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 <sys/queue.h>
18 #include <sys/stat.h>
19 
20 #include <stdio.h>
21 #include <stdlib.h>
22 #include <string.h>
23 #include <limits.h>
24 #include <sha1.h>
25 #include <zlib.h>
26 
27 #include "got_repository.h"
28 #include "got_object.h"
29 #include "got_error.h"
30 
31 #include "diff.h"
32 
33 static FILE *
34 opentemp(void)
35 {
36 	char name[PATH_MAX];
37 	int fd;
38 
39 	if (strlcpy(name, "/tmp/got.XXXXXXXX", sizeof(name)) >= sizeof(name))
40 		return NULL;
41 
42 	fd = mkstemp(name);
43 	if (fd < 0)
44 		return NULL;
45 
46 	unlink(name);
47 	return (fdopen(fd, "w+"));
48 }
49 
50 const struct got_error *
51 got_diff_blob(struct got_blob_object *blob1, struct got_blob_object *blob2,
52     const char *label1, const char *label2, FILE *outfile)
53 {
54 	struct got_diff_state ds;
55 	struct got_diff_args args;
56 	const struct got_error *err = NULL;
57 	FILE *f1 = NULL, *f2 = NULL;
58 	char hex1[SHA1_DIGEST_STRING_LENGTH];
59 	char hex2[SHA1_DIGEST_STRING_LENGTH];
60 	size_t len, hdrlen;
61 	int res, flags = 0;
62 
63 	if (blob1) {
64 		f1 = opentemp();
65 		if (f1 == NULL)
66 			return got_error(GOT_ERR_FILE_OPEN);
67 	} else
68 		flags |= D_EMPTY1;
69 
70 	if (blob2) {
71 		f2 = opentemp();
72 		if (f2 == NULL) {
73 			fclose(f1);
74 			return got_error(GOT_ERR_FILE_OPEN);
75 		}
76 	} else
77 		flags |= D_EMPTY2;
78 
79 	if (blob1 == NULL) {
80 		f1 = NULL;
81 	} else {
82 		hdrlen = blob1->hdrlen;
83 		do {
84 			err = got_object_blob_read_block(blob1, &len);
85 			if (err)
86 				goto done;
87 			/* Skip blob object header first time around. */
88 			fwrite(blob1->zb.outbuf + hdrlen, len - hdrlen, 1, f1);
89 			hdrlen = 0;
90 		} while (len != 0);
91 	}
92 
93 	hdrlen = blob2->hdrlen;
94 	do {
95 		err = got_object_blob_read_block(blob2, &len);
96 		if (err)
97 			goto done;
98 		/* Skip blob object header first time around. */
99 		fwrite(blob2->zb.outbuf + hdrlen, len - hdrlen, 1, f2);
100 		hdrlen = 0;
101 	} while (len != 0);
102 
103 	fflush(f1);
104 	fflush(f2);
105 	/* rewind(f1); */
106 	/* rewind(f2);*/
107 
108 	memset(&ds, 0, sizeof(ds));
109 	/* XXX should stat buffers be passed in args instead of ds? */
110 	ds.stb1.st_mode = S_IFREG;
111 	ds.stb1.st_size = blob1->zb.z.total_out;
112 	ds.stb1.st_mtime = 0; /* XXX */
113 
114 	ds.stb2.st_mode = S_IFREG;
115 	ds.stb2.st_size = blob2->zb.z.total_out;
116 	ds.stb2.st_mtime = 0; /* XXX */
117 
118 	memset(&args, 0, sizeof(args));
119 	args.diff_format = D_UNIFIED;
120 	args.label[0] = label1 ?
121 	    label1 : got_object_id_str(&blob1->id, hex1, sizeof(hex1));
122 	args.label[1] = label2 ?
123 	    label2 : got_object_id_str(&blob2->id, hex2, sizeof(hex2));
124 
125 	err = got_diffreg(&res, f1, f2, flags, &args, &ds, outfile);
126 done:
127 	fclose(f1);
128 	fclose(f2);
129 	return err;
130 }
131 
132 static const struct got_error *
133 match_entry_by_name(struct got_tree_entry **te, struct got_tree_entry *te1,
134     struct got_tree_object *tree2)
135 {
136 	*te = NULL;
137 	return NULL;
138 }
139 
140 static int
141 same_id(struct got_object_id *id1, struct got_object_id *id2)
142 {
143 	return (memcmp(id1->sha1, id2->sha1, SHA1_DIGEST_LENGTH) == 0);
144 }
145 
146 static const struct got_error *
147 diff_added_blob(struct got_object_id *id, struct got_repository *repo)
148 {
149 	const struct got_error *err;
150 	struct got_blob_object  *blob;
151 	struct got_object *obj;
152 
153 	err = got_object_open(&obj, repo, id);
154 	if (err)
155 		return err;
156 	err = got_object_blob_open(&blob, repo, obj, 512);
157 	if (err != NULL)
158 		return err;
159 
160 	return got_diff_blob(NULL, blob, NULL, NULL, stdout);
161 }
162 
163 static const struct got_error *
164 diff_modified_blob(struct got_object_id *id1, struct got_object_id *id2,
165     struct got_repository *repo)
166 {
167 	const struct got_error *err;
168 	struct got_object *obj1 = NULL;
169 	struct got_object *obj2 = NULL;
170 	struct got_blob_object *blob1 = NULL;
171 	struct got_blob_object *blob2 = NULL;
172 
173 	err = got_object_open(&obj1, repo, id1);
174 	if (err)
175 		return got_error(GOT_ERR_BAD_OBJ_HDR);
176 	if (obj1->type != GOT_OBJ_TYPE_BLOB) {
177 		err = got_error(GOT_ERR_OBJ_TYPE);
178 		goto done;
179 	}
180 
181 	err = got_object_open(&obj2, repo, id2);
182 	if (err) {
183 		err= got_error(GOT_ERR_BAD_OBJ_HDR);
184 		goto done;
185 	}
186 	if (obj2->type != GOT_OBJ_TYPE_BLOB) {
187 		err = got_error(GOT_ERR_BAD_OBJ_DATA);
188 		goto done;
189 	}
190 
191 	err = got_object_blob_open(&blob1, repo, obj1, 512);
192 	if (err != NULL) {
193 		err = got_error(GOT_ERR_FILE_OPEN);
194 		goto done;
195 	}
196 
197 	err = got_object_blob_open(&blob2, repo, obj2, 512);
198 	if (err != NULL) {
199 		err = got_error(GOT_ERR_FILE_OPEN);
200 		goto done;
201 	}
202 
203 	err = got_diff_blob(blob1, blob2, NULL, NULL, stdout);
204 
205 done:
206 	got_object_close(obj1);
207 	got_object_close(obj2);
208 	got_object_blob_close(blob1);
209 	got_object_blob_close(blob2);
210 	return err;
211 }
212 
213 static const struct got_error *
214 diff_deleted_blob(struct got_object_id *id)
215 {
216 	return NULL;
217 }
218 
219 static const struct got_error *
220 diff_added_tree(struct got_object_id *id)
221 {
222 	return NULL;
223 }
224 
225 static const struct got_error *
226 diff_modified_tree(struct got_object_id *id1, struct got_object_id *id2)
227 {
228 	return NULL;
229 }
230 
231 static const struct got_error *
232 diff_deleted_tree(struct got_object_id *id)
233 {
234 	return NULL;
235 }
236 
237 static const struct got_error *
238 diff_kind_mismatch(struct got_object_id *id1, struct got_object_id *id2)
239 {
240 	return NULL;
241 }
242 
243 static const struct got_error *
244 diff_entry_old_new(struct got_tree_entry *te1, struct got_tree_object *tree2,
245     struct got_repository *repo)
246 {
247 	const struct got_error *err;
248 	struct got_tree_entry *te2;
249 
250 	err = match_entry_by_name(&te2, te1, tree2);
251 	if (err)
252 		return err;
253 	if (te2 == NULL) {
254 		if (S_ISDIR(te1->mode))
255 			return diff_deleted_tree(&te1->id);
256 		return diff_deleted_blob(&te1->id);
257 	}
258 
259 	if (S_ISDIR(te1->mode) && S_ISDIR(te2->mode)) {
260 		if (!same_id(&te1->id, &te2->id))
261 			return diff_modified_tree(&te1->id, &te2->id);
262 	} else if (S_ISREG(te1->mode) && S_ISREG(te2->mode)) {
263 		if (!same_id(&te1->id, &te2->id))
264 			return diff_modified_blob(&te1->id, &te2->id, repo);
265 	}
266 
267 	return diff_kind_mismatch(&te1->id, &te2->id);
268 }
269 
270 static const struct got_error *
271 diff_entry_new_old(struct got_tree_entry *te2, struct got_tree_object *tree1,
272     struct got_repository *repo)
273 {
274 	const struct got_error *err;
275 	struct got_tree_entry *te1;
276 
277 	err = match_entry_by_name(&te1, te2, tree1);
278 	if (err)
279 		return err;
280 	if (te1 != NULL) /* handled by diff_entry_old_new() */
281 		return NULL;
282 
283 	if (S_ISDIR(te2->mode))
284 		return diff_added_tree(&te2->id);
285 	return diff_added_blob(&te2->id, repo);
286 }
287 
288 const struct got_error *
289 got_diff_tree(struct got_tree_object *tree1, struct got_tree_object *tree2,
290     struct got_repository *repo)
291 {
292 	const struct got_error *err = NULL;
293 	struct got_tree_entry *te1;
294 	struct got_tree_entry *te2;
295 
296 	if (tree1->nentries == 0 && tree2->nentries == 0)
297 		return NULL;
298 
299 	te1 = SIMPLEQ_FIRST(&tree1->entries);
300 	te2 = SIMPLEQ_FIRST(&tree2->entries);
301 
302 	do {
303 		if (te1) {
304 			err = diff_entry_old_new(te1, tree2, repo);
305 			if (err)
306 				break;
307 		}
308 
309 		if (te2) {
310 			err = diff_entry_new_old(te2, tree1, repo);
311 			if (err)
312 				break;
313 		}
314 
315 		if (te1)
316 			te1 = SIMPLEQ_NEXT(te1, entry);
317 		if (te2)
318 			te2 = SIMPLEQ_NEXT(te2, entry);
319 	} while (te1 || te2);
320 
321 	return err;
322 }