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.
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
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.
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
29 * ASCII magic -- try to detect text encoding.
31 * Extensively modified by Eric Fischer <enf@pobox.com> in July, 2000,
32 * to handle character codes other than ASCII on a unified basis.
38 FILE_RCSID("@(#)$File: ascmagic.c,v 1.91 2014/11/28 02:46:39 christos Exp $")
50 #define MAXLINELEN 300 /* longest sane line length */
51 #define ISSPC(x) ((x) == ' ' || (x) == '\t' || (x) == '\r' || (x) == '\n' \
52 || (x) == 0x85 || (x) == '\f')
54 private unsigned char *encode_utf8(unsigned char *, size_t, unichar
*, size_t);
55 private size_t trim_nuls(const unsigned char *, size_t);
58 * Undo the NUL-termination kindly provided by process()
59 * but leave at least one byte to look at
62 trim_nuls(const unsigned char *buf
, size_t nbytes
)
64 while (nbytes
> 1 && buf
[nbytes
- 1] == '\0')
71 file_ascmagic(struct magic_set
*ms
, const unsigned char *buf
, size_t nbytes
,
78 const char *code
= NULL
;
79 const char *code_mime
= NULL
;
80 const char *type
= NULL
;
82 if (ms
->flags
& MAGIC_APPLE
)
85 nbytes
= trim_nuls(buf
, nbytes
);
87 /* If file doesn't look like any sort of text, give up. */
88 if (file_encoding(ms
, buf
, nbytes
, &ubuf
, &ulen
, &code
, &code_mime
,
92 rv
= file_ascmagic_with_encoding(ms
, buf
, nbytes
, ubuf
, ulen
, code
,
101 file_ascmagic_with_encoding(struct magic_set
*ms
, const unsigned char *buf
,
102 size_t nbytes
, unichar
*ubuf
, size_t ulen
, const char *code
,
103 const char *type
, int text
)
105 unsigned char *utf8_buf
= NULL
, *utf8_end
;
108 int mime
= ms
->flags
& MAGIC_MIME
;
110 const char *subtype
= NULL
;
111 const char *subtype_mime
= NULL
;
114 int has_backspace
= 0;
123 size_t last_line_end
= (size_t)-1;
124 int has_long_lines
= 0;
126 if (ms
->flags
& MAGIC_APPLE
)
129 nbytes
= trim_nuls(buf
, nbytes
);
131 /* If we have fewer than 2 bytes, give up. */
137 if (ulen
> 0 && (ms
->flags
& MAGIC_NO_CHECK_SOFT
) == 0) {
138 /* Convert ubuf to UTF-8 and try text soft magic */
139 /* malloc size is a conservative overestimate; could be
140 improved, or at least realloced after conversion. */
142 if ((utf8_buf
= CAST(unsigned char *, malloc(mlen
))) == NULL
) {
143 file_oomem(ms
, mlen
);
146 if ((utf8_end
= encode_utf8(utf8_buf
, mlen
, ubuf
, ulen
))
149 if ((rv
= file_softmagic(ms
, utf8_buf
,
150 (size_t)(utf8_end
- utf8_buf
), 0, NULL
,
151 TEXTTEST
, text
)) == 0)
155 /* Now try to discover other details about the file. */
156 for (i
= 0; i
< ulen
; i
++) {
157 if (ubuf
[i
] == '\n') {
166 seen_cr
= (ubuf
[i
] == '\r');
170 if (ubuf
[i
] == 0x85) { /* X3.64/ECMA-43 "next line" character */
175 /* If this line is _longer_ than MAXLINELEN, remember it. */
176 if (i
> last_line_end
+ MAXLINELEN
)
179 if (ubuf
[i
] == '\033')
185 /* Beware, if the data has been truncated, the final CR could have
186 been followed by a LF. If we have HOWMANY bytes, it indicates
187 that the data might have been truncated, probably even before
188 this function was called. */
189 if (seen_cr
&& nbytes
< HOWMANY
)
192 if (strcmp(type
, "binary") == 0) {
197 if (!file_printedlen(ms
) && (mime
& MAGIC_MIME_TYPE
) != 0) {
199 if (file_printf(ms
, "%s", subtype_mime
) == -1)
202 if (file_printf(ms
, "text/plain") == -1)
207 if (file_printedlen(ms
)) {
208 switch (file_replace(ms
, " text$", ", ")) {
210 switch (file_replace(ms
, " text executable$",
213 if (file_printf(ms
, ", ") == -1)
230 if (file_printf(ms
, "%s", code
) == -1)
234 if (file_printf(ms
, " %s", subtype
) == -1)
238 if (file_printf(ms
, " %s", type
) == -1)
242 if (file_printf(ms
, " executable") == -1)
246 if (file_printf(ms
, ", with very long lines") == -1)
250 * Only report line terminators if we find one other than LF,
251 * or if we find none at all.
253 if ((n_crlf
== 0 && n_cr
== 0 && n_nel
== 0 && n_lf
== 0) ||
254 (n_crlf
!= 0 || n_cr
!= 0 || n_nel
!= 0)) {
255 if (file_printf(ms
, ", with") == -1)
258 if (n_crlf
== 0 && n_cr
== 0 && n_nel
== 0 && n_lf
== 0) {
259 if (file_printf(ms
, " no") == -1)
263 if (file_printf(ms
, " CRLF") == -1)
265 if (n_cr
|| n_lf
|| n_nel
)
266 if (file_printf(ms
, ",") == -1)
270 if (file_printf(ms
, " CR") == -1)
273 if (file_printf(ms
, ",") == -1)
277 if (file_printf(ms
, " LF") == -1)
280 if (file_printf(ms
, ",") == -1)
284 if (file_printf(ms
, " NEL") == -1)
288 if (file_printf(ms
, " line terminators") == -1)
293 if (file_printf(ms
, ", with escape sequences") == -1)
296 if (file_printf(ms
, ", with overstriking") == -1)
307 * Encode Unicode string as UTF-8, returning pointer to character
308 * after end of string, or NULL if an invalid character is found.
310 private unsigned char *
311 encode_utf8(unsigned char *buf
, size_t len
, unichar
*ubuf
, size_t ulen
)
314 unsigned char *end
= buf
+ len
;
316 for (i
= 0; i
< ulen
; i
++) {
317 if (ubuf
[i
] <= 0x7f) {
320 *buf
++ = (unsigned char)ubuf
[i
];
321 } else if (ubuf
[i
] <= 0x7ff) {
324 *buf
++ = (unsigned char)((ubuf
[i
] >> 6) + 0xc0);
325 *buf
++ = (unsigned char)((ubuf
[i
] & 0x3f) + 0x80);
326 } else if (ubuf
[i
] <= 0xffff) {
329 *buf
++ = (unsigned char)((ubuf
[i
] >> 12) + 0xe0);
330 *buf
++ = (unsigned char)(((ubuf
[i
] >> 6) & 0x3f) + 0x80);
331 *buf
++ = (unsigned char)((ubuf
[i
] & 0x3f) + 0x80);
332 } else if (ubuf
[i
] <= 0x1fffff) {
335 *buf
++ = (unsigned char)((ubuf
[i
] >> 18) + 0xf0);
336 *buf
++ = (unsigned char)(((ubuf
[i
] >> 12) & 0x3f) + 0x80);
337 *buf
++ = (unsigned char)(((ubuf
[i
] >> 6) & 0x3f) + 0x80);
338 *buf
++ = (unsigned char)((ubuf
[i
] & 0x3f) + 0x80);
339 } else if (ubuf
[i
] <= 0x3ffffff) {
342 *buf
++ = (unsigned char)((ubuf
[i
] >> 24) + 0xf8);
343 *buf
++ = (unsigned char)(((ubuf
[i
] >> 18) & 0x3f) + 0x80);
344 *buf
++ = (unsigned char)(((ubuf
[i
] >> 12) & 0x3f) + 0x80);
345 *buf
++ = (unsigned char)(((ubuf
[i
] >> 6) & 0x3f) + 0x80);
346 *buf
++ = (unsigned char)((ubuf
[i
] & 0x3f) + 0x80);
347 } else if (ubuf
[i
] <= 0x7fffffff) {
350 *buf
++ = (unsigned char)((ubuf
[i
] >> 30) + 0xfc);
351 *buf
++ = (unsigned char)(((ubuf
[i
] >> 24) & 0x3f) + 0x80);
352 *buf
++ = (unsigned char)(((ubuf
[i
] >> 18) & 0x3f) + 0x80);
353 *buf
++ = (unsigned char)(((ubuf
[i
] >> 12) & 0x3f) + 0x80);
354 *buf
++ = (unsigned char)(((ubuf
[i
] >> 6) & 0x3f) + 0x80);
355 *buf
++ = (unsigned char)((ubuf
[i
] & 0x3f) + 0x80);
356 } else /* Invalid character */