1 /*
2  * Copyright (c) 2018, 2019 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 <sha1.h>
21 #include <string.h>
22 #include <stdio.h>
23 #include <stdlib.h>
24 #include <time.h>
25 #include <util.h>
26 #include <zlib.h>
27 
28 #include "got_error.h"
29 #include "got_object.h"
30 #include "got_blame.h"
31 #include "got_opentemp.h"
32 
33 #include "got_lib_inflate.h"
34 #include "got_lib_delta.h"
35 #include "got_lib_object.h"
36 #include "got_lib_diff.h"
37 #include "got_commit_graph.h"
38 
39 struct got_blame_line {
40 	int annotated;
41 	struct got_object_id id;
42 };
43 
44 struct got_blame {
45 	FILE *f;
46 	size_t filesize;
47 	int nlines;
48 	int nannotated;
49 	struct got_blame_line *lines; /* one per line */
50 	off_t *line_offsets;		/* one per line */
51 	int ncommits;
52 };
53 
54 static const struct got_error *
55 annotate_line(struct got_blame *blame, int lineno, struct got_object_id *id,
56     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
57     void *arg)
58 {
59 	const struct got_error *err = NULL;
60 	struct got_blame_line *line;
61 
62 	if (lineno < 1 || lineno > blame->nlines)
63 		return NULL;
64 	
65 	line = &blame->lines[lineno - 1];
66 	if (line->annotated)
67 		return NULL;
68 
69 	memcpy(&line->id, id, sizeof(line->id));
70 	line->annotated = 1;
71 	blame->nannotated++;
72 	if (cb)
73 		err = cb(arg, blame->nlines, lineno, id);
74 	return err;
75 }
76 
77 static const struct got_error *
78 blame_changes(struct got_blame *blame, struct got_diff_changes *changes,
79     struct got_object_id *commit_id,
80     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
81     void *arg)
82 {
83 	const struct got_error *err = NULL;
84 	struct got_diff_change *change;
85 
86 	SIMPLEQ_FOREACH(change, &changes->entries, entry) {
87 		int c = change->cv.c;
88 		int d = change->cv.d;
89 		int new_lineno = (c < d ? c : d);
90 		int new_length = (c < d ? d - c + 1 : (c == d ? 1 : 0));
91 		int ln;
92 
93 		for (ln = new_lineno; ln < new_lineno + new_length; ln++) {
94 			err = annotate_line(blame, ln, commit_id, cb, arg);
95 			if (err)
96 				return err;
97 			if (blame->nlines == blame->nannotated)
98 				break;
99 		}
100 	}
101 
102 	return NULL;
103 }
104 
105 static const struct got_error *
106 blame_commit(struct got_blame *blame, struct got_object_id *parent_id,
107     struct got_object_id *id, const char *path, struct got_repository *repo,
108     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
109     void *arg)
110 {
111 	const struct got_error *err = NULL;
112 	struct got_object *obj = NULL, *pobj = NULL;
113 	struct got_object_id *obj_id = NULL;
114 	struct got_commit_object *commit = NULL;
115 	struct got_blob_object *blob = NULL;
116 	struct got_diff_changes *changes = NULL;
117 
118 	err = got_object_open_as_commit(&commit, repo, parent_id);
119 	if (err)
120 		return err;
121 
122 	err = got_object_id_by_path(&obj_id, repo, parent_id, path);
123 	if (err) {
124 		if (err->code == GOT_ERR_NO_TREE_ENTRY)
125 			err = NULL;
126 		goto done;
127 	}
128 
129 	err = got_object_open(&obj, repo, obj_id);
130 	if (err)
131 		goto done;
132 
133 	if (obj->type != GOT_OBJ_TYPE_BLOB) {
134 		err = got_error(GOT_ERR_OBJ_TYPE);
135 		goto done;
136 	}
137 
138 	err = got_object_blob_open(&blob, repo, obj, 8192);
139 	if (err)
140 		goto done;
141 
142 	if (fseek(blame->f, 0L, SEEK_SET) == -1) {
143 		err = got_ferror(blame->f, GOT_ERR_IO);
144 		goto done;
145 	}
146 
147 	err = got_diff_blob_file_lines_changed(&changes, blob, blame->f,
148 	    blame->filesize);
149 	if (err)
150 		goto done;
151 
152 	if (changes) {
153 		err = blame_changes(blame, changes, id, cb, arg);
154 		got_diff_free_changes(changes);
155 	} else if (cb)
156 		err = cb(arg, blame->nlines, -1, id);
157 done:
158 	if (commit)
159 		got_object_commit_close(commit);
160 	free(obj_id);
161 	if (obj)
162 		got_object_close(obj);
163 	if (pobj)
164 		got_object_close(pobj);
165 	if (blob)
166 		got_object_blob_close(blob);
167 	return err;
168 }
169 
170 static const struct got_error *
171 blame_close(struct got_blame *blame)
172 {
173 	const struct got_error *err = NULL;
174 
175 	if (blame->f && fclose(blame->f) != 0)
176 		err = got_error_from_errno("fclose");
177 	free(blame->lines);
178 	free(blame);
179 	return err;
180 }
181 
182 static const struct got_error *
183 blame_open(struct got_blame **blamep, const char *path,
184     struct got_object_id *start_commit_id, struct got_repository *repo,
185     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
186     void *arg)
187 {
188 	const struct got_error *err = NULL;
189 	struct got_object *obj = NULL;
190 	struct got_object_id *obj_id = NULL;
191 	struct got_blob_object *blob = NULL;
192 	struct got_blame *blame = NULL;
193 	struct got_object_id *id = NULL, *pid = NULL;
194 	int lineno;
195 	struct got_commit_graph *graph = NULL;
196 
197 	*blamep = NULL;
198 
199 	err = got_object_id_by_path(&obj_id, repo, start_commit_id, path);
200 	if (err)
201 		return err;
202 
203 	err = got_object_open(&obj, repo, obj_id);
204 	if (err)
205 		goto done;
206 
207 	if (obj->type != GOT_OBJ_TYPE_BLOB) {
208 		err = got_error(GOT_ERR_OBJ_TYPE);
209 		goto done;
210 	}
211 
212 	err = got_object_blob_open(&blob, repo, obj, 8192);
213 	if (err)
214 		goto done;
215 
216 	blame = calloc(1, sizeof(*blame));
217 	if (blame == NULL)
218 		return got_error_from_errno("calloc");
219 
220 	blame->f = got_opentemp();
221 	if (blame->f == NULL) {
222 		err = got_error_from_errno("got_opentemp");
223 		goto done;
224 	}
225 	err = got_object_blob_dump_to_file(&blame->filesize, &blame->nlines,
226 	   &blame->line_offsets, blame->f, blob);
227 	if (err || blame->nlines == 0)
228 		goto done;
229 
230 	/* Don't include \n at EOF in the blame line count. */
231 	if (blame->line_offsets[blame->nlines - 1] == blame->filesize)
232 		blame->nlines--;
233 
234 	blame->lines = calloc(blame->nlines, sizeof(*blame->lines));
235 	if (blame->lines == NULL) {
236 		err = got_error_from_errno("calloc");
237 		goto done;
238 	}
239 
240 	err = got_commit_graph_open(&graph, start_commit_id, path, 1, repo);
241 	if (err)
242 		return err;
243 	err = got_commit_graph_iter_start(graph, start_commit_id, repo);
244 	if (err)
245 		goto done;
246 	id = start_commit_id;
247 	for (;;) {
248 		err = got_commit_graph_iter_next(&pid, graph);
249 		if (err) {
250 			if (err->code == GOT_ERR_ITER_COMPLETED) {
251 				err = NULL;
252 				break;
253 			}
254 			if (err->code != GOT_ERR_ITER_NEED_MORE)
255 				break;
256 			err = got_commit_graph_fetch_commits(graph, 1, repo);
257 			if (err)
258 				break;
259 			continue;
260 		}
261 		if (pid) {
262 			err = blame_commit(blame, pid, id, path, repo, cb, arg);
263 			if (err) {
264 				if (err->code == GOT_ERR_ITER_COMPLETED)
265 					err = NULL;
266 				break;
267 			}
268 			if (blame->nannotated == blame->nlines)
269 				break;
270 		}
271 		id = pid;
272 	}
273 
274 	if (id && blame->nannotated < blame->nlines) {
275 		/* Annotate remaining non-annotated lines with last commit. */
276 		for (lineno = 1; lineno <= blame->nlines; lineno++) {
277 			err = annotate_line(blame, lineno, id, cb, arg);
278 			if (err)
279 				goto done;
280 		}
281 	}
282 
283 done:
284 	if (graph)
285 		got_commit_graph_close(graph);
286 	free(obj_id);
287 	if (obj)
288 		got_object_close(obj);
289 	if (blob)
290 		got_object_blob_close(blob);
291 	if (err) {
292 		if (blame)
293 			blame_close(blame);
294 	} else
295 		*blamep = blame;
296 
297 	return err;
298 }
299 
300 const struct got_error *
301 got_blame(const char *path, struct got_object_id *commit_id,
302     struct got_repository *repo,
303     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
304     void *arg)
305 {
306 	const struct got_error *err = NULL, *close_err = NULL;
307 	struct got_blame *blame;
308 	char *abspath;
309 
310 	if (asprintf(&abspath, "%s%s", path[0] == '/' ? "" : "/", path) == -1)
311 		return got_error_from_errno2("asprintf", path);
312 
313 	err = blame_open(&blame, abspath, commit_id, repo, cb, arg);
314 	free(abspath);
315 	if (blame)
316 		close_err = blame_close(blame);
317 	return err ? err : close_err;
318 }