Infrastructure to write revisions in fast-export format
repo_tree maintains the exporter's state and provides a facility to to call fast_export, which writes objects to stdout suitable for consumption by fast-import. The exported functions roughly correspond to Subversion FS operations. . repo_add, repo_modify, repo_copy, repo_replace, and repo_delete update the current commit, based roughly on the corresponding Subversion FS operation. . repo_commit calls out to fast_export to write the current commit to the fast-import stream in stdout. . repo_diff is used by the fast_export module to write the changes for a commit. . repo_reset erases the exporter's state, so valgrind can be happy. [rr: squelched compiler warnings] [jn: removed support for maintaining state on-disk, though we may want to add it back later] Signed-off-by: David Barr <david.barr@cordelta.com> Signed-off-by: Ramkumar Ramachandra <artagnon@gmail.com> Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
David Barr committed
Aug 9, 2010 at 17:48 UTC
c0e6c23dca84227167a6fe1077503ddf32208919
5 files changed
+443
-2
Makefile
+3
-2
@@ -1744,7 +1744,8 @@ ifndef NO_CURL
1744
endif
1745
XDIFF_OBJS = xdiff/xdiffi.o xdiff/xprepare.o xdiff/xutils.o xdiff/xemit.o \
1746
xdiff/xmerge.o xdiff/xpatience.o
1747
-VCSSVN_OBJS = vcs-svn/string_pool.o vcs-svn/line_buffer.o
1747
+VCSSVN_OBJS = vcs-svn/string_pool.o vcs-svn/line_buffer.o \
1748
+ vcs-svn/repo_tree.o vcs-svn/fast_export.o
1749
OBJECTS := $(GIT_OBJS) $(XDIFF_OBJS) $(VCSSVN_OBJS)
1750
1751
dep_files := $(foreach f,$(OBJECTS),$(dir $f).depend/$(notdir $f).d)
@@ -1870,7 +1871,7 @@ xdiff-interface.o $(XDIFF_OBJS): \
1871
1872
$(VCSSVN_OBJS): \
1873
vcs-svn/obj_pool.h vcs-svn/trp.h vcs-svn/string_pool.h \
1873
- vcs-svn/line_buffer.h
1874
+ vcs-svn/line_buffer.h vcs-svn/repo_tree.h vcs-svn/fast_export.h
1875
endif
1876
1877
exec_cmd.s exec_cmd.o: EXTRA_CPPFLAGS = \
vcs-svn/fast_export.c
new
+74
@@ -0,0 +1,74 @@
1
+/*
2
+ * Licensed under a two-clause BSD-style license.
3
+ * See LICENSE for details.
4
+ */
5
+
6
+#include "git-compat-util.h"
7
+#include "fast_export.h"
8
+#include "line_buffer.h"
9
+#include "repo_tree.h"
10
+#include "string_pool.h"
11
+
12
+#define MAX_GITSVN_LINE_LEN 4096
13
+
14
+static uint32_t first_commit_done;
15
+
16
+void fast_export_delete(uint32_t depth, uint32_t *path)
17
+{
18
+ putchar('D');
19
+ putchar(' ');
20
+ pool_print_seq(depth, path, '/', stdout);
21
+ putchar('\n');
22
+}
23
+
24
+void fast_export_modify(uint32_t depth, uint32_t *path, uint32_t mode,
25
+ uint32_t mark)
26
+{
27
+ /* Mode must be 100644, 100755, 120000, or 160000. */
28
+ printf("M %06o :%d ", mode, mark);
29
+ pool_print_seq(depth, path, '/', stdout);
30
+ putchar('\n');
31
+}
32
+
33
+static char gitsvnline[MAX_GITSVN_LINE_LEN];
34
+void fast_export_commit(uint32_t revision, uint32_t author, char *log,
35
+ uint32_t uuid, uint32_t url,
36
+ unsigned long timestamp)
37
+{
38
+ if (!log)
39
+ log = "";
40
+ if (~uuid && ~url) {
41
+ snprintf(gitsvnline, MAX_GITSVN_LINE_LEN, "\n\ngit-svn-id: %s@%d %s\n",
42
+ pool_fetch(url), revision, pool_fetch(uuid));
43
+ } else {
44
+ *gitsvnline = '\0';
45
+ }
46
+ printf("commit refs/heads/master\n");
47
+ printf("committer %s <%s@%s> %ld +0000\n",
48
+ ~author ? pool_fetch(author) : "nobody",
49
+ ~author ? pool_fetch(author) : "nobody",
50
+ ~uuid ? pool_fetch(uuid) : "local", timestamp);
51
+ printf("data %zd\n%s%s\n",
52
+ strlen(log) + strlen(gitsvnline), log, gitsvnline);
53
+ if (!first_commit_done) {
54
+ if (revision > 1)
55
+ printf("from refs/heads/master^0\n");
56
+ first_commit_done = 1;
57
+ }
58
+ repo_diff(revision - 1, revision);
59
+ fputc('\n', stdout);
60
+
61
+ printf("progress Imported commit %d.\n\n", revision);
62
+}
63
+
64
+void fast_export_blob(uint32_t mode, uint32_t mark, uint32_t len)
65
+{
66
+ if (mode == REPO_MODE_LNK) {
67
+ /* svn symlink blobs start with "link " */
68
+ buffer_skip_bytes(5);
69
+ len -= 5;
70
+ }
71
+ printf("blob\nmark :%d\ndata %d\n", mark, len);
72
+ buffer_copy_bytes(len);
73
+ fputc('\n', stdout);
74
+}
vcs-svn/fast_export.h
new
+11
@@ -0,0 +1,11 @@
1
+#ifndef FAST_EXPORT_H_
2
+#define FAST_EXPORT_H_
3
+
4
+void fast_export_delete(uint32_t depth, uint32_t *path);
5
+void fast_export_modify(uint32_t depth, uint32_t *path, uint32_t mode,
6
+ uint32_t mark);
7
+void fast_export_commit(uint32_t revision, uint32_t author, char *log,
8
+ uint32_t uuid, uint32_t url, unsigned long timestamp);
9
+void fast_export_blob(uint32_t mode, uint32_t mark, uint32_t len);
10
+
11
+#endif
vcs-svn/repo_tree.c
new
+329
@@ -0,0 +1,329 @@
1
+/*
2
+ * Licensed under a two-clause BSD-style license.
3
+ * See LICENSE for details.
4
+ */
5
+
6
+#include "git-compat-util.h"
7
+
8
+#include "string_pool.h"
9
+#include "repo_tree.h"
10
+#include "obj_pool.h"
11
+#include "fast_export.h"
12
+
13
+#include "trp.h"
14
+
15
+struct repo_dirent {
16
+ uint32_t name_offset;
17
+ struct trp_node children;
18
+ uint32_t mode;
19
+ uint32_t content_offset;
20
+};
21
+
22
+struct repo_dir {
23
+ struct trp_root entries;
24
+};
25
+
26
+struct repo_commit {
27
+ uint32_t root_dir_offset;
28
+};
29
+
30
+/* Memory pools for commit, dir and dirent */
31
+obj_pool_gen(commit, struct repo_commit, 4096)
32
+obj_pool_gen(dir, struct repo_dir, 4096)
33
+obj_pool_gen(dirent, struct repo_dirent, 4096)
34
+
35
+static uint32_t active_commit;
36
+static uint32_t mark;
37
+
38
+static int repo_dirent_name_cmp(const void *a, const void *b);
39
+
40
+/* Treap for directory entries */
41
+trp_gen(static, dirent_, struct repo_dirent, children, dirent, repo_dirent_name_cmp);
42
+
43
+uint32_t next_blob_mark(void)
44
+{
45
+ return mark++;
46
+}
47
+
48
+static struct repo_dir *repo_commit_root_dir(struct repo_commit *commit)
49
+{
50
+ return dir_pointer(commit->root_dir_offset);
51
+}
52
+
53
+static struct repo_dirent *repo_first_dirent(struct repo_dir *dir)
54
+{
55
+ return dirent_first(&dir->entries);
56
+}
57
+
58
+static int repo_dirent_name_cmp(const void *a, const void *b)
59
+{
60
+ const struct repo_dirent *dirent1 = a, *dirent2 = b;
61
+ uint32_t a_offset = dirent1->name_offset;
62
+ uint32_t b_offset = dirent2->name_offset;
63
+ return (a_offset > b_offset) - (a_offset < b_offset);
64
+}
65
+
66
+static int repo_dirent_is_dir(struct repo_dirent *dirent)
67
+{
68
+ return dirent != NULL && dirent->mode == REPO_MODE_DIR;
69
+}
70
+
71
+static struct repo_dir *repo_dir_from_dirent(struct repo_dirent *dirent)
72
+{
73
+ if (!repo_dirent_is_dir(dirent))
74
+ return NULL;
75
+ return dir_pointer(dirent->content_offset);
76
+}
77
+
78
+static struct repo_dir *repo_clone_dir(struct repo_dir *orig_dir)
79
+{
80
+ uint32_t orig_o, new_o;
81
+ orig_o = dir_offset(orig_dir);
82
+ if (orig_o >= dir_pool.committed)
83
+ return orig_dir;
84
+ new_o = dir_alloc(1);
85
+ orig_dir = dir_pointer(orig_o);
86
+ *dir_pointer(new_o) = *orig_dir;
87
+ return dir_pointer(new_o);
88
+}
89
+
90
+static struct repo_dirent *repo_read_dirent(uint32_t revision, uint32_t *path)
91
+{
92
+ uint32_t name = 0;
93
+ struct repo_dirent *key = dirent_pointer(dirent_alloc(1));
94
+ struct repo_dir *dir = NULL;
95
+ struct repo_dirent *dirent = NULL;
96
+ dir = repo_commit_root_dir(commit_pointer(revision));
97
+ while (~(name = *path++)) {
98
+ key->name_offset = name;
99
+ dirent = dirent_search(&dir->entries, key);
100
+ if (dirent == NULL || !repo_dirent_is_dir(dirent))
101
+ break;
102
+ dir = repo_dir_from_dirent(dirent);
103
+ }
104
+ dirent_free(1);
105
+ return dirent;
106
+}
107
+
108
+static void repo_write_dirent(uint32_t *path, uint32_t mode,
109
+ uint32_t content_offset, uint32_t del)
110
+{
111
+ uint32_t name, revision, dir_o = ~0, parent_dir_o = ~0;
112
+ struct repo_dir *dir;
113
+ struct repo_dirent *key;
114
+ struct repo_dirent *dirent = NULL;
115
+ revision = active_commit;
116
+ dir = repo_commit_root_dir(commit_pointer(revision));
117
+ dir = repo_clone_dir(dir);
118
+ commit_pointer(revision)->root_dir_offset = dir_offset(dir);
119
+ while (~(name = *path++)) {
120
+ parent_dir_o = dir_offset(dir);
121
+
122
+ key = dirent_pointer(dirent_alloc(1));
123
+ key->name_offset = name;
124
+
125
+ dirent = dirent_search(&dir->entries, key);
126
+ if (dirent == NULL)
127
+ dirent = key;
128
+ else
129
+ dirent_free(1);
130
+
131
+ if (dirent == key) {
132
+ dirent->mode = REPO_MODE_DIR;
133
+ dirent->content_offset = 0;
134
+ dirent_insert(&dir->entries, dirent);
135
+ }
136
+
137
+ if (dirent_offset(dirent) < dirent_pool.committed) {
138
+ dir_o = repo_dirent_is_dir(dirent) ?
139
+ dirent->content_offset : ~0;
140
+ dirent_remove(&dir->entries, dirent);
141
+ dirent = dirent_pointer(dirent_alloc(1));
142
+ dirent->name_offset = name;
143
+ dirent->mode = REPO_MODE_DIR;
144
+ dirent->content_offset = dir_o;
145
+ dirent_insert(&dir->entries, dirent);
146
+ }
147
+
148
+ dir = repo_dir_from_dirent(dirent);
149
+ dir = repo_clone_dir(dir);
150
+ dirent->content_offset = dir_offset(dir);
151
+ }
152
+ if (dirent == NULL)
153
+ return;
154
+ dirent->mode = mode;
155
+ dirent->content_offset = content_offset;
156
+ if (del && ~parent_dir_o)
157
+ dirent_remove(&dir_pointer(parent_dir_o)->entries, dirent);
158
+}
159
+
160
+uint32_t repo_copy(uint32_t revision, uint32_t *src, uint32_t *dst)
161
+{
162
+ uint32_t mode = 0, content_offset = 0;
163
+ struct repo_dirent *src_dirent;
164
+ src_dirent = repo_read_dirent(revision, src);
165
+ if (src_dirent != NULL) {
166
+ mode = src_dirent->mode;
167
+ content_offset = src_dirent->content_offset;
168
+ repo_write_dirent(dst, mode, content_offset, 0);
169
+ }
170
+ return mode;
171
+}
172
+
173
+void repo_add(uint32_t *path, uint32_t mode, uint32_t blob_mark)
174
+{
175
+ repo_write_dirent(path, mode, blob_mark, 0);
176
+}
177
+
178
+uint32_t repo_replace(uint32_t *path, uint32_t blob_mark)
179
+{
180
+ uint32_t mode = 0;
181
+ struct repo_dirent *src_dirent;
182
+ src_dirent = repo_read_dirent(active_commit, path);
183
+ if (src_dirent != NULL) {
184
+ mode = src_dirent->mode;
185
+ repo_write_dirent(path, mode, blob_mark, 0);
186
+ }
187
+ return mode;
188
+}
189
+
190
+void repo_modify(uint32_t *path, uint32_t mode, uint32_t blob_mark)
191
+{
192
+ struct repo_dirent *src_dirent;
193
+ src_dirent = repo_read_dirent(active_commit, path);
194
+ if (src_dirent != NULL && blob_mark == 0)
195
+ blob_mark = src_dirent->content_offset;
196
+ repo_write_dirent(path, mode, blob_mark, 0);
197
+}
198
+
199
+void repo_delete(uint32_t *path)
200
+{
201
+ repo_write_dirent(path, 0, 0, 1);
202
+}
203
+
204
+static void repo_git_add_r(uint32_t depth, uint32_t *path, struct repo_dir *dir);
205
+
206
+static void repo_git_add(uint32_t depth, uint32_t *path, struct repo_dirent *dirent)
207
+{
208
+ if (repo_dirent_is_dir(dirent))
209
+ repo_git_add_r(depth, path, repo_dir_from_dirent(dirent));
210
+ else
211
+ fast_export_modify(depth, path,
212
+ dirent->mode, dirent->content_offset);
213
+}
214
+
215
+static void repo_git_add_r(uint32_t depth, uint32_t *path, struct repo_dir *dir)
216
+{
217
+ struct repo_dirent *de = repo_first_dirent(dir);
218
+ while (de) {
219
+ path[depth] = de->name_offset;
220
+ repo_git_add(depth + 1, path, de);
221
+ de = dirent_next(&dir->entries, de);
222
+ }
223
+}
224
+
225
+static void repo_diff_r(uint32_t depth, uint32_t *path, struct repo_dir *dir1,
226
+ struct repo_dir *dir2)
227
+{
228
+ struct repo_dirent *de1, *de2;
229
+ de1 = repo_first_dirent(dir1);
230
+ de2 = repo_first_dirent(dir2);
231
+
232
+ while (de1 && de2) {
233
+ if (de1->name_offset < de2->name_offset) {
234
+ path[depth] = de1->name_offset;
235
+ fast_export_delete(depth + 1, path);
236
+ de1 = dirent_next(&dir1->entries, de1);
237
+ continue;
238
+ }
239
+ if (de1->name_offset > de2->name_offset) {
240
+ path[depth] = de2->name_offset;
241
+ repo_git_add(depth + 1, path, de2);
242
+ de2 = dirent_next(&dir2->entries, de2);
243
+ continue;
244
+ }
245
+ path[depth] = de1->name_offset;
246
+
247
+ if (de1->mode == de2->mode &&
248
+ de1->content_offset == de2->content_offset) {
249
+ ; /* No change. */
250
+ } else if (repo_dirent_is_dir(de1) && repo_dirent_is_dir(de2)) {
251
+ repo_diff_r(depth + 1, path,
252
+ repo_dir_from_dirent(de1),
253
+ repo_dir_from_dirent(de2));
254
+ } else if (!repo_dirent_is_dir(de1) && !repo_dirent_is_dir(de2)) {
255
+ repo_git_add(depth + 1, path, de2);
256
+ } else {
257
+ fast_export_delete(depth + 1, path);
258
+ repo_git_add(depth + 1, path, de2);
259
+ }
260
+ de1 = dirent_next(&dir1->entries, de1);
261
+ de2 = dirent_next(&dir2->entries, de2);
262
+ }
263
+ while (de1) {
264
+ path[depth] = de1->name_offset;
265
+ fast_export_delete(depth + 1, path);
266
+ de1 = dirent_next(&dir1->entries, de1);
267
+ }
268
+ while (de2) {
269
+ path[depth] = de2->name_offset;
270
+ repo_git_add(depth + 1, path, de2);
271
+ de2 = dirent_next(&dir2->entries, de2);
272
+ }
273
+}
274
+
275
+static uint32_t path_stack[REPO_MAX_PATH_DEPTH];
276
+
277
+void repo_diff(uint32_t r1, uint32_t r2)
278
+{
279
+ repo_diff_r(0,
280
+ path_stack,
281
+ repo_commit_root_dir(commit_pointer(r1)),
282
+ repo_commit_root_dir(commit_pointer(r2)));
283
+}
284
+
285
+void repo_commit(uint32_t revision, uint32_t author, char *log, uint32_t uuid,
286
+ uint32_t url, unsigned long timestamp)
287
+{
288
+ fast_export_commit(revision, author, log, uuid, url, timestamp);
289
+ dirent_commit();
290
+ dir_commit();
291
+ active_commit = commit_alloc(1);
292
+ commit_pointer(active_commit)->root_dir_offset =
293
+ commit_pointer(active_commit - 1)->root_dir_offset;
294
+}
295
+
296
+static void mark_init(void)
297
+{
298
+ uint32_t i;
299
+ mark = 0;
300
+ for (i = 0; i < dirent_pool.size; i++)
301
+ if (!repo_dirent_is_dir(dirent_pointer(i)) &&
302
+ dirent_pointer(i)->content_offset > mark)
303
+ mark = dirent_pointer(i)->content_offset;
304
+ mark++;
305
+}
306
+
307
+void repo_init(void)
308
+{
309
+ mark_init();
310
+ if (commit_pool.size == 0) {
311
+ /* Create empty tree for commit 0. */
312
+ commit_alloc(1);
313
+ commit_pointer(0)->root_dir_offset = dir_alloc(1);
314
+ dir_pointer(0)->entries.trp_root = ~0;
315
+ dir_commit();
316
+ }
317
+ /* Preallocate next commit, ready for changes. */
318
+ active_commit = commit_alloc(1);
319
+ commit_pointer(active_commit)->root_dir_offset =
320
+ commit_pointer(active_commit - 1)->root_dir_offset;
321
+}
322
+
323
+void repo_reset(void)
324
+{
325
+ pool_reset();
326
+ commit_reset();
327
+ dir_reset();
328
+ dirent_reset();
329
+}
vcs-svn/repo_tree.h
new
+26
@@ -0,0 +1,26 @@
1
+#ifndef REPO_TREE_H_
2
+#define REPO_TREE_H_
3
+
4
+#include "git-compat-util.h"
5
+
6
+#define REPO_MODE_DIR 0040000
7
+#define REPO_MODE_BLB 0100644
8
+#define REPO_MODE_EXE 0100755
9
+#define REPO_MODE_LNK 0120000
10
+
11
+#define REPO_MAX_PATH_LEN 4096
12
+#define REPO_MAX_PATH_DEPTH 1000
13
+
14
+uint32_t next_blob_mark(void);
15
+uint32_t repo_copy(uint32_t revision, uint32_t *src, uint32_t *dst);
16
+void repo_add(uint32_t *path, uint32_t mode, uint32_t blob_mark);
17
+uint32_t repo_replace(uint32_t *path, uint32_t blob_mark);
18
+void repo_modify(uint32_t *path, uint32_t mode, uint32_t blob_mark);
19
+void repo_delete(uint32_t *path);
20
+void repo_commit(uint32_t revision, uint32_t author, char *log, uint32_t uuid,
21
+ uint32_t url, long unsigned timestamp);
22
+void repo_diff(uint32_t r1, uint32_t r2);
23
+void repo_init(void);
24
+void repo_reset(void);
25
+
26
+#endif