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_cancel.h"
31 #include "got_blame.h"
32 #include "got_commit_graph.h"
33 #include "got_opentemp.h"
34 
35 #include "got_lib_inflate.h"
36 #include "got_lib_delta.h"
37 #include "got_lib_object.h"
38 #include "got_lib_diff.h"
39 
40 struct got_blame_line {
41 	int annotated;
42 	struct got_object_id id;
43 };
44 
45 struct got_blame {
46 	FILE *f;
47 	size_t filesize;
48 	int nlines;
49 	int nannotated;
50 	struct got_blame_line *lines; /* one per line */
51 	off_t *line_offsets;		/* one per line */
52 	int ncommits;
53 };
54 
55 static const struct got_error *
56 annotate_line(struct got_blame *blame, int lineno, struct got_object_id *id,
57     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
58     void *arg)
59 {
60 	const struct got_error *err = NULL;
61 	struct got_blame_line *line;
62 
63 	if (lineno < 1 || lineno > blame->nlines)
64 		return NULL;
65 	
66 	line = &blame->lines[lineno - 1];
67 	if (line->annotated)
68 		return NULL;
69 
70 	memcpy(&line->id, id, sizeof(line->id));
71 	line->annotated = 1;
72 	blame->nannotated++;
73 	if (cb)
74 		err = cb(arg, blame->nlines, lineno, id);
75 	return err;
76 }
77 
78 static const struct got_error *
79 blame_changes(struct got_blame *blame, struct got_diff_changes *changes,
80     struct got_object_id *commit_id,
81     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
82     void *arg)
83 {
84 	const struct got_error *err = NULL;
85 	struct got_diff_change *change;
86 
87 	SIMPLEQ_FOREACH(change, &changes->entries, entry) {
88 		int c = change->cv.c;
89 		int d = change->cv.d;
90 		int new_lineno = (c < d ? c : d);
91 		int new_length = (c < d ? d - c + 1 : (c == d ? 1 : 0));
92 		int ln;
93 
94 		for (ln = new_lineno; ln < new_lineno + new_length; ln++) {
95 			err = annotate_line(blame, ln, commit_id, cb, arg);
96 			if (err)
97 				return err;
98 			if (blame->nlines == blame->nannotated)
99 				break;
100 		}
101 	}
102 
103 	return NULL;
104 }
105 
106 static const struct got_error *
107 blame_commit(struct got_blame *blame, struct got_object_id *parent_id,
108     struct got_object_id *id, const char *path, struct got_repository *repo,
109     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
110     void *arg)
111 {
112 	const struct got_error *err = NULL;
113 	struct got_object *obj = NULL;
114 	struct got_object_id *obj_id = NULL;
115 	struct got_commit_object *commit = NULL;
116 	struct got_blob_object *blob = NULL;
117 	struct got_diff_changes *changes = NULL;
118 
119 	err = got_object_open_as_commit(&commit, repo, parent_id);
120 	if (err)
121 		return err;
122 
123 	err = got_object_id_by_path(&obj_id, repo, parent_id, path);
124 	if (err) {
125 		if (err->code == GOT_ERR_NO_TREE_ENTRY)
126 			err = NULL;
127 		goto done;
128 	}
129 
130 	err = got_object_open(&obj, repo, obj_id);
131 	if (err)
132 		goto done;
133 
134 	if (obj->type != GOT_OBJ_TYPE_BLOB) {
135 		err = got_error(GOT_ERR_OBJ_TYPE);
136 		goto done;
137 	}
138 
139 	err = got_object_blob_open(&blob, repo, obj, 8192);
140 	if (err)
141 		goto done;
142 
143 	if (fseek(blame->f, 0L, SEEK_SET) == -1) {
144 		err = got_ferror(blame->f, GOT_ERR_IO);
145 		goto done;
146 	}
147 
148 	err = got_diff_blob_file_lines_changed(&changes, blob, blame->f,
149 	    blame->filesize);
150 	if (err)
151 		goto done;
152 
153 	if (changes) {
154 		err = blame_changes(blame, changes, id, cb, arg);
155 		got_diff_free_changes(changes);
156 	} else if (cb)
157 		err = cb(arg, blame->nlines, -1, id);
158 done:
159 	if (commit)
160 		got_object_commit_close(commit);
161 	free(obj_id);
162 	if (obj)
163 		got_object_close(obj);
164 	if (blob)
165 		got_object_blob_close(blob);
166 	return err;
167 }
168 
169 static const struct got_error *
170 blame_close(struct got_blame *blame)
171 {
172 	const struct got_error *err = NULL;
173 
174 	if (blame->f && fclose(blame->f) != 0)
175 		err = got_error_from_errno("fclose");
176 	free(blame->lines);
177 	free(blame);
178 	return err;
179 }
180 
181 static const struct got_error *
182 blame_open(struct got_blame **blamep, const char *path,
183     struct got_object_id *start_commit_id, struct got_repository *repo,
184     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
185     void *arg, got_cancel_cb cancel_cb, void *cancel_arg)
186 {
187 	const struct got_error *err = NULL;
188 	struct got_object *obj = NULL;
189 	struct got_object_id *obj_id = NULL;
190 	struct got_blob_object *blob = NULL;
191 	struct got_blame *blame = NULL;
192 	struct got_object_id *id = NULL, *pid = NULL;
193 	int lineno;
194 	struct got_commit_graph *graph = NULL;
195 
196 	*blamep = NULL;
197 
198 	err = got_object_id_by_path(&obj_id, repo, start_commit_id, path);
199 	if (err)
200 		return err;
201 
202 	err = got_object_open(&obj, repo, obj_id);
203 	if (err)
204 		goto done;
205 
206 	if (obj->type != GOT_OBJ_TYPE_BLOB) {
207 		err = got_error(GOT_ERR_OBJ_TYPE);
208 		goto done;
209 	}
210 
211 	err = got_object_blob_open(&blob, repo, obj, 8192);
212 	if (err)
213 		goto done;
214 
215 	blame = calloc(1, sizeof(*blame));
216 	if (blame == NULL)
217 		return got_error_from_errno("calloc");
218 
219 	blame->f = got_opentemp();
220 	if (blame->f == NULL) {
221 		err = got_error_from_errno("got_opentemp");
222 		goto done;
223 	}
224 	err = got_object_blob_dump_to_file(&blame->filesize, &blame->nlines,
225 	   &blame->line_offsets, blame->f, blob);
226 	if (err || blame->nlines == 0)
227 		goto done;
228 
229 	/* Don't include \n at EOF in the blame line count. */
230 	if (blame->line_offsets[blame->nlines - 1] == blame->filesize)
231 		blame->nlines--;
232 
233 	blame->lines = calloc(blame->nlines, sizeof(*blame->lines));
234 	if (blame->lines == NULL) {
235 		err = got_error_from_errno("calloc");
236 		goto done;
237 	}
238 
239 	err = got_commit_graph_open(&graph, start_commit_id, path, 1, repo);
240 	if (err)
241 		return err;
242 	err = got_commit_graph_iter_start(graph, start_commit_id, repo,
243 	    cancel_cb, cancel_arg);
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 			    cancel_cb, cancel_arg);
258 			if (err)
259 				break;
260 			continue;
261 		}
262 		if (pid) {
263 			err = blame_commit(blame, pid, id, path, repo, cb, arg);
264 			if (err) {
265 				if (err->code == GOT_ERR_ITER_COMPLETED)
266 					err = NULL;
267 				break;
268 			}
269 			if (blame->nannotated == blame->nlines)
270 				break;
271 		}
272 		id = pid;
273 	}
274 
275 	if (id && blame->nannotated < blame->nlines) {
276 		/* Annotate remaining non-annotated lines with last commit. */
277 		for (lineno = 1; lineno <= blame->nlines; lineno++) {
278 			err = annotate_line(blame, lineno, id, cb, arg);
279 			if (err)
280 				goto done;
281 		}
282 	}
283 
284 done:
285 	if (graph)
286 		got_commit_graph_close(graph);
287 	free(obj_id);
288 	if (obj)
289 		got_object_close(obj);
290 	if (blob)
291 		got_object_blob_close(blob);
292 	if (err) {
293 		if (blame)
294 			blame_close(blame);
295 	} else
296 		*blamep = blame;
297 
298 	return err;
299 }
300 
301 const struct got_error *
302 got_blame(const char *path, struct got_object_id *commit_id,
303     struct got_repository *repo,
304     const struct got_error *(*cb)(void *, int, int, struct got_object_id *),
305     void *arg, got_cancel_cb cancel_cb, void* cancel_arg)
306 {
307 	const struct got_error *err = NULL, *close_err = NULL;
308 	struct got_blame *blame;
309 	char *abspath;
310 
311 	if (asprintf(&abspath, "%s%s", path[0] == '/' ? "" : "/", path) == -1)
312 		return got_error_from_errno2("asprintf", path);
313 
314 	err = blame_open(&blame, abspath, commit_id, repo, cb, arg,
315 	    cancel_cb, cancel_arg);
316 	free(abspath);
317 	if (blame)
318 		close_err = blame_close(blame);
319 	return err ? err : close_err;
320 }