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;
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 (blob)
164 		got_object_blob_close(blob);
165 	return err;
166 }
167 
168 static const struct got_error *
169 blame_close(struct got_blame *blame)
170 {
171 	const struct got_error *err = NULL;
172 
173 	if (blame->f && fclose(blame->f) != 0)
174 		err = got_error_from_errno("fclose");
175 	free(blame->lines);
176 	free(blame);
177 	return err;
178 }
179 
180 static const struct got_error *
181 blame_open(struct got_blame **blamep, const char *path,
182     struct got_object_id *start_commit_id, struct got_repository *repo,
183     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
184     void *arg)
185 {
186 	const struct got_error *err = NULL;
187 	struct got_object *obj = NULL;
188 	struct got_object_id *obj_id = NULL;
189 	struct got_blob_object *blob = NULL;
190 	struct got_blame *blame = NULL;
191 	struct got_object_id *id = NULL, *pid = NULL;
192 	int lineno;
193 	struct got_commit_graph *graph = NULL;
194 
195 	*blamep = NULL;
196 
197 	err = got_object_id_by_path(&obj_id, repo, start_commit_id, path);
198 	if (err)
199 		return err;
200 
201 	err = got_object_open(&obj, repo, obj_id);
202 	if (err)
203 		goto done;
204 
205 	if (obj->type != GOT_OBJ_TYPE_BLOB) {
206 		err = got_error(GOT_ERR_OBJ_TYPE);
207 		goto done;
208 	}
209 
210 	err = got_object_blob_open(&blob, repo, obj, 8192);
211 	if (err)
212 		goto done;
213 
214 	blame = calloc(1, sizeof(*blame));
215 	if (blame == NULL)
216 		return got_error_from_errno("calloc");
217 
218 	blame->f = got_opentemp();
219 	if (blame->f == NULL) {
220 		err = got_error_from_errno("got_opentemp");
221 		goto done;
222 	}
223 	err = got_object_blob_dump_to_file(&blame->filesize, &blame->nlines,
224 	   &blame->line_offsets, blame->f, blob);
225 	if (err || blame->nlines == 0)
226 		goto done;
227 
228 	/* Don't include \n at EOF in the blame line count. */
229 	if (blame->line_offsets[blame->nlines - 1] == blame->filesize)
230 		blame->nlines--;
231 
232 	blame->lines = calloc(blame->nlines, sizeof(*blame->lines));
233 	if (blame->lines == NULL) {
234 		err = got_error_from_errno("calloc");
235 		goto done;
236 	}
237 
238 	err = got_commit_graph_open(&graph, start_commit_id, path, 1, repo);
239 	if (err)
240 		return err;
241 	err = got_commit_graph_iter_start(graph, start_commit_id, repo);
242 	if (err)
243 		goto done;
244 	id = start_commit_id;
245 	for (;;) {
246 		err = got_commit_graph_iter_next(&pid, graph);
247 		if (err) {
248 			if (err->code == GOT_ERR_ITER_COMPLETED) {
249 				err = NULL;
250 				break;
251 			}
252 			if (err->code != GOT_ERR_ITER_NEED_MORE)
253 				break;
254 			err = got_commit_graph_fetch_commits(graph, 1, repo);
255 			if (err)
256 				break;
257 			continue;
258 		}
259 		if (pid) {
260 			err = blame_commit(blame, pid, id, path, repo, cb, arg);
261 			if (err) {
262 				if (err->code == GOT_ERR_ITER_COMPLETED)
263 					err = NULL;
264 				break;
265 			}
266 			if (blame->nannotated == blame->nlines)
267 				break;
268 		}
269 		id = pid;
270 	}
271 
272 	if (id && blame->nannotated < blame->nlines) {
273 		/* Annotate remaining non-annotated lines with last commit. */
274 		for (lineno = 1; lineno <= blame->nlines; lineno++) {
275 			err = annotate_line(blame, lineno, id, cb, arg);
276 			if (err)
277 				goto done;
278 		}
279 	}
280 
281 done:
282 	if (graph)
283 		got_commit_graph_close(graph);
284 	free(obj_id);
285 	if (obj)
286 		got_object_close(obj);
287 	if (blob)
288 		got_object_blob_close(blob);
289 	if (err) {
290 		if (blame)
291 			blame_close(blame);
292 	} else
293 		*blamep = blame;
294 
295 	return err;
296 }
297 
298 const struct got_error *
299 got_blame(const char *path, struct got_object_id *commit_id,
300     struct got_repository *repo,
301     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
302     void *arg)
303 {
304 	const struct got_error *err = NULL, *close_err = NULL;
305 	struct got_blame *blame;
306 	char *abspath;
307 
308 	if (asprintf(&abspath, "%s%s", path[0] == '/' ? "" : "/", path) == -1)
309 		return got_error_from_errno2("asprintf", path);
310 
311 	err = blame_open(&blame, abspath, commit_id, repo, cb, arg);
312 	free(abspath);
313 	if (blame)
314 		close_err = blame_close(blame);
315 	return err ? err : close_err;
316 }