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: head/lib/libarchive/test/test_write_format_zip.c 201247 2009-12-30 05:59:21Z kientzle $");
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;
79 #if defined(HAVE_LOCALTIME_R) || defined(HAVE__LOCALTIME64_S)
80 	struct tm tmbuf;
81 #endif
82 #if defined(HAVE__LOCALTIME64_S)
83 	errno_t terr;
84 	__time64_t tmptime;
85 #endif
86 	size_t used, buffsize = 1000000;
87 	unsigned long crc;
88 	int file_perm = 00644;
89 	int zip_version = 45;
90 	int zip_compression = 8;
91 	short file_uid = 10, file_gid = 20;
92 	unsigned char *buff, *buffend, *p;
93 	unsigned char *central_header, *local_header, *eocd, *eocd_record;
94 	unsigned char *extension_start, *extension_end;
95 	char file_data[] = {'1', '2', '3', '4', '5', '6', '7', '8'};
96 	const char *file_name = "file";
97 
98 #ifndef HAVE_ZLIB_H
99 	zip_compression = 0;
100 #endif
101 
102 #if defined(HAVE_LOCALTIME_R)
103 	tm = localtime_r(&t, &tmbuf);
104 #elif defined(HAVE__LOCALTIME64_S)
105 	tmptime = t;
106 	terr = _localtime64_s(&tmbuf, &tmptime);
107 	if (terr)
108 		tm = NULL;
109 	else
110 		tm = &tmbuf;
111 #else
112 	tm = localtime(&t);
113 #endif
114 	buff = malloc(buffsize);
115 
116 	/* Create a new archive in memory. */
117 	assert((a = archive_write_new()) != NULL);
118 	assertEqualIntA(a, ARCHIVE_OK, archive_write_set_format_zip(a));
119 	assertEqualIntA(a, ARCHIVE_OK,
120 	    archive_write_set_options(a, "zip:zip64"));
121 	assertEqualIntA(a, ARCHIVE_OK,
122 	    archive_write_set_options(a, "zip:experimental"));
123 	assertEqualIntA(a, ARCHIVE_OK,
124 	    archive_write_open_memory(a, buff, buffsize, &used));
125 
126 	assert((ae = archive_entry_new()) != NULL);
127 	archive_entry_copy_pathname(ae, file_name);
128 	archive_entry_set_mode(ae, AE_IFREG | file_perm);
129 	archive_entry_set_size(ae, sizeof(file_data));
130 	archive_entry_set_uid(ae, file_uid);
131 	archive_entry_set_gid(ae, file_gid);
132 	archive_entry_set_mtime(ae, t, 0);
133 	assertEqualInt(0, archive_write_header(a, ae));
134 	archive_entry_free(ae);
135 	assertEqualInt(8, archive_write_data(a, file_data, sizeof(file_data)));
136 	assertEqualIntA(a, ARCHIVE_OK, archive_write_close(a));
137 	assertEqualInt(ARCHIVE_OK, archive_write_free(a));
138 	buffend = buff + used;
139 	dumpfile("constructed.zip", buff, used);
140 
141 	/* Verify "End of Central Directory" record. */
142 	/* Get address of end-of-central-directory record. */
143 	eocd_record = p = buffend - 22; /* Assumes there is no zip comment field. */
144 	failure("End-of-central-directory begins with PK\\005\\006 signature");
145 	assertEqualMem(p, "PK\005\006", 4);
146 	failure("This must be disk 0");
147 	assertEqualInt(i2(p + 4), 0);
148 	failure("Central dir must start on disk 0");
149 	assertEqualInt(i2(p + 6), 0);
150 	failure("All central dir entries are on this disk");
151 	assertEqualInt(i2(p + 8), i2(p + 10));
152 	eocd = buff + i4(p + 12) + i4(p + 16);
153 	failure("no zip comment");
154 	assertEqualInt(i2(p + 20), 0);
155 
156 	/* Get address of first entry in central directory. */
157 	central_header = p = buff + i4(buffend - 6);
158 	failure("Central file record at offset %d should begin with"
159 	    " PK\\001\\002 signature",
160 	    i4(buffend - 10));
161 
162 	/* Verify file entry in central directory. */
163 	assertEqualMem(p, "PK\001\002", 4); /* Signature */
164 	assertEqualInt(i2(p + 4), 3 * 256 + zip_version); /* Version made by */
165 	assertEqualInt(i2(p + 6), zip_version); /* Version needed to extract */
166 	assertEqualInt(i2(p + 8), 8); /* Flags */
167 	assertEqualInt(i2(p + 10), zip_compression); /* Compression method */
168 	assertEqualInt(i2(p + 12), (tm->tm_hour * 2048) + (tm->tm_min * 32) + (tm->tm_sec / 2)); /* File time */
169 	assertEqualInt(i2(p + 14), ((tm->tm_year - 80) * 512) + ((tm->tm_mon + 1) * 32) + tm->tm_mday); /* File date */
170 	crc = bitcrc32(0, file_data, sizeof(file_data));
171 	assertEqualInt(i4(p + 16), crc); /* CRC-32 */
172 	/* assertEqualInt(i4(p + 20), sizeof(file_data)); */ /* Compressed size */
173 	assertEqualInt(i4(p + 24), sizeof(file_data)); /* Uncompressed size */
174 	assertEqualInt(i2(p + 28), strlen(file_name)); /* Pathname length */
175 	/* assertEqualInt(i2(p + 30), 28); */ /* Extra field length: See below */
176 	assertEqualInt(i2(p + 32), 0); /* File comment length */
177 	assertEqualInt(i2(p + 34), 0); /* Disk number start */
178 	assertEqualInt(i2(p + 36), 0); /* Internal file attrs */
179 	assertEqualInt(i4(p + 38) >> 16 & 01777, file_perm); /* External file attrs */
180 	assertEqualInt(i4(p + 42), 0); /* Offset of local header */
181 	assertEqualMem(p + 46, file_name, strlen(file_name)); /* Pathname */
182 	p = extension_start = central_header + 46 + strlen(file_name);
183 	extension_end = extension_start + i2(central_header + 30);
184 
185 	assertEqualInt(i2(p), 0x5455);  /* 'UT' extension header */
186 	assertEqualInt(i2(p + 2), 5); /* 'UT' size */
187 	assertEqualInt(p[4], 1); /* 'UT' flags */
188 	assertEqualInt(i4(p + 5), t); /* 'UT' mtime */
189 	p += 4 + i2(p + 2);
190 
191 	assertEqualInt(i2(p), 0x7875);  /* 'ux' extension header */
192 	assertEqualInt(i2(p + 2), 11); /* 'ux' size */
193 	/* TODO: verify 'ux' contents */
194 	p += 4 + i2(p + 2);
195 
196 	/* Note: We don't expect to see zip64 extension in the central
197 	 * directory, since the writer knows the actual full size by
198 	 * the time it is ready to write the central directory and has
199 	 * no reason to insert it then.  Info-Zip seems to do the same
200 	 * thing. */
201 
202 	/* Just in case: Report any extra extensions. */
203 	while (p < extension_end) {
204 		failure("Unexpected extension 0x%04X", i2(p));
205 		assert(0);
206 		p += 4 + i2(p + 2);
207 	}
208 
209 	/* Should have run exactly to end of extra data. */
210 	assert(p == extension_end);
211 
212 	assert(p == eocd);
213 
214 	/* After Central dir, we find Zip64 eocd and Zip64 eocd locator. */
215 	assertEqualMem(p, "PK\006\006", 4); /* Zip64 eocd */
216 	assertEqualInt(i8(p + 4), 44); /* We're using v1 Zip64 eocd */
217 	assertEqualInt(i2(p + 12), 45); /* Written by Version 4.5 */
218 	assertEqualInt(i2(p + 14), 45); /* Needs version 4.5 to extract */
219 	assertEqualInt(i4(p + 16), 0); /* This is disk #0 */
220 	assertEqualInt(i4(p + 20), 0); /* Dir starts on disk #0 */
221 	assertEqualInt(i8(p + 24), 1); /* 1 entry on this disk */
222 	assertEqualInt(i8(p + 32), 1); /* 1 entry total */
223 	assertEqualInt(i8(p + 40), eocd - central_header); /* size of cd */
224 	assertEqualInt(i8(p + 48), central_header - buff); /* start of cd */
225 	p += 12 + i8(p + 4);
226 
227 	assertEqualMem(p, "PK\006\007", 4); /* Zip64 eocd locator */
228 	assertEqualInt(i4(p + 4), 0); /* Zip64 eocd is on disk #0 */
229 	assertEqualInt(i8(p + 8), eocd - buff); /* Offset of Zip64 eocd */
230 	assertEqualInt(i4(p + 16), 1); /* 1 disk */
231 	p += 20;
232 
233 	/* Regular EOCD immediately follows Zip64 records. */
234 	assert(p == eocd_record);
235 
236 	/* Verify local header of file entry. */
237 	p = local_header = buff;
238 	assertEqualMem(p, "PK\003\004", 4); /* Signature */
239 	assertEqualInt(i2(p + 4), zip_version); /* Version needed to extract */
240 	assertEqualInt(i2(p + 6), 8); /* Flags */
241 	assertEqualInt(i2(p + 8), zip_compression); /* Compression method */
242 	assertEqualInt(i2(p + 10), (tm->tm_hour * 2048) + (tm->tm_min * 32) + (tm->tm_sec / 2)); /* File time */
243 	assertEqualInt(i2(p + 12), ((tm->tm_year - 80) * 512) + ((tm->tm_mon + 1) * 32) + tm->tm_mday); /* File date */
244 	assertEqualInt(i4(p + 14), 0); /* CRC-32 */
245 	/* assertEqualInt(i4(p + 18), sizeof(file_data)); */ /* Compressed size */
246 	/* assertEqualInt(i4(p + 22), sizeof(file_data)); */ /* Uncompressed size not stored because we're using length-at-end. */
247 	assertEqualInt(i2(p + 26), strlen(file_name)); /* Pathname length */
248 	assertEqualInt(i2(p + 28), 57); /* Extra field length */
249 	assertEqualMem(p + 30, file_name, strlen(file_name)); /* Pathname */
250 	p = extension_start = local_header + 30 + strlen(file_name);
251 	extension_end = extension_start + i2(local_header + 28);
252 
253 	assertEqualInt(i2(p), 0x5455);  /* 'UT' extension header */
254 	assertEqualInt(i2(p + 2), 5); /* 'UT' size */
255 	assertEqualInt(p[4], 1); /* 'UT' flags */
256 	assertEqualInt(i4(p + 5), t); /* 'UT' mtime */
257 	p += 4 + i2(p + 2);
258 
259 	assertEqualInt(i2(p), 0x7875);  /* 'ux' extension header */
260 	assertEqualInt(i2(p + 2), 11); /* 'ux' size */
261 	assertEqualInt(p[4], 1); /* 'ux' version */
262 	assertEqualInt(p[5], 4); /* 'ux' uid size */
263 	assertEqualInt(i4(p + 6), file_uid); /* 'Ux' UID */
264 	assertEqualInt(p[10], 4); /* 'ux' gid size */
265 	assertEqualInt(i4(p + 11), file_gid); /* 'Ux' GID */
266 	p += 4 + i2(p + 2);
267 
268 	assertEqualInt(i2(p), 0x0001);  /* Zip64 extension header */
269 	assertEqualInt(i2(p + 2), 16); /* size */
270 	assertEqualInt(i8(p + 4), 8); /* uncompressed file size */
271 	/* compressed file size we can't verify here */
272 	p += 4 + i2(p + 2);
273 
274 	assertEqualInt(i2(p), 0x6c78); /* 'xl' experimental extension header */
275 	assertEqualInt(i2(p + 2), 9); /* size */
276 	assertEqualInt(p[4], 7); /* bitmap of included fields */
277 	assertEqualInt(i2(p + 5) >> 8, 3); /* system & version made by */
278 	assertEqualInt(i2(p + 7), 0); /* internal file attributes */
279 	assertEqualInt(i4(p + 9) >> 16 & 01777, file_perm); /* external file attributes */
280 	p += 4 + i2(p + 2);
281 
282 	/* Just in case: Report any extra extensions. */
283 	while (p < extension_end) {
284 		failure("Unexpected extension 0x%04X", i2(p));
285 		assert(0);
286 		p += 4 + i2(p + 2);
287 	}
288 
289 	/* Should have run exactly to end of extra data. */
290 	assert(p == extension_end);
291 
292 	/* Data descriptor should follow compressed data. */
293 	while (p < central_header && memcmp(p, "PK\007\010", 4) != 0)
294 		++p;
295 	assertEqualMem(p, "PK\007\010", 4);
296 	assertEqualInt(i4(p + 4), crc); /* CRC-32 */
297 	/* assertEqualInt(i8(p + 8), ???); */ /* compressed size */
298 	assertEqualInt(i8(p + 16), sizeof(file_data)); /* uncompressed size */
299 
300 	/* Central directory should immediately follow the only entry. */
301 	assert(p + 24 == central_header);
302 
303 	free(buff);
304 }
305