fast-import.c 58.9 KB
Newer Older
1 2 3 4 5 6
/*
Format of STDIN stream:

  stream ::= cmd*;

  cmd ::= new_blob
7
        | new_commit
8
        | new_tag
9
        | reset_branch
10
        | checkpoint
11
        | progress
12 13
        ;

14
  new_blob ::= 'blob' lf
15
    mark?
16 17
    file_content;
  file_content ::= data;
18

19
  new_commit ::= 'commit' sp ref_str lf
20
    mark?
21 22
    ('author' sp name '<' email '>' when lf)?
    'committer' sp name '<' email '>' when lf
23
    commit_msg
24
    ('from' sp (ref_str | hexsha1 | sha1exp_str | idnum) lf)?
25
    ('merge' sp (ref_str | hexsha1 | sha1exp_str | idnum) lf)*
26
    file_change*
27
    lf?;
28
  commit_msg ::= data;
29

30 31 32 33 34 35
  file_change ::= file_clr
    | file_del
    | file_rnm
    | file_cpy
    | file_obm
    | file_inm;
36
  file_clr ::= 'deleteall' lf;
37
  file_del ::= 'D' sp path_str lf;
38
  file_rnm ::= 'R' sp path_str sp path_str lf;
39
  file_cpy ::= 'C' sp path_str sp path_str lf;
40 41 42
  file_obm ::= 'M' sp mode sp (hexsha1 | idnum) sp path_str lf;
  file_inm ::= 'M' sp mode sp 'inline' sp path_str lf
    data;
43 44 45

  new_tag ::= 'tag' sp tag_str lf
    'from' sp (ref_str | hexsha1 | sha1exp_str | idnum) lf
46
    'tagger' sp name '<' email '>' when lf
47 48 49
    tag_msg;
  tag_msg ::= data;

50 51
  reset_branch ::= 'reset' sp ref_str lf
    ('from' sp (ref_str | hexsha1 | sha1exp_str | idnum) lf)?
52
    lf?;
53

54
  checkpoint ::= 'checkpoint' lf
55
    lf?;
56

57 58 59
  progress ::= 'progress' sp not_lf* lf
    lf?;

60 61 62
     # 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
63
     # idnum can be updated in the future to a new object by issuing
64
     # a new mark directive with the old idnum.
65
     #
66
  mark ::= 'mark' sp idnum lf;
67
  data ::= (delimited_data | exact_data)
68
    lf?;
69 70 71 72 73 74

    # 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)*
75
    delim lf;
76 77

     # note: declen indicates the length of binary_data in bytes.
78
     # declen does not include the lf preceeding the binary data.
79
     #
80 81
  exact_data ::= 'data' sp declen lf
    binary_data;
82 83 84

     # note: quoted strings are C-style quoting supporting \c for
     # common escapes of 'c' (e..g \n, \t, \\, \") or \nnn where nnn
85
     # is the signed byte value in octal.  Note that the only
86 87
     # characters which must actually be escaped to protect the
     # stream formatting is: \, " and LF.  Otherwise these values
88
     # are UTF8.
89
     #
90 91 92
  ref_str     ::= ref;
  sha1exp_str ::= sha1exp;
  tag_str     ::= tag;
93
  path_str    ::= path    | '"' quoted(path)    '"' ;
94 95
  mode        ::= '100644' | '644'
                | '100755' | '755'
J
Junio C Hamano 已提交
96
                | '120000'
97
                ;
98 99

  declen ::= # unsigned 32 bit value, ascii base10 notation;
100
  bigint ::= # unsigned integer value, ascii base10 notation;
101
  binary_data ::= # file content, not interpreted;
102

103 104 105 106
  when         ::= raw_when | rfc2822_when;
  raw_when     ::= ts sp tz;
  rfc2822_when ::= # Valid RFC 2822 date and time;

107 108
  sp ::= # ASCII space character;
  lf ::= # ASCII newline (LF) character;
109 110

     # note: a colon (':') must precede the numerical value assigned to
111
     # an idnum.  This is to distinguish it from a ref or tag name as
112
     # GIT does not permit ':' in ref or tag strings.
113
     #
114
  idnum   ::= ':' bigint;
115 116 117
  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";
118 119
  sha1exp ::= # Any valid GIT SHA1 expression;
  hexsha1 ::= # SHA1 in hexadecimal format;
120 121

     # note: name and email are UTF8 strings, however name must not
122
     # contain '<' or lf and email must not contain any of the
123
     # following: '<', '>', lf.
124
     #
125
  name  ::= # valid GIT author/committer name;
126
  email ::= # valid GIT author/committer email;
127 128
  ts    ::= # time since the epoch in seconds, ascii base10 notation;
  tz    ::= # GIT style timezone;
129 130 131 132 133 134 135 136 137 138 139

     # note: comments 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.
     #
     # In case it is not clear, the '#' that starts the comment
     # must be the first character on that the line (an lf have
     # preceeded it).
     #
  comment ::= '#' not_lf* lf;
  not_lf  ::= # Any byte that is not ASCII newline (LF);
140 141
*/

142 143 144 145
#include "builtin.h"
#include "cache.h"
#include "object.h"
#include "blob.h"
146
#include "tree.h"
147
#include "commit.h"
148 149
#include "delta.h"
#include "pack.h"
150
#include "refs.h"
151
#include "csum-file.h"
152
#include "quote.h"
153

154 155
#define PACK_ID_BITS 16
#define MAX_PACK_ID ((1<<PACK_ID_BITS)-1)
156 157
#define DEPTH_BITS 13
#define MAX_DEPTH ((1<<DEPTH_BITS)-1)
158

159 160 161
struct object_entry
{
	struct object_entry *next;
162
	uint32_t offset;
163 164 165
	uint32_t type : TYPE_BITS,
		pack_id : PACK_ID_BITS,
		depth : DEPTH_BITS;
166 167 168
	unsigned char sha1[20];
};

169
struct object_entry_pool
170
{
171
	struct object_entry_pool *next_pool;
172 173
	struct object_entry *next_free;
	struct object_entry *end;
174
	struct object_entry entries[FLEX_ARRAY]; /* more */
175 176
};

177 178 179 180 181 182
struct mark_set
{
	union {
		struct object_entry *marked[1024];
		struct mark_set *sets[1024];
	} data;
183
	unsigned int shift;
184 185
};

186 187
struct last_object
{
P
Pierre Habouzit 已提交
188
	struct strbuf data;
189
	uint32_t offset;
190
	unsigned int depth;
P
Pierre Habouzit 已提交
191
	unsigned no_swap : 1;
192 193
};

194 195 196 197 198
struct mem_pool
{
	struct mem_pool *next_pool;
	char *next_free;
	char *end;
199
	uintmax_t space[FLEX_ARRAY]; /* more */
200 201 202 203 204
};

struct atom_str
{
	struct atom_str *next_atom;
205
	unsigned short str_len;
206 207 208 209
	char str_dat[FLEX_ARRAY]; /* more */
};

struct tree_content;
210 211
struct tree_entry
{
212 213
	struct tree_content *tree;
	struct atom_str* name;
214 215
	struct tree_entry_ms
	{
216
		uint16_t mode;
217 218
		unsigned char sha1[20];
	} versions[2];
219 220
};

221
struct tree_content
222
{
223 224
	unsigned int entry_capacity; /* must match avail_tree_content */
	unsigned int entry_count;
225
	unsigned int delta_depth;
226 227 228 229 230 231 232
	struct tree_entry *entries[FLEX_ARRAY]; /* more */
};

struct avail_tree_content
{
	unsigned int entry_capacity; /* must match tree_content */
	struct avail_tree_content *next_avail;
233 234 235 236
};

struct branch
{
237 238
	struct branch *table_next_branch;
	struct branch *active_next_branch;
239
	const char *name;
240
	struct tree_entry branch_tree;
241
	uintmax_t last_commit;
242 243
	unsigned active : 1;
	unsigned pack_id : PACK_ID_BITS;
244
	unsigned char sha1[20];
245 246
};

247 248 249 250
struct tag
{
	struct tag *next_tag;
	const char *name;
251
	unsigned int pack_id;
252 253 254
	unsigned char sha1[20];
};

255 256 257 258 259
struct hash_list
{
	struct hash_list *next;
	unsigned char sha1[20];
};
260

261 262 263 264 265 266
typedef enum {
	WHENSPEC_RAW = 1,
	WHENSPEC_RFC2822,
	WHENSPEC_NOW,
} whenspec_type;

267 268 269 270 271 272 273
struct recent_command
{
	struct recent_command *prev;
	struct recent_command *next;
	char *buf;
};

274
/* Configured limits on output */
275
static unsigned long max_depth = 10;
276
static off_t max_packsize = (1LL << 32) - 1;
277
static int force_update;
278 279 280 281 282 283 284

/* 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];
285
static unsigned long object_count;
286
static unsigned long branch_count;
287
static unsigned long branch_load_count;
288
static int failure;
289
static FILE *pack_edges;
290

291 292 293 294 295
/* 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;

296
/* Atom management */
297 298 299 300 301
static unsigned int atom_table_sz = 4451;
static unsigned int atom_cnt;
static struct atom_str **atom_table;

/* The .pack file being generated */
302
static unsigned int pack_id;
303
static struct packed_git *pack_data;
304
static struct packed_git **all_packs;
305
static unsigned long pack_size;
306 307

/* Table of objects we've written. */
308
static unsigned int object_entry_alloc = 5000;
309 310
static struct object_entry_pool *blocks;
static struct object_entry *object_table[1 << 16];
311
static struct mark_set *marks;
312
static const char* mark_file;
313 314

/* Our last blob */
P
Pierre Habouzit 已提交
315
static struct last_object last_blob = { STRBUF_INIT, 0, 0, 0 };
316 317 318 319 320 321

/* 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;
322 323
static struct strbuf old_tree = STRBUF_INIT;
static struct strbuf new_tree = STRBUF_INIT;
324

325
/* Branch data */
326 327 328
static unsigned long max_active_branches = 5;
static unsigned long cur_active_branches;
static unsigned long branch_table_sz = 1039;
329 330 331
static struct branch **branch_table;
static struct branch *active_branches;

332 333 334 335
/* Tag data */
static struct tag *first_tag;
static struct tag *last_tag;

336
/* Input stream parsing */
337
static whenspec_type whenspec = WHENSPEC_RAW;
338
static struct strbuf command_buf = STRBUF_INIT;
339
static int unread_command_buf;
340 341 342 343
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;
344
static uintmax_t next_mark;
345
static struct strbuf new_data = STRBUF_INIT;
346

347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372
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);
}

373
static void write_crash_report(const char *err)
374 375 376 377 378
{
	char *loc = git_path("fast_import_crash_%d", getpid());
	FILE *rpt = fopen(loc, "w");
	struct branch *b;
	unsigned long lu;
379
	struct recent_command *rc;
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394

	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");
	fprintf(rpt, "    fast-import process: %d\n", getpid());
	fprintf(rpt, "    parent process     : %d\n", getppid());
	fprintf(rpt, "    at %s\n", show_date(time(NULL), 0, DATE_LOCAL));
	fputc('\n', rpt);

	fputs("fatal: ", rpt);
395
	fputs(err, rpt);
396 397
	fputc('\n', rpt);

398 399 400 401 402 403 404 405 406 407 408 409
	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);
	}

410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439
	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);
	}

	fputc('\n', rpt);
	fputs("-------------------\n", rpt);
	fputs("END OF CRASH REPORT\n", rpt);
	fclose(rpt);
}

static NORETURN void die_nicely(const char *err, va_list params)
{
	static int zombie;
440
	char message[2 * PATH_MAX];
441

442
	vsnprintf(message, sizeof(message), err, params);
443
	fputs("fatal: ", stderr);
444
	fputs(message, stderr);
445 446 447 448
	fputc('\n', stderr);

	if (!zombie) {
		zombie = 1;
449
		write_crash_report(message);
450 451 452
	}
	exit(128);
}
453

454
static void alloc_objects(unsigned int cnt)
455
{
456
	struct object_entry_pool *b;
457

458
	b = xmalloc(sizeof(struct object_entry_pool)
459
		+ cnt * sizeof(struct object_entry));
460
	b->next_pool = blocks;
461 462 463 464 465
	b->next_free = b->entries;
	b->end = b->entries + cnt;
	blocks = b;
	alloc_count += cnt;
}
466

467
static struct object_entry *new_object(unsigned char *sha1)
468
{
469
	struct object_entry *e;
470

471
	if (blocks->next_free == blocks->end)
472
		alloc_objects(object_entry_alloc);
473

474
	e = blocks->next_free++;
475
	hashcpy(e->sha1, sha1);
476
	return e;
477 478
}

479
static struct object_entry *find_object(unsigned char *sha1)
480 481 482 483
{
	unsigned int h = sha1[0] << 8 | sha1[1];
	struct object_entry *e;
	for (e = object_table[h]; e; e = e->next)
484
		if (!hashcmp(sha1, e->sha1))
485 486 487 488
			return e;
	return NULL;
}

489
static struct object_entry *insert_object(unsigned char *sha1)
490 491
{
	unsigned int h = sha1[0] << 8 | sha1[1];
492
	struct object_entry *e = object_table[h];
493
	struct object_entry *p = NULL;
494 495

	while (e) {
496
		if (!hashcmp(sha1, e->sha1))
497 498 499 500 501 502
			return e;
		p = e;
		e = e->next;
	}

	e = new_object(sha1);
503
	e->next = NULL;
504 505 506 507
	e->offset = 0;
	if (p)
		p->next = e;
	else
508
		object_table[h] = e;
509 510
	return e;
}
511

512 513 514 515 516 517 518 519
static unsigned int hc_str(const char *s, size_t len)
{
	unsigned int r = 0;
	while (len-- > 0)
		r = r * 31 + *s++;
	return r;
}

520
static void *pool_alloc(size_t len)
521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536
{
	struct mem_pool *p;
	void *r;

	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;
537
		p->next_free = (char *) p->space;
538 539 540 541 542
		p->end = p->next_free + mem_pool_alloc;
		mem_pool = p;
	}

	r = p->next_free;
543 544 545
	/* round out to a 'uintmax_t' alignment */
	if (len & (sizeof(uintmax_t) - 1))
		len += sizeof(uintmax_t) - (len & (sizeof(uintmax_t) - 1));
546 547 548 549
	p->next_free += len;
	return r;
}

550
static void *pool_calloc(size_t count, size_t size)
551 552 553 554 555 556 557
{
	size_t len = count * size;
	void *r = pool_alloc(len);
	memset(r, 0, len);
	return r;
}

558
static char *pool_strdup(const char *s)
559 560 561 562 563 564
{
	char *r = pool_alloc(strlen(s) + 1);
	strcpy(r, s);
	return r;
}

565
static void insert_mark(uintmax_t idnum, struct object_entry *oe)
566 567 568 569 570 571 572 573 574
{
	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) {
575
		uintmax_t i = idnum >> s->shift;
576 577 578 579 580 581 582 583 584 585 586 587
		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;
}

588
static struct object_entry *find_mark(uintmax_t idnum)
589
{
590
	uintmax_t orig_idnum = idnum;
591 592 593 594
	struct mark_set *s = marks;
	struct object_entry *oe = NULL;
	if ((idnum >> s->shift) < 1024) {
		while (s && s->shift) {
595
			uintmax_t i = idnum >> s->shift;
596 597 598 599 600 601 602
			idnum -= i << s->shift;
			s = s->data.sets[i];
		}
		if (s)
			oe = s->data.marked[idnum];
	}
	if (!oe)
603
		die("mark :%" PRIuMAX " not declared", orig_idnum);
604 605 606
	return oe;
}

607
static struct atom_str *to_atom(const char *s, unsigned short len)
608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625
{
	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;
}

626
static struct branch *lookup_branch(const char *name)
627 628 629 630 631 632 633 634 635 636
{
	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;
}

637
static struct branch *new_branch(const char *name)
638 639 640 641 642 643
{
	unsigned int hc = hc_str(name, strlen(name)) % branch_table_sz;
	struct branch* b = lookup_branch(name);

	if (b)
		die("Invalid attempt to create duplicate branch: %s", name);
644
	switch (check_ref_format(name)) {
645 646 647
	case 0: break; /* its valid */
	case CHECK_REF_FORMAT_ONELEVEL:
		break; /* valid, but too few '/', allow anyway */
648
	default:
649
		die("Branch name doesn't conform to GIT standards: %s", name);
650
	}
651 652 653 654

	b = pool_calloc(1, sizeof(struct branch));
	b->name = pool_strdup(name);
	b->table_next_branch = branch_table[hc];
655 656
	b->branch_tree.versions[0].mode = S_IFDIR;
	b->branch_tree.versions[1].mode = S_IFDIR;
657
	b->active = 0;
658
	b->pack_id = MAX_PACK_ID;
659 660 661 662 663 664 665 666 667 668 669
	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;
}

670
static struct tree_content *new_tree_content(unsigned int cnt)
671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692
{
	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;
693
	t->delta_depth = 0;
694 695 696 697 698 699 700 701
	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);
702 703 704 705 706 707
	f->next_avail = avail_tree_table[hc];
	avail_tree_table[hc] = f;
}

static void release_tree_content_recursive(struct tree_content *t)
{
708 709 710
	unsigned int i;
	for (i = 0; i < t->entry_count; i++)
		release_tree_entry(t->entries[i]);
711
	release_tree_content(t);
712 713
}

714
static struct tree_content *grow_tree_content(
715 716 717 718 719
	struct tree_content *t,
	int amt)
{
	struct tree_content *r = new_tree_content(t->entry_count + amt);
	r->entry_count = t->entry_count;
720
	r->delta_depth = t->delta_depth;
721 722 723 724 725
	memcpy(r->entries,t->entries,t->entry_count*sizeof(t->entries[0]));
	release_tree_content(t);
	return r;
}

726
static struct tree_entry *new_tree_entry(void)
727 728 729 730 731
{
	struct tree_entry *e;

	if (!avail_tree_entry) {
		unsigned int n = tree_entry_alloc;
732
		total_allocd += n * sizeof(struct tree_entry);
733
		avail_tree_entry = e = xmalloc(n * sizeof(struct tree_entry));
734
		while (n-- > 1) {
735 736 737
			*((void**)e) = e + 1;
			e++;
		}
738
		*((void**)e) = NULL;
739 740 741 742 743 744 745 746 747 748
	}

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

static void release_tree_entry(struct tree_entry *e)
{
	if (e->tree)
749
		release_tree_content_recursive(e->tree);
750 751 752 753
	*((void**)e) = avail_tree_entry;
	avail_tree_entry = e;
}

754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778
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;
}

779
static void start_packfile(void)
780
{
781
	static char tmpfile[PATH_MAX];
782
	struct packed_git *p;
783
	struct pack_header hdr;
784
	int pack_fd;
785

786
	snprintf(tmpfile, sizeof(tmpfile),
787
		"%s/tmp_pack_XXXXXX", get_object_directory());
788
	pack_fd = xmkstemp(tmpfile);
789 790
	p = xcalloc(1, sizeof(*p) + strlen(tmpfile) + 2);
	strcpy(p->pack_name, tmpfile);
791
	p->pack_fd = pack_fd;
792 793 794 795

	hdr.hdr_signature = htonl(PACK_SIGNATURE);
	hdr.hdr_version = htonl(2);
	hdr.hdr_entries = 0;
796
	write_or_die(p->pack_fd, &hdr, sizeof(hdr));
797 798

	pack_data = p;
799 800
	pack_size = sizeof(hdr);
	object_count = 0;
801 802 803

	all_packs = xrealloc(all_packs, sizeof(*all_packs) * (pack_id + 1));
	all_packs[pack_id] = p;
804 805 806 807 808 809 810 811 812
}

static int oecmp (const void *a_, const void *b_)
{
	struct object_entry *a = *((struct object_entry**)a_);
	struct object_entry *b = *((struct object_entry**)b_);
	return hashcmp(a->sha1, b->sha1);
}

813
static char *create_index(void)
814
{
815 816
	static char tmpfile[PATH_MAX];
	SHA_CTX ctx;
817 818 819
	struct sha1file *f;
	struct object_entry **idx, **c, **last, *e;
	struct object_entry_pool *o;
820
	uint32_t array[256];
821
	int i, idx_fd;
822 823 824 825 826

	/* Build the sorted table of object IDs. */
	idx = xmalloc(object_count * sizeof(struct object_entry*));
	c = idx;
	for (o = blocks; o; o = o->next_pool)
827 828 829
		for (e = o->next_free; e-- != o->entries;)
			if (pack_id == e->pack_id)
				*c++ = e;
830
	last = idx + object_count;
831 832
	if (c != last)
		die("internal consistency error creating the index");
833 834 835 836 837 838 839 840 841 842 843 844 845 846 847
	qsort(idx, object_count, sizeof(struct object_entry*), oecmp);

	/* Generate the fan-out array. */
	c = idx;
	for (i = 0; i < 256; i++) {
		struct object_entry **next = c;;
		while (next < last) {
			if ((*next)->sha1[0] != i)
				break;
			next++;
		}
		array[i] = htonl(next - idx);
		c = next;
	}

848
	snprintf(tmpfile, sizeof(tmpfile),
849
		"%s/tmp_idx_XXXXXX", get_object_directory());
850
	idx_fd = xmkstemp(tmpfile);
851
	f = sha1fd(idx_fd, tmpfile);
852
	sha1write(f, array, 256 * sizeof(int));
853
	SHA1_Init(&ctx);
854
	for (c = idx; c != last; c++) {
855
		uint32_t offset = htonl((*c)->offset);
856 857
		sha1write(f, &offset, 4);
		sha1write(f, (*c)->sha1, sizeof((*c)->sha1));
858
		SHA1_Update(&ctx, (*c)->sha1, 20);
859
	}
860
	sha1write(f, pack_data->sha1, sizeof(pack_data->sha1));
861
	sha1close(f, NULL, 1);
862
	free(idx);
863 864 865 866
	SHA1_Final(pack_data->sha1, &ctx);
	return tmpfile;
}

867
static char *keep_pack(char *curr_index_name)
868 869
{
	static char name[PATH_MAX];
870
	static const char *keep_msg = "fast-import";
871 872 873 874 875 876 877 878 879 880
	int keep_fd;

	chmod(pack_data->pack_name, 0444);
	chmod(curr_index_name, 0444);

	snprintf(name, sizeof(name), "%s/pack/pack-%s.keep",
		 get_object_directory(), sha1_to_hex(pack_data->sha1));
	keep_fd = open(name, O_RDWR|O_CREAT|O_EXCL, 0600);
	if (keep_fd < 0)
		die("cannot create keep file");
881 882 883
	write_or_die(keep_fd, keep_msg, strlen(keep_msg));
	if (close(keep_fd))
		die("failed to write keep file");
884 885 886 887 888 889 890 891 892 893 894 895 896

	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");
	return name;
}

897
static void unkeep_all_packs(void)
898 899 900 901 902 903 904 905 906 907
{
	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));
		unlink(name);
	}
908 909
}

910
static void end_packfile(void)
911
{
912 913
	struct packed_git *old_p = pack_data, *new_p;

914
	if (object_count) {
915
		char *idx_name;
916 917 918
		int i;
		struct branch *b;
		struct tag *t;
919

920
		close_pack_windows(pack_data);
921 922 923
		fixup_pack_header_footer(pack_data->pack_fd, pack_data->sha1,
				    pack_data->pack_name, object_count);
		close(pack_data->pack_fd);
924
		idx_name = keep_pack(create_index());
925 926 927 928 929

		/* Register the packfile with core git's machinary. */
		new_p = add_packed_git(idx_name, strlen(idx_name), 1);
		if (!new_p)
			die("core git rejected index %s", idx_name);
930
		all_packs[pack_id] = new_p;
931
		install_packed_git(new_p);
932 933

		/* Print the boundary */
934 935 936 937 938 939 940
		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));
				}
941
			}
942 943 944 945 946 947
			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);
948 949 950
		}

		pack_id++;
951
	}
952
	else
953
		unlink(old_p->pack_name);
954 955 956
	free(old_p);

	/* We can't carry a delta across packfiles. */
P
Pierre Habouzit 已提交
957
	strbuf_release(&last_blob.data);
958 959
	last_blob.offset = 0;
	last_blob.depth = 0;
960 961
}

962
static void cycle_packfile(void)
963 964 965 966 967
{
	end_packfile();
	start_packfile();
}

968
static size_t encode_header(
969
	enum object_type type,
970
	size_t size,
971
	unsigned char *hdr)
972 973 974 975
{
	int n = 1;
	unsigned char c;

976
	if (type < OBJ_COMMIT || type > OBJ_REF_DELTA)
977 978 979 980 981 982 983 984 985 986 987 988 989 990
		die("bad type %d", type);

	c = (type << 4) | (size & 15);
	size >>= 4;
	while (size) {
		*hdr++ = c | 0x80;
		c = size & 0x7f;
		size >>= 7;
		n++;
	}
	*hdr = c;
	return n;
}

991 992
static int store_object(
	enum object_type type,
993
	struct strbuf *dat,
994
	struct last_object *last,
995
	unsigned char *sha1out,
996
	uintmax_t mark)
997 998
{
	void *out, *delta;
999 1000 1001
	struct object_entry *e;
	unsigned char hdr[96];
	unsigned char sha1[20];
1002
	unsigned long hdrlen, deltalen;
1003 1004 1005
	SHA_CTX c;
	z_stream s;

1006
	hdrlen = sprintf((char*)hdr,"%s %lu", typename(type),
1007
		(unsigned long)dat->len) + 1;
1008 1009
	SHA1_Init(&c);
	SHA1_Update(&c, hdr, hdrlen);
1010
	SHA1_Update(&c, dat->buf, dat->len);
1011
	SHA1_Final(sha1, &c);
1012
	if (sha1out)
1013
		hashcpy(sha1out, sha1);
1014 1015

	e = insert_object(sha1);
1016 1017
	if (mark)
		insert_mark(mark, e);
1018
	if (e->offset) {
1019
		duplicate_count_by_type[type]++;
1020
		return 1;
1021 1022 1023 1024 1025 1026
	} else if (find_sha1_pack(sha1, packed_git)) {
		e->type = type;
		e->pack_id = MAX_PACK_ID;
		e->offset = 1; /* just not zero! */
		duplicate_count_by_type[type]++;
		return 1;
1027
	}
1028

P
Pierre Habouzit 已提交
1029 1030
	if (last && last->data.buf && last->depth < max_depth) {
		delta = diff_delta(last->data.buf, last->data.len,
1031
			dat->buf, dat->len,
1032
			&deltalen, 0);
1033
		if (delta && deltalen >= dat->len) {
1034 1035 1036 1037 1038
			free(delta);
			delta = NULL;
		}
	} else
		delta = NULL;
1039 1040 1041

	memset(&s, 0, sizeof(s));
	deflateInit(&s, zlib_compression_level);
1042 1043 1044 1045
	if (delta) {
		s.next_in = delta;
		s.avail_in = deltalen;
	} else {
1046 1047
		s.next_in = (void *)dat->buf;
		s.avail_in = dat->len;
1048 1049 1050 1051 1052 1053 1054 1055
	}
	s.avail_out = deflateBound(&s, s.avail_in);
	s.next_out = out = xmalloc(s.avail_out);
	while (deflate(&s, Z_FINISH) == Z_OK)
		/* nothing */;
	deflateEnd(&s);

	/* Determine if we should auto-checkpoint. */
1056
	if ((pack_size + 60 + s.total_out) > max_packsize
1057 1058 1059 1060
		|| (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;
1061
		cycle_packfile();
1062 1063 1064 1065 1066

		/* We cannot carry a delta into the new pack. */
		if (delta) {
			free(delta);
			delta = NULL;
1067 1068 1069

			memset(&s, 0, sizeof(s));
			deflateInit(&s, zlib_compression_level);
1070 1071
			s.next_in = (void *)dat->buf;
			s.avail_in = dat->len;
1072 1073 1074 1075 1076
			s.avail_out = deflateBound(&s, s.avail_in);
			s.next_out = out = xrealloc(out, s.avail_out);
			while (deflate(&s, Z_FINISH) == Z_OK)
				/* nothing */;
			deflateEnd(&s);
1077 1078 1079 1080 1081 1082 1083 1084
		}
	}

	e->type = type;
	e->pack_id = pack_id;
	e->offset = pack_size;
	object_count++;
	object_count_by_type[type]++;
1085 1086

	if (delta) {
1087 1088 1089
		unsigned long ofs = e->offset - last->offset;
		unsigned pos = sizeof(hdr) - 1;

1090
		delta_count_by_type[type]++;
1091
		e->depth = last->depth + 1;
1092 1093

		hdrlen = encode_header(OBJ_OFS_DELTA, deltalen, hdr);
1094
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1095 1096 1097 1098 1099
		pack_size += hdrlen;

		hdr[pos] = ofs & 127;
		while (ofs >>= 7)
			hdr[--pos] = 128 | (--ofs & 127);
1100
		write_or_die(pack_data->pack_fd, hdr + pos, sizeof(hdr) - pos);
1101
		pack_size += sizeof(hdr) - pos;
1102
	} else {
1103
		e->depth = 0;
1104
		hdrlen = encode_header(type, dat->len, hdr);
1105
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1106
		pack_size += hdrlen;
1107 1108
	}

1109
	write_or_die(pack_data->pack_fd, out, s.total_out);
1110
	pack_size += s.total_out;
1111 1112

	free(out);
1113
	free(delta);
1114
	if (last) {
P
Pierre Habouzit 已提交
1115 1116 1117
		if (last->no_swap) {
			last->data = *dat;
		} else {
1118
			strbuf_swap(&last->data, dat);
P
Pierre Habouzit 已提交
1119
		}
1120
		last->offset = e->offset;
1121
		last->depth = e->depth;
1122 1123 1124 1125
	}
	return 0;
}

1126 1127 1128
static void *gfi_unpack_entry(
	struct object_entry *oe,
	unsigned long *sizep)
1129
{
1130
	enum object_type type;
1131
	struct packed_git *p = all_packs[oe->pack_id];
1132 1133
	if (p == pack_data && p->pack_size < (pack_size + 20)) {
		close_pack_windows(p);
1134
		p->pack_size = pack_size + 20;
1135
	}
1136
	return unpack_entry(p, oe->offset, &type, sizep);
1137 1138
}

1139
static const char *get_mode(const char *str, uint16_t *modep)
1140 1141
{
	unsigned char c;
1142
	uint16_t mode = 0;
1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154

	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)
{
1155
	unsigned char* sha1 = root->versions[1].sha1;
1156 1157 1158 1159 1160 1161 1162
	struct object_entry *myoe;
	struct tree_content *t;
	unsigned long size;
	char *buf;
	const char *c;

	root->tree = t = new_tree_content(8);
1163
	if (is_null_sha1(sha1))
1164 1165
		return;

1166
	myoe = find_object(sha1);
1167
	if (myoe && myoe->pack_id != MAX_PACK_ID) {
1168
		if (myoe->type != OBJ_TREE)
1169
			die("Not a tree: %s", sha1_to_hex(sha1));
1170
		t->delta_depth = myoe->depth;
1171
		buf = gfi_unpack_entry(myoe, &size);
1172
	} else {
1173 1174 1175
		enum object_type type;
		buf = read_sha1_file(sha1, &type, &size);
		if (!buf || type != OBJ_TREE)
1176
			die("Can't load tree %s", sha1_to_hex(sha1));
1177 1178 1179 1180 1181 1182 1183
	}

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

		if (t->entry_count == t->entry_capacity)
1184
			root->tree = t = grow_tree_content(t, t->entry_count);
1185 1186 1187
		t->entries[t->entry_count++] = e;

		e->tree = NULL;
1188
		c = get_mode(c, &e->versions[1].mode);
1189
		if (!c)
1190 1191
			die("Corrupt mode in %s", sha1_to_hex(sha1));
		e->versions[0].mode = e->versions[1].mode;
1192
		e->name = to_atom(c, strlen(c));
1193
		c += e->name->str_len + 1;
1194 1195
		hashcpy(e->versions[0].sha1, (unsigned char*)c);
		hashcpy(e->versions[1].sha1, (unsigned char*)c);
1196 1197 1198 1199 1200
		c += 20;
	}
	free(buf);
}

1201
static int tecmp0 (const void *_a, const void *_b)
1202 1203 1204 1205
{
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
1206 1207
		a->name->str_dat, a->name->str_len, a->versions[0].mode,
		b->name->str_dat, b->name->str_len, b->versions[0].mode);
1208 1209
}

1210
static int tecmp1 (const void *_a, const void *_b)
1211
{
1212 1213 1214 1215 1216 1217 1218
	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);
}

1219
static void mktree(struct tree_content *t, int v, struct strbuf *b)
1220 1221
{
	size_t maxlen = 0;
1222 1223
	unsigned int i;

1224 1225 1226 1227
	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);
1228 1229

	for (i = 0; i < t->entry_count; i++) {
1230 1231
		if (t->entries[i]->versions[v].mode)
			maxlen += t->entries[i]->name->str_len + 34;
1232 1233
	}

1234 1235
	strbuf_reset(b);
	strbuf_grow(b, maxlen);
1236 1237
	for (i = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
1238 1239
		if (!e->versions[v].mode)
			continue;
1240 1241 1242
		strbuf_addf(b, "%o %s%c", (unsigned int)e->versions[v].mode,
					e->name->str_dat, '\0');
		strbuf_add(b, e->versions[v].sha1, 20);
1243
	}
1244 1245 1246 1247 1248 1249
}

static void store_tree(struct tree_entry *root)
{
	struct tree_content *t = root->tree;
	unsigned int i, j, del;
P
Pierre Habouzit 已提交
1250
	struct last_object lo = { STRBUF_INIT, 0, 0, /* no_swap */ 1 };
1251
	struct object_entry *le;
1252 1253 1254 1255 1256 1257 1258 1259 1260

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

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

1261
	le = find_object(root->versions[0].sha1);
P
Pierre Habouzit 已提交
1262
	if (S_ISDIR(root->versions[0].mode) && le && le->pack_id == pack_id) {
1263
		mktree(t, 0, &old_tree);
P
Pierre Habouzit 已提交
1264
		lo.data = old_tree;
1265
		lo.offset = le->offset;
1266 1267 1268
		lo.depth = t->delta_depth;
	}

1269 1270
	mktree(t, 1, &new_tree);
	store_object(OBJ_TREE, &new_tree, &lo, root->versions[1].sha1, 0);
1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284

	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;
1285 1286 1287 1288 1289 1290
}

static int tree_content_set(
	struct tree_entry *root,
	const char *p,
	const unsigned char *sha1,
1291 1292
	const uint16_t mode,
	struct tree_content *subtree)
1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303
{
	struct tree_content *t = root->tree;
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

	slash1 = strchr(p, '/');
	if (slash1)
		n = slash1 - p;
	else
		n = strlen(p);
1304 1305
	if (!n)
		die("Empty path component found in input");
1306 1307
	if (!slash1 && !S_ISDIR(mode) && subtree)
		die("Non-directories cannot have subtrees");
1308 1309 1310 1311 1312

	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
		if (e->name->str_len == n && !strncmp(p, e->name->str_dat, n)) {
			if (!slash1) {
1313 1314
				if (!S_ISDIR(mode)
						&& e->versions[1].mode == mode
1315
						&& !hashcmp(e->versions[1].sha1, sha1))
1316
					return 0;
1317 1318
				e->versions[1].mode = mode;
				hashcpy(e->versions[1].sha1, sha1);
1319
				if (e->tree)
1320
					release_tree_content_recursive(e->tree);
1321
				e->tree = subtree;
1322
				hashclr(root->versions[1].sha1);
1323 1324
				return 1;
			}
1325
			if (!S_ISDIR(e->versions[1].mode)) {
1326
				e->tree = new_tree_content(8);
1327
				e->versions[1].mode = S_IFDIR;
1328 1329 1330
			}
			if (!e->tree)
				load_tree(e);
1331
			if (tree_content_set(e, slash1 + 1, sha1, mode, subtree)) {
1332
				hashclr(root->versions[1].sha1);
1333 1334 1335 1336 1337 1338 1339
				return 1;
			}
			return 0;
		}
	}

	if (t->entry_count == t->entry_capacity)
1340
		root->tree = t = grow_tree_content(t, t->entry_count);
1341
	e = new_tree_entry();
1342
	e->name = to_atom(p, n);
1343 1344
	e->versions[0].mode = 0;
	hashclr(e->versions[0].sha1);
1345 1346 1347
	t->entries[t->entry_count++] = e;
	if (slash1) {
		e->tree = new_tree_content(8);
1348
		e->versions[1].mode = S_IFDIR;
1349
		tree_content_set(e, slash1 + 1, sha1, mode, subtree);
1350
	} else {
1351
		e->tree = subtree;
1352 1353
		e->versions[1].mode = mode;
		hashcpy(e->versions[1].sha1, sha1);
1354
	}
1355
	hashclr(root->versions[1].sha1);
1356 1357 1358
	return 1;
}

1359 1360 1361 1362
static int tree_content_remove(
	struct tree_entry *root,
	const char *p,
	struct tree_entry *backup_leaf)
1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377
{
	struct tree_content *t = root->tree;
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

	slash1 = strchr(p, '/');
	if (slash1)
		n = slash1 - p;
	else
		n = strlen(p);

	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
		if (e->name->str_len == n && !strncmp(p, e->name->str_dat, n)) {
1378
			if (!slash1 || !S_ISDIR(e->versions[1].mode))
1379 1380 1381
				goto del_entry;
			if (!e->tree)
				load_tree(e);
1382
			if (tree_content_remove(e, slash1 + 1, backup_leaf)) {
1383 1384 1385 1386 1387 1388
				for (n = 0; n < e->tree->entry_count; n++) {
					if (e->tree->entries[n]->versions[1].mode) {
						hashclr(root->versions[1].sha1);
						return 1;
					}
				}
1389
				backup_leaf = NULL;
1390
				goto del_entry;
1391 1392 1393 1394 1395 1396 1397
			}
			return 0;
		}
	}
	return 0;

del_entry:
1398 1399 1400
	if (backup_leaf)
		memcpy(backup_leaf, e, sizeof(*backup_leaf));
	else if (e->tree)
1401
		release_tree_content_recursive(e->tree);
1402
	e->tree = NULL;
1403 1404 1405
	e->versions[1].mode = 0;
	hashclr(e->versions[1].sha1);
	hashclr(root->versions[1].sha1);
1406
	return 1;
1407 1408
}

1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445
static int tree_content_get(
	struct tree_entry *root,
	const char *p,
	struct tree_entry *leaf)
{
	struct tree_content *t = root->tree;
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

	slash1 = strchr(p, '/');
	if (slash1)
		n = slash1 - p;
	else
		n = strlen(p);

	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
		if (e->name->str_len == n && !strncmp(p, e->name->str_dat, n)) {
			if (!slash1) {
				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;
			}
			if (!S_ISDIR(e->versions[1].mode))
				return 0;
			if (!e->tree)
				load_tree(e);
			return tree_content_get(e, slash1 + 1, leaf);
		}
	}
	return 0;
}

1446
static int update_branch(struct branch *b)
1447 1448
{
	static const char *msg = "fast-import";
1449 1450 1451 1452 1453
	struct ref_lock *lock;
	unsigned char old_sha1[20];

	if (read_ref(b->name, old_sha1))
		hashclr(old_sha1);
1454
	lock = lock_any_ref_for_update(b->name, old_sha1, 0);
1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466
	if (!lock)
		return error("Unable to lock %s", b->name);
	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);
		if (!old_cmit || !new_cmit) {
			unlock_ref(lock);
			return error("Branch %s is missing commits.", b->name);
		}

1467
		if (!in_merge_bases(old_cmit, &new_cmit, 1)) {
1468
			unlock_ref(lock);
1469
			warning("Not updating %s"
1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481
				" (new tip %s does not contain %s)",
				b->name, sha1_to_hex(b->sha1), sha1_to_hex(old_sha1));
			return -1;
		}
	}
	if (write_ref_sha1(lock, b->sha1, msg) < 0)
		return error("Unable to update %s", b->name);
	return 0;
}

static void dump_branches(void)
{
1482 1483 1484 1485
	unsigned int i;
	struct branch *b;

	for (i = 0; i < branch_table_sz; i++) {
1486 1487
		for (b = branch_table[i]; b; b = b->table_next_branch)
			failure |= update_branch(b);
1488 1489 1490
	}
}

1491
static void dump_tags(void)
1492 1493 1494 1495
{
	static const char *msg = "fast-import";
	struct tag *t;
	struct ref_lock *lock;
1496
	char ref_name[PATH_MAX];
1497 1498

	for (t = first_tag; t; t = t->next_tag) {
1499 1500
		sprintf(ref_name, "tags/%s", t->name);
		lock = lock_ref_sha1(ref_name, NULL);
1501
		if (!lock || write_ref_sha1(lock, t->sha1, msg) < 0)
1502
			failure |= error("Unable to update %s", ref_name);
1503 1504 1505
	}
}

1506
static void dump_marks_helper(FILE *f,
1507
	uintmax_t base,
1508 1509
	struct mark_set *m)
{
1510
	uintmax_t k;
1511 1512 1513 1514 1515 1516 1517 1518 1519
	if (m->shift) {
		for (k = 0; k < 1024; k++) {
			if (m->data.sets[k])
				dump_marks_helper(f, (base + k) << m->shift,
					m->data.sets[k]);
		}
	} else {
		for (k = 0; k < 1024; k++) {
			if (m->data.marked[k])
1520
				fprintf(f, ":%" PRIuMAX " %s\n", base + k,
1521 1522 1523 1524 1525
					sha1_to_hex(m->data.marked[k]->sha1));
		}
	}
}

1526
static void dump_marks(void)
1527
{
1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539
	static struct lock_file mark_lock;
	int mark_fd;
	FILE *f;

	if (!mark_file)
		return;

	mark_fd = hold_lock_file_for_update(&mark_lock, mark_file, 0);
	if (mark_fd < 0) {
		failure |= error("Unable to write marks file %s: %s",
			mark_file, strerror(errno));
		return;
1540
	}
1541 1542 1543

	f = fdopen(mark_fd, "w");
	if (!f) {
1544
		int saved_errno = errno;
1545 1546
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1547
			mark_file, strerror(saved_errno));
1548
		return;
1549
	}
1550

B
Brandon Casey 已提交
1551
	/*
1552 1553
	 * Since the lock file was fdopen()'ed, it should not be close()'ed.
	 * Assign -1 to the lock file descriptor so that commit_lock_file()
B
Brandon Casey 已提交
1554 1555 1556
	 * won't try to close() it.
	 */
	mark_lock.fd = -1;
1557 1558 1559

	dump_marks_helper(f, 0, marks);
	if (ferror(f) || fclose(f)) {
1560
		int saved_errno = errno;
1561 1562
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1563
			mark_file, strerror(saved_errno));
1564 1565 1566 1567
		return;
	}

	if (commit_lock_file(&mark_lock)) {
1568
		int saved_errno = errno;
1569 1570
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to commit marks file %s: %s",
1571
			mark_file, strerror(saved_errno));
1572 1573
		return;
	}
1574 1575
}

1576
static int read_next_command(void)
1577
{
1578 1579 1580 1581 1582 1583 1584
	static int stdin_eof = 0;

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

1585
	do {
1586
		if (unread_command_buf) {
1587
			unread_command_buf = 0;
1588 1589 1590
		} else {
			struct recent_command *rc;

1591
			strbuf_detach(&command_buf, NULL);
1592 1593 1594
			stdin_eof = strbuf_getline(&command_buf, stdin, '\n');
			if (stdin_eof)
				return EOF;
1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612

			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;
		}
	} while (command_buf.buf[0] == '#');
1613 1614

	return 0;
1615 1616
}

J
Junio C Hamano 已提交
1617
static void skip_optional_lf(void)
1618 1619 1620 1621 1622 1623
{
	int term_char = fgetc(stdin);
	if (term_char != '\n' && term_char != EOF)
		ungetc(term_char, stdin);
}

1624
static void cmd_mark(void)
1625
{
1626
	if (!prefixcmp(command_buf.buf, "mark :")) {
1627
		next_mark = strtoumax(command_buf.buf + 6, NULL, 10);
1628 1629 1630
		read_next_command();
	}
	else
1631
		next_mark = 0;
1632 1633
}

1634
static void cmd_data(struct strbuf *sb)
1635
{
1636
	strbuf_reset(sb);
1637

1638
	if (prefixcmp(command_buf.buf, "data "))
1639 1640
		die("Expected 'data n' command, found: %s", command_buf.buf);

1641
	if (!prefixcmp(command_buf.buf + 5, "<<")) {
1642
		char *term = xstrdup(command_buf.buf + 5 + 2);
1643 1644
		size_t term_len = command_buf.len - 5 - 2;

1645
		strbuf_detach(&command_buf, NULL);
1646
		for (;;) {
1647
			if (strbuf_getline(&command_buf, stdin, '\n') == EOF)
1648 1649 1650 1651
				die("EOF in data (terminator '%s' not found)", term);
			if (term_len == command_buf.len
				&& !strcmp(term, command_buf.buf))
				break;
1652 1653
			strbuf_addbuf(sb, &command_buf);
			strbuf_addch(sb, '\n');
1654 1655 1656 1657
		}
		free(term);
	}
	else {
1658 1659
		size_t n = 0, length;

1660
		length = strtoul(command_buf.buf + 5, NULL, 10);
1661

1662
		while (n < length) {
1663
			size_t s = strbuf_fread(sb, length - n, stdin);
1664
			if (!s && feof(stdin))
1665 1666
				die("EOF in data (%lu bytes remaining)",
					(unsigned long)(length - n));
1667 1668
			n += s;
		}
1669 1670
	}

1671
	skip_optional_lf();
1672 1673
}

1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731
static int validate_raw_date(const char *src, char *result, int maxlen)
{
	const char *orig_src = src;
	char *endp, sign;

	strtoul(src, &endp, 10);
	if (endp == src || *endp != ' ')
		return -1;

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

	strtoul(src + 1, &endp, 10);
	if (endp == src || *endp || (endp - orig_src) >= maxlen)
		return -1;

	strcpy(result, orig_src);
	return 0;
}

static char *parse_ident(const char *buf)
{
	const char *gt;
	size_t name_len;
	char *ident;

	gt = strrchr(buf, '>');
	if (!gt)
		die("Missing > in ident string: %s", buf);
	gt++;
	if (*gt != ' ')
		die("Missing space after > in ident string: %s", buf);
	gt++;
	name_len = gt - buf;
	ident = xmalloc(name_len + 24);
	strncpy(ident, buf, name_len);

	switch (whenspec) {
	case WHENSPEC_RAW:
		if (validate_raw_date(gt, ident + name_len, 24) < 0)
			die("Invalid raw date \"%s\" in ident: %s", gt, buf);
		break;
	case WHENSPEC_RFC2822:
		if (parse_date(gt, ident + name_len, 24) < 0)
			die("Invalid rfc2822 date \"%s\" in ident: %s", gt, buf);
		break;
	case WHENSPEC_NOW:
		if (strcmp("now", gt))
			die("Date in ident must be 'now': %s", buf);
		datestamp(ident + name_len, 24);
		break;
	}

	return ident;
}

1732
static void cmd_new_blob(void)
1733
{
P
Pierre Habouzit 已提交
1734
	static struct strbuf buf = STRBUF_INIT;
1735 1736 1737

	read_next_command();
	cmd_mark();
1738
	cmd_data(&buf);
P
Pierre Habouzit 已提交
1739
	store_object(OBJ_BLOB, &buf, &last_blob, NULL, next_mark);
1740 1741
}

1742
static void unload_one_branch(void)
1743
{
1744 1745
	while (cur_active_branches
		&& cur_active_branches >= max_active_branches) {
1746
		uintmax_t min_commit = ULONG_MAX;
1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763
		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;
		}
1764
		e->active = 0;
1765 1766
		e->active_next_branch = NULL;
		if (e->branch_tree.tree) {
1767
			release_tree_content_recursive(e->branch_tree.tree);
1768 1769 1770
			e->branch_tree.tree = NULL;
		}
		cur_active_branches--;
1771 1772 1773
	}
}

1774
static void load_branch(struct branch *b)
1775
{
1776
	load_tree(&b->branch_tree);
1777 1778 1779 1780 1781 1782 1783
	if (!b->active) {
		b->active = 1;
		b->active_next_branch = active_branches;
		active_branches = b;
		cur_active_branches++;
		branch_load_count++;
	}
1784 1785
}

1786
static void file_change_m(struct branch *b)
1787
{
1788
	const char *p = command_buf.buf + 2;
1789
	static struct strbuf uq = STRBUF_INIT;
1790
	const char *endp;
1791
	struct object_entry *oe = oe;
1792
	unsigned char sha1[20];
1793
	uint16_t mode, inline_data = 0;
1794

1795 1796 1797 1798 1799 1800
	p = get_mode(p, &mode);
	if (!p)
		die("Corrupt mode: %s", command_buf.buf);
	switch (mode) {
	case S_IFREG | 0644:
	case S_IFREG | 0755:
1801
	case S_IFLNK:
1802 1803 1804 1805 1806 1807 1808 1809
	case 0644:
	case 0755:
		/* ok */
		break;
	default:
		die("Corrupt mode: %s", command_buf.buf);
	}

1810 1811
	if (*p == ':') {
		char *x;
1812
		oe = find_mark(strtoumax(p + 1, &x, 10));
1813
		hashcpy(sha1, oe->sha1);
1814
		p = x;
1815
	} else if (!prefixcmp(p, "inline")) {
1816 1817
		inline_data = 1;
		p += 6;
1818 1819 1820 1821 1822 1823
	} else {
		if (get_sha1_hex(p, sha1))
			die("Invalid SHA1: %s", command_buf.buf);
		oe = find_object(sha1);
		p += 40;
	}
1824 1825 1826
	if (*p++ != ' ')
		die("Missing space after SHA1: %s", command_buf.buf);

1827 1828
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1829 1830
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1831
		p = uq.buf;
1832
	}
1833

1834
	if (inline_data) {
P
Pierre Habouzit 已提交
1835
		static struct strbuf buf = STRBUF_INIT;
1836

1837 1838 1839 1840
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
1841
		read_next_command();
1842
		cmd_data(&buf);
P
Pierre Habouzit 已提交
1843
		store_object(OBJ_BLOB, &buf, &last_blob, sha1, 0);
1844
	} else if (oe) {
1845
		if (oe->type != OBJ_BLOB)
1846
			die("Not a blob (actually a %s): %s",
1847
				typename(oe->type), command_buf.buf);
1848
	} else {
1849 1850
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
1851
			die("Blob not found: %s", command_buf.buf);
1852
		if (type != OBJ_BLOB)
1853
			die("Not a blob (actually a %s): %s",
1854
			    typename(type), command_buf.buf);
1855
	}
1856

1857
	tree_content_set(&b->branch_tree, p, sha1, S_IFREG | mode, NULL);
1858
}
1859

1860 1861
static void file_change_d(struct branch *b)
{
1862
	const char *p = command_buf.buf + 2;
1863
	static struct strbuf uq = STRBUF_INIT;
1864 1865
	const char *endp;

1866 1867
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1868 1869
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1870
		p = uq.buf;
1871
	}
1872
	tree_content_remove(&b->branch_tree, p, NULL);
1873 1874
}

1875
static void file_change_cr(struct branch *b, int rename)
1876 1877
{
	const char *s, *d;
1878 1879
	static struct strbuf s_uq = STRBUF_INIT;
	static struct strbuf d_uq = STRBUF_INIT;
1880 1881 1882 1883
	const char *endp;
	struct tree_entry leaf;

	s = command_buf.buf + 2;
1884 1885
	strbuf_reset(&s_uq);
	if (!unquote_c_style(&s_uq, s, &endp)) {
1886 1887
		if (*endp != ' ')
			die("Missing space after source: %s", command_buf.buf);
1888
	} else {
1889 1890 1891
		endp = strchr(s, ' ');
		if (!endp)
			die("Missing space after source: %s", command_buf.buf);
1892
		strbuf_add(&s_uq, s, endp - s);
1893
	}
1894
	s = s_uq.buf;
1895 1896 1897 1898 1899 1900

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

	d = endp;
1901 1902
	strbuf_reset(&d_uq);
	if (!unquote_c_style(&d_uq, d, &endp)) {
1903 1904
		if (*endp)
			die("Garbage after dest in: %s", command_buf.buf);
1905
		d = d_uq.buf;
1906 1907 1908
	}

	memset(&leaf, 0, sizeof(leaf));
1909 1910 1911 1912
	if (rename)
		tree_content_remove(&b->branch_tree, s, &leaf);
	else
		tree_content_get(&b->branch_tree, s, &leaf);
1913 1914 1915 1916 1917 1918 1919 1920
	if (!leaf.versions[1].mode)
		die("Path %s not in branch", s);
	tree_content_set(&b->branch_tree, d,
		leaf.versions[1].sha1,
		leaf.versions[1].mode,
		leaf.tree);
}

1921 1922 1923 1924 1925 1926 1927 1928
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);
}

1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955
static void cmd_from_commit(struct branch *b, char *buf, unsigned long size)
{
	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);
}

static void cmd_from_existing(struct branch *b)
{
	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);
		cmd_from_commit(b, buf, size);
		free(buf);
	}
}

1956
static int cmd_from(struct branch *b)
1957
{
1958
	const char *from;
1959 1960
	struct branch *s;

1961
	if (prefixcmp(command_buf.buf, "from "))
1962
		return 0;
1963

1964 1965 1966 1967
	if (b->branch_tree.tree) {
		release_tree_content_recursive(b->branch_tree.tree);
		b->branch_tree.tree = NULL;
	}
1968 1969 1970 1971 1972 1973

	from = strchr(command_buf.buf, ' ') + 1;
	s = lookup_branch(from);
	if (b == s)
		die("Can't create a branch from itself: %s", b->name);
	else if (s) {
1974
		unsigned char *t = s->branch_tree.versions[1].sha1;
1975
		hashcpy(b->sha1, s->sha1);
1976 1977
		hashcpy(b->branch_tree.versions[0].sha1, t);
		hashcpy(b->branch_tree.versions[1].sha1, t);
1978
	} else if (*from == ':') {
1979
		uintmax_t idnum = strtoumax(from + 1, NULL, 10);
1980 1981
		struct object_entry *oe = find_mark(idnum);
		if (oe->type != OBJ_COMMIT)
1982
			die("Mark :%" PRIuMAX " not a commit", idnum);
1983
		hashcpy(b->sha1, oe->sha1);
1984
		if (oe->pack_id != MAX_PACK_ID) {
1985
			unsigned long size;
1986 1987
			char *buf = gfi_unpack_entry(oe, &size);
			cmd_from_commit(b, buf, size);
1988
			free(buf);
1989 1990
		} else
			cmd_from_existing(b);
1991 1992 1993
	} else if (!get_sha1(from, b->sha1))
		cmd_from_existing(b);
	else
1994 1995 1996
		die("Invalid ref name or SHA1 expression: %s", from);

	read_next_command();
1997
	return 1;
1998 1999
}

2000
static struct hash_list *cmd_merge(unsigned int *count)
2001
{
2002
	struct hash_list *list = NULL, *n, *e = e;
2003
	const char *from;
2004 2005 2006
	struct branch *s;

	*count = 0;
2007
	while (!prefixcmp(command_buf.buf, "merge ")) {
2008 2009 2010 2011 2012 2013
		from = strchr(command_buf.buf, ' ') + 1;
		n = xmalloc(sizeof(*n));
		s = lookup_branch(from);
		if (s)
			hashcpy(n->sha1, s->sha1);
		else if (*from == ':') {
2014
			uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2015 2016
			struct object_entry *oe = find_mark(idnum);
			if (oe->type != OBJ_COMMIT)
2017
				die("Mark :%" PRIuMAX " not a commit", idnum);
2018
			hashcpy(n->sha1, oe->sha1);
2019 2020 2021
		} 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
2022
				commit_type, &size, n->sha1);
2023 2024 2025 2026
			if (!buf || size < 46)
				die("Not a valid commit: %s", from);
			free(buf);
		} else
2027 2028 2029 2030 2031 2032 2033 2034
			die("Invalid ref name or SHA1 expression: %s", from);

		n->next = NULL;
		if (list)
			e->next = n;
		else
			list = n;
		e = n;
2035
		(*count)++;
2036 2037 2038 2039 2040
		read_next_command();
	}
	return list;
}

2041
static void cmd_new_commit(void)
2042
{
2043
	static struct strbuf msg = STRBUF_INIT;
2044 2045 2046 2047
	struct branch *b;
	char *sp;
	char *author = NULL;
	char *committer = NULL;
2048 2049
	struct hash_list *merge_list = NULL;
	unsigned int merge_count;
2050 2051 2052 2053

	/* Obtain the branch name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	b = lookup_branch(sp);
2054
	if (!b)
2055
		b = new_branch(sp);
2056 2057 2058

	read_next_command();
	cmd_mark();
2059
	if (!prefixcmp(command_buf.buf, "author ")) {
2060
		author = parse_ident(command_buf.buf + 7);
2061 2062
		read_next_command();
	}
2063
	if (!prefixcmp(command_buf.buf, "committer ")) {
2064
		committer = parse_ident(command_buf.buf + 10);
2065 2066 2067 2068
		read_next_command();
	}
	if (!committer)
		die("Expected committer but didn't get one");
2069
	cmd_data(&msg);
2070 2071
	read_next_command();
	cmd_from(b);
2072
	merge_list = cmd_merge(&merge_count);
2073 2074

	/* ensure the branch is active/loaded */
2075
	if (!b->branch_tree.tree || !max_active_branches) {
2076 2077 2078
		unload_one_branch();
		load_branch(b);
	}
2079

2080
	/* file_change* */
2081
	while (command_buf.len > 0) {
2082
		if (!prefixcmp(command_buf.buf, "M "))
2083
			file_change_m(b);
2084
		else if (!prefixcmp(command_buf.buf, "D "))
2085
			file_change_d(b);
2086
		else if (!prefixcmp(command_buf.buf, "R "))
2087 2088 2089
			file_change_cr(b, 1);
		else if (!prefixcmp(command_buf.buf, "C "))
			file_change_cr(b, 0);
2090 2091
		else if (!strcmp("deleteall", command_buf.buf))
			file_change_deleteall(b);
2092 2093 2094 2095
		else {
			unread_command_buf = 1;
			break;
		}
2096 2097
		if (read_next_command() == EOF)
			break;
2098 2099
	}

2100
	/* build the tree and the commit */
2101
	store_tree(&b->branch_tree);
2102 2103
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
2104 2105 2106

	strbuf_reset(&new_data);
	strbuf_addf(&new_data, "tree %s\n",
2107
		sha1_to_hex(b->branch_tree.versions[1].sha1));
2108
	if (!is_null_sha1(b->sha1))
2109
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(b->sha1));
2110 2111
	while (merge_list) {
		struct hash_list *next = merge_list->next;
2112
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(merge_list->sha1));
2113 2114 2115
		free(merge_list);
		merge_list = next;
	}
2116 2117 2118 2119 2120 2121
	strbuf_addf(&new_data,
		"author %s\n"
		"committer %s\n"
		"\n",
		author ? author : committer, committer);
	strbuf_addbuf(&new_data, &msg);
2122
	free(author);
2123 2124
	free(committer);

2125
	if (!store_object(OBJ_COMMIT, &new_data, NULL, b->sha1, next_mark))
2126
		b->pack_id = pack_id;
2127
	b->last_commit = object_count_by_type[OBJ_COMMIT];
2128 2129
}

2130
static void cmd_new_tag(void)
2131
{
2132
	static struct strbuf msg = STRBUF_INIT;
2133 2134 2135 2136 2137
	char *sp;
	const char *from;
	char *tagger;
	struct branch *s;
	struct tag *t;
2138
	uintmax_t from_mark = 0;
2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153
	unsigned char sha1[20];

	/* Obtain the new tag name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	t = pool_alloc(sizeof(struct tag));
	t->next_tag = NULL;
	t->name = pool_strdup(sp);
	if (last_tag)
		last_tag->next_tag = t;
	else
		first_tag = t;
	last_tag = t;
	read_next_command();

	/* from ... */
2154
	if (prefixcmp(command_buf.buf, "from "))
2155 2156 2157 2158
		die("Expected from command, got %s", command_buf.buf);
	from = strchr(command_buf.buf, ' ') + 1;
	s = lookup_branch(from);
	if (s) {
2159
		hashcpy(sha1, s->sha1);
2160
	} else if (*from == ':') {
2161
		struct object_entry *oe;
2162
		from_mark = strtoumax(from + 1, NULL, 10);
2163
		oe = find_mark(from_mark);
2164
		if (oe->type != OBJ_COMMIT)
2165
			die("Mark :%" PRIuMAX " not a commit", from_mark);
2166
		hashcpy(sha1, oe->sha1);
2167 2168 2169 2170 2171
	} else if (!get_sha1(from, sha1)) {
		unsigned long size;
		char *buf;

		buf = read_object_with_reference(sha1,
2172
			commit_type, &size, sha1);
2173 2174 2175 2176 2177 2178 2179 2180
		if (!buf || size < 46)
			die("Not a valid commit: %s", from);
		free(buf);
	} else
		die("Invalid ref name or SHA1 expression: %s", from);
	read_next_command();

	/* tagger ... */
2181
	if (prefixcmp(command_buf.buf, "tagger "))
2182
		die("Expected tagger command, got %s", command_buf.buf);
2183
	tagger = parse_ident(command_buf.buf + 7);
2184 2185 2186

	/* tag payload/message */
	read_next_command();
2187
	cmd_data(&msg);
2188 2189

	/* build the tag object */
2190 2191 2192 2193 2194 2195 2196 2197 2198
	strbuf_reset(&new_data);
	strbuf_addf(&new_data,
		"object %s\n"
		"type %s\n"
		"tag %s\n"
		"tagger %s\n"
		"\n",
		sha1_to_hex(sha1), commit_type, t->name, tagger);
	strbuf_addbuf(&new_data, &msg);
2199 2200
	free(tagger);

2201
	if (store_object(OBJ_TAG, &new_data, NULL, t->sha1, 0))
2202 2203 2204
		t->pack_id = MAX_PACK_ID;
	else
		t->pack_id = pack_id;
2205 2206
}

2207
static void cmd_reset_branch(void)
2208 2209 2210 2211 2212 2213 2214 2215
{
	struct branch *b;
	char *sp;

	/* Obtain the branch name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	b = lookup_branch(sp);
	if (b) {
2216 2217 2218
		hashclr(b->sha1);
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
2219 2220 2221 2222 2223
		if (b->branch_tree.tree) {
			release_tree_content_recursive(b->branch_tree.tree);
			b->branch_tree.tree = NULL;
		}
	}
2224 2225 2226
	else
		b = new_branch(sp);
	read_next_command();
2227
	if (!cmd_from(b) && command_buf.len > 0)
2228
		unread_command_buf = 1;
2229 2230
}

2231
static void cmd_checkpoint(void)
2232
{
2233 2234 2235 2236 2237 2238
	if (object_count) {
		cycle_packfile();
		dump_branches();
		dump_tags();
		dump_marks();
	}
2239
	skip_optional_lf();
2240 2241
}

2242 2243
static void cmd_progress(void)
{
2244
	fwrite(command_buf.buf, 1, command_buf.len, stdout);
2245 2246 2247 2248 2249
	fputc('\n', stdout);
	fflush(stdout);
	skip_optional_lf();
}

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
static void import_marks(const char *input_file)
{
	char line[512];
	FILE *f = fopen(input_file, "r");
	if (!f)
		die("cannot read %s: %s", input_file, strerror(errno));
	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
			|| *end != ' ' || get_sha1(end + 1, sha1))
			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;
2278
			e->offset = 1; /* just not zero! */
2279 2280 2281 2282 2283 2284
		}
		insert_mark(mark, e);
	}
	fclose(f);
}

2285
static const char fast_import_usage[] =
2286
"git-fast-import [--date-format=f] [--max-pack-size=n] [--depth=n] [--active-branches=n] [--export-marks=marks.file]";
2287

2288 2289
int main(int argc, const char **argv)
{
2290
	unsigned int i, show_stats = 1;
2291

2292
	git_config(git_default_config);
2293
	alloc_objects(object_entry_alloc);
2294
	strbuf_init(&command_buf, 0);
2295 2296 2297 2298
	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));
2299

2300 2301 2302 2303 2304
	for (i = 1; i < argc; i++) {
		const char *a = argv[i];

		if (*a != '-' || !strcmp(a, "--"))
			break;
2305
		else if (!prefixcmp(a, "--date-format=")) {
2306 2307 2308 2309 2310 2311 2312 2313 2314 2315
			const char *fmt = a + 14;
			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);
		}
2316
		else if (!prefixcmp(a, "--max-pack-size="))
2317
			max_packsize = strtoumax(a + 16, NULL, 0) * 1024 * 1024;
2318
		else if (!prefixcmp(a, "--depth=")) {
2319
			max_depth = strtoul(a + 8, NULL, 0);
2320 2321 2322
			if (max_depth > MAX_DEPTH)
				die("--depth cannot exceed %u", MAX_DEPTH);
		}
2323
		else if (!prefixcmp(a, "--active-branches="))
2324
			max_active_branches = strtoul(a + 18, NULL, 0);
2325 2326
		else if (!prefixcmp(a, "--import-marks="))
			import_marks(a + 15);
2327
		else if (!prefixcmp(a, "--export-marks="))
2328
			mark_file = a + 15;
2329
		else if (!prefixcmp(a, "--export-pack-edges=")) {
2330 2331 2332 2333 2334 2335
			if (pack_edges)
				fclose(pack_edges);
			pack_edges = fopen(a + 20, "a");
			if (!pack_edges)
				die("Cannot open %s: %s", a + 20, strerror(errno));
		} else if (!strcmp(a, "--force"))
2336
			force_update = 1;
2337 2338 2339 2340
		else if (!strcmp(a, "--quiet"))
			show_stats = 0;
		else if (!strcmp(a, "--stats"))
			show_stats = 1;
2341 2342 2343
		else
			die("unknown option %s", a);
	}
2344
	if (i != argc)
2345 2346
		usage(fast_import_usage);

2347 2348 2349 2350 2351
	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;

2352
	prepare_packed_git();
2353
	start_packfile();
2354
	set_die_routine(die_nicely);
2355 2356
	while (read_next_command() != EOF) {
		if (!strcmp("blob", command_buf.buf))
2357
			cmd_new_blob();
2358
		else if (!prefixcmp(command_buf.buf, "commit "))
2359
			cmd_new_commit();
2360
		else if (!prefixcmp(command_buf.buf, "tag "))
2361
			cmd_new_tag();
2362
		else if (!prefixcmp(command_buf.buf, "reset "))
2363
			cmd_reset_branch();
2364 2365
		else if (!strcmp("checkpoint", command_buf.buf))
			cmd_checkpoint();
2366 2367
		else if (!prefixcmp(command_buf.buf, "progress "))
			cmd_progress();
2368 2369
		else
			die("Unsupported command: %s", command_buf.buf);
2370
	}
2371
	end_packfile();
2372

2373
	dump_branches();
2374
	dump_tags();
2375
	unkeep_all_packs();
2376
	dump_marks();
2377

2378 2379 2380
	if (pack_edges)
		fclose(pack_edges);

2381 2382 2383 2384 2385 2386 2387 2388 2389
	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");
2390 2391 2392 2393 2394 2395
		fprintf(stderr, "Alloc'd objects: %10" PRIuMAX "\n", alloc_count);
		fprintf(stderr, "Total objects:   %10" PRIuMAX " (%10" PRIuMAX " duplicates                  )\n", total_count, duplicate_count);
		fprintf(stderr, "      blobs  :   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas)\n", object_count_by_type[OBJ_BLOB], duplicate_count_by_type[OBJ_BLOB], delta_count_by_type[OBJ_BLOB]);
		fprintf(stderr, "      trees  :   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas)\n", object_count_by_type[OBJ_TREE], duplicate_count_by_type[OBJ_TREE], delta_count_by_type[OBJ_TREE]);
		fprintf(stderr, "      commits:   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas)\n", object_count_by_type[OBJ_COMMIT], duplicate_count_by_type[OBJ_COMMIT], delta_count_by_type[OBJ_COMMIT]);
		fprintf(stderr, "      tags   :   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas)\n", object_count_by_type[OBJ_TAG], duplicate_count_by_type[OBJ_TAG], delta_count_by_type[OBJ_TAG]);
2396
		fprintf(stderr, "Total branches:  %10lu (%10lu loads     )\n", branch_count, branch_load_count);
2397
		fprintf(stderr, "      marks:     %10" PRIuMAX " (%10" PRIuMAX " unique    )\n", (((uintmax_t)1) << marks->shift) * 1024, marks_set_count);
2398
		fprintf(stderr, "      atoms:     %10u\n", atom_cnt);
2399
		fprintf(stderr, "Memory total:    %10" PRIuMAX " KiB\n", (total_allocd + alloc_count*sizeof(struct object_entry))/1024);
2400
		fprintf(stderr, "       pools:    %10lu KiB\n", (unsigned long)(total_allocd/1024));
2401
		fprintf(stderr, "     objects:    %10" PRIuMAX " KiB\n", (alloc_count*sizeof(struct object_entry))/1024);
2402 2403 2404 2405 2406
		fprintf(stderr, "---------------------------------------------------------------------\n");
		pack_report();
		fprintf(stderr, "---------------------------------------------------------------------\n");
		fprintf(stderr, "\n");
	}
2407

2408
	return failure ? 1 : 0;
2409
}