1 /*-
2  * Copyright (c) 2003-2008 Tim Kientzle
3  * Copyright (c) 2008 Anselm Strauss
4  * All rights reserved.
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, this list of conditions and the following disclaimer.
11  * 2. Redistributions in binary form must reproduce the above copyright
12  *    notice, this list of conditions and the following disclaimer in the
13  *    documentation and/or other materials provided with the distribution.
14  *
15  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR(S) ``AS IS'' AND ANY EXPRESS OR
16  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
17  * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
18  * IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT, INDIRECT,
19  * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
20  * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
21  * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
22  * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
23  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
24  * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
25  */
26 
27 /*
28  * Development supported by Google Summer of Code 2008.
29  */
30 
31 #include "test.h"
32 __FBSDID("$FreeBSD$");
33 
34 /*
35  * Detailed byte-for-byte verification of the format of a zip archive
36  * with a single file written to it that uses Zip64 extensions.
37  */
38 
39 static unsigned long
40 bitcrc32(unsigned long c, void *_p, size_t s)
41 {
42 	/* This is a drop-in replacement for crc32() from zlib.
43 	 * Libarchive should be able to correctly generate
44 	 * uncompressed zip archives (including correct CRCs) even
45 	 * when zlib is unavailable, and this function helps us verify
46 	 * that.  Yes, this is very, very slow and unsuitable for
47 	 * production use, but it's correct, compact, and works well
48 	 * enough for this particular usage.  Libarchive internally
49 	 * uses a much more efficient implementation.  */
50 	const unsigned char *p = _p;
51 	int bitctr;
52 
53 	if (p == NULL)
54 		return (0);
55 
56 	for (; s > 0; --s) {
57 		c ^= *p++;
58 		for (bitctr = 8; bitctr > 0; --bitctr) {
59 			if (c & 1) c = (c >> 1);
60 			else	   c = (c >> 1) ^ 0xedb88320;
61 			c ^= 0x80000000;
62 		}
63 	}
64 	return (c);
65 }
66 
67 /* Quick and dirty: Read 2-byte and 4-byte integers from Zip file. */
68 static unsigned i2(const unsigned char *p) { return ((p[0] & 0xff) | ((p[1] & 0xff) << 8)); }
69 static unsigned i4(const unsigned char *p) { return (i2(p) | (i2(p + 2) << 16)); }
70 /* We're only working with small values here; ignore the 4 high bytes. */
71 static unsigned i8(const unsigned char *p) { return (i4(p)); }
72 
73 DEFINE_TEST(test_write_format_zip_file_zip64)
74 {
75 	struct archive *a;
76 	struct archive_entry *ae;
77 	time_t t = 1234567890;
78 	struct tm *tm = localtime(&t);
79 	size_t used, buffsize = 1000000;
80 	unsigned long crc;
81 	int file_perm = 00644;
82 	int zip_version = 45;
83 	int zip_compression = 8;
84 	short file_uid = 10, file_gid = 20;
85 	unsigned char *buff, *buffend, *p;
86 	unsigned char *central_header, *local_header, *eocd, *eocd_record;
87 	unsigned char *extension_start, *extension_end;
88 	char file_data[] = {'1', '2', '3', '4', '5', '6', '7', '8'};
89 	char *file_name = "file";
90 
91 #ifndef HAVE_ZLIB_H
92 	zip_compression = 0;
93 #endif
94 
95 	buff = malloc(buffsize);
96 
97 	/* Create a new archive in memory. */
98 	assert((a = archive_write_new()) != NULL);
99 	assertEqualIntA(a, ARCHIVE_OK, archive_write_set_format_zip(a));
100 	assertEqualIntA(a, ARCHIVE_OK,
101 	    archive_write_set_options(a, "zip:zip64"));
102 	assertEqualIntA(a, ARCHIVE_OK,
103 	    archive_write_set_options(a, "zip:experimental"));
104 	assertEqualIntA(a, ARCHIVE_OK,
105 	    archive_write_open_memory(a, buff, buffsize, &used));
106 
107 	assert((ae = archive_entry_new()) != NULL);
108 	archive_entry_copy_pathname(ae, file_name);
109 	archive_entry_set_mode(ae, AE_IFREG | file_perm);
110 	archive_entry_set_size(ae, sizeof(file_data));
111 	archive_entry_set_uid(ae, file_uid);
112 	archive_entry_set_gid(ae, file_gid);
113 	archive_entry_set_mtime(ae, t, 0);
114 	assertEqualInt(0, archive_write_header(a, ae));
115 	archive_entry_free(ae);
116 	assertEqualInt(8, archive_write_data(a, file_data, sizeof(file_data)));
117 	assertEqualIntA(a, ARCHIVE_OK, archive_write_close(a));
118 	assertEqualInt(ARCHIVE_OK, archive_write_free(a));
119 	buffend = buff + used;
120 	dumpfile("constructed.zip", buff, used);
121 
122 	/* Verify "End of Central Directory" record. */
123 	/* Get address of end-of-central-directory record. */
124 	eocd_record = p = buffend - 22; /* Assumes there is no zip comment field. */
125 	failure("End-of-central-directory begins with PK\\005\\006 signature");
126 	assertEqualMem(p, "PK\005\006", 4);
127 	failure("This must be disk 0");
128 	assertEqualInt(i2(p + 4), 0);
129 	failure("Central dir must start on disk 0");
130 	assertEqualInt(i2(p + 6), 0);
131 	failure("All central dir entries are on this disk");
132 	assertEqualInt(i2(p + 8), i2(p + 10));
133 	eocd = buff + i4(p + 12) + i4(p + 16);
134 	failure("no zip comment");
135 	assertEqualInt(i2(p + 20), 0);
136 
137 	/* Get address of first entry in central directory. */
138 	central_header = p = buff + i4(buffend - 6);
139 	failure("Central file record at offset %d should begin with"
140 	    " PK\\001\\002 signature",
141 	    i4(buffend - 10));
142 
143 	/* Verify file entry in central directory. */
144 	assertEqualMem(p, "PK\001\002", 4); /* Signature */
145 	assertEqualInt(i2(p + 4), 3 * 256 + zip_version); /* Version made by */
146 	assertEqualInt(i2(p + 6), zip_version); /* Version needed to extract */
147 	assertEqualInt(i2(p + 8), 8); /* Flags */
148 	assertEqualInt(i2(p + 10), zip_compression); /* Compression method */
149 	assertEqualInt(i2(p + 12), (tm->tm_hour * 2048) + (tm->tm_min * 32) + (tm->tm_sec / 2)); /* File time */
150 	assertEqualInt(i2(p + 14), ((tm->tm_year - 80) * 512) + ((tm->tm_mon + 1) * 32) + tm->tm_mday); /* File date */
151 	crc = bitcrc32(0, file_data, sizeof(file_data));
152 	assertEqualInt(i4(p + 16), crc); /* CRC-32 */
153 	/* assertEqualInt(i4(p + 20), sizeof(file_data)); */ /* Compressed size */
154 	assertEqualInt(i4(p + 24), sizeof(file_data)); /* Uncompressed size */
155 	assertEqualInt(i2(p + 28), strlen(file_name)); /* Pathname length */
156 	/* assertEqualInt(i2(p + 30), 28); */ /* Extra field length: See below */
157 	assertEqualInt(i2(p + 32), 0); /* File comment length */
158 	assertEqualInt(i2(p + 34), 0); /* Disk number start */
159 	assertEqualInt(i2(p + 36), 0); /* Internal file attrs */
160 	assertEqualInt(i4(p + 38) >> 16 & 01777, file_perm); /* External file attrs */
161 	assertEqualInt(i4(p + 42), 0); /* Offset of local header */
162 	assertEqualMem(p + 46, file_name, strlen(file_name)); /* Pathname */
163 	p = extension_start = central_header + 46 + strlen(file_name);
164 	extension_end = extension_start + i2(central_header + 30);
165 
166 	assertEqualInt(i2(p), 0x5455);  /* 'UT' extension header */
167 	assertEqualInt(i2(p + 2), 5); /* 'UT' size */
168 	assertEqualInt(p[4], 1); /* 'UT' flags */
169 	assertEqualInt(i4(p + 5), t); /* 'UT' mtime */
170 	p += 4 + i2(p + 2);
171 
172 	assertEqualInt(i2(p), 0x7875);  /* 'ux' extension header */
173 	assertEqualInt(i2(p + 2), 11); /* 'ux' size */
174 	/* TODO: verify 'ux' contents */
175 	p += 4 + i2(p + 2);
176 
177 	/* Note: We don't expect to see zip64 extension in the central
178 	 * directory, since the writer knows the actual full size by
179 	 * the time it is ready to write the central directory and has
180 	 * no reason to insert it then.  Info-Zip seems to do the same
181 	 * thing. */
182 
183 	/* Just in case: Report any extra extensions. */
184 	while (p < extension_end) {
185 		failure("Unexpected extension 0x%04X", i2(p));
186 		assert(0);
187 		p += 4 + i2(p + 2);
188 	}
189 
190 	/* Should have run exactly to end of extra data. */
191 	assert(p == extension_end);
192 
193 	assert(p == eocd);
194 
195 	/* After Central dir, we find Zip64 eocd and Zip64 eocd locator. */
196 	assertEqualMem(p, "PK\006\006", 4); /* Zip64 eocd */
197 	assertEqualInt(i8(p + 4), 44); /* We're using v1 Zip64 eocd */
198 	assertEqualInt(i2(p + 12), 45); /* Written by Version 4.5 */
199 	assertEqualInt(i2(p + 14), 45); /* Needs version 4.5 to extract */
200 	assertEqualInt(i4(p + 16), 0); /* This is disk #0 */
201 	assertEqualInt(i4(p + 20), 0); /* Dir starts on disk #0 */
202 	assertEqualInt(i8(p + 24), 1); /* 1 entry on this disk */
203 	assertEqualInt(i8(p + 32), 1); /* 1 entry total */
204 	assertEqualInt(i8(p + 40), eocd - central_header); /* size of cd */
205 	assertEqualInt(i8(p + 48), central_header - buff); /* start of cd */
206 	p += 12 + i8(p + 4);
207 
208 	assertEqualMem(p, "PK\006\007", 4); /* Zip64 eocd locator */
209 	assertEqualInt(i4(p + 4), 0); /* Zip64 eocd is on disk #0 */
210 	assertEqualInt(i8(p + 8), eocd - buff); /* Offset of Zip64 eocd */
211 	assertEqualInt(i4(p + 16), 1); /* 1 disk */
212 	p += 20;
213 
214 	/* Regular EOCD immediately follows Zip64 records. */
215 	assert(p == eocd_record);
216 
217 	/* Verify local header of file entry. */
218 	p = local_header = buff;
219 	assertEqualMem(p, "PK\003\004", 4); /* Signature */
220 	assertEqualInt(i2(p + 4), zip_version); /* Version needed to extract */
221 	assertEqualInt(i2(p + 6), 8); /* Flags */
222 	assertEqualInt(i2(p + 8), zip_compression); /* Compression method */
223 	assertEqualInt(i2(p + 10), (tm->tm_hour * 2048) + (tm->tm_min * 32) + (tm->tm_sec / 2)); /* File time */
224 	assertEqualInt(i2(p + 12), ((tm->tm_year - 80) * 512) + ((tm->tm_mon + 1) * 32) + tm->tm_mday); /* File date */
225 	assertEqualInt(i4(p + 14), 0); /* CRC-32 */
226 	/* assertEqualInt(i4(p + 18), sizeof(file_data)); */ /* Compressed size */
227 	/* assertEqualInt(i4(p + 22), sizeof(file_data)); */ /* Uncompressed size not stored because we're using length-at-end. */
228 	assertEqualInt(i2(p + 26), strlen(file_name)); /* Pathname length */
229 	assertEqualInt(i2(p + 28), 57); /* Extra field length */
230 	assertEqualMem(p + 30, file_name, strlen(file_name)); /* Pathname */
231 	p = extension_start = local_header + 30 + strlen(file_name);
232 	extension_end = extension_start + i2(local_header + 28);
233 
234 	assertEqualInt(i2(p), 0x5455);  /* 'UT' extension header */
235 	assertEqualInt(i2(p + 2), 5); /* 'UT' size */
236 	assertEqualInt(p[4], 1); /* 'UT' flags */
237 	assertEqualInt(i4(p + 5), t); /* 'UT' mtime */
238 	p += 4 + i2(p + 2);
239 
240 	assertEqualInt(i2(p), 0x7875);  /* 'ux' extension header */
241 	assertEqualInt(i2(p + 2), 11); /* 'ux' size */
242 	assertEqualInt(p[4], 1); /* 'ux' version */
243 	assertEqualInt(p[5], 4); /* 'ux' uid size */
244 	assertEqualInt(i4(p + 6), file_uid); /* 'Ux' UID */
245 	assertEqualInt(p[10], 4); /* 'ux' gid size */
246 	assertEqualInt(i4(p + 11), file_gid); /* 'Ux' GID */
247 	p += 4 + i2(p + 2);
248 
249 	assertEqualInt(i2(p), 0x0001);  /* Zip64 extension header */
250 	assertEqualInt(i2(p + 2), 16); /* size */
251 	assertEqualInt(i8(p + 4), 8); /* uncompressed file size */
252 	/* compressed file size we can't verify here */
253 	p += 4 + i2(p + 2);
254 
255 	assertEqualInt(i2(p), 0x6c78); /* 'xl' experimental extension header */
256 	assertEqualInt(i2(p + 2), 9); /* size */
257 	assertEqualInt(p[4], 7); /* bitmap of included fields */
258 	assertEqualInt(i2(p + 5) >> 8, 3); /* system & version made by */
259 	assertEqualInt(i2(p + 7), 0); /* internal file attributes */
260 	assertEqualInt(i4(p + 9) >> 16 & 01777, file_perm); /* external file attributes */
261 	p += 4 + i2(p + 2);
262 
263 	/* Just in case: Report any extra extensions. */
264 	while (p < extension_end) {
265 		failure("Unexpected extension 0x%04X", i2(p));
266 		assert(0);
267 		p += 4 + i2(p + 2);
268 	}
269 
270 	/* Should have run exactly to end of extra data. */
271 	assert(p == extension_end);
272 
273 	/* Data descriptor should follow compressed data. */
274 	while (p < central_header && memcmp(p, "PK\007\010", 4) != 0)
275 		++p;
276 	assertEqualMem(p, "PK\007\010", 4);
277 	assertEqualInt(i4(p + 4), crc); /* CRC-32 */
278 	/* assertEqualInt(i8(p + 8), ???); */ /* compressed size */
279 	assertEqualInt(i8(p + 16), sizeof(file_data)); /* uncompressed size */
280 
281 	/* Central directory should immediately follow the only entry. */
282 	assert(p + 24 == central_header);
283 
284 	free(buff);
285 }
286