fast-import.c 88.0 KB
Newer Older
1
/*
2
(See Documentation/git-fast-import.txt for maintained documentation.)
3 4 5 6 7
Format of STDIN stream:

  stream ::= cmd*;

  cmd ::= new_blob
8
        | new_commit
9
        | new_tag
10
        | reset_branch
11
        | checkpoint
12
        | progress
13 14
        ;

15
  new_blob ::= 'blob' lf
16
    mark?
17 18
    file_content;
  file_content ::= data;
19

20
  new_commit ::= 'commit' sp ref_str lf
21
    mark?
22 23
    ('author' (sp name)? sp '<' email '>' sp when lf)?
    'committer' (sp name)? sp '<' email '>' sp when lf
24
    commit_msg
25 26
    ('from' sp commit-ish lf)?
    ('merge' sp commit-ish lf)*
D
David Barr 已提交
27
    (file_change | ls)*
28
    lf?;
29
  commit_msg ::= data;
30

D
David Barr 已提交
31 32
  ls ::= 'ls' sp '"' quoted(path) '"' lf;

33 34 35 36 37 38
  file_change ::= file_clr
    | file_del
    | file_rnm
    | file_cpy
    | file_obm
    | file_inm;
39
  file_clr ::= 'deleteall' lf;
40
  file_del ::= 'D' sp path_str lf;
41
  file_rnm ::= 'R' sp path_str sp path_str lf;
42
  file_cpy ::= 'C' sp path_str sp path_str lf;
43 44 45
  file_obm ::= 'M' sp mode sp (hexsha1 | idnum) sp path_str lf;
  file_inm ::= 'M' sp mode sp 'inline' sp path_str lf
    data;
46 47
  note_obm ::= 'N' sp (hexsha1 | idnum) sp commit-ish lf;
  note_inm ::= 'N' sp 'inline' sp commit-ish lf
48
    data;
49 50

  new_tag ::= 'tag' sp tag_str lf
51
    'from' sp commit-ish lf
52
    ('tagger' (sp name)? sp '<' email '>' sp when lf)?
53 54 55
    tag_msg;
  tag_msg ::= data;

56
  reset_branch ::= 'reset' sp ref_str lf
57
    ('from' sp commit-ish lf)?
58
    lf?;
59

60
  checkpoint ::= 'checkpoint' lf
61
    lf?;
62

63 64 65
  progress ::= 'progress' sp not_lf* lf
    lf?;

66 67 68
     # note: the first idnum in a stream should be 1 and subsequent
     # idnums should not have gaps between values as this will cause
     # the stream parser to reserve space for the gapped values.  An
69
     # idnum can be updated in the future to a new object by issuing
70
     # a new mark directive with the old idnum.
71
     #
72
  mark ::= 'mark' sp idnum lf;
73
  data ::= (delimited_data | exact_data)
74
    lf?;
75 76 77 78 79 80

    # note: delim may be any string but must not contain lf.
    # data_line may contain any data but must not be exactly
    # delim.
  delimited_data ::= 'data' sp '<<' delim lf
    (data_line lf)*
81
    delim lf;
82 83

     # note: declen indicates the length of binary_data in bytes.
M
Mike Ralphson 已提交
84
     # declen does not include the lf preceding the binary data.
85
     #
86 87
  exact_data ::= 'data' sp declen lf
    binary_data;
88 89 90

     # note: quoted strings are C-style quoting supporting \c for
     # common escapes of 'c' (e..g \n, \t, \\, \") or \nnn where nnn
91
     # is the signed byte value in octal.  Note that the only
92 93
     # characters which must actually be escaped to protect the
     # stream formatting is: \, " and LF.  Otherwise these values
94
     # are UTF8.
95
     #
96
  commit-ish  ::= (ref_str | hexsha1 | sha1exp_str | idnum);
97 98 99
  ref_str     ::= ref;
  sha1exp_str ::= sha1exp;
  tag_str     ::= tag;
100
  path_str    ::= path    | '"' quoted(path)    '"' ;
101 102
  mode        ::= '100644' | '644'
                | '100755' | '755'
J
Junio C Hamano 已提交
103
                | '120000'
104
                ;
105 106

  declen ::= # unsigned 32 bit value, ascii base10 notation;
107
  bigint ::= # unsigned integer value, ascii base10 notation;
108
  binary_data ::= # file content, not interpreted;
109

110 111 112 113
  when         ::= raw_when | rfc2822_when;
  raw_when     ::= ts sp tz;
  rfc2822_when ::= # Valid RFC 2822 date and time;

114 115
  sp ::= # ASCII space character;
  lf ::= # ASCII newline (LF) character;
116 117

     # note: a colon (':') must precede the numerical value assigned to
118
     # an idnum.  This is to distinguish it from a ref or tag name as
119
     # GIT does not permit ':' in ref or tag strings.
120
     #
121
  idnum   ::= ':' bigint;
122 123 124
  path    ::= # GIT style file path, e.g. "a/b/c";
  ref     ::= # GIT ref name, e.g. "refs/heads/MOZ_GECKO_EXPERIMENT";
  tag     ::= # GIT tag name, e.g. "FIREFOX_1_5";
125 126
  sha1exp ::= # Any valid GIT SHA1 expression;
  hexsha1 ::= # SHA1 in hexadecimal format;
127 128

     # note: name and email are UTF8 strings, however name must not
129
     # contain '<' or lf and email must not contain any of the
130
     # following: '<', '>', lf.
131
     #
132
  name  ::= # valid GIT author/committer name;
133
  email ::= # valid GIT author/committer email;
134 135
  ts    ::= # time since the epoch in seconds, ascii base10 notation;
  tz    ::= # GIT style timezone;
136

137 138 139 140
     # note: comments, get-mark, ls-tree, and cat-blob requests may
     # appear anywhere in the input, except within a data command. Any
     # form of the data command always escapes the related input from
     # comment processing.
141 142
     #
     # In case it is not clear, the '#' that starts the comment
143
     # must be the first character on that line (an lf
M
Mike Ralphson 已提交
144
     # preceded it).
145
     #
D
David Barr 已提交
146

147
  get_mark ::= 'get-mark' sp idnum lf;
148
  cat_blob ::= 'cat-blob' sp (hexsha1 | idnum) lf;
D
David Barr 已提交
149
  ls_tree  ::= 'ls' sp (hexsha1 | idnum) sp path_str lf;
150

151 152
  comment ::= '#' not_lf* lf;
  not_lf  ::= # Any byte that is not ASCII newline (LF);
153 154
*/

155 156
#include "builtin.h"
#include "cache.h"
157
#include "lockfile.h"
158 159
#include "object.h"
#include "blob.h"
160
#include "tree.h"
161
#include "commit.h"
162 163
#include "delta.h"
#include "pack.h"
164
#include "refs.h"
165
#include "csum-file.h"
166
#include "quote.h"
167
#include "exec_cmd.h"
168
#include "dir.h"
169

170 171
#define PACK_ID_BITS 16
#define MAX_PACK_ID ((1<<PACK_ID_BITS)-1)
172 173
#define DEPTH_BITS 13
#define MAX_DEPTH ((1<<DEPTH_BITS)-1)
174

175 176 177 178 179
/*
 * We abuse the setuid bit on directories to mean "do not delta".
 */
#define NO_DELTA S_ISUID

180
struct object_entry {
181
	struct pack_idx_entry idx;
182
	struct object_entry *next;
183 184 185
	uint32_t type : TYPE_BITS,
		pack_id : PACK_ID_BITS,
		depth : DEPTH_BITS;
186 187
};

188
struct object_entry_pool {
189
	struct object_entry_pool *next_pool;
190 191
	struct object_entry *next_free;
	struct object_entry *end;
192
	struct object_entry entries[FLEX_ARRAY]; /* more */
193 194
};

195
struct mark_set {
196 197 198 199
	union {
		struct object_entry *marked[1024];
		struct mark_set *sets[1024];
	} data;
200
	unsigned int shift;
201 202
};

203
struct last_object {
P
Pierre Habouzit 已提交
204
	struct strbuf data;
205
	off_t offset;
206
	unsigned int depth;
P
Pierre Habouzit 已提交
207
	unsigned no_swap : 1;
208 209
};

210
struct mem_pool {
211 212 213
	struct mem_pool *next_pool;
	char *next_free;
	char *end;
214
	uintmax_t space[FLEX_ARRAY]; /* more */
215 216
};

217
struct atom_str {
218
	struct atom_str *next_atom;
219
	unsigned short str_len;
220 221 222 223
	char str_dat[FLEX_ARRAY]; /* more */
};

struct tree_content;
224
struct tree_entry {
225
	struct tree_content *tree;
226
	struct atom_str *name;
227
	struct tree_entry_ms {
228
		uint16_t mode;
229 230
		unsigned char sha1[20];
	} versions[2];
231 232
};

233
struct tree_content {
234 235
	unsigned int entry_capacity; /* must match avail_tree_content */
	unsigned int entry_count;
236
	unsigned int delta_depth;
237 238 239
	struct tree_entry *entries[FLEX_ARRAY]; /* more */
};

240
struct avail_tree_content {
241 242
	unsigned int entry_capacity; /* must match tree_content */
	struct avail_tree_content *next_avail;
243 244
};

245
struct branch {
246 247
	struct branch *table_next_branch;
	struct branch *active_next_branch;
248
	const char *name;
249
	struct tree_entry branch_tree;
250
	uintmax_t last_commit;
251
	uintmax_t num_notes;
252
	unsigned active : 1;
253
	unsigned delete : 1;
254
	unsigned pack_id : PACK_ID_BITS;
255
	unsigned char sha1[20];
256 257
};

258
struct tag {
259 260
	struct tag *next_tag;
	const char *name;
261
	unsigned int pack_id;
262 263 264
	unsigned char sha1[20];
};

265
struct hash_list {
266 267 268
	struct hash_list *next;
	unsigned char sha1[20];
};
269

270 271 272
typedef enum {
	WHENSPEC_RAW = 1,
	WHENSPEC_RFC2822,
273
	WHENSPEC_NOW
274 275
} whenspec_type;

276
struct recent_command {
277 278 279 280 281
	struct recent_command *prev;
	struct recent_command *next;
	char *buf;
};

282
/* Configured limits on output */
283
static unsigned long max_depth = 10;
284
static off_t max_packsize;
285
static int force_update;
286 287
static int pack_compression_level = Z_DEFAULT_COMPRESSION;
static int pack_compression_seen;
288 289 290 291 292 293 294

/* Stats and misc. counters */
static uintmax_t alloc_count;
static uintmax_t marks_set_count;
static uintmax_t object_count_by_type[1 << TYPE_BITS];
static uintmax_t duplicate_count_by_type[1 << TYPE_BITS];
static uintmax_t delta_count_by_type[1 << TYPE_BITS];
295
static uintmax_t delta_count_attempts_by_type[1 << TYPE_BITS];
296
static unsigned long object_count;
297
static unsigned long branch_count;
298
static unsigned long branch_load_count;
299
static int failure;
300
static FILE *pack_edges;
301
static unsigned int show_stats = 1;
S
Sverre Rabbelier 已提交
302
static int global_argc;
303
static char **global_argv;
304

305 306 307 308 309
/* Memory pools */
static size_t mem_pool_alloc = 2*1024*1024 - sizeof(struct mem_pool);
static size_t total_allocd;
static struct mem_pool *mem_pool;

310
/* Atom management */
311 312 313 314 315
static unsigned int atom_table_sz = 4451;
static unsigned int atom_cnt;
static struct atom_str **atom_table;

/* The .pack file being generated */
316
static struct pack_idx_option pack_idx_opts;
317
static unsigned int pack_id;
318
static struct sha1file *pack_file;
319
static struct packed_git *pack_data;
320
static struct packed_git **all_packs;
321
static off_t pack_size;
322 323

/* Table of objects we've written. */
324
static unsigned int object_entry_alloc = 5000;
325 326
static struct object_entry_pool *blocks;
static struct object_entry *object_table[1 << 16];
327
static struct mark_set *marks;
328 329
static const char *export_marks_file;
static const char *import_marks_file;
330
static int import_marks_file_from_stream;
331
static int import_marks_file_ignore_missing;
332
static int relative_marks_paths;
333 334

/* Our last blob */
P
Pierre Habouzit 已提交
335
static struct last_object last_blob = { STRBUF_INIT, 0, 0, 0 };
336 337 338 339 340 341

/* Tree management */
static unsigned int tree_entry_alloc = 1000;
static void *avail_tree_entry;
static unsigned int avail_tree_table_sz = 100;
static struct avail_tree_content **avail_tree_table;
342 343
static struct strbuf old_tree = STRBUF_INIT;
static struct strbuf new_tree = STRBUF_INIT;
344

345
/* Branch data */
346 347 348
static unsigned long max_active_branches = 5;
static unsigned long cur_active_branches;
static unsigned long branch_table_sz = 1039;
349 350 351
static struct branch **branch_table;
static struct branch *active_branches;

352 353 354 355
/* Tag data */
static struct tag *first_tag;
static struct tag *last_tag;

356
/* Input stream parsing */
357
static whenspec_type whenspec = WHENSPEC_RAW;
358
static struct strbuf command_buf = STRBUF_INIT;
359
static int unread_command_buf;
360 361 362 363
static struct recent_command cmd_hist = {&cmd_hist, &cmd_hist, NULL};
static struct recent_command *cmd_tail = &cmd_hist;
static struct recent_command *rc_free;
static unsigned int cmd_save = 100;
364
static uintmax_t next_mark;
365
static struct strbuf new_data = STRBUF_INIT;
366
static int seen_data_command;
367
static int require_explicit_termination;
368

369 370 371
/* Signal handling */
static volatile sig_atomic_t checkpoint_requested;

372 373 374
/* Where to write output of cat-blob commands */
static int cat_blob_fd = STDOUT_FILENO;

S
Sverre Rabbelier 已提交
375
static void parse_argv(void);
376
static void parse_get_mark(const char *p);
377 378
static void parse_cat_blob(const char *p);
static void parse_ls(const char *p, struct branch *b);
379

380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405
static void write_branch_report(FILE *rpt, struct branch *b)
{
	fprintf(rpt, "%s:\n", b->name);

	fprintf(rpt, "  status      :");
	if (b->active)
		fputs(" active", rpt);
	if (b->branch_tree.tree)
		fputs(" loaded", rpt);
	if (is_null_sha1(b->branch_tree.versions[1].sha1))
		fputs(" dirty", rpt);
	fputc('\n', rpt);

	fprintf(rpt, "  tip commit  : %s\n", sha1_to_hex(b->sha1));
	fprintf(rpt, "  old tree    : %s\n", sha1_to_hex(b->branch_tree.versions[0].sha1));
	fprintf(rpt, "  cur tree    : %s\n", sha1_to_hex(b->branch_tree.versions[1].sha1));
	fprintf(rpt, "  commit clock: %" PRIuMAX "\n", b->last_commit);

	fputs("  last pack   : ", rpt);
	if (b->pack_id < MAX_PACK_ID)
		fprintf(rpt, "%u", b->pack_id);
	fputc('\n', rpt);

	fputc('\n', rpt);
}

406 407
static void dump_marks_helper(FILE *, uintmax_t, struct mark_set *);

408
static void write_crash_report(const char *err)
409
{
410
	const char *loc = git_path("fast_import_crash_%"PRIuMAX, (uintmax_t) getpid());
411 412 413
	FILE *rpt = fopen(loc, "w");
	struct branch *b;
	unsigned long lu;
414
	struct recent_command *rc;
415 416 417 418 419 420 421 422 423

	if (!rpt) {
		error("can't write crash report %s: %s", loc, strerror(errno));
		return;
	}

	fprintf(stderr, "fast-import: dumping crash report to %s\n", loc);

	fprintf(rpt, "fast-import crash report:\n");
424 425
	fprintf(rpt, "    fast-import process: %"PRIuMAX"\n", (uintmax_t) getpid());
	fprintf(rpt, "    parent process     : %"PRIuMAX"\n", (uintmax_t) getppid());
426
	fprintf(rpt, "    at %s\n", show_date(time(NULL), 0, DATE_MODE(LOCAL)));
427 428 429
	fputc('\n', rpt);

	fputs("fatal: ", rpt);
430
	fputs(err, rpt);
431 432
	fputc('\n', rpt);

433 434 435 436 437 438 439 440 441 442 443 444
	fputc('\n', rpt);
	fputs("Most Recent Commands Before Crash\n", rpt);
	fputs("---------------------------------\n", rpt);
	for (rc = cmd_hist.next; rc != &cmd_hist; rc = rc->next) {
		if (rc->next == &cmd_hist)
			fputs("* ", rpt);
		else
			fputs("  ", rpt);
		fputs(rc->buf, rpt);
		fputc('\n', rpt);
	}

445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465
	fputc('\n', rpt);
	fputs("Active Branch LRU\n", rpt);
	fputs("-----------------\n", rpt);
	fprintf(rpt, "    active_branches = %lu cur, %lu max\n",
		cur_active_branches,
		max_active_branches);
	fputc('\n', rpt);
	fputs("  pos  clock name\n", rpt);
	fputs("  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n", rpt);
	for (b = active_branches, lu = 0; b; b = b->active_next_branch)
		fprintf(rpt, "  %2lu) %6" PRIuMAX" %s\n",
			++lu, b->last_commit, b->name);

	fputc('\n', rpt);
	fputs("Inactive Branches\n", rpt);
	fputs("-----------------\n", rpt);
	for (lu = 0; lu < branch_table_sz; lu++) {
		for (b = branch_table[lu]; b; b = b->table_next_branch)
			write_branch_report(rpt, b);
	}

466 467 468 469 470 471 472 473 474 475 476 477 478
	if (first_tag) {
		struct tag *tg;
		fputc('\n', rpt);
		fputs("Annotated Tags\n", rpt);
		fputs("--------------\n", rpt);
		for (tg = first_tag; tg; tg = tg->next_tag) {
			fputs(sha1_to_hex(tg->sha1), rpt);
			fputc(' ', rpt);
			fputs(tg->name, rpt);
			fputc('\n', rpt);
		}
	}

479 480 481
	fputc('\n', rpt);
	fputs("Marks\n", rpt);
	fputs("-----\n", rpt);
482 483
	if (export_marks_file)
		fprintf(rpt, "  exported to %s\n", export_marks_file);
484 485 486
	else
		dump_marks_helper(rpt, 0, marks);

487 488 489 490 491 492
	fputc('\n', rpt);
	fputs("-------------------\n", rpt);
	fputs("END OF CRASH REPORT\n", rpt);
	fclose(rpt);
}

493 494 495 496
static void end_packfile(void);
static void unkeep_all_packs(void);
static void dump_marks(void);

497 498 499
static NORETURN void die_nicely(const char *err, va_list params)
{
	static int zombie;
500
	char message[2 * PATH_MAX];
501

502
	vsnprintf(message, sizeof(message), err, params);
503
	fputs("fatal: ", stderr);
504
	fputs(message, stderr);
505 506 507 508
	fputc('\n', stderr);

	if (!zombie) {
		zombie = 1;
509
		write_crash_report(message);
510 511 512
		end_packfile();
		unkeep_all_packs();
		dump_marks();
513 514 515
	}
	exit(128);
}
516

517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542
#ifndef SIGUSR1	/* Windows, for example */

static void set_checkpoint_signal(void)
{
}

#else

static void checkpoint_signal(int signo)
{
	checkpoint_requested = 1;
}

static void set_checkpoint_signal(void)
{
	struct sigaction sa;

	memset(&sa, 0, sizeof(sa));
	sa.sa_handler = checkpoint_signal;
	sigemptyset(&sa.sa_mask);
	sa.sa_flags = SA_RESTART;
	sigaction(SIGUSR1, &sa, NULL);
}

#endif

543
static void alloc_objects(unsigned int cnt)
544
{
545
	struct object_entry_pool *b;
546

547
	b = xmalloc(sizeof(struct object_entry_pool)
548
		+ cnt * sizeof(struct object_entry));
549
	b->next_pool = blocks;
550 551 552 553 554
	b->next_free = b->entries;
	b->end = b->entries + cnt;
	blocks = b;
	alloc_count += cnt;
}
555

556
static struct object_entry *new_object(unsigned char *sha1)
557
{
558
	struct object_entry *e;
559

560
	if (blocks->next_free == blocks->end)
561
		alloc_objects(object_entry_alloc);
562

563
	e = blocks->next_free++;
564
	hashcpy(e->idx.sha1, sha1);
565
	return e;
566 567
}

568
static struct object_entry *find_object(unsigned char *sha1)
569 570 571 572
{
	unsigned int h = sha1[0] << 8 | sha1[1];
	struct object_entry *e;
	for (e = object_table[h]; e; e = e->next)
573
		if (!hashcmp(sha1, e->idx.sha1))
574 575 576 577
			return e;
	return NULL;
}

578
static struct object_entry *insert_object(unsigned char *sha1)
579 580
{
	unsigned int h = sha1[0] << 8 | sha1[1];
581
	struct object_entry *e = object_table[h];
582 583

	while (e) {
584
		if (!hashcmp(sha1, e->idx.sha1))
585 586 587 588 589
			return e;
		e = e->next;
	}

	e = new_object(sha1);
590
	e->next = object_table[h];
591
	e->idx.offset = 0;
592
	object_table[h] = e;
593 594
	return e;
}
595

596 597 598 599 600 601 602 603
static unsigned int hc_str(const char *s, size_t len)
{
	unsigned int r = 0;
	while (len-- > 0)
		r = r * 31 + *s++;
	return r;
}

604
static void *pool_alloc(size_t len)
605 606 607 608
{
	struct mem_pool *p;
	void *r;

609 610 611 612
	/* round up to a 'uintmax_t' alignment */
	if (len & (sizeof(uintmax_t) - 1))
		len += sizeof(uintmax_t) - (len & (sizeof(uintmax_t) - 1));

613 614 615 616 617 618 619 620 621 622 623 624
	for (p = mem_pool; p; p = p->next_pool)
		if ((p->end - p->next_free >= len))
			break;

	if (!p) {
		if (len >= (mem_pool_alloc/2)) {
			total_allocd += len;
			return xmalloc(len);
		}
		total_allocd += sizeof(struct mem_pool) + mem_pool_alloc;
		p = xmalloc(sizeof(struct mem_pool) + mem_pool_alloc);
		p->next_pool = mem_pool;
625
		p->next_free = (char *) p->space;
626 627 628 629 630 631 632 633 634
		p->end = p->next_free + mem_pool_alloc;
		mem_pool = p;
	}

	r = p->next_free;
	p->next_free += len;
	return r;
}

635
static void *pool_calloc(size_t count, size_t size)
636 637 638 639 640 641 642
{
	size_t len = count * size;
	void *r = pool_alloc(len);
	memset(r, 0, len);
	return r;
}

643
static char *pool_strdup(const char *s)
644 645 646 647 648 649
{
	char *r = pool_alloc(strlen(s) + 1);
	strcpy(r, s);
	return r;
}

650
static void insert_mark(uintmax_t idnum, struct object_entry *oe)
651 652 653 654 655 656 657 658 659
{
	struct mark_set *s = marks;
	while ((idnum >> s->shift) >= 1024) {
		s = pool_calloc(1, sizeof(struct mark_set));
		s->shift = marks->shift + 10;
		s->data.sets[0] = marks;
		marks = s;
	}
	while (s->shift) {
660
		uintmax_t i = idnum >> s->shift;
661 662 663 664 665 666 667 668 669 670 671 672
		idnum -= i << s->shift;
		if (!s->data.sets[i]) {
			s->data.sets[i] = pool_calloc(1, sizeof(struct mark_set));
			s->data.sets[i]->shift = s->shift - 10;
		}
		s = s->data.sets[i];
	}
	if (!s->data.marked[idnum])
		marks_set_count++;
	s->data.marked[idnum] = oe;
}

673
static struct object_entry *find_mark(uintmax_t idnum)
674
{
675
	uintmax_t orig_idnum = idnum;
676 677 678 679
	struct mark_set *s = marks;
	struct object_entry *oe = NULL;
	if ((idnum >> s->shift) < 1024) {
		while (s && s->shift) {
680
			uintmax_t i = idnum >> s->shift;
681 682 683 684 685 686 687
			idnum -= i << s->shift;
			s = s->data.sets[i];
		}
		if (s)
			oe = s->data.marked[idnum];
	}
	if (!oe)
688
		die("mark :%" PRIuMAX " not declared", orig_idnum);
689 690 691
	return oe;
}

692
static struct atom_str *to_atom(const char *s, unsigned short len)
693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710
{
	unsigned int hc = hc_str(s, len) % atom_table_sz;
	struct atom_str *c;

	for (c = atom_table[hc]; c; c = c->next_atom)
		if (c->str_len == len && !strncmp(s, c->str_dat, len))
			return c;

	c = pool_alloc(sizeof(struct atom_str) + len + 1);
	c->str_len = len;
	strncpy(c->str_dat, s, len);
	c->str_dat[len] = 0;
	c->next_atom = atom_table[hc];
	atom_table[hc] = c;
	atom_cnt++;
	return c;
}

711
static struct branch *lookup_branch(const char *name)
712 713 714 715 716 717 718 719 720 721
{
	unsigned int hc = hc_str(name, strlen(name)) % branch_table_sz;
	struct branch *b;

	for (b = branch_table[hc]; b; b = b->table_next_branch)
		if (!strcmp(name, b->name))
			return b;
	return NULL;
}

722
static struct branch *new_branch(const char *name)
723 724
{
	unsigned int hc = hc_str(name, strlen(name)) % branch_table_sz;
725
	struct branch *b = lookup_branch(name);
726 727 728

	if (b)
		die("Invalid attempt to create duplicate branch: %s", name);
729
	if (check_refname_format(name, REFNAME_ALLOW_ONELEVEL))
730
		die("Branch name doesn't conform to GIT standards: %s", name);
731 732 733 734

	b = pool_calloc(1, sizeof(struct branch));
	b->name = pool_strdup(name);
	b->table_next_branch = branch_table[hc];
735 736
	b->branch_tree.versions[0].mode = S_IFDIR;
	b->branch_tree.versions[1].mode = S_IFDIR;
737
	b->num_notes = 0;
738
	b->active = 0;
739
	b->pack_id = MAX_PACK_ID;
740 741 742 743 744 745 746 747 748 749 750
	branch_table[hc] = b;
	branch_count++;
	return b;
}

static unsigned int hc_entries(unsigned int cnt)
{
	cnt = cnt & 7 ? (cnt / 8) + 1 : cnt / 8;
	return cnt < avail_tree_table_sz ? cnt : avail_tree_table_sz - 1;
}

751
static struct tree_content *new_tree_content(unsigned int cnt)
752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773
{
	struct avail_tree_content *f, *l = NULL;
	struct tree_content *t;
	unsigned int hc = hc_entries(cnt);

	for (f = avail_tree_table[hc]; f; l = f, f = f->next_avail)
		if (f->entry_capacity >= cnt)
			break;

	if (f) {
		if (l)
			l->next_avail = f->next_avail;
		else
			avail_tree_table[hc] = f->next_avail;
	} else {
		cnt = cnt & 7 ? ((cnt / 8) + 1) * 8 : cnt;
		f = pool_alloc(sizeof(*t) + sizeof(t->entries[0]) * cnt);
		f->entry_capacity = cnt;
	}

	t = (struct tree_content*)f;
	t->entry_count = 0;
774
	t->delta_depth = 0;
775 776 777 778 779 780 781 782
	return t;
}

static void release_tree_entry(struct tree_entry *e);
static void release_tree_content(struct tree_content *t)
{
	struct avail_tree_content *f = (struct avail_tree_content*)t;
	unsigned int hc = hc_entries(f->entry_capacity);
783 784 785 786 787 788
	f->next_avail = avail_tree_table[hc];
	avail_tree_table[hc] = f;
}

static void release_tree_content_recursive(struct tree_content *t)
{
789 790 791
	unsigned int i;
	for (i = 0; i < t->entry_count; i++)
		release_tree_entry(t->entries[i]);
792
	release_tree_content(t);
793 794
}

795
static struct tree_content *grow_tree_content(
796 797 798 799 800
	struct tree_content *t,
	int amt)
{
	struct tree_content *r = new_tree_content(t->entry_count + amt);
	r->entry_count = t->entry_count;
801
	r->delta_depth = t->delta_depth;
802 803 804 805 806
	memcpy(r->entries,t->entries,t->entry_count*sizeof(t->entries[0]));
	release_tree_content(t);
	return r;
}

807
static struct tree_entry *new_tree_entry(void)
808 809 810 811 812
{
	struct tree_entry *e;

	if (!avail_tree_entry) {
		unsigned int n = tree_entry_alloc;
813
		total_allocd += n * sizeof(struct tree_entry);
814
		avail_tree_entry = e = xmalloc(n * sizeof(struct tree_entry));
815
		while (n-- > 1) {
816 817 818
			*((void**)e) = e + 1;
			e++;
		}
819
		*((void**)e) = NULL;
820 821 822 823 824 825 826 827 828 829
	}

	e = avail_tree_entry;
	avail_tree_entry = *((void**)e);
	return e;
}

static void release_tree_entry(struct tree_entry *e)
{
	if (e->tree)
830
		release_tree_content_recursive(e->tree);
831 832 833 834
	*((void**)e) = avail_tree_entry;
	avail_tree_entry = e;
}

835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859
static struct tree_content *dup_tree_content(struct tree_content *s)
{
	struct tree_content *d;
	struct tree_entry *a, *b;
	unsigned int i;

	if (!s)
		return NULL;
	d = new_tree_content(s->entry_count);
	for (i = 0; i < s->entry_count; i++) {
		a = s->entries[i];
		b = new_tree_entry();
		memcpy(b, a, sizeof(*a));
		if (a->tree && is_null_sha1(b->versions[1].sha1))
			b->tree = dup_tree_content(a->tree);
		else
			b->tree = NULL;
		d->entries[i] = b;
	}
	d->entry_count = s->entry_count;
	d->delta_depth = s->delta_depth;

	return d;
}

860
static void start_packfile(void)
861
{
862
	static char tmp_file[PATH_MAX];
863
	struct packed_git *p;
864
	struct pack_header hdr;
865
	int pack_fd;
866

867
	pack_fd = odb_mkstemp(tmp_file, sizeof(tmp_file),
868
			      "pack/tmp_pack_XXXXXX");
869 870
	p = xcalloc(1, sizeof(*p) + strlen(tmp_file) + 2);
	strcpy(p->pack_name, tmp_file);
871
	p->pack_fd = pack_fd;
872
	p->do_not_close = 1;
873
	pack_file = sha1fd(pack_fd, p->pack_name);
874 875 876 877

	hdr.hdr_signature = htonl(PACK_SIGNATURE);
	hdr.hdr_version = htonl(2);
	hdr.hdr_entries = 0;
878
	sha1write(pack_file, &hdr, sizeof(hdr));
879 880

	pack_data = p;
881 882
	pack_size = sizeof(hdr);
	object_count = 0;
883

884
	REALLOC_ARRAY(all_packs, pack_id + 1);
885
	all_packs[pack_id] = p;
886 887
}

888
static const char *create_index(void)
889
{
890 891 892
	const char *tmpfile;
	struct pack_idx_entry **idx, **c, **last;
	struct object_entry *e;
893 894
	struct object_entry_pool *o;

895 896
	/* Build the table of object IDs. */
	idx = xmalloc(object_count * sizeof(*idx));
897 898
	c = idx;
	for (o = blocks; o; o = o->next_pool)
899 900
		for (e = o->next_free; e-- != o->entries;)
			if (pack_id == e->pack_id)
901
				*c++ = &e->idx;
902
	last = idx + object_count;
903 904
	if (c != last)
		die("internal consistency error creating the index");
905

906
	tmpfile = write_idx_file(NULL, idx, object_count, &pack_idx_opts, pack_data->sha1);
907
	free(idx);
908 909 910
	return tmpfile;
}

911
static char *keep_pack(const char *curr_index_name)
912 913
{
	static char name[PATH_MAX];
914
	static const char *keep_msg = "fast-import";
915 916
	int keep_fd;

917
	keep_fd = odb_pack_keep(name, sizeof(name), pack_data->sha1);
918
	if (keep_fd < 0)
919
		die_errno("cannot create keep file");
920 921
	write_or_die(keep_fd, keep_msg, strlen(keep_msg));
	if (close(keep_fd))
922
		die_errno("failed to write keep file");
923 924 925 926 927 928 929 930 931 932

	snprintf(name, sizeof(name), "%s/pack/pack-%s.pack",
		 get_object_directory(), sha1_to_hex(pack_data->sha1));
	if (move_temp_to_file(pack_data->pack_name, name))
		die("cannot store pack file");

	snprintf(name, sizeof(name), "%s/pack/pack-%s.idx",
		 get_object_directory(), sha1_to_hex(pack_data->sha1));
	if (move_temp_to_file(curr_index_name, name))
		die("cannot store index file");
933
	free((void *)curr_index_name);
934 935 936
	return name;
}

937
static void unkeep_all_packs(void)
938 939 940 941 942 943 944 945
{
	static char name[PATH_MAX];
	int k;

	for (k = 0; k < pack_id; k++) {
		struct packed_git *p = all_packs[k];
		snprintf(name, sizeof(name), "%s/pack/pack-%s.keep",
			 get_object_directory(), sha1_to_hex(p->sha1));
946
		unlink_or_warn(name);
947
	}
948 949
}

950
static void end_packfile(void)
951
{
952 953 954
	static int running;

	if (running || !pack_data)
955
		return;
956

957
	running = 1;
958
	clear_delta_base_cache();
959
	if (object_count) {
960
		struct packed_git *new_p;
961
		unsigned char cur_pack_sha1[20];
962
		char *idx_name;
963 964 965
		int i;
		struct branch *b;
		struct tag *t;
966

967
		close_pack_windows(pack_data);
968
		sha1close(pack_file, cur_pack_sha1, 0);
969
		fixup_pack_header_footer(pack_data->pack_fd, pack_data->sha1,
970
				    pack_data->pack_name, object_count,
971
				    cur_pack_sha1, pack_size);
972
		close(pack_data->pack_fd);
973
		idx_name = keep_pack(create_index());
974

M
Mike Ralphson 已提交
975
		/* Register the packfile with core git's machinery. */
976 977 978
		new_p = add_packed_git(idx_name, strlen(idx_name), 1);
		if (!new_p)
			die("core git rejected index %s", idx_name);
979
		all_packs[pack_id] = new_p;
980
		install_packed_git(new_p);
981 982

		/* Print the boundary */
983 984 985 986 987 988 989
		if (pack_edges) {
			fprintf(pack_edges, "%s:", new_p->pack_name);
			for (i = 0; i < branch_table_sz; i++) {
				for (b = branch_table[i]; b; b = b->table_next_branch) {
					if (b->pack_id == pack_id)
						fprintf(pack_edges, " %s", sha1_to_hex(b->sha1));
				}
990
			}
991 992 993 994 995 996
			for (t = first_tag; t; t = t->next_tag) {
				if (t->pack_id == pack_id)
					fprintf(pack_edges, " %s", sha1_to_hex(t->sha1));
			}
			fputc('\n', pack_edges);
			fflush(pack_edges);
997 998 999
		}

		pack_id++;
1000
	}
1001
	else {
1002 1003
		close(pack_data->pack_fd);
		unlink_or_warn(pack_data->pack_name);
1004
	}
1005 1006
	free(pack_data);
	pack_data = NULL;
1007
	running = 0;
1008 1009

	/* We can't carry a delta across packfiles. */
P
Pierre Habouzit 已提交
1010
	strbuf_release(&last_blob.data);
1011 1012
	last_blob.offset = 0;
	last_blob.depth = 0;
1013 1014
}

1015
static void cycle_packfile(void)
1016 1017 1018 1019 1020
{
	end_packfile();
	start_packfile();
}

1021 1022
static int store_object(
	enum object_type type,
1023
	struct strbuf *dat,
1024
	struct last_object *last,
1025
	unsigned char *sha1out,
1026
	uintmax_t mark)
1027 1028
{
	void *out, *delta;
1029 1030 1031
	struct object_entry *e;
	unsigned char hdr[96];
	unsigned char sha1[20];
1032
	unsigned long hdrlen, deltalen;
1033
	git_SHA_CTX c;
1034
	git_zstream s;
1035

1036
	hdrlen = sprintf((char *)hdr,"%s %lu", typename(type),
1037
		(unsigned long)dat->len) + 1;
1038 1039 1040 1041
	git_SHA1_Init(&c);
	git_SHA1_Update(&c, hdr, hdrlen);
	git_SHA1_Update(&c, dat->buf, dat->len);
	git_SHA1_Final(sha1, &c);
1042
	if (sha1out)
1043
		hashcpy(sha1out, sha1);
1044 1045

	e = insert_object(sha1);
1046 1047
	if (mark)
		insert_mark(mark, e);
1048
	if (e->idx.offset) {
1049
		duplicate_count_by_type[type]++;
1050
		return 1;
1051 1052 1053
	} else if (find_sha1_pack(sha1, packed_git)) {
		e->type = type;
		e->pack_id = MAX_PACK_ID;
1054
		e->idx.offset = 1; /* just not zero! */
1055 1056
		duplicate_count_by_type[type]++;
		return 1;
1057
	}
1058

1059
	if (last && last->data.buf && last->depth < max_depth && dat->len > 20) {
1060
		delta_count_attempts_by_type[type]++;
P
Pierre Habouzit 已提交
1061
		delta = diff_delta(last->data.buf, last->data.len,
1062
			dat->buf, dat->len,
1063
			&deltalen, dat->len - 20);
1064 1065
	} else
		delta = NULL;
1066

1067
	git_deflate_init(&s, pack_compression_level);
1068 1069 1070 1071
	if (delta) {
		s.next_in = delta;
		s.avail_in = deltalen;
	} else {
1072 1073
		s.next_in = (void *)dat->buf;
		s.avail_in = dat->len;
1074
	}
J
Junio C Hamano 已提交
1075
	s.avail_out = git_deflate_bound(&s, s.avail_in);
1076
	s.next_out = out = xmalloc(s.avail_out);
1077 1078 1079
	while (git_deflate(&s, Z_FINISH) == Z_OK)
		; /* nothing */
	git_deflate_end(&s);
1080 1081

	/* Determine if we should auto-checkpoint. */
1082
	if ((max_packsize && (pack_size + 60 + s.total_out) > max_packsize)
1083 1084 1085 1086
		|| (pack_size + 60 + s.total_out) < pack_size) {

		/* This new object needs to *not* have the current pack_id. */
		e->pack_id = pack_id + 1;
1087
		cycle_packfile();
1088 1089 1090 1091 1092

		/* We cannot carry a delta into the new pack. */
		if (delta) {
			free(delta);
			delta = NULL;
1093

1094
			git_deflate_init(&s, pack_compression_level);
1095 1096
			s.next_in = (void *)dat->buf;
			s.avail_in = dat->len;
J
Junio C Hamano 已提交
1097
			s.avail_out = git_deflate_bound(&s, s.avail_in);
1098
			s.next_out = out = xrealloc(out, s.avail_out);
1099 1100 1101
			while (git_deflate(&s, Z_FINISH) == Z_OK)
				; /* nothing */
			git_deflate_end(&s);
1102 1103 1104 1105 1106
		}
	}

	e->type = type;
	e->pack_id = pack_id;
1107
	e->idx.offset = pack_size;
1108 1109
	object_count++;
	object_count_by_type[type]++;
1110

1111 1112
	crc32_begin(pack_file);

1113
	if (delta) {
1114
		off_t ofs = e->idx.offset - last->offset;
1115 1116
		unsigned pos = sizeof(hdr) - 1;

1117
		delta_count_by_type[type]++;
1118
		e->depth = last->depth + 1;
1119

1120
		hdrlen = encode_in_pack_object_header(OBJ_OFS_DELTA, deltalen, hdr);
1121
		sha1write(pack_file, hdr, hdrlen);
1122 1123 1124 1125 1126
		pack_size += hdrlen;

		hdr[pos] = ofs & 127;
		while (ofs >>= 7)
			hdr[--pos] = 128 | (--ofs & 127);
1127
		sha1write(pack_file, hdr + pos, sizeof(hdr) - pos);
1128
		pack_size += sizeof(hdr) - pos;
1129
	} else {
1130
		e->depth = 0;
1131
		hdrlen = encode_in_pack_object_header(type, dat->len, hdr);
1132
		sha1write(pack_file, hdr, hdrlen);
1133
		pack_size += hdrlen;
1134 1135
	}

1136
	sha1write(pack_file, out, s.total_out);
1137
	pack_size += s.total_out;
1138

1139 1140
	e->idx.crc32 = crc32_end(pack_file);

1141
	free(out);
1142
	free(delta);
1143
	if (last) {
P
Pierre Habouzit 已提交
1144 1145 1146
		if (last->no_swap) {
			last->data = *dat;
		} else {
1147
			strbuf_swap(&last->data, dat);
P
Pierre Habouzit 已提交
1148
		}
1149
		last->offset = e->idx.offset;
1150
		last->depth = e->depth;
1151 1152 1153 1154
	}
	return 0;
}

1155
static void truncate_pack(struct sha1file_checkpoint *checkpoint)
1156
{
1157
	if (sha1file_truncate(pack_file, checkpoint))
1158
		die_errno("cannot truncate pack to skip duplicate");
1159
	pack_size = checkpoint->offset;
1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171
}

static void stream_blob(uintmax_t len, unsigned char *sha1out, uintmax_t mark)
{
	size_t in_sz = 64 * 1024, out_sz = 64 * 1024;
	unsigned char *in_buf = xmalloc(in_sz);
	unsigned char *out_buf = xmalloc(out_sz);
	struct object_entry *e;
	unsigned char sha1[20];
	unsigned long hdrlen;
	off_t offset;
	git_SHA_CTX c;
1172
	git_zstream s;
1173
	struct sha1file_checkpoint checkpoint;
1174 1175 1176
	int status = Z_OK;

	/* Determine if we should auto-checkpoint. */
1177
	if ((max_packsize && (pack_size + 60 + len) > max_packsize)
1178 1179 1180
		|| (pack_size + 60 + len) < pack_size)
		cycle_packfile();

1181 1182
	sha1file_checkpoint(pack_file, &checkpoint);
	offset = checkpoint.offset;
1183

1184 1185 1186 1187 1188 1189 1190
	hdrlen = snprintf((char *)out_buf, out_sz, "blob %" PRIuMAX, len) + 1;
	if (out_sz <= hdrlen)
		die("impossibly large object header");

	git_SHA1_Init(&c);
	git_SHA1_Update(&c, out_buf, hdrlen);

1191 1192
	crc32_begin(pack_file);

1193
	git_deflate_init(&s, pack_compression_level);
1194

1195
	hdrlen = encode_in_pack_object_header(OBJ_BLOB, len, out_buf);
1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214
	if (out_sz <= hdrlen)
		die("impossibly large object header");

	s.next_out = out_buf + hdrlen;
	s.avail_out = out_sz - hdrlen;

	while (status != Z_STREAM_END) {
		if (0 < len && !s.avail_in) {
			size_t cnt = in_sz < len ? in_sz : (size_t)len;
			size_t n = fread(in_buf, 1, cnt, stdin);
			if (!n && feof(stdin))
				die("EOF in data (%" PRIuMAX " bytes remaining)", len);

			git_SHA1_Update(&c, in_buf, n);
			s.next_in = in_buf;
			s.avail_in = n;
			len -= n;
		}

1215
		status = git_deflate(&s, len ? 0 : Z_FINISH);
1216 1217 1218

		if (!s.avail_out || status == Z_STREAM_END) {
			size_t n = s.next_out - out_buf;
1219
			sha1write(pack_file, out_buf, n);
1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233
			pack_size += n;
			s.next_out = out_buf;
			s.avail_out = out_sz;
		}

		switch (status) {
		case Z_OK:
		case Z_BUF_ERROR:
		case Z_STREAM_END:
			continue;
		default:
			die("unexpected deflate failure: %d", status);
		}
	}
1234
	git_deflate_end(&s);
1235 1236 1237 1238 1239 1240 1241 1242 1243 1244
	git_SHA1_Final(sha1, &c);

	if (sha1out)
		hashcpy(sha1out, sha1);

	e = insert_object(sha1);

	if (mark)
		insert_mark(mark, e);

1245
	if (e->idx.offset) {
1246
		duplicate_count_by_type[OBJ_BLOB]++;
1247
		truncate_pack(&checkpoint);
1248 1249 1250 1251

	} else if (find_sha1_pack(sha1, packed_git)) {
		e->type = OBJ_BLOB;
		e->pack_id = MAX_PACK_ID;
1252
		e->idx.offset = 1; /* just not zero! */
1253
		duplicate_count_by_type[OBJ_BLOB]++;
1254
		truncate_pack(&checkpoint);
1255 1256 1257 1258 1259

	} else {
		e->depth = 0;
		e->type = OBJ_BLOB;
		e->pack_id = pack_id;
1260
		e->idx.offset = offset;
1261
		e->idx.crc32 = crc32_end(pack_file);
1262 1263 1264 1265 1266 1267 1268 1269
		object_count++;
		object_count_by_type[OBJ_BLOB]++;
	}

	free(in_buf);
	free(out_buf);
}

1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287
/* All calls must be guarded by find_object() or find_mark() to
 * ensure the 'struct object_entry' passed was written by this
 * process instance.  We unpack the entry by the offset, avoiding
 * the need for the corresponding .idx file.  This unpacking rule
 * works because we only use OBJ_REF_DELTA within the packfiles
 * created by fast-import.
 *
 * oe must not be NULL.  Such an oe usually comes from giving
 * an unknown SHA-1 to find_object() or an undefined mark to
 * find_mark().  Callers must test for this condition and use
 * the standard read_sha1_file() when it happens.
 *
 * oe->pack_id must not be MAX_PACK_ID.  Such an oe is usually from
 * find_mark(), where the mark was reloaded from an existing marks
 * file and is referencing an object that this fast-import process
 * instance did not write out to a packfile.  Callers must test for
 * this condition and use read_sha1_file() instead.
 */
1288 1289 1290
static void *gfi_unpack_entry(
	struct object_entry *oe,
	unsigned long *sizep)
1291
{
1292
	enum object_type type;
1293
	struct packed_git *p = all_packs[oe->pack_id];
1294
	if (p == pack_data && p->pack_size < (pack_size + 20)) {
1295 1296 1297 1298 1299 1300 1301 1302
		/* The object is stored in the packfile we are writing to
		 * and we have modified it since the last time we scanned
		 * back to read a previously written object.  If an old
		 * window covered [p->pack_size, p->pack_size + 20) its
		 * data is stale and is not valid.  Closing all windows
		 * and updating the packfile length ensures we can read
		 * the newly written data.
		 */
1303
		close_pack_windows(p);
1304
		sha1flush(pack_file);
1305 1306 1307 1308 1309 1310 1311

		/* We have to offer 20 bytes additional on the end of
		 * the packfile as the core unpacker code assumes the
		 * footer is present at the file end and must promise
		 * at least 20 bytes within any window it maps.  But
		 * we don't actually create the footer here.
		 */
1312
		p->pack_size = pack_size + 20;
1313
	}
1314
	return unpack_entry(p, oe->idx.offset, &type, sizep);
1315 1316
}

1317
static const char *get_mode(const char *str, uint16_t *modep)
1318 1319
{
	unsigned char c;
1320
	uint16_t mode = 0;
1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332

	while ((c = *str++) != ' ') {
		if (c < '0' || c > '7')
			return NULL;
		mode = (mode << 3) + (c - '0');
	}
	*modep = mode;
	return str;
}

static void load_tree(struct tree_entry *root)
{
1333
	unsigned char *sha1 = root->versions[1].sha1;
1334 1335 1336 1337 1338 1339 1340
	struct object_entry *myoe;
	struct tree_content *t;
	unsigned long size;
	char *buf;
	const char *c;

	root->tree = t = new_tree_content(8);
1341
	if (is_null_sha1(sha1))
1342 1343
		return;

1344
	myoe = find_object(sha1);
1345
	if (myoe && myoe->pack_id != MAX_PACK_ID) {
1346
		if (myoe->type != OBJ_TREE)
1347
			die("Not a tree: %s", sha1_to_hex(sha1));
1348
		t->delta_depth = myoe->depth;
1349
		buf = gfi_unpack_entry(myoe, &size);
1350 1351
		if (!buf)
			die("Can't load tree %s", sha1_to_hex(sha1));
1352
	} else {
1353 1354 1355
		enum object_type type;
		buf = read_sha1_file(sha1, &type, &size);
		if (!buf || type != OBJ_TREE)
1356
			die("Can't load tree %s", sha1_to_hex(sha1));
1357 1358 1359 1360 1361 1362 1363
	}

	c = buf;
	while (c != (buf + size)) {
		struct tree_entry *e = new_tree_entry();

		if (t->entry_count == t->entry_capacity)
1364
			root->tree = t = grow_tree_content(t, t->entry_count);
1365 1366 1367
		t->entries[t->entry_count++] = e;

		e->tree = NULL;
1368
		c = get_mode(c, &e->versions[1].mode);
1369
		if (!c)
1370 1371
			die("Corrupt mode in %s", sha1_to_hex(sha1));
		e->versions[0].mode = e->versions[1].mode;
1372
		e->name = to_atom(c, strlen(c));
1373
		c += e->name->str_len + 1;
1374 1375
		hashcpy(e->versions[0].sha1, (unsigned char *)c);
		hashcpy(e->versions[1].sha1, (unsigned char *)c);
1376 1377 1378 1379 1380
		c += 20;
	}
	free(buf);
}

1381
static int tecmp0 (const void *_a, const void *_b)
1382 1383 1384 1385
{
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
1386 1387
		a->name->str_dat, a->name->str_len, a->versions[0].mode,
		b->name->str_dat, b->name->str_len, b->versions[0].mode);
1388 1389
}

1390
static int tecmp1 (const void *_a, const void *_b)
1391
{
1392 1393 1394 1395 1396 1397 1398
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
		a->name->str_dat, a->name->str_len, a->versions[1].mode,
		b->name->str_dat, b->name->str_len, b->versions[1].mode);
}

1399
static void mktree(struct tree_content *t, int v, struct strbuf *b)
1400 1401
{
	size_t maxlen = 0;
1402 1403
	unsigned int i;

1404 1405 1406 1407
	if (!v)
		qsort(t->entries,t->entry_count,sizeof(t->entries[0]),tecmp0);
	else
		qsort(t->entries,t->entry_count,sizeof(t->entries[0]),tecmp1);
1408 1409

	for (i = 0; i < t->entry_count; i++) {
1410 1411
		if (t->entries[i]->versions[v].mode)
			maxlen += t->entries[i]->name->str_len + 34;
1412 1413
	}

1414 1415
	strbuf_reset(b);
	strbuf_grow(b, maxlen);
1416 1417
	for (i = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
1418 1419
		if (!e->versions[v].mode)
			continue;
1420 1421 1422
		strbuf_addf(b, "%o %s%c",
			(unsigned int)(e->versions[v].mode & ~NO_DELTA),
			e->name->str_dat, '\0');
1423
		strbuf_add(b, e->versions[v].sha1, 20);
1424
	}
1425 1426 1427 1428
}

static void store_tree(struct tree_entry *root)
{
1429
	struct tree_content *t;
1430
	unsigned int i, j, del;
P
Pierre Habouzit 已提交
1431
	struct last_object lo = { STRBUF_INIT, 0, 0, /* no_swap */ 1 };
1432
	struct object_entry *le = NULL;
1433 1434 1435 1436

	if (!is_null_sha1(root->versions[1].sha1))
		return;

1437 1438 1439 1440
	if (!root->tree)
		load_tree(root);
	t = root->tree;

1441 1442 1443 1444 1445
	for (i = 0; i < t->entry_count; i++) {
		if (t->entries[i]->tree)
			store_tree(t->entries[i]);
	}

1446 1447
	if (!(root->versions[0].mode & NO_DELTA))
		le = find_object(root->versions[0].sha1);
P
Pierre Habouzit 已提交
1448
	if (S_ISDIR(root->versions[0].mode) && le && le->pack_id == pack_id) {
1449
		mktree(t, 0, &old_tree);
P
Pierre Habouzit 已提交
1450
		lo.data = old_tree;
1451
		lo.offset = le->idx.offset;
1452 1453 1454
		lo.depth = t->delta_depth;
	}

1455 1456
	mktree(t, 1, &new_tree);
	store_object(OBJ_TREE, &new_tree, &lo, root->versions[1].sha1, 0);
1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470

	t->delta_depth = lo.depth;
	for (i = 0, j = 0, del = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
		if (e->versions[1].mode) {
			e->versions[0].mode = e->versions[1].mode;
			hashcpy(e->versions[0].sha1, e->versions[1].sha1);
			t->entries[j++] = e;
		} else {
			release_tree_entry(e);
			del++;
		}
	}
	t->entry_count -= del;
1471 1472
}

1473 1474 1475 1476 1477 1478 1479 1480
static void tree_content_replace(
	struct tree_entry *root,
	const unsigned char *sha1,
	const uint16_t mode,
	struct tree_content *newtree)
{
	if (!S_ISDIR(mode))
		die("Root cannot be a non-directory");
1481
	hashclr(root->versions[0].sha1);
1482 1483 1484 1485 1486 1487
	hashcpy(root->versions[1].sha1, sha1);
	if (root->tree)
		release_tree_content_recursive(root->tree);
	root->tree = newtree;
}

1488 1489 1490 1491
static int tree_content_set(
	struct tree_entry *root,
	const char *p,
	const unsigned char *sha1,
1492 1493
	const uint16_t mode,
	struct tree_content *subtree)
1494
{
1495
	struct tree_content *t;
1496 1497 1498 1499
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

1500 1501
	slash1 = strchrnul(p, '/');
	n = slash1 - p;
1502 1503
	if (!n)
		die("Empty path component found in input");
1504
	if (!*slash1 && !S_ISDIR(mode) && subtree)
1505
		die("Non-directories cannot have subtrees");
1506

1507 1508 1509
	if (!root->tree)
		load_tree(root);
	t = root->tree;
1510 1511
	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
1512
		if (e->name->str_len == n && !strncmp_icase(p, e->name->str_dat, n)) {
1513
			if (!*slash1) {
1514 1515
				if (!S_ISDIR(mode)
						&& e->versions[1].mode == mode
1516
						&& !hashcmp(e->versions[1].sha1, sha1))
1517
					return 0;
1518 1519
				e->versions[1].mode = mode;
				hashcpy(e->versions[1].sha1, sha1);
1520
				if (e->tree)
1521
					release_tree_content_recursive(e->tree);
1522
				e->tree = subtree;
1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539

				/*
				 * We need to leave e->versions[0].sha1 alone
				 * to avoid modifying the preimage tree used
				 * when writing out the parent directory.
				 * But after replacing the subdir with a
				 * completely different one, it's not a good
				 * delta base any more, and besides, we've
				 * thrown away the tree entries needed to
				 * make a delta against it.
				 *
				 * So let's just explicitly disable deltas
				 * for the subtree.
				 */
				if (S_ISDIR(e->versions[0].mode))
					e->versions[0].mode |= NO_DELTA;

1540
				hashclr(root->versions[1].sha1);
1541 1542
				return 1;
			}
1543
			if (!S_ISDIR(e->versions[1].mode)) {
1544
				e->tree = new_tree_content(8);
1545
				e->versions[1].mode = S_IFDIR;
1546 1547 1548
			}
			if (!e->tree)
				load_tree(e);
1549
			if (tree_content_set(e, slash1 + 1, sha1, mode, subtree)) {
1550
				hashclr(root->versions[1].sha1);
1551 1552 1553 1554 1555 1556 1557
				return 1;
			}
			return 0;
		}
	}

	if (t->entry_count == t->entry_capacity)
1558
		root->tree = t = grow_tree_content(t, t->entry_count);
1559
	e = new_tree_entry();
1560
	e->name = to_atom(p, n);
1561 1562
	e->versions[0].mode = 0;
	hashclr(e->versions[0].sha1);
1563
	t->entries[t->entry_count++] = e;
1564
	if (*slash1) {
1565
		e->tree = new_tree_content(8);
1566
		e->versions[1].mode = S_IFDIR;
1567
		tree_content_set(e, slash1 + 1, sha1, mode, subtree);
1568
	} else {
1569
		e->tree = subtree;
1570 1571
		e->versions[1].mode = mode;
		hashcpy(e->versions[1].sha1, sha1);
1572
	}
1573
	hashclr(root->versions[1].sha1);
1574 1575 1576
	return 1;
}

1577 1578 1579
static int tree_content_remove(
	struct tree_entry *root,
	const char *p,
1580 1581
	struct tree_entry *backup_leaf,
	int allow_root)
1582
{
1583
	struct tree_content *t;
1584 1585 1586 1587
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

1588 1589
	slash1 = strchrnul(p, '/');
	n = slash1 - p;
1590

1591 1592
	if (!root->tree)
		load_tree(root);
1593 1594 1595 1596 1597 1598

	if (!*p && allow_root) {
		e = root;
		goto del_entry;
	}

1599
	t = root->tree;
1600 1601
	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
1602
		if (e->name->str_len == n && !strncmp_icase(p, e->name->str_dat, n)) {
1603
			if (*slash1 && !S_ISDIR(e->versions[1].mode))
1604 1605 1606 1607 1608 1609 1610
				/*
				 * If p names a file in some subdirectory, and a
				 * file or symlink matching the name of the
				 * parent directory of p exists, then p cannot
				 * exist and need not be deleted.
				 */
				return 1;
1611
			if (!*slash1 || !S_ISDIR(e->versions[1].mode))
1612 1613 1614
				goto del_entry;
			if (!e->tree)
				load_tree(e);
1615
			if (tree_content_remove(e, slash1 + 1, backup_leaf, 0)) {
1616 1617 1618 1619 1620 1621
				for (n = 0; n < e->tree->entry_count; n++) {
					if (e->tree->entries[n]->versions[1].mode) {
						hashclr(root->versions[1].sha1);
						return 1;
					}
				}
1622
				backup_leaf = NULL;
1623
				goto del_entry;
1624 1625 1626 1627 1628 1629 1630
			}
			return 0;
		}
	}
	return 0;

del_entry:
1631 1632 1633
	if (backup_leaf)
		memcpy(backup_leaf, e, sizeof(*backup_leaf));
	else if (e->tree)
1634
		release_tree_content_recursive(e->tree);
1635
	e->tree = NULL;
1636 1637 1638
	e->versions[1].mode = 0;
	hashclr(e->versions[1].sha1);
	hashclr(root->versions[1].sha1);
1639
	return 1;
1640 1641
}

1642 1643 1644
static int tree_content_get(
	struct tree_entry *root,
	const char *p,
1645 1646
	struct tree_entry *leaf,
	int allow_root)
1647
{
1648
	struct tree_content *t;
1649 1650 1651 1652
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

1653 1654
	slash1 = strchrnul(p, '/');
	n = slash1 - p;
1655
	if (!n && !allow_root)
1656
		die("Empty path component found in input");
1657

1658 1659
	if (!root->tree)
		load_tree(root);
1660 1661 1662 1663 1664 1665

	if (!n) {
		e = root;
		goto found_entry;
	}

1666
	t = root->tree;
1667 1668
	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
1669
		if (e->name->str_len == n && !strncmp_icase(p, e->name->str_dat, n)) {
1670
			if (!*slash1)
1671
				goto found_entry;
1672 1673 1674 1675
			if (!S_ISDIR(e->versions[1].mode))
				return 0;
			if (!e->tree)
				load_tree(e);
1676
			return tree_content_get(e, slash1 + 1, leaf, 0);
1677 1678 1679
		}
	}
	return 0;
1680 1681 1682 1683 1684 1685 1686 1687

found_entry:
	memcpy(leaf, e, sizeof(*leaf));
	if (e->tree && is_null_sha1(e->versions[1].sha1))
		leaf->tree = dup_tree_content(e->tree);
	else
		leaf->tree = NULL;
	return 1;
1688 1689
}

1690
static int update_branch(struct branch *b)
1691 1692
{
	static const char *msg = "fast-import";
1693
	struct ref_transaction *transaction;
1694
	unsigned char old_sha1[20];
1695
	struct strbuf err = STRBUF_INIT;
1696

1697 1698
	if (is_null_sha1(b->sha1)) {
		if (b->delete)
1699
			delete_ref(b->name, NULL, 0);
1700 1701
		return 0;
	}
1702 1703
	if (read_ref(b->name, old_sha1))
		hashclr(old_sha1);
1704 1705 1706 1707 1708
	if (!force_update && !is_null_sha1(old_sha1)) {
		struct commit *old_cmit, *new_cmit;

		old_cmit = lookup_commit_reference_gently(old_sha1, 0);
		new_cmit = lookup_commit_reference_gently(b->sha1, 0);
1709
		if (!old_cmit || !new_cmit)
1710 1711
			return error("Branch %s is missing commits.", b->name);

1712
		if (!in_merge_bases(old_cmit, new_cmit)) {
1713
			warning("Not updating %s"
1714 1715 1716 1717 1718
				" (new tip %s does not contain %s)",
				b->name, sha1_to_hex(b->sha1), sha1_to_hex(old_sha1));
			return -1;
		}
	}
1719 1720 1721
	transaction = ref_transaction_begin(&err);
	if (!transaction ||
	    ref_transaction_update(transaction, b->name, b->sha1, old_sha1,
1722
				   0, msg, &err) ||
1723
	    ref_transaction_commit(transaction, &err)) {
1724 1725 1726 1727 1728 1729 1730
		ref_transaction_free(transaction);
		error("%s", err.buf);
		strbuf_release(&err);
		return -1;
	}
	ref_transaction_free(transaction);
	strbuf_release(&err);
1731 1732 1733 1734 1735
	return 0;
}

static void dump_branches(void)
{
1736 1737 1738 1739
	unsigned int i;
	struct branch *b;

	for (i = 0; i < branch_table_sz; i++) {
1740 1741
		for (b = branch_table[i]; b; b = b->table_next_branch)
			failure |= update_branch(b);
1742 1743 1744
	}
}

1745
static void dump_tags(void)
1746 1747 1748
{
	static const char *msg = "fast-import";
	struct tag *t;
1749 1750 1751
	struct strbuf ref_name = STRBUF_INIT;
	struct strbuf err = STRBUF_INIT;
	struct ref_transaction *transaction;
1752

1753 1754 1755 1756 1757
	transaction = ref_transaction_begin(&err);
	if (!transaction) {
		failure |= error("%s", err.buf);
		goto cleanup;
	}
1758
	for (t = first_tag; t; t = t->next_tag) {
1759 1760 1761
		strbuf_reset(&ref_name);
		strbuf_addf(&ref_name, "refs/tags/%s", t->name);

1762 1763
		if (ref_transaction_update(transaction, ref_name.buf,
					   t->sha1, NULL, 0, msg, &err)) {
1764 1765 1766
			failure |= error("%s", err.buf);
			goto cleanup;
		}
1767
	}
1768
	if (ref_transaction_commit(transaction, &err))
1769 1770 1771 1772 1773 1774
		failure |= error("%s", err.buf);

 cleanup:
	ref_transaction_free(transaction);
	strbuf_release(&ref_name);
	strbuf_release(&err);
1775 1776
}

1777
static void dump_marks_helper(FILE *f,
1778
	uintmax_t base,
1779 1780
	struct mark_set *m)
{
1781
	uintmax_t k;
1782 1783 1784
	if (m->shift) {
		for (k = 0; k < 1024; k++) {
			if (m->data.sets[k])
1785
				dump_marks_helper(f, base + (k << m->shift),
1786 1787 1788 1789 1790
					m->data.sets[k]);
		}
	} else {
		for (k = 0; k < 1024; k++) {
			if (m->data.marked[k])
1791
				fprintf(f, ":%" PRIuMAX " %s\n", base + k,
1792
					sha1_to_hex(m->data.marked[k]->idx.sha1));
1793 1794 1795 1796
		}
	}
}

1797
static void dump_marks(void)
1798
{
1799 1800 1801
	static struct lock_file mark_lock;
	FILE *f;

1802
	if (!export_marks_file)
1803 1804
		return;

1805
	if (hold_lock_file_for_update(&mark_lock, export_marks_file, 0) < 0) {
1806
		failure |= error("Unable to write marks file %s: %s",
1807
			export_marks_file, strerror(errno));
1808
		return;
1809
	}
1810

1811
	f = fdopen_lock_file(&mark_lock, "w");
1812
	if (!f) {
1813
		int saved_errno = errno;
1814 1815
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1816
			export_marks_file, strerror(saved_errno));
1817
		return;
1818
	}
1819

1820 1821 1822
	dump_marks_helper(f, 0, marks);
	if (commit_lock_file(&mark_lock)) {
		failure |= error("Unable to commit marks file %s: %s",
1823
			export_marks_file, strerror(errno));
1824 1825
		return;
	}
1826 1827
}

1828 1829 1830 1831
static void read_marks(void)
{
	char line[512];
	FILE *f = fopen(import_marks_file, "r");
1832 1833 1834 1835 1836
	if (f)
		;
	else if (import_marks_file_ignore_missing && errno == ENOENT)
		return; /* Marks file does not exist */
	else
1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849
		die_errno("cannot read '%s'", import_marks_file);
	while (fgets(line, sizeof(line), f)) {
		uintmax_t mark;
		char *end;
		unsigned char sha1[20];
		struct object_entry *e;

		end = strchr(line, '\n');
		if (line[0] != ':' || !end)
			die("corrupt mark line: %s", line);
		*end = 0;
		mark = strtoumax(line + 1, &end, 10);
		if (!mark || end == line + 1
1850
			|| *end != ' ' || get_sha1_hex(end + 1, sha1))
1851 1852 1853 1854 1855 1856 1857 1858 1859
			die("corrupt mark line: %s", line);
		e = find_object(sha1);
		if (!e) {
			enum object_type type = sha1_object_info(sha1, NULL);
			if (type < 0)
				die("object not found: %s", sha1_to_hex(sha1));
			e = insert_object(sha1);
			e->type = type;
			e->pack_id = MAX_PACK_ID;
1860
			e->idx.offset = 1; /* just not zero! */
1861 1862 1863 1864 1865 1866 1867
		}
		insert_mark(mark, e);
	}
	fclose(f);
}


1868
static int read_next_command(void)
1869
{
1870 1871 1872 1873 1874 1875 1876
	static int stdin_eof = 0;

	if (stdin_eof) {
		unread_command_buf = 0;
		return EOF;
	}

1877
	for (;;) {
1878 1879
		const char *p;

1880
		if (unread_command_buf) {
1881
			unread_command_buf = 0;
1882 1883 1884
		} else {
			struct recent_command *rc;

1885
			strbuf_detach(&command_buf, NULL);
1886 1887 1888
			stdin_eof = strbuf_getline(&command_buf, stdin, '\n');
			if (stdin_eof)
				return EOF;
1889

1890
			if (!seen_data_command
1891 1892
				&& !starts_with(command_buf.buf, "feature ")
				&& !starts_with(command_buf.buf, "option ")) {
S
Sverre Rabbelier 已提交
1893
				parse_argv();
1894 1895
			}

1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911
			rc = rc_free;
			if (rc)
				rc_free = rc->next;
			else {
				rc = cmd_hist.next;
				cmd_hist.next = rc->next;
				cmd_hist.next->prev = &cmd_hist;
				free(rc->buf);
			}

			rc->buf = command_buf.buf;
			rc->prev = cmd_tail;
			rc->next = cmd_hist.prev;
			rc->prev->next = rc;
			cmd_tail = rc;
		}
1912 1913 1914 1915
		if (skip_prefix(command_buf.buf, "get-mark ", &p)) {
			parse_get_mark(p);
			continue;
		}
1916 1917
		if (skip_prefix(command_buf.buf, "cat-blob ", &p)) {
			parse_cat_blob(p);
1918 1919 1920 1921 1922 1923
			continue;
		}
		if (command_buf.buf[0] == '#')
			continue;
		return 0;
	}
1924 1925
}

J
Junio C Hamano 已提交
1926
static void skip_optional_lf(void)
1927 1928 1929 1930 1931 1932
{
	int term_char = fgetc(stdin);
	if (term_char != '\n' && term_char != EOF)
		ungetc(term_char, stdin);
}

1933
static void parse_mark(void)
1934
{
1935 1936 1937
	const char *v;
	if (skip_prefix(command_buf.buf, "mark :", &v)) {
		next_mark = strtoumax(v, NULL, 10);
1938 1939 1940
		read_next_command();
	}
	else
1941
		next_mark = 0;
1942 1943
}

1944
static int parse_data(struct strbuf *sb, uintmax_t limit, uintmax_t *len_res)
1945
{
1946
	const char *data;
1947
	strbuf_reset(sb);
1948

1949
	if (!skip_prefix(command_buf.buf, "data ", &data))
1950 1951
		die("Expected 'data n' command, found: %s", command_buf.buf);

1952 1953 1954
	if (skip_prefix(data, "<<", &data)) {
		char *term = xstrdup(data);
		size_t term_len = command_buf.len - (data - command_buf.buf);
1955

1956
		strbuf_detach(&command_buf, NULL);
1957
		for (;;) {
1958
			if (strbuf_getline(&command_buf, stdin, '\n') == EOF)
1959 1960 1961 1962
				die("EOF in data (terminator '%s' not found)", term);
			if (term_len == command_buf.len
				&& !strcmp(term, command_buf.buf))
				break;
1963 1964
			strbuf_addbuf(sb, &command_buf);
			strbuf_addch(sb, '\n');
1965 1966 1967 1968
		}
		free(term);
	}
	else {
1969
		uintmax_t len = strtoumax(data, NULL, 10);
1970
		size_t n = 0, length = (size_t)len;
1971

1972 1973 1974 1975 1976 1977
		if (limit && limit < len) {
			*len_res = len;
			return 0;
		}
		if (length < len)
			die("data is too large to use in this context");
1978

1979
		while (n < length) {
1980
			size_t s = strbuf_fread(sb, length - n, stdin);
1981
			if (!s && feof(stdin))
1982 1983
				die("EOF in data (%lu bytes remaining)",
					(unsigned long)(length - n));
1984 1985
			n += s;
		}
1986 1987
	}

1988
	skip_optional_lf();
1989
	return 1;
1990 1991
}

1992
static int validate_raw_date(const char *src, struct strbuf *result)
1993 1994
{
	const char *orig_src = src;
1995
	char *endp;
1996
	unsigned long num;
1997

1998 1999
	errno = 0;

2000 2001
	num = strtoul(src, &endp, 10);
	/* NEEDSWORK: perhaps check for reasonable values? */
2002
	if (errno || endp == src || *endp != ' ')
2003 2004 2005 2006 2007 2008
		return -1;

	src = endp + 1;
	if (*src != '-' && *src != '+')
		return -1;

2009
	num = strtoul(src + 1, &endp, 10);
2010
	if (errno || endp == src + 1 || *endp || 1400 < num)
2011 2012
		return -1;

2013
	strbuf_addstr(result, orig_src);
2014 2015 2016 2017 2018
	return 0;
}

static char *parse_ident(const char *buf)
{
2019
	const char *ltgt;
2020
	size_t name_len;
2021
	struct strbuf ident = STRBUF_INIT;
2022

2023 2024 2025 2026
	/* ensure there is a space delimiter even if there is no name */
	if (*buf == '<')
		--buf;

2027 2028 2029 2030 2031 2032 2033
	ltgt = buf + strcspn(buf, "<>");
	if (*ltgt != '<')
		die("Missing < in ident string: %s", buf);
	if (ltgt != buf && ltgt[-1] != ' ')
		die("Missing space before < in ident string: %s", buf);
	ltgt = ltgt + 1 + strcspn(ltgt + 1, "<>");
	if (*ltgt != '>')
2034
		die("Missing > in ident string: %s", buf);
2035 2036
	ltgt++;
	if (*ltgt != ' ')
2037
		die("Missing space after > in ident string: %s", buf);
2038 2039
	ltgt++;
	name_len = ltgt - buf;
2040
	strbuf_add(&ident, buf, name_len);
2041 2042 2043

	switch (whenspec) {
	case WHENSPEC_RAW:
2044
		if (validate_raw_date(ltgt, &ident) < 0)
2045
			die("Invalid raw date \"%s\" in ident: %s", ltgt, buf);
2046 2047
		break;
	case WHENSPEC_RFC2822:
2048
		if (parse_date(ltgt, &ident) < 0)
2049
			die("Invalid rfc2822 date \"%s\" in ident: %s", ltgt, buf);
2050 2051
		break;
	case WHENSPEC_NOW:
2052
		if (strcmp("now", ltgt))
2053
			die("Date in ident must be 'now': %s", buf);
2054
		datestamp(&ident);
2055 2056 2057
		break;
	}

2058
	return strbuf_detach(&ident, NULL);
2059 2060
}

2061 2062 2063 2064
static void parse_and_store_blob(
	struct last_object *last,
	unsigned char *sha1out,
	uintmax_t mark)
2065
{
P
Pierre Habouzit 已提交
2066
	static struct strbuf buf = STRBUF_INIT;
2067
	uintmax_t len;
2068

2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083
	if (parse_data(&buf, big_file_threshold, &len))
		store_object(OBJ_BLOB, &buf, last, sha1out, mark);
	else {
		if (last) {
			strbuf_release(&last->data);
			last->offset = 0;
			last->depth = 0;
		}
		stream_blob(len, sha1out, mark);
		skip_optional_lf();
	}
}

static void parse_new_blob(void)
{
2084
	read_next_command();
2085
	parse_mark();
2086
	parse_and_store_blob(&last_blob, NULL, next_mark);
2087 2088
}

2089
static void unload_one_branch(void)
2090
{
2091 2092
	while (cur_active_branches
		&& cur_active_branches >= max_active_branches) {
2093
		uintmax_t min_commit = ULONG_MAX;
2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110
		struct branch *e, *l = NULL, *p = NULL;

		for (e = active_branches; e; e = e->active_next_branch) {
			if (e->last_commit < min_commit) {
				p = l;
				min_commit = e->last_commit;
			}
			l = e;
		}

		if (p) {
			e = p->active_next_branch;
			p->active_next_branch = e->active_next_branch;
		} else {
			e = active_branches;
			active_branches = e->active_next_branch;
		}
2111
		e->active = 0;
2112 2113
		e->active_next_branch = NULL;
		if (e->branch_tree.tree) {
2114
			release_tree_content_recursive(e->branch_tree.tree);
2115 2116 2117
			e->branch_tree.tree = NULL;
		}
		cur_active_branches--;
2118 2119 2120
	}
}

2121
static void load_branch(struct branch *b)
2122
{
2123
	load_tree(&b->branch_tree);
2124 2125 2126 2127 2128 2129 2130
	if (!b->active) {
		b->active = 1;
		b->active_next_branch = active_branches;
		active_branches = b;
		cur_active_branches++;
		branch_load_count++;
	}
2131 2132
}

2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199
static unsigned char convert_num_notes_to_fanout(uintmax_t num_notes)
{
	unsigned char fanout = 0;
	while ((num_notes >>= 8))
		fanout++;
	return fanout;
}

static void construct_path_with_fanout(const char *hex_sha1,
		unsigned char fanout, char *path)
{
	unsigned int i = 0, j = 0;
	if (fanout >= 20)
		die("Too large fanout (%u)", fanout);
	while (fanout) {
		path[i++] = hex_sha1[j++];
		path[i++] = hex_sha1[j++];
		path[i++] = '/';
		fanout--;
	}
	memcpy(path + i, hex_sha1 + j, 40 - j);
	path[i + 40 - j] = '\0';
}

static uintmax_t do_change_note_fanout(
		struct tree_entry *orig_root, struct tree_entry *root,
		char *hex_sha1, unsigned int hex_sha1_len,
		char *fullpath, unsigned int fullpath_len,
		unsigned char fanout)
{
	struct tree_content *t = root->tree;
	struct tree_entry *e, leaf;
	unsigned int i, tmp_hex_sha1_len, tmp_fullpath_len;
	uintmax_t num_notes = 0;
	unsigned char sha1[20];
	char realpath[60];

	for (i = 0; t && i < t->entry_count; i++) {
		e = t->entries[i];
		tmp_hex_sha1_len = hex_sha1_len + e->name->str_len;
		tmp_fullpath_len = fullpath_len;

		/*
		 * We're interested in EITHER existing note entries (entries
		 * with exactly 40 hex chars in path, not including directory
		 * separators), OR directory entries that may contain note
		 * entries (with < 40 hex chars in path).
		 * Also, each path component in a note entry must be a multiple
		 * of 2 chars.
		 */
		if (!e->versions[1].mode ||
		    tmp_hex_sha1_len > 40 ||
		    e->name->str_len % 2)
			continue;

		/* This _may_ be a note entry, or a subdir containing notes */
		memcpy(hex_sha1 + hex_sha1_len, e->name->str_dat,
		       e->name->str_len);
		if (tmp_fullpath_len)
			fullpath[tmp_fullpath_len++] = '/';
		memcpy(fullpath + tmp_fullpath_len, e->name->str_dat,
		       e->name->str_len);
		tmp_fullpath_len += e->name->str_len;
		fullpath[tmp_fullpath_len] = '\0';

		if (tmp_hex_sha1_len == 40 && !get_sha1_hex(hex_sha1, sha1)) {
			/* This is a note entry */
2200 2201 2202 2203 2204
			if (fanout == 0xff) {
				/* Counting mode, no rename */
				num_notes++;
				continue;
			}
2205 2206 2207 2208 2209 2210 2211 2212
			construct_path_with_fanout(hex_sha1, fanout, realpath);
			if (!strcmp(fullpath, realpath)) {
				/* Note entry is in correct location */
				num_notes++;
				continue;
			}

			/* Rename fullpath to realpath */
2213
			if (!tree_content_remove(orig_root, fullpath, &leaf, 0))
2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240
				die("Failed to remove path %s", fullpath);
			tree_content_set(orig_root, realpath,
				leaf.versions[1].sha1,
				leaf.versions[1].mode,
				leaf.tree);
		} else if (S_ISDIR(e->versions[1].mode)) {
			/* This is a subdir that may contain note entries */
			if (!e->tree)
				load_tree(e);
			num_notes += do_change_note_fanout(orig_root, e,
				hex_sha1, tmp_hex_sha1_len,
				fullpath, tmp_fullpath_len, fanout);
		}

		/* The above may have reallocated the current tree_content */
		t = root->tree;
	}
	return num_notes;
}

static uintmax_t change_note_fanout(struct tree_entry *root,
		unsigned char fanout)
{
	char hex_sha1[40], path[60];
	return do_change_note_fanout(root, root, hex_sha1, 0, path, 0, fanout);
}

2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287
/*
 * Given a pointer into a string, parse a mark reference:
 *
 *   idnum ::= ':' bigint;
 *
 * Return the first character after the value in *endptr.
 *
 * Complain if the following character is not what is expected,
 * either a space or end of the string.
 */
static uintmax_t parse_mark_ref(const char *p, char **endptr)
{
	uintmax_t mark;

	assert(*p == ':');
	p++;
	mark = strtoumax(p, endptr, 10);
	if (*endptr == p)
		die("No value after ':' in mark: %s", command_buf.buf);
	return mark;
}

/*
 * Parse the mark reference, and complain if this is not the end of
 * the string.
 */
static uintmax_t parse_mark_ref_eol(const char *p)
{
	char *end;
	uintmax_t mark;

	mark = parse_mark_ref(p, &end);
	if (*end != '\0')
		die("Garbage after mark: %s", command_buf.buf);
	return mark;
}

/*
 * Parse the mark reference, demanding a trailing space.  Return a
 * pointer to the space.
 */
static uintmax_t parse_mark_ref_space(const char **p)
{
	uintmax_t mark;
	char *end;

	mark = parse_mark_ref(*p, &end);
2288
	if (*end++ != ' ')
2289 2290 2291 2292 2293
		die("Missing space after mark: %s", command_buf.buf);
	*p = end;
	return mark;
}

2294
static void file_change_m(const char *p, struct branch *b)
2295
{
2296
	static struct strbuf uq = STRBUF_INIT;
2297
	const char *endp;
2298
	struct object_entry *oe;
2299
	unsigned char sha1[20];
2300
	uint16_t mode, inline_data = 0;
2301

2302 2303 2304 2305
	p = get_mode(p, &mode);
	if (!p)
		die("Corrupt mode: %s", command_buf.buf);
	switch (mode) {
2306 2307 2308
	case 0644:
	case 0755:
		mode |= S_IFREG;
2309 2310
	case S_IFREG | 0644:
	case S_IFREG | 0755:
2311
	case S_IFLNK:
2312
	case S_IFDIR:
2313
	case S_IFGITLINK:
2314 2315 2316 2317 2318 2319
		/* ok */
		break;
	default:
		die("Corrupt mode: %s", command_buf.buf);
	}

2320
	if (*p == ':') {
2321
		oe = find_mark(parse_mark_ref_space(&p));
2322
		hashcpy(sha1, oe->idx.sha1);
2323
	} else if (skip_prefix(p, "inline ", &p)) {
2324
		inline_data = 1;
2325
		oe = NULL; /* not used with inline_data, but makes gcc happy */
2326 2327
	} else {
		if (get_sha1_hex(p, sha1))
2328
			die("Invalid dataref: %s", command_buf.buf);
2329 2330
		oe = find_object(sha1);
		p += 40;
2331
		if (*p++ != ' ')
2332
			die("Missing space after SHA1: %s", command_buf.buf);
2333
	}
2334

2335 2336
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
2337 2338
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
2339
		p = uq.buf;
2340
	}
2341

2342
	/* Git does not track empty, non-toplevel directories. */
2343
	if (S_ISDIR(mode) && !hashcmp(sha1, EMPTY_TREE_SHA1_BIN) && *p) {
2344
		tree_content_remove(&b->branch_tree, p, NULL, 0);
2345 2346 2347
		return;
	}

2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361
	if (S_ISGITLINK(mode)) {
		if (inline_data)
			die("Git links cannot be specified 'inline': %s",
				command_buf.buf);
		else if (oe) {
			if (oe->type != OBJ_COMMIT)
				die("Not a commit (actually a %s): %s",
					typename(oe->type), command_buf.buf);
		}
		/*
		 * Accept the sha1 without checking; it expected to be in
		 * another repository.
		 */
	} else if (inline_data) {
2362 2363 2364
		if (S_ISDIR(mode))
			die("Directories cannot be specified 'inline': %s",
				command_buf.buf);
2365 2366 2367 2368
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
2369
		read_next_command();
2370
		parse_and_store_blob(&last_blob, sha1, 0);
2371
	} else {
2372 2373 2374 2375
		enum object_type expected = S_ISDIR(mode) ?
						OBJ_TREE: OBJ_BLOB;
		enum object_type type = oe ? oe->type :
					sha1_object_info(sha1, NULL);
2376
		if (type < 0)
2377 2378 2379 2380 2381 2382 2383
			die("%s not found: %s",
					S_ISDIR(mode) ?  "Tree" : "Blob",
					command_buf.buf);
		if (type != expected)
			die("Not a %s (actually a %s): %s",
				typename(expected), typename(type),
				command_buf.buf);
2384
	}
2385

2386 2387 2388 2389
	if (!*p) {
		tree_content_replace(&b->branch_tree, sha1, mode, NULL);
		return;
	}
2390
	tree_content_set(&b->branch_tree, p, sha1, mode, NULL);
2391
}
2392

2393
static void file_change_d(const char *p, struct branch *b)
2394
{
2395
	static struct strbuf uq = STRBUF_INIT;
2396 2397
	const char *endp;

2398 2399
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
2400 2401
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
2402
		p = uq.buf;
2403
	}
2404
	tree_content_remove(&b->branch_tree, p, NULL, 1);
2405 2406
}

2407
static void file_change_cr(const char *s, struct branch *b, int rename)
2408
{
2409
	const char *d;
2410 2411
	static struct strbuf s_uq = STRBUF_INIT;
	static struct strbuf d_uq = STRBUF_INIT;
2412 2413 2414
	const char *endp;
	struct tree_entry leaf;

2415 2416
	strbuf_reset(&s_uq);
	if (!unquote_c_style(&s_uq, s, &endp)) {
2417 2418
		if (*endp != ' ')
			die("Missing space after source: %s", command_buf.buf);
2419
	} else {
2420 2421 2422
		endp = strchr(s, ' ');
		if (!endp)
			die("Missing space after source: %s", command_buf.buf);
2423
		strbuf_add(&s_uq, s, endp - s);
2424
	}
2425
	s = s_uq.buf;
2426 2427 2428 2429 2430 2431

	endp++;
	if (!*endp)
		die("Missing dest: %s", command_buf.buf);

	d = endp;
2432 2433
	strbuf_reset(&d_uq);
	if (!unquote_c_style(&d_uq, d, &endp)) {
2434 2435
		if (*endp)
			die("Garbage after dest in: %s", command_buf.buf);
2436
		d = d_uq.buf;
2437 2438 2439
	}

	memset(&leaf, 0, sizeof(leaf));
2440
	if (rename)
2441
		tree_content_remove(&b->branch_tree, s, &leaf, 1);
2442
	else
2443
		tree_content_get(&b->branch_tree, s, &leaf, 1);
2444 2445
	if (!leaf.versions[1].mode)
		die("Path %s not in branch", s);
2446 2447 2448 2449 2450 2451 2452
	if (!*d) {	/* C "path/to/subdir" "" */
		tree_content_replace(&b->branch_tree,
			leaf.versions[1].sha1,
			leaf.versions[1].mode,
			leaf.tree);
		return;
	}
2453 2454 2455 2456 2457 2458
	tree_content_set(&b->branch_tree, d,
		leaf.versions[1].sha1,
		leaf.versions[1].mode,
		leaf.tree);
}

2459
static void note_change_n(const char *p, struct branch *b, unsigned char *old_fanout)
2460 2461
{
	static struct strbuf uq = STRBUF_INIT;
2462
	struct object_entry *oe;
2463 2464
	struct branch *s;
	unsigned char sha1[20], commit_sha1[20];
2465
	char path[60];
2466
	uint16_t inline_data = 0;
2467
	unsigned char new_fanout;
2468

2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485
	/*
	 * When loading a branch, we don't traverse its tree to count the real
	 * number of notes (too expensive to do this for all non-note refs).
	 * This means that recently loaded notes refs might incorrectly have
	 * b->num_notes == 0, and consequently, old_fanout might be wrong.
	 *
	 * Fix this by traversing the tree and counting the number of notes
	 * when b->num_notes == 0. If the notes tree is truly empty, the
	 * calculation should not take long.
	 */
	if (b->num_notes == 0 && *old_fanout == 0) {
		/* Invoke change_note_fanout() in "counting mode". */
		b->num_notes = change_note_fanout(&b->branch_tree, 0xff);
		*old_fanout = convert_num_notes_to_fanout(b->num_notes);
	}

	/* Now parse the notemodify command. */
2486 2487
	/* <dataref> or 'inline' */
	if (*p == ':') {
2488
		oe = find_mark(parse_mark_ref_space(&p));
2489
		hashcpy(sha1, oe->idx.sha1);
2490
	} else if (skip_prefix(p, "inline ", &p)) {
2491
		inline_data = 1;
2492
		oe = NULL; /* not used with inline_data, but makes gcc happy */
2493 2494
	} else {
		if (get_sha1_hex(p, sha1))
2495
			die("Invalid dataref: %s", command_buf.buf);
2496 2497
		oe = find_object(sha1);
		p += 40;
2498
		if (*p++ != ' ')
2499
			die("Missing space after SHA1: %s", command_buf.buf);
2500 2501
	}

2502
	/* <commit-ish> */
2503 2504
	s = lookup_branch(p);
	if (s) {
2505 2506
		if (is_null_sha1(s->sha1))
			die("Can't add a note on empty branch.");
2507 2508
		hashcpy(commit_sha1, s->sha1);
	} else if (*p == ':') {
2509
		uintmax_t commit_mark = parse_mark_ref_eol(p);
2510 2511 2512
		struct object_entry *commit_oe = find_mark(commit_mark);
		if (commit_oe->type != OBJ_COMMIT)
			die("Mark :%" PRIuMAX " not a commit", commit_mark);
2513
		hashcpy(commit_sha1, commit_oe->idx.sha1);
2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529
	} else if (!get_sha1(p, commit_sha1)) {
		unsigned long size;
		char *buf = read_object_with_reference(commit_sha1,
			commit_type, &size, commit_sha1);
		if (!buf || size < 46)
			die("Not a valid commit: %s", p);
		free(buf);
	} else
		die("Invalid ref name or SHA1 expression: %s", p);

	if (inline_data) {
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
		read_next_command();
2530
		parse_and_store_blob(&last_blob, sha1, 0);
2531 2532 2533 2534
	} else if (oe) {
		if (oe->type != OBJ_BLOB)
			die("Not a blob (actually a %s): %s",
				typename(oe->type), command_buf.buf);
2535
	} else if (!is_null_sha1(sha1)) {
2536 2537 2538 2539 2540 2541 2542 2543
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
			die("Blob not found: %s", command_buf.buf);
		if (type != OBJ_BLOB)
			die("Not a blob (actually a %s): %s",
			    typename(type), command_buf.buf);
	}

2544
	construct_path_with_fanout(sha1_to_hex(commit_sha1), *old_fanout, path);
2545
	if (tree_content_remove(&b->branch_tree, path, NULL, 0))
2546 2547 2548 2549 2550 2551 2552 2553 2554
		b->num_notes--;

	if (is_null_sha1(sha1))
		return; /* nothing to insert */

	b->num_notes++;
	new_fanout = convert_num_notes_to_fanout(b->num_notes);
	construct_path_with_fanout(sha1_to_hex(commit_sha1), new_fanout, path);
	tree_content_set(&b->branch_tree, path, sha1, S_IFREG | 0644, NULL);
2555 2556
}

2557 2558 2559 2560 2561 2562
static void file_change_deleteall(struct branch *b)
{
	release_tree_content_recursive(b->branch_tree.tree);
	hashclr(b->branch_tree.versions[0].sha1);
	hashclr(b->branch_tree.versions[1].sha1);
	load_tree(&b->branch_tree);
2563
	b->num_notes = 0;
2564 2565
}

2566
static void parse_from_commit(struct branch *b, char *buf, unsigned long size)
2567 2568 2569 2570 2571 2572 2573 2574 2575 2576
{
	if (!buf || size < 46)
		die("Not a valid commit: %s", sha1_to_hex(b->sha1));
	if (memcmp("tree ", buf, 5)
		|| get_sha1_hex(buf + 5, b->branch_tree.versions[1].sha1))
		die("The commit %s is corrupt", sha1_to_hex(b->sha1));
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
}

2577
static void parse_from_existing(struct branch *b)
2578 2579 2580 2581 2582 2583 2584 2585 2586 2587
{
	if (is_null_sha1(b->sha1)) {
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
	} else {
		unsigned long size;
		char *buf;

		buf = read_object_with_reference(b->sha1,
			commit_type, &size, b->sha1);
2588
		parse_from_commit(b, buf, size);
2589 2590 2591 2592
		free(buf);
	}
}

2593
static int parse_from(struct branch *b)
2594
{
2595
	const char *from;
2596
	struct branch *s;
2597
	unsigned char sha1[20];
2598

2599
	if (!skip_prefix(command_buf.buf, "from ", &from))
2600
		return 0;
2601

2602
	hashcpy(sha1, b->branch_tree.versions[1].sha1);
2603 2604 2605 2606 2607

	s = lookup_branch(from);
	if (b == s)
		die("Can't create a branch from itself: %s", b->name);
	else if (s) {
2608
		unsigned char *t = s->branch_tree.versions[1].sha1;
2609
		hashcpy(b->sha1, s->sha1);
2610 2611
		hashcpy(b->branch_tree.versions[0].sha1, t);
		hashcpy(b->branch_tree.versions[1].sha1, t);
2612
	} else if (*from == ':') {
2613
		uintmax_t idnum = parse_mark_ref_eol(from);
2614 2615
		struct object_entry *oe = find_mark(idnum);
		if (oe->type != OBJ_COMMIT)
2616
			die("Mark :%" PRIuMAX " not a commit", idnum);
2617 2618 2619 2620 2621 2622 2623 2624 2625 2626
		if (hashcmp(b->sha1, oe->idx.sha1)) {
			hashcpy(b->sha1, oe->idx.sha1);
			if (oe->pack_id != MAX_PACK_ID) {
				unsigned long size;
				char *buf = gfi_unpack_entry(oe, &size);
				parse_from_commit(b, buf, size);
				free(buf);
			} else
				parse_from_existing(b);
		}
2627
	} else if (!get_sha1(from, b->sha1)) {
2628
		parse_from_existing(b);
2629 2630 2631
		if (is_null_sha1(b->sha1))
			b->delete = 1;
	}
2632
	else
2633 2634
		die("Invalid ref name or SHA1 expression: %s", from);

2635 2636 2637 2638 2639
	if (b->branch_tree.tree && hashcmp(sha1, b->branch_tree.versions[1].sha1)) {
		release_tree_content_recursive(b->branch_tree.tree);
		b->branch_tree.tree = NULL;
	}

2640
	read_next_command();
2641
	return 1;
2642 2643
}

2644
static struct hash_list *parse_merge(unsigned int *count)
2645
{
2646
	struct hash_list *list = NULL, **tail = &list, *n;
2647
	const char *from;
2648 2649 2650
	struct branch *s;

	*count = 0;
2651
	while (skip_prefix(command_buf.buf, "merge ", &from)) {
2652 2653 2654 2655 2656
		n = xmalloc(sizeof(*n));
		s = lookup_branch(from);
		if (s)
			hashcpy(n->sha1, s->sha1);
		else if (*from == ':') {
2657
			uintmax_t idnum = parse_mark_ref_eol(from);
2658 2659
			struct object_entry *oe = find_mark(idnum);
			if (oe->type != OBJ_COMMIT)
2660
				die("Mark :%" PRIuMAX " not a commit", idnum);
2661
			hashcpy(n->sha1, oe->idx.sha1);
2662 2663 2664
		} else if (!get_sha1(from, n->sha1)) {
			unsigned long size;
			char *buf = read_object_with_reference(n->sha1,
S
Shawn O. Pearce 已提交
2665
				commit_type, &size, n->sha1);
2666 2667 2668 2669
			if (!buf || size < 46)
				die("Not a valid commit: %s", from);
			free(buf);
		} else
2670 2671 2672
			die("Invalid ref name or SHA1 expression: %s", from);

		n->next = NULL;
2673 2674 2675
		*tail = n;
		tail = &n->next;

2676
		(*count)++;
2677 2678 2679 2680 2681
		read_next_command();
	}
	return list;
}

2682
static void parse_new_commit(const char *arg)
2683
{
2684
	static struct strbuf msg = STRBUF_INIT;
2685 2686 2687
	struct branch *b;
	char *author = NULL;
	char *committer = NULL;
2688 2689
	struct hash_list *merge_list = NULL;
	unsigned int merge_count;
2690
	unsigned char prev_fanout, new_fanout;
2691
	const char *v;
2692

2693
	b = lookup_branch(arg);
2694
	if (!b)
2695
		b = new_branch(arg);
2696 2697

	read_next_command();
2698
	parse_mark();
2699 2700
	if (skip_prefix(command_buf.buf, "author ", &v)) {
		author = parse_ident(v);
2701 2702
		read_next_command();
	}
2703 2704
	if (skip_prefix(command_buf.buf, "committer ", &v)) {
		committer = parse_ident(v);
2705 2706 2707 2708
		read_next_command();
	}
	if (!committer)
		die("Expected committer but didn't get one");
2709
	parse_data(&msg, 0, NULL);
2710
	read_next_command();
2711 2712
	parse_from(b);
	merge_list = parse_merge(&merge_count);
2713 2714

	/* ensure the branch is active/loaded */
2715
	if (!b->branch_tree.tree || !max_active_branches) {
2716 2717 2718
		unload_one_branch();
		load_branch(b);
	}
2719

2720 2721
	prev_fanout = convert_num_notes_to_fanout(b->num_notes);

2722
	/* file_change* */
2723
	while (command_buf.len > 0) {
2724 2725 2726 2727 2728 2729 2730 2731 2732 2733
		if (skip_prefix(command_buf.buf, "M ", &v))
			file_change_m(v, b);
		else if (skip_prefix(command_buf.buf, "D ", &v))
			file_change_d(v, b);
		else if (skip_prefix(command_buf.buf, "R ", &v))
			file_change_cr(v, b, 1);
		else if (skip_prefix(command_buf.buf, "C ", &v))
			file_change_cr(v, b, 0);
		else if (skip_prefix(command_buf.buf, "N ", &v))
			note_change_n(v, b, &prev_fanout);
2734 2735
		else if (!strcmp("deleteall", command_buf.buf))
			file_change_deleteall(b);
2736 2737
		else if (skip_prefix(command_buf.buf, "ls ", &v))
			parse_ls(v, b);
2738 2739 2740 2741
		else {
			unread_command_buf = 1;
			break;
		}
2742 2743
		if (read_next_command() == EOF)
			break;
2744 2745
	}

2746 2747 2748 2749
	new_fanout = convert_num_notes_to_fanout(b->num_notes);
	if (new_fanout != prev_fanout)
		b->num_notes = change_note_fanout(&b->branch_tree, new_fanout);

2750
	/* build the tree and the commit */
2751
	store_tree(&b->branch_tree);
2752 2753
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
2754 2755 2756

	strbuf_reset(&new_data);
	strbuf_addf(&new_data, "tree %s\n",
2757
		sha1_to_hex(b->branch_tree.versions[1].sha1));
2758
	if (!is_null_sha1(b->sha1))
2759
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(b->sha1));
2760 2761
	while (merge_list) {
		struct hash_list *next = merge_list->next;
2762
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(merge_list->sha1));
2763 2764 2765
		free(merge_list);
		merge_list = next;
	}
2766 2767 2768 2769 2770 2771
	strbuf_addf(&new_data,
		"author %s\n"
		"committer %s\n"
		"\n",
		author ? author : committer, committer);
	strbuf_addbuf(&new_data, &msg);
2772
	free(author);
2773 2774
	free(committer);

2775
	if (!store_object(OBJ_COMMIT, &new_data, NULL, b->sha1, next_mark))
2776
		b->pack_id = pack_id;
2777
	b->last_commit = object_count_by_type[OBJ_COMMIT];
2778 2779
}

2780
static void parse_new_tag(const char *arg)
2781
{
2782
	static struct strbuf msg = STRBUF_INIT;
2783 2784 2785 2786
	const char *from;
	char *tagger;
	struct branch *s;
	struct tag *t;
2787
	uintmax_t from_mark = 0;
2788
	unsigned char sha1[20];
2789
	enum object_type type;
2790
	const char *v;
2791 2792

	t = pool_alloc(sizeof(struct tag));
2793
	memset(t, 0, sizeof(struct tag));
2794
	t->name = pool_strdup(arg);
2795 2796 2797 2798 2799 2800 2801 2802
	if (last_tag)
		last_tag->next_tag = t;
	else
		first_tag = t;
	last_tag = t;
	read_next_command();

	/* from ... */
2803
	if (!skip_prefix(command_buf.buf, "from ", &from))
2804 2805 2806
		die("Expected from command, got %s", command_buf.buf);
	s = lookup_branch(from);
	if (s) {
2807 2808
		if (is_null_sha1(s->sha1))
			die("Can't tag an empty branch.");
2809
		hashcpy(sha1, s->sha1);
2810
		type = OBJ_COMMIT;
2811
	} else if (*from == ':') {
2812
		struct object_entry *oe;
2813
		from_mark = parse_mark_ref_eol(from);
2814
		oe = find_mark(from_mark);
2815
		type = oe->type;
2816
		hashcpy(sha1, oe->idx.sha1);
2817
	} else if (!get_sha1(from, sha1)) {
2818 2819 2820 2821 2822 2823 2824
		struct object_entry *oe = find_object(sha1);
		if (!oe) {
			type = sha1_object_info(sha1, NULL);
			if (type < 0)
				die("Not a valid object: %s", from);
		} else
			type = oe->type;
2825 2826 2827 2828 2829
	} else
		die("Invalid ref name or SHA1 expression: %s", from);
	read_next_command();

	/* tagger ... */
2830 2831
	if (skip_prefix(command_buf.buf, "tagger ", &v)) {
		tagger = parse_ident(v);
2832 2833 2834
		read_next_command();
	} else
		tagger = NULL;
2835 2836

	/* tag payload/message */
2837
	parse_data(&msg, 0, NULL);
2838 2839

	/* build the tag object */
2840
	strbuf_reset(&new_data);
2841

2842
	strbuf_addf(&new_data,
2843 2844 2845
		    "object %s\n"
		    "type %s\n"
		    "tag %s\n",
2846
		    sha1_to_hex(sha1), typename(type), t->name);
2847 2848 2849 2850
	if (tagger)
		strbuf_addf(&new_data,
			    "tagger %s\n", tagger);
	strbuf_addch(&new_data, '\n');
2851
	strbuf_addbuf(&new_data, &msg);
2852 2853
	free(tagger);

2854
	if (store_object(OBJ_TAG, &new_data, NULL, t->sha1, 0))
2855 2856 2857
		t->pack_id = MAX_PACK_ID;
	else
		t->pack_id = pack_id;
2858 2859
}

2860
static void parse_reset_branch(const char *arg)
2861 2862 2863
{
	struct branch *b;

2864
	b = lookup_branch(arg);
2865
	if (b) {
2866 2867 2868
		hashclr(b->sha1);
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
2869 2870 2871 2872 2873
		if (b->branch_tree.tree) {
			release_tree_content_recursive(b->branch_tree.tree);
			b->branch_tree.tree = NULL;
		}
	}
2874
	else
2875
		b = new_branch(arg);
2876
	read_next_command();
2877
	parse_from(b);
2878
	if (command_buf.len > 0)
2879
		unread_command_buf = 1;
2880 2881
}

2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924
static void cat_blob_write(const char *buf, unsigned long size)
{
	if (write_in_full(cat_blob_fd, buf, size) != size)
		die_errno("Write to frontend failed");
}

static void cat_blob(struct object_entry *oe, unsigned char sha1[20])
{
	struct strbuf line = STRBUF_INIT;
	unsigned long size;
	enum object_type type = 0;
	char *buf;

	if (!oe || oe->pack_id == MAX_PACK_ID) {
		buf = read_sha1_file(sha1, &type, &size);
	} else {
		type = oe->type;
		buf = gfi_unpack_entry(oe, &size);
	}

	/*
	 * Output based on batch_one_object() from cat-file.c.
	 */
	if (type <= 0) {
		strbuf_reset(&line);
		strbuf_addf(&line, "%s missing\n", sha1_to_hex(sha1));
		cat_blob_write(line.buf, line.len);
		strbuf_release(&line);
		free(buf);
		return;
	}
	if (!buf)
		die("Can't read object %s", sha1_to_hex(sha1));
	if (type != OBJ_BLOB)
		die("Object %s is a %s but a blob was expected.",
		    sha1_to_hex(sha1), typename(type));
	strbuf_reset(&line);
	strbuf_addf(&line, "%s %s %lu\n", sha1_to_hex(sha1),
						typename(type), size);
	cat_blob_write(line.buf, line.len);
	strbuf_release(&line);
	cat_blob_write(buf, size);
	cat_blob_write("\n", 1);
2925 2926 2927 2928 2929 2930
	if (oe && oe->pack_id == pack_id) {
		last_blob.offset = oe->idx.offset;
		strbuf_attach(&last_blob.data, buf, size, size);
		last_blob.depth = oe->depth;
	} else
		free(buf);
2931 2932
}

2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949
static void parse_get_mark(const char *p)
{
	struct object_entry *oe = oe;
	char output[42];

	/* get-mark SP <object> LF */
	if (*p != ':')
		die("Not a mark: %s", p);

	oe = find_mark(parse_mark_ref_eol(p));
	if (!oe)
		die("Unknown mark: %s", command_buf.buf);

	snprintf(output, sizeof(output), "%s\n", sha1_to_hex(oe->idx.sha1));
	cat_blob_write(output, 41);
}

2950
static void parse_cat_blob(const char *p)
2951 2952 2953 2954 2955 2956
{
	struct object_entry *oe = oe;
	unsigned char sha1[20];

	/* cat-blob SP <object> LF */
	if (*p == ':') {
2957
		oe = find_mark(parse_mark_ref_eol(p));
2958 2959 2960 2961 2962
		if (!oe)
			die("Unknown mark: %s", command_buf.buf);
		hashcpy(sha1, oe->idx.sha1);
	} else {
		if (get_sha1_hex(p, sha1))
2963
			die("Invalid dataref: %s", command_buf.buf);
2964 2965 2966 2967 2968 2969 2970 2971
		if (p[40])
			die("Garbage after SHA1: %s", command_buf.buf);
		oe = find_object(sha1);
	}

	cat_blob(oe, sha1);
}

D
David Barr 已提交
2972 2973 2974 2975
static struct object_entry *dereference(struct object_entry *oe,
					unsigned char sha1[20])
{
	unsigned long size;
2976
	char *buf = NULL;
D
David Barr 已提交
2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993
	if (!oe) {
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
			die("object not found: %s", sha1_to_hex(sha1));
		/* cache it! */
		oe = insert_object(sha1);
		oe->type = type;
		oe->pack_id = MAX_PACK_ID;
		oe->idx.offset = 1;
	}
	switch (oe->type) {
	case OBJ_TREE:	/* easy case. */
		return oe;
	case OBJ_COMMIT:
	case OBJ_TAG:
		break;
	default:
2994
		die("Not a tree-ish: %s", command_buf.buf);
D
David Barr 已提交
2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028
	}

	if (oe->pack_id != MAX_PACK_ID) {	/* in a pack being written */
		buf = gfi_unpack_entry(oe, &size);
	} else {
		enum object_type unused;
		buf = read_sha1_file(sha1, &unused, &size);
	}
	if (!buf)
		die("Can't load object %s", sha1_to_hex(sha1));

	/* Peel one layer. */
	switch (oe->type) {
	case OBJ_TAG:
		if (size < 40 + strlen("object ") ||
		    get_sha1_hex(buf + strlen("object "), sha1))
			die("Invalid SHA1 in tag: %s", command_buf.buf);
		break;
	case OBJ_COMMIT:
		if (size < 40 + strlen("tree ") ||
		    get_sha1_hex(buf + strlen("tree "), sha1))
			die("Invalid SHA1 in commit: %s", command_buf.buf);
	}

	free(buf);
	return find_object(sha1);
}

static struct object_entry *parse_treeish_dataref(const char **p)
{
	unsigned char sha1[20];
	struct object_entry *e;

	if (**p == ':') {	/* <mark> */
3029
		e = find_mark(parse_mark_ref_space(p));
D
David Barr 已提交
3030 3031 3032 3033 3034
		if (!e)
			die("Unknown mark: %s", command_buf.buf);
		hashcpy(sha1, e->idx.sha1);
	} else {	/* <sha1> */
		if (get_sha1_hex(*p, sha1))
3035
			die("Invalid dataref: %s", command_buf.buf);
D
David Barr 已提交
3036 3037
		e = find_object(sha1);
		*p += 40;
3038 3039
		if (*(*p)++ != ' ')
			die("Missing space after tree-ish: %s", command_buf.buf);
D
David Barr 已提交
3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066
	}

	while (!e || e->type != OBJ_TREE)
		e = dereference(e, sha1);
	return e;
}

static void print_ls(int mode, const unsigned char *sha1, const char *path)
{
	static struct strbuf line = STRBUF_INIT;

	/* See show_tree(). */
	const char *type =
		S_ISGITLINK(mode) ? commit_type :
		S_ISDIR(mode) ? tree_type :
		blob_type;

	if (!mode) {
		/* missing SP path LF */
		strbuf_reset(&line);
		strbuf_addstr(&line, "missing ");
		quote_c_style(path, &line, NULL, 0);
		strbuf_addch(&line, '\n');
	} else {
		/* mode SP type SP object_name TAB path LF */
		strbuf_reset(&line);
		strbuf_addf(&line, "%06o %s %s\t",
3067
				mode & ~NO_DELTA, type, sha1_to_hex(sha1));
D
David Barr 已提交
3068 3069 3070 3071 3072 3073
		quote_c_style(path, &line, NULL, 0);
		strbuf_addch(&line, '\n');
	}
	cat_blob_write(line.buf, line.len);
}

3074
static void parse_ls(const char *p, struct branch *b)
D
David Barr 已提交
3075 3076
{
	struct tree_entry *root = NULL;
S
Stephen Boyd 已提交
3077
	struct tree_entry leaf = {NULL};
D
David Barr 已提交
3078

3079
	/* ls SP (<tree-ish> SP)? <path> */
D
David Barr 已提交
3080 3081 3082 3083 3084 3085 3086 3087
	if (*p == '"') {
		if (!b)
			die("Not in a commit: %s", command_buf.buf);
		root = &b->branch_tree;
	} else {
		struct object_entry *e = parse_treeish_dataref(&p);
		root = new_tree_entry();
		hashcpy(root->versions[1].sha1, e->idx.sha1);
3088 3089
		if (!is_null_sha1(root->versions[1].sha1))
			root->versions[1].mode = S_IFDIR;
D
David Barr 已提交
3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101
		load_tree(root);
	}
	if (*p == '"') {
		static struct strbuf uq = STRBUF_INIT;
		const char *endp;
		strbuf_reset(&uq);
		if (unquote_c_style(&uq, p, &endp))
			die("Invalid path: %s", command_buf.buf);
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
		p = uq.buf;
	}
3102
	tree_content_get(root, p, &leaf, 1);
D
David Barr 已提交
3103 3104 3105 3106 3107 3108 3109 3110
	/*
	 * A directory in preparation would have a sha1 of zero
	 * until it is saved.  Save, for simplicity.
	 */
	if (S_ISDIR(leaf.versions[1].mode))
		store_tree(&leaf);

	print_ls(leaf.versions[1].mode, leaf.versions[1].sha1, p);
3111 3112
	if (leaf.tree)
		release_tree_content_recursive(leaf.tree);
D
David Barr 已提交
3113 3114 3115 3116
	if (!b || root != &b->branch_tree)
		release_tree_entry(root);
}

3117
static void checkpoint(void)
3118
{
3119
	checkpoint_requested = 0;
3120 3121 3122 3123 3124 3125
	if (object_count) {
		cycle_packfile();
		dump_branches();
		dump_tags();
		dump_marks();
	}
3126 3127 3128 3129 3130
}

static void parse_checkpoint(void)
{
	checkpoint_requested = 1;
3131
	skip_optional_lf();
3132 3133
}

3134
static void parse_progress(void)
3135
{
3136
	fwrite(command_buf.buf, 1, command_buf.len, stdout);
3137 3138 3139 3140 3141
	fputc('\n', stdout);
	fflush(stdout);
	skip_optional_lf();
}

3142
static char* make_fast_import_path(const char *path)
3143
{
3144 3145
	if (!relative_marks_paths || is_absolute_path(path))
		return xstrdup(path);
3146
	return xstrdup(git_path("info/fast-import/%s", path));
3147 3148
}

3149 3150
static void option_import_marks(const char *marks,
					int from_stream, int ignore_missing)
3151
{
3152 3153 3154 3155 3156 3157 3158
	if (import_marks_file) {
		if (from_stream)
			die("Only one import-marks command allowed per stream");

		/* read previous mark file */
		if(!import_marks_file_from_stream)
			read_marks();
3159
	}
3160

3161
	import_marks_file = make_fast_import_path(marks);
3162
	safe_create_leading_directories_const(import_marks_file);
3163
	import_marks_file_from_stream = from_stream;
3164
	import_marks_file_ignore_missing = ignore_missing;
3165 3166
}

3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178
static void option_date_format(const char *fmt)
{
	if (!strcmp(fmt, "raw"))
		whenspec = WHENSPEC_RAW;
	else if (!strcmp(fmt, "rfc2822"))
		whenspec = WHENSPEC_RFC2822;
	else if (!strcmp(fmt, "now"))
		whenspec = WHENSPEC_NOW;
	else
		die("unknown --date-format argument %s", fmt);
}

3179 3180 3181 3182 3183 3184 3185 3186 3187
static unsigned long ulong_arg(const char *option, const char *arg)
{
	char *endptr;
	unsigned long rv = strtoul(arg, &endptr, 0);
	if (strchr(arg, '-') || endptr == arg || *endptr)
		die("%s: argument must be a non-negative integer", option);
	return rv;
}

3188 3189
static void option_depth(const char *depth)
{
3190
	max_depth = ulong_arg("--depth", depth);
3191 3192 3193 3194 3195 3196
	if (max_depth > MAX_DEPTH)
		die("--depth cannot exceed %u", MAX_DEPTH);
}

static void option_active_branches(const char *branches)
{
3197
	max_active_branches = ulong_arg("--active-branches", branches);
3198 3199 3200 3201
}

static void option_export_marks(const char *marks)
{
3202
	export_marks_file = make_fast_import_path(marks);
3203
	safe_create_leading_directories_const(export_marks_file);
3204 3205
}

3206 3207 3208 3209 3210 3211 3212 3213
static void option_cat_blob_fd(const char *fd)
{
	unsigned long n = ulong_arg("--cat-blob-fd", fd);
	if (n > (unsigned long) INT_MAX)
		die("--cat-blob-fd cannot exceed %d", INT_MAX);
	cat_blob_fd = (int) n;
}

3214 3215 3216 3217 3218 3219 3220 3221 3222
static void option_export_pack_edges(const char *edges)
{
	if (pack_edges)
		fclose(pack_edges);
	pack_edges = fopen(edges, "a");
	if (!pack_edges)
		die_errno("Cannot open '%s'", edges);
}

S
Sverre Rabbelier 已提交
3223
static int parse_one_option(const char *option)
3224
{
3225
	if (skip_prefix(option, "max-pack-size=", &option)) {
3226
		unsigned long v;
3227
		if (!git_parse_ulong(option, &v))
3228 3229 3230 3231 3232 3233 3234 3235 3236
			return 0;
		if (v < 8192) {
			warning("max-pack-size is now in bytes, assuming --max-pack-size=%lum", v);
			v *= 1024 * 1024;
		} else if (v < 1024 * 1024) {
			warning("minimum max-pack-size is 1 MiB");
			v = 1024 * 1024;
		}
		max_packsize = v;
3237
	} else if (skip_prefix(option, "big-file-threshold=", &option)) {
3238
		unsigned long v;
3239
		if (!git_parse_ulong(option, &v))
3240 3241
			return 0;
		big_file_threshold = v;
3242 3243 3244 3245 3246 3247
	} else if (skip_prefix(option, "depth=", &option)) {
		option_depth(option);
	} else if (skip_prefix(option, "active-branches=", &option)) {
		option_active_branches(option);
	} else if (skip_prefix(option, "export-pack-edges=", &option)) {
		option_export_pack_edges(option);
3248
	} else if (starts_with(option, "quiet")) {
3249
		show_stats = 0;
3250
	} else if (starts_with(option, "stats")) {
3251 3252
		show_stats = 1;
	} else {
S
Sverre Rabbelier 已提交
3253
		return 0;
3254
	}
S
Sverre Rabbelier 已提交
3255 3256

	return 1;
3257 3258
}

3259
static int parse_one_feature(const char *feature, int from_stream)
3260
{
3261 3262 3263 3264 3265 3266 3267 3268 3269 3270
	const char *arg;

	if (skip_prefix(feature, "date-format=", &arg)) {
		option_date_format(arg);
	} else if (skip_prefix(feature, "import-marks=", &arg)) {
		option_import_marks(arg, from_stream, 0);
	} else if (skip_prefix(feature, "import-marks-if-exists=", &arg)) {
		option_import_marks(arg, from_stream, 1);
	} else if (skip_prefix(feature, "export-marks=", &arg)) {
		option_export_marks(arg);
3271 3272
	} else if (!strcmp(feature, "get-mark")) {
		; /* Don't die - this feature is supported */
3273 3274
	} else if (!strcmp(feature, "cat-blob")) {
		; /* Don't die - this feature is supported */
3275
	} else if (!strcmp(feature, "relative-marks")) {
3276
		relative_marks_paths = 1;
3277
	} else if (!strcmp(feature, "no-relative-marks")) {
3278
		relative_marks_paths = 0;
3279 3280
	} else if (!strcmp(feature, "done")) {
		require_explicit_termination = 1;
3281
	} else if (!strcmp(feature, "force")) {
3282
		force_update = 1;
D
David Barr 已提交
3283
	} else if (!strcmp(feature, "notes") || !strcmp(feature, "ls")) {
3284
		; /* do nothing; we have the feature */
3285 3286 3287 3288 3289 3290 3291
	} else {
		return 0;
	}

	return 1;
}

3292
static void parse_feature(const char *feature)
3293 3294 3295 3296
{
	if (seen_data_command)
		die("Got feature command '%s' after data command", feature);

3297
	if (parse_one_feature(feature, 1))
3298 3299 3300 3301 3302
		return;

	die("This version of fast-import does not support feature %s.", feature);
}

3303
static void parse_option(const char *option)
S
Sverre Rabbelier 已提交
3304 3305 3306 3307 3308 3309 3310 3311
{
	if (seen_data_command)
		die("Got option command '%s' after data command", option);

	if (parse_one_option(option))
		return;

	die("This version of fast-import does not support option: %s", option);
3312 3313
}

3314
static void git_pack_config(void)
3315
{
3316 3317 3318 3319
	int indexversion_value;
	unsigned long packsizelimit_value;

	if (!git_config_get_ulong("pack.depth", &max_depth)) {
3320 3321 3322
		if (max_depth > MAX_DEPTH)
			max_depth = MAX_DEPTH;
	}
3323 3324 3325 3326 3327 3328 3329
	if (!git_config_get_int("pack.compression", &pack_compression_level)) {
		if (pack_compression_level == -1)
			pack_compression_level = Z_DEFAULT_COMPRESSION;
		else if (pack_compression_level < 0 ||
			 pack_compression_level > Z_BEST_COMPRESSION)
			git_die_config("pack.compression",
					"bad pack compression level %d", pack_compression_level);
3330 3331
		pack_compression_seen = 1;
	}
3332 3333
	if (!git_config_get_int("pack.indexversion", &indexversion_value)) {
		pack_idx_opts.version = indexversion_value;
3334
		if (pack_idx_opts.version > 2)
3335 3336
			git_die_config("pack.indexversion",
					"bad pack.indexversion=%"PRIu32, pack_idx_opts.version);
3337
	}
3338 3339 3340 3341
	if (!git_config_get_ulong("pack.packsizelimit", &packsizelimit_value))
		max_packsize = packsizelimit_value;

	git_config(git_default_config, NULL);
3342 3343
}

3344
static const char fast_import_usage[] =
3345
"git fast-import [--date-format=<f>] [--max-pack-size=<n>] [--big-file-threshold=<n>] [--depth=<n>] [--active-branches=<n>] [--export-marks=<marks.file>]";
3346

S
Sverre Rabbelier 已提交
3347 3348 3349 3350 3351 3352 3353 3354 3355 3356
static void parse_argv(void)
{
	unsigned int i;

	for (i = 1; i < global_argc; i++) {
		const char *a = global_argv[i];

		if (*a != '-' || !strcmp(a, "--"))
			break;

3357 3358 3359 3360
		if (!skip_prefix(a, "--", &a))
			die("unknown option %s", a);

		if (parse_one_option(a))
S
Sverre Rabbelier 已提交
3361 3362
			continue;

3363
		if (parse_one_feature(a, 0))
S
Sverre Rabbelier 已提交
3364 3365
			continue;

3366 3367
		if (skip_prefix(a, "cat-blob-fd=", &a)) {
			option_cat_blob_fd(a);
3368 3369 3370
			continue;
		}

3371
		die("unknown option --%s", a);
S
Sverre Rabbelier 已提交
3372 3373 3374 3375 3376 3377 3378 3379 3380
	}
	if (i != global_argc)
		usage(fast_import_usage);

	seen_data_command = 1;
	if (import_marks_file)
		read_marks();
}

3381
int main(int argc, char **argv)
3382
{
3383
	unsigned int i;
3384

3385 3386
	git_extract_argv0_path(argv[0]);

3387 3388
	git_setup_gettext();

3389 3390 3391
	if (argc == 2 && !strcmp(argv[1], "-h"))
		usage(fast_import_usage);

3392
	setup_git_directory();
3393
	reset_pack_idx_option(&pack_idx_opts);
3394
	git_pack_config();
3395 3396 3397
	if (!pack_compression_seen && core_compression_seen)
		pack_compression_level = core_compression_level;

3398
	alloc_objects(object_entry_alloc);
3399
	strbuf_init(&command_buf, 0);
3400 3401 3402 3403
	atom_table = xcalloc(atom_table_sz, sizeof(struct atom_str*));
	branch_table = xcalloc(branch_table_sz, sizeof(struct branch*));
	avail_tree_table = xcalloc(avail_tree_table_sz, sizeof(struct avail_tree_content*));
	marks = pool_calloc(1, sizeof(struct mark_set));
3404

S
Sverre Rabbelier 已提交
3405 3406
	global_argc = argc;
	global_argv = argv;
3407

3408 3409 3410 3411 3412
	rc_free = pool_alloc(cmd_save * sizeof(*rc_free));
	for (i = 0; i < (cmd_save - 1); i++)
		rc_free[i].next = &rc_free[i + 1];
	rc_free[cmd_save - 1].next = NULL;

3413
	prepare_packed_git();
3414
	start_packfile();
3415
	set_die_routine(die_nicely);
3416
	set_checkpoint_signal();
3417
	while (read_next_command() != EOF) {
3418
		const char *v;
3419
		if (!strcmp("blob", command_buf.buf))
3420
			parse_new_blob();
3421 3422 3423 3424 3425 3426 3427 3428
		else if (skip_prefix(command_buf.buf, "ls ", &v))
			parse_ls(v, NULL);
		else if (skip_prefix(command_buf.buf, "commit ", &v))
			parse_new_commit(v);
		else if (skip_prefix(command_buf.buf, "tag ", &v))
			parse_new_tag(v);
		else if (skip_prefix(command_buf.buf, "reset ", &v))
			parse_reset_branch(v);
3429
		else if (!strcmp("checkpoint", command_buf.buf))
3430
			parse_checkpoint();
3431 3432
		else if (!strcmp("done", command_buf.buf))
			break;
3433
		else if (starts_with(command_buf.buf, "progress "))
3434
			parse_progress();
3435 3436 3437 3438
		else if (skip_prefix(command_buf.buf, "feature ", &v))
			parse_feature(v);
		else if (skip_prefix(command_buf.buf, "option git ", &v))
			parse_option(v);
3439
		else if (starts_with(command_buf.buf, "option "))
S
Sverre Rabbelier 已提交
3440
			/* ignore non-git options*/;
3441 3442
		else
			die("Unsupported command: %s", command_buf.buf);
3443 3444 3445

		if (checkpoint_requested)
			checkpoint();
3446
	}
S
Sverre Rabbelier 已提交
3447 3448 3449 3450 3451

	/* argv hasn't been parsed yet, do so */
	if (!seen_data_command)
		parse_argv();

3452 3453 3454
	if (require_explicit_termination && feof(stdin))
		die("stream ends early");

3455
	end_packfile();
3456

3457
	dump_branches();
3458
	dump_tags();
3459
	unkeep_all_packs();
3460
	dump_marks();
3461

3462 3463 3464
	if (pack_edges)
		fclose(pack_edges);

3465 3466 3467 3468 3469 3470 3471 3472 3473
	if (show_stats) {
		uintmax_t total_count = 0, duplicate_count = 0;
		for (i = 0; i < ARRAY_SIZE(object_count_by_type); i++)
			total_count += object_count_by_type[i];
		for (i = 0; i < ARRAY_SIZE(duplicate_count_by_type); i++)
			duplicate_count += duplicate_count_by_type[i];

		fprintf(stderr, "%s statistics:\n", argv[0]);
		fprintf(stderr, "---------------------------------------------------------------------\n");
3474 3475
		fprintf(stderr, "Alloc'd objects: %10" PRIuMAX "\n", alloc_count);
		fprintf(stderr, "Total objects:   %10" PRIuMAX " (%10" PRIuMAX " duplicates                  )\n", total_count, duplicate_count);
3476 3477 3478 3479
		fprintf(stderr, "      blobs  :   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas of %10" PRIuMAX" attempts)\n", object_count_by_type[OBJ_BLOB], duplicate_count_by_type[OBJ_BLOB], delta_count_by_type[OBJ_BLOB], delta_count_attempts_by_type[OBJ_BLOB]);
		fprintf(stderr, "      trees  :   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas of %10" PRIuMAX" attempts)\n", object_count_by_type[OBJ_TREE], duplicate_count_by_type[OBJ_TREE], delta_count_by_type[OBJ_TREE], delta_count_attempts_by_type[OBJ_TREE]);
		fprintf(stderr, "      commits:   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas of %10" PRIuMAX" attempts)\n", object_count_by_type[OBJ_COMMIT], duplicate_count_by_type[OBJ_COMMIT], delta_count_by_type[OBJ_COMMIT], delta_count_attempts_by_type[OBJ_COMMIT]);
		fprintf(stderr, "      tags   :   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas of %10" PRIuMAX" attempts)\n", object_count_by_type[OBJ_TAG], duplicate_count_by_type[OBJ_TAG], delta_count_by_type[OBJ_TAG], delta_count_attempts_by_type[OBJ_TAG]);
3480
		fprintf(stderr, "Total branches:  %10lu (%10lu loads     )\n", branch_count, branch_load_count);
3481
		fprintf(stderr, "      marks:     %10" PRIuMAX " (%10" PRIuMAX " unique    )\n", (((uintmax_t)1) << marks->shift) * 1024, marks_set_count);
3482
		fprintf(stderr, "      atoms:     %10u\n", atom_cnt);
3483
		fprintf(stderr, "Memory total:    %10" PRIuMAX " KiB\n", (total_allocd + alloc_count*sizeof(struct object_entry))/1024);
3484
		fprintf(stderr, "       pools:    %10lu KiB\n", (unsigned long)(total_allocd/1024));
3485
		fprintf(stderr, "     objects:    %10" PRIuMAX " KiB\n", (alloc_count*sizeof(struct object_entry))/1024);
3486 3487 3488 3489 3490
		fprintf(stderr, "---------------------------------------------------------------------\n");
		pack_report();
		fprintf(stderr, "---------------------------------------------------------------------\n");
		fprintf(stderr, "\n");
	}
3491

3492
	return failure ? 1 : 0;
3493
}