xref: /PHP-8.4/ext/fileinfo/libmagic/is_tar.c (revision b7c5813c)
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  * is_tar() -- figure out whether file is a tar archive.
30  *
31  * Stolen (by the author!) from the file_public domain tar program:
32  * Public Domain version written 26 Aug 1985 John Gilmore (ihnp4!hoptoad!gnu).
33  *
34  * @(#)list.c 1.18 9/23/86 Public Domain - gnu
35  *
36  * Comments changed and some code/comments reformatted
37  * for file command by Ian Darwin.
38  */
39 
40 #include "file.h"
41 
42 #ifndef lint
43 FILE_RCSID("@(#)$File: is_tar.c,v 1.50 2022/12/26 17:31:14 christos Exp $")
44 #endif
45 
46 #include "magic.h"
47 #include <string.h>
48 #include <ctype.h>
49 #include "tar.h"
50 
51 #define	isodigit(c)	( ((c) >= '0') && ((c) <= '7') )
52 
53 file_private int is_tar(const unsigned char *, size_t);
54 file_private int from_oct(const char *, size_t);	/* Decode octal number */
55 
56 static const char tartype[][32] = {	/* should be equal to messages */
57 	"tar archive",			/* found in ../magic/Magdir/archive */
58 	"POSIX tar archive",
59 	"POSIX tar archive (GNU)",	/*  */
60 };
61 
62 file_protected int
file_is_tar(struct magic_set * ms,const struct buffer * b)63 file_is_tar(struct magic_set *ms, const struct buffer *b)
64 {
65 	const unsigned char *buf = CAST(const unsigned char *, b->fbuf);
66 	size_t nbytes = b->flen;
67 	/*
68 	 * Do the tar test first, because if the first file in the tar
69 	 * archive starts with a dot, we can confuse it with an nroff file.
70 	 */
71 	int tar;
72 	int mime = ms->flags & MAGIC_MIME;
73 
74 	if ((ms->flags & (MAGIC_APPLE|MAGIC_EXTENSION)) != 0)
75 		return 0;
76 
77 	tar = is_tar(buf, nbytes);
78 	if (tar < 1 || tar > 3)
79 		return 0;
80 
81 	if (mime == MAGIC_MIME_ENCODING)
82 		return 1;
83 
84 	if (file_printf(ms, "%s", mime ? "application/x-tar" :
85 	    tartype[tar - 1]) == -1)
86 		return -1;
87 
88 	return 1;
89 }
90 
91 /*
92  * Return
93  *	0 if the checksum is bad (i.e., probably not a tar archive),
94  *	1 for old UNIX tar file,
95  *	2 for Unix Std (POSIX) tar file,
96  *	3 for GNU tar file.
97  */
98 file_private int
is_tar(const unsigned char * buf,size_t nbytes)99 is_tar(const unsigned char *buf, size_t nbytes)
100 {
101 	static const char gpkg_match[] = "/gpkg-1";
102 
103 	const union record *header = RCAST(const union record *,
104 	    RCAST(const void *, buf));
105 	size_t i;
106 	int sum, recsum;
107 	const unsigned char *p, *ep;
108 	const char *nulp;
109 
110 	if (nbytes < sizeof(*header))
111 		return 0;
112 
113 	/* If the file looks like Gentoo GLEP 78 binary package (GPKG),
114 	 * don't waste time on further checks and fall back to magic rules.
115 	 */
116 	nulp = CAST(const char *,
117 	    memchr(header->header.name, 0, sizeof(header->header.name)));
118 	if (nulp != NULL && nulp >= header->header.name + sizeof(gpkg_match) &&
119 	    memcmp(nulp - sizeof(gpkg_match) + 1, gpkg_match,
120 	    sizeof(gpkg_match)) == 0)
121 	    return 0;
122 
123 	recsum = from_oct(header->header.chksum, sizeof(header->header.chksum));
124 
125 	sum = 0;
126 	p = header->charptr;
127 	ep = header->charptr + sizeof(*header);
128 	while (p < ep)
129 		sum += *p++;
130 
131 	/* Adjust checksum to count the "chksum" field as blanks. */
132 	for (i = 0; i < sizeof(header->header.chksum); i++)
133 		sum -= header->header.chksum[i];
134 	sum += ' ' * sizeof(header->header.chksum);
135 
136 	if (sum != recsum)
137 		return 0;	/* Not a tar archive */
138 
139 	if (strncmp(header->header.magic, GNUTMAGIC,
140 	    sizeof(header->header.magic)) == 0)
141 		return 3;		/* GNU Unix Standard tar archive */
142 
143 	if (strncmp(header->header.magic, TMAGIC,
144 	    sizeof(header->header.magic)) == 0)
145 		return 2;		/* Unix Standard tar archive */
146 
147 	return 1;			/* Old fashioned tar archive */
148 }
149 
150 
151 /*
152  * Quick and dirty octal conversion.
153  *
154  * Result is -1 if the field is invalid (all blank, or non-octal).
155  */
156 file_private int
from_oct(const char * where,size_t digs)157 from_oct(const char *where, size_t digs)
158 {
159 	int	value;
160 
161 	if (digs == 0)
162 		return -1;
163 
164 	while (isspace(CAST(unsigned char, *where))) {	/* Skip spaces */
165 		where++;
166 		if (digs-- == 0)
167 			return -1;		/* All blank field */
168 	}
169 	value = 0;
170 	while (digs > 0 && isodigit(*where)) {	/* Scan til non-octal */
171 		value = (value << 3) | (*where++ - '0');
172 		digs--;
173 	}
174 
175 	if (digs > 0 && *where && !isspace(CAST(unsigned char, *where)))
176 		return -1;			/* Ended on non-(space/NUL) */
177 
178 	return value;
179 }
180