, i;
113	const struct names *p;
114	int rv = -1;
115	int mime = ms->flags & MAGIC_MIME;
116
117	const char *subtype = NULL;
118	const char *subtype_mime = NULL;
119
120	int has_escapes = 0;
121	int has_backspace = 0;
122	int seen_cr = 0;
123
124	int n_crlf = 0;
125	int n_lf = 0;
126	int n_cr = 0;
127	int n_nel = 0;
128
129	size_t last_line_end = (size_t)-1;
130	int has_long_lines = 0;
131
132	if (ms->flags & MAGIC_APPLE)
133		return 0;
134
135	nbytes = trim_nuls(buf, nbytes);
136
137	/* If we have fewer than 2 bytes, give up. */
138	if (nbytes <= 1) {
139		rv = 0;
140		goto done;
141	}
142
143	/* Convert ubuf to UTF-8 and try text soft magic */
144	/* malloc size is a conservative overestimate; could be
145	   improved, or at least realloced after conversion. */
146	mlen = ulen * 6;
147	if ((utf8_buf = CAST(unsigned char *, malloc(mlen))) == NULL) {
148		file_oomem(ms, mlen);
149		goto done;
150	}
151	if ((utf8_end = encode_utf8(utf8_buf, mlen, ubuf, ulen)) == NULL)
152		goto done;
153	if ((rv = file_softmagic(ms, utf8_buf, (size_t)(utf8_end - utf8_buf),
154	    TEXTTEST)) != 0)
155		goto done;
156	else
157		rv = -1;
158
159	/* look for tokens from names.h - this is expensive! */
160	if ((ms->flags & MAGIC_NO_CHECK_TOKENS) != 0)
161		goto subtype_identified;
162
163	i = 0;
164	while (i < ulen) {
165		size_t end;
166
167		/* skip past any leading space */
168		while (i < ulen && ISSPC(ubuf[i]))
169			i++;
170		if (i >= ulen)
171			break;
172
173		/* find the next whitespace */
174		for (end = i + 1; end < nbytes; end++)
175			if (ISSPC(ubuf[end]))
176				break;
177
178		/* compare the word thus isolated against the token list */
179		for (p = names; p < names + NNAMES; p++) {
180			if (ascmatch((const unsigned char *)p->name, ubuf + i,
181			    end - i)) {
182				subtype = types[p->type].human;
183				subtype_mime = types[p->type].mime;
184				goto subtype_identified;
185			}
186		}
187
188		i = end;
189	}
190
191subtype_identified:
192
193	/* Now try to discover other details about the file. */
194	for (i = 0; i < ulen; i++) {
195		if (ubuf[i] == '\n') {
196			if (seen_cr)
197				n_crlf++;
198			else
199				n_lf++;
200			last_line_end = i;
201		} else if (seen_cr)
202			n_cr++;
203
204		seen_cr = (ubuf[i] == '\r');
205		if (seen_cr)
206			last_line_end = i;
207
208		if (ubuf[i] == 0x85) { /* X3.64/ECMA-43 "next line" character */
209			n_nel++;
210			last_line_end = i;
211		}
212
213		/* If this line is _longer_ than MAXLINELEN, remember it. */
214		if (i > last_line_end + MAXLINELEN)
215			has_long_lines = 1;
216
217		if (ubuf[i] == '\033')
218			has_escapes = 1;
219		if (ubuf[i] == '\b')
220			has_backspace = 1;
221	}
222
223	/* Beware, if the data has been truncated, the final CR could have
224	   been followed by a LF.  If we have HOWMANY bytes, it indicates
225	   that the data might have been truncated, probably even before
226	   this function was called. */
227	if (seen_cr && nbytes < HOWMANY)
228		n_cr++;
229
230	if (strcmp(type, "binary") == 0) {
231		rv = 0;
232		goto done;
233	}
234	if (mime) {
235		if ((mime & MAGIC_MIME_TYPE) != 0) {
236			if (subtype_mime) {
237				if (file_printf(ms, "%s", subtype_mime) == -1)
238					goto done;
239			} else {
240				if (file_printf(ms, "text/plain") == -1)
241					goto done;
242			}
243		}
244	} else {
245		if (file_printf(ms, "%s", code) == -1)
246			goto done;
247
248		if (subtype) {
249			if (file_printf(ms, " %s", subtype) == -1)
250				goto done;
251		}
252
253		if (file_printf(ms, " %s", type) == -1)
254			goto done;
255
256		if (has_long_lines)
257			if (file_printf(ms, ", with very long lines") == -1)
258				goto done;
259
260		/*
261		 * Only report line terminators if we find one other than LF,
262		 * or if we find none at all.
263		 */
264		if ((n_crlf == 0 && n_cr == 0 && n_nel == 0 && n_lf == 0) ||
265		    (n_crlf != 0 || n_cr != 0 || n_nel != 0)) {
266			if (file_printf(ms, ", with") == -1)
267				goto done;
268
269			if (n_crlf == 0 && n_cr == 0 && n_nel == 0 && n_lf == 0) {
270				if (file_printf(ms, " no") == -1)
271					goto done;
272			} else {
273				if (n_crlf) {
274					if (file_printf(ms, " CRLF") == -1)
275						goto done;
276					if (n_cr || n_lf || n_nel)
277						if (file_printf(ms, ",") == -1)
278							goto done;
279				}
280				if (n_cr) {
281					if (file_printf(ms, " CR") == -1)
282						goto done;
283					if (n_lf || n_nel)
284						if (file_printf(ms, ",") == -1)
285							goto done;
286				}
287				if (n_lf) {
288					if (file_printf(ms, " LF") == -1)
289						goto done;
290					if (n_nel)
291						if (file_printf(ms, ",") == -1)
292							goto done;
293				}
294				if (n_nel)
295					if (file_printf(ms, " NEL") == -1)
296						goto done;
297			}
298
299			if (file_printf(ms, " line terminators") == -1)
300				goto done;
301		}
302
303		if (has_escapes)
304			if (file_printf(ms, ", with escape sequences") == -1)
305				goto done;
306		if (has_backspace)
307			if (file_printf(ms, ", with overstriking") == -1)
308				goto done;
309	}
310	rv = 1;
311done:
312	if (utf8_buf)
313		free(utf8_buf);
314
315	return rv;
316}
317
318private int
319ascmatch(const unsigned char *s, const unichar *us, size_t ulen)
320{
321	size_t i;
322
323	for (i = 0; i < ulen; i++) {
324		if (s[i] != us[i])
325			return 0;
326	}
327
328	if (s[i])
329		return 0;
330	else
331		return 1;
332}
333
334/*
335 * Encode Unicode string as UTF-8, returning pointer to character
336 * after end of string, or NULL if an invalid character is found.
337 */
338private unsigned char *
339encode_utf8(unsigned char *buf, size_t len, unichar *ubuf, size_t ulen)
340{
341	size_t i;
342	unsigned char *end = buf + len;
343
344	for (i = 0; i < ulen; i++) {
345		if (ubuf[i] <= 0x7f) {
346			if (end - buf < 1)
347				return NULL;
348			*buf++ = (unsigned char)ubuf[i];
349		} else if (ubuf[i] <= 0x7ff) {
350			if (end - buf < 2)
351				return NULL;
352			*buf++ = (unsigned char)((ubuf[i] >> 6) + 0xc0);
353			*buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80);
354		} else if (ubuf[i] <= 0xffff) {
355			if (end - buf < 3)
356				return NULL;
357			*buf++ = (unsigned char)((ubuf[i] >> 12) + 0xe0);
358			*buf++ = (unsigned char)(((ubuf[i] >> 6) & 0x3f) + 0x80);
359			*buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80);
360		} else if (ubuf[i] <= 0x1fffff) {
361			if (end - buf < 4)
362				return NULL;
363			*buf++ = (unsigned char)((ubuf[i] >> 18) + 0xf0);
364			*buf++ = (unsigned char)(((ubuf[i] >> 12) & 0x3f) + 0x80);
365			*buf++ = (unsigned char)(((ubuf[i] >>  6) & 0x3f) + 0x80);
366			*buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80);
367		} else if (ubuf[i] <= 0x3ffffff) {
368			if (end - buf < 5)
369				return NULL;
370			*buf++ = (unsigned char)((ubuf[i] >> 24) + 0xf8);
371			*buf++ = (unsigned char)(((ubuf[i] >> 18) & 0x3f) + 0x80);
372			*buf++ = (unsigned char)(((ubuf[i] >> 12) & 0x3f) + 0x80);
373			*buf++ = (unsigned char)(((ubuf[i] >>  6) & 0x3f) + 0x80);
374			*buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80);
375		} else if (ubuf[i] <= 0x7fffffff) {
376			if (end - buf < 6)
377				return NULL;
378			*buf++ = (unsigned char)((ubuf[i] >> 30) + 0xfc);
379			*buf++ = (unsigned char)(((ubuf[i] >> 24) & 0x3f) + 0x80);
380			*buf++ = (unsigned char)(((ubuf[i] >> 18) & 0x3f) + 0x80);
381			*buf++ = (unsigned char)(((ubuf[i] >> 12) & 0x3f) + 0x80);
382			*buf++ = (unsigned char)(((ubuf[i] >>  6) & 0x3f) + 0x80);
383			*buf++ = (unsigned char)((ubuf[i] & 0x3f) + 0x80);
384		} else /* Invalid character */
385			return NULL;
386	}
387
388	return buf;
389}
390