1 /*
2  * Copyright (c) 2017 Stefan Sperling <stsp@openbsd.org>
3  *
4  * Permission to use, copy, modify, and distribute this software for any
5  * purpose with or without fee is hereby granted, provided that the above
6  * copyright notice and this permission notice appear in all copies.
7  *
8  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
9  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
10  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
11  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
12  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
13  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
14  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
15  */
16 
17 #include <stdio.h>
18 #include <stdlib.h>
19 #include <string.h>
20 #include <sha1.h>
21 #include <zlib.h>
22 #include <ctype.h>
23 #include <limits.h>
24 
25 #include "got_error.h"
26 #include "got_object.h"
27 #include "got_repository.h"
28 
29 #ifndef MIN
30 #define	MIN(_a,_b) ((_a) < (_b) ? (_a) : (_b))
31 #endif
32 
33 #ifndef nitems
34 #define nitems(_a) (sizeof(_a) / sizeof((_a)[0]))
35 #endif
36 
37 #define GOT_OBJ_TAG_COMMIT	"commit"
38 #define GOT_OBJ_TAG_TREE	"tree"
39 #define GOT_OBJ_TAG_BLOB	"blob"
40 
41 const char *
42 got_object_id_str(struct got_object_id *id, char *buf, size_t size)
43 {
44 	char *p = buf;
45 	char hex[3];
46 	int i;
47 
48 	if (size < SHA1_DIGEST_STRING_LENGTH)
49 		return NULL;
50 
51 	for (i = 0; i < SHA1_DIGEST_LENGTH; i++) {
52 		snprintf(hex, sizeof(hex), "%.2x", id->sha1[i]);
53 		p[0] = hex[0];
54 		p[1] = hex[1];
55 		p += 2;
56 	}
57 	p[0] = '\0';
58 
59 	return buf;
60 }
61 
62 struct got_zstream_buf {
63 	z_stream z;
64 	char *inbuf;
65 	size_t inlen;
66 	char *outbuf;
67 	size_t outlen;
68 	int flags;
69 #define GOT_ZSTREAM_F_HAVE_MORE 0x01
70 };
71 
72 static void
73 inflate_end(struct got_zstream_buf *zb)
74 {
75 	free(zb->inbuf);
76 	free(zb->outbuf);
77 	inflateEnd(&zb->z);
78 }
79 
80 static const struct got_error *
81 inflate_init(struct got_zstream_buf *zb, size_t bufsize)
82 {
83 	const struct got_error *err = NULL;
84 
85 	memset(zb, 0, sizeof(*zb));
86 
87 	zb->z.zalloc = Z_NULL;
88 	zb->z.zfree = Z_NULL;
89 	if (inflateInit(&zb->z) != Z_OK) {
90 		err = got_error(GOT_ERR_IO);
91 		goto done;
92 	}
93 
94 	zb->inlen = zb->outlen = bufsize;
95 
96 	zb->inbuf = calloc(1, zb->inlen);
97 	if (zb->inbuf == NULL) {
98 		err = got_error(GOT_ERR_NO_MEM);
99 		goto done;
100 	}
101 
102 	zb->outbuf = calloc(1, zb->outlen);
103 	if (zb->outbuf == NULL) {
104 		err = got_error(GOT_ERR_NO_MEM);
105 		goto done;
106 	}
107 
108 done:
109 	if (err)
110 		inflate_end(zb);
111 	return err;
112 }
113 
114 static const struct got_error *
115 inflate_read(struct got_zstream_buf *zb, FILE *f, size_t *outlenp)
116 {
117 	size_t last_total_out = zb->z.total_out;
118 	z_stream *z = &zb->z;
119 	int n, ret;
120 
121 	z->next_out = zb->outbuf;
122 	z->avail_out = zb->outlen;
123 
124 	if (z->avail_in == 0 && (zb->flags & GOT_ZSTREAM_F_HAVE_MORE) == 0) {
125 		int i;
126 		n = fread(zb->inbuf, 1, zb->inlen, f);
127 		if (n == 0) {
128 			if (ferror(f))
129 				return got_error(GOT_ERR_IO);
130 			*outlenp = 0;
131 			return NULL;
132 		}
133 		z->next_in = zb->inbuf;
134 		z->avail_in = n;
135 	}
136 
137 	ret = inflate(z, Z_SYNC_FLUSH);
138 	if (ret == Z_OK) {
139 		if (z->avail_out == 0)
140 			zb->flags |= GOT_ZSTREAM_F_HAVE_MORE;
141 		else
142 			zb->flags &= ~GOT_ZSTREAM_F_HAVE_MORE;
143 	} else if (ret != Z_STREAM_END)
144 		return got_error(GOT_ERR_DECOMPRESSION);
145 
146 	*outlenp = z->total_out - last_total_out;
147 	return NULL;
148 }
149 
150 static const struct got_error *
151 parse_obj_header(struct got_object **obj, char *buf, size_t len)
152 {
153 	const char *obj_tags[] = {
154 		GOT_OBJ_TAG_COMMIT,
155 		GOT_OBJ_TAG_TREE,
156 		GOT_OBJ_TAG_BLOB
157 	};
158 	const int obj_types[] = {
159 		GOT_OBJ_TYPE_COMMIT,
160 		GOT_OBJ_TYPE_TREE,
161 		GOT_OBJ_TYPE_BLOB,
162 	};
163 	int type = 0;
164 	size_t size = 0;
165 	int i;
166 	char *p = strchr(buf, '\0');
167 
168 	if (p == NULL)
169 		return got_error(GOT_ERR_BAD_OBJ_HDR);
170 
171 	for (i = 0; i < nitems(obj_tags); i++) {
172 		const char *tag = obj_tags[i];
173 		const char *errstr;
174 
175 		if (strncmp(buf, tag, strlen(tag)) != 0)
176 			continue;
177 
178 		type = obj_types[i];
179 		if (len <= strlen(tag))
180 			return got_error(GOT_ERR_BAD_OBJ_HDR);
181 		size = strtonum(buf + strlen(tag), 0, LONG_MAX, &errstr);
182 		if (errstr != NULL)
183 			return got_error(GOT_ERR_BAD_OBJ_HDR);
184 		break;
185 	}
186 
187 	if (type == 0)
188 		return got_error(GOT_ERR_BAD_OBJ_HDR);
189 
190 	*obj = calloc(1, sizeof(**obj));
191 	(*obj)->type = type;
192 	(*obj)->size = size;
193 	return NULL;
194 }
195 
196 static const struct got_error *
197 read_object_header(struct got_object **obj, struct got_repository *repo,
198     const char *path)
199 {
200 	const struct got_error *err;
201 	FILE *f;
202 	struct got_zstream_buf zb;
203 	char *p;
204 	size_t outlen;
205 	int i, ret;
206 
207 	f = fopen(path, "rb");
208 	if (f == NULL)
209 		return got_error(GOT_ERR_BAD_PATH);
210 
211 	err = inflate_init(&zb, 64);
212 	if (err) {
213 		fclose(f);
214 		return err;
215 	}
216 
217 	err = inflate_read(&zb, f, &outlen);
218 	if (err)
219 		goto done;
220 
221 	err = parse_obj_header(obj, zb.outbuf, outlen);
222 done:
223 	inflate_end(&zb);
224 	fclose(f);
225 	return err;
226 }
227 
228 const struct got_error *
229 got_object_open(struct got_object **obj, struct got_repository *repo,
230     struct got_object_id *id)
231 {
232 	const struct got_error *err = NULL;
233 	char *path_objects = got_repo_get_path_objects(repo);
234 	char hex[SHA1_DIGEST_STRING_LENGTH];
235 	char *path = NULL;
236 
237 	if (path_objects == NULL)
238 		return got_error(GOT_ERR_NO_MEM);
239 
240 	got_object_id_str(id, hex, sizeof(hex));
241 
242 	if (asprintf(&path, "%s/%.2x/%s",
243 	    path_objects, id->sha1[0], hex + 2) == -1) {
244 		err = got_error(GOT_ERR_NO_MEM);
245 		goto done;
246 	}
247 
248 	err = read_object_header(obj, repo, path);
249 	if (err == NULL)
250 		memcpy((*obj)->id.sha1, id->sha1, SHA1_DIGEST_LENGTH);
251 
252 done:
253 	free(path);
254 	free(path_objects);
255 	return err;
256 }
257 
258 void
259 got_object_close(struct got_object *obj)
260 {
261 	free(obj);
262 }