fast-import.c 61.7 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
static int pack_compression_level = Z_DEFAULT_COMPRESSION;
static int pack_compression_seen;
280 281 282 283 284 285 286

/* 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];
287
static unsigned long object_count;
288
static unsigned long branch_count;
289
static unsigned long branch_load_count;
290
static int failure;
291
static FILE *pack_edges;
292

293 294 295 296 297
/* 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;

298
/* Atom management */
299 300 301 302 303
static unsigned int atom_table_sz = 4451;
static unsigned int atom_cnt;
static struct atom_str **atom_table;

/* The .pack file being generated */
304
static unsigned int pack_id;
305
static struct packed_git *pack_data;
306
static struct packed_git **all_packs;
307
static unsigned long pack_size;
308 309

/* Table of objects we've written. */
310
static unsigned int object_entry_alloc = 5000;
311 312
static struct object_entry_pool *blocks;
static struct object_entry *object_table[1 << 16];
313
static struct mark_set *marks;
314
static const char* mark_file;
315 316

/* Our last blob */
P
Pierre Habouzit 已提交
317
static struct last_object last_blob = { STRBUF_INIT, 0, 0, 0 };
318 319 320 321 322 323

/* 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;
324 325
static struct strbuf old_tree = STRBUF_INIT;
static struct strbuf new_tree = STRBUF_INIT;
326

327
/* Branch data */
328 329 330
static unsigned long max_active_branches = 5;
static unsigned long cur_active_branches;
static unsigned long branch_table_sz = 1039;
331 332 333
static struct branch **branch_table;
static struct branch *active_branches;

334 335 336 337
/* Tag data */
static struct tag *first_tag;
static struct tag *last_tag;

338
/* Input stream parsing */
339
static whenspec_type whenspec = WHENSPEC_RAW;
340
static struct strbuf command_buf = STRBUF_INIT;
341
static int unread_command_buf;
342 343 344 345
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;
346
static uintmax_t next_mark;
347
static struct strbuf new_data = STRBUF_INIT;
348

349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
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);
}

375 376
static void dump_marks_helper(FILE *, uintmax_t, struct mark_set *);

377
static void write_crash_report(const char *err)
378 379 380 381 382
{
	char *loc = git_path("fast_import_crash_%d", getpid());
	FILE *rpt = fopen(loc, "w");
	struct branch *b;
	unsigned long lu;
383
	struct recent_command *rc;
384 385 386 387 388 389 390 391 392 393 394 395 396 397 398

	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);
399
	fputs(err, rpt);
400 401
	fputc('\n', rpt);

402 403 404 405 406 407 408 409 410 411 412 413
	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);
	}

414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434
	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);
	}

435 436 437 438 439 440 441 442 443 444 445 446 447
	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);
		}
	}

448 449 450 451 452 453 454 455
	fputc('\n', rpt);
	fputs("Marks\n", rpt);
	fputs("-----\n", rpt);
	if (mark_file)
		fprintf(rpt, "  exported to %s\n", mark_file);
	else
		dump_marks_helper(rpt, 0, marks);

456 457 458 459 460 461 462 463 464
	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;
465
	char message[2 * PATH_MAX];
466

467
	vsnprintf(message, sizeof(message), err, params);
468
	fputs("fatal: ", stderr);
469
	fputs(message, stderr);
470 471 472 473
	fputc('\n', stderr);

	if (!zombie) {
		zombie = 1;
474
		write_crash_report(message);
475 476 477
	}
	exit(128);
}
478

479
static void alloc_objects(unsigned int cnt)
480
{
481
	struct object_entry_pool *b;
482

483
	b = xmalloc(sizeof(struct object_entry_pool)
484
		+ cnt * sizeof(struct object_entry));
485
	b->next_pool = blocks;
486 487 488 489 490
	b->next_free = b->entries;
	b->end = b->entries + cnt;
	blocks = b;
	alloc_count += cnt;
}
491

492
static struct object_entry *new_object(unsigned char *sha1)
493
{
494
	struct object_entry *e;
495

496
	if (blocks->next_free == blocks->end)
497
		alloc_objects(object_entry_alloc);
498

499
	e = blocks->next_free++;
500
	hashcpy(e->sha1, sha1);
501
	return e;
502 503
}

504
static struct object_entry *find_object(unsigned char *sha1)
505 506 507 508
{
	unsigned int h = sha1[0] << 8 | sha1[1];
	struct object_entry *e;
	for (e = object_table[h]; e; e = e->next)
509
		if (!hashcmp(sha1, e->sha1))
510 511 512 513
			return e;
	return NULL;
}

514
static struct object_entry *insert_object(unsigned char *sha1)
515 516
{
	unsigned int h = sha1[0] << 8 | sha1[1];
517
	struct object_entry *e = object_table[h];
518
	struct object_entry *p = NULL;
519 520

	while (e) {
521
		if (!hashcmp(sha1, e->sha1))
522 523 524 525 526 527
			return e;
		p = e;
		e = e->next;
	}

	e = new_object(sha1);
528
	e->next = NULL;
529 530 531 532
	e->offset = 0;
	if (p)
		p->next = e;
	else
533
		object_table[h] = e;
534 535
	return e;
}
536

537 538 539 540 541 542 543 544
static unsigned int hc_str(const char *s, size_t len)
{
	unsigned int r = 0;
	while (len-- > 0)
		r = r * 31 + *s++;
	return r;
}

545
static void *pool_alloc(size_t len)
546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561
{
	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;
562
		p->next_free = (char *) p->space;
563 564 565 566 567
		p->end = p->next_free + mem_pool_alloc;
		mem_pool = p;
	}

	r = p->next_free;
568 569 570
	/* round out to a 'uintmax_t' alignment */
	if (len & (sizeof(uintmax_t) - 1))
		len += sizeof(uintmax_t) - (len & (sizeof(uintmax_t) - 1));
571 572 573 574
	p->next_free += len;
	return r;
}

575
static void *pool_calloc(size_t count, size_t size)
576 577 578 579 580 581 582
{
	size_t len = count * size;
	void *r = pool_alloc(len);
	memset(r, 0, len);
	return r;
}

583
static char *pool_strdup(const char *s)
584 585 586 587 588 589
{
	char *r = pool_alloc(strlen(s) + 1);
	strcpy(r, s);
	return r;
}

590
static void insert_mark(uintmax_t idnum, struct object_entry *oe)
591 592 593 594 595 596 597 598 599
{
	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) {
600
		uintmax_t i = idnum >> s->shift;
601 602 603 604 605 606 607 608 609 610 611 612
		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;
}

613
static struct object_entry *find_mark(uintmax_t idnum)
614
{
615
	uintmax_t orig_idnum = idnum;
616 617 618 619
	struct mark_set *s = marks;
	struct object_entry *oe = NULL;
	if ((idnum >> s->shift) < 1024) {
		while (s && s->shift) {
620
			uintmax_t i = idnum >> s->shift;
621 622 623 624 625 626 627
			idnum -= i << s->shift;
			s = s->data.sets[i];
		}
		if (s)
			oe = s->data.marked[idnum];
	}
	if (!oe)
628
		die("mark :%" PRIuMAX " not declared", orig_idnum);
629 630 631
	return oe;
}

632
static struct atom_str *to_atom(const char *s, unsigned short len)
633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650
{
	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;
}

651
static struct branch *lookup_branch(const char *name)
652 653 654 655 656 657 658 659 660 661
{
	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;
}

662
static struct branch *new_branch(const char *name)
663 664 665 666 667 668
{
	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);
669
	switch (check_ref_format(name)) {
670 671 672
	case 0: break; /* its valid */
	case CHECK_REF_FORMAT_ONELEVEL:
		break; /* valid, but too few '/', allow anyway */
673
	default:
674
		die("Branch name doesn't conform to GIT standards: %s", name);
675
	}
676 677 678 679

	b = pool_calloc(1, sizeof(struct branch));
	b->name = pool_strdup(name);
	b->table_next_branch = branch_table[hc];
680 681
	b->branch_tree.versions[0].mode = S_IFDIR;
	b->branch_tree.versions[1].mode = S_IFDIR;
682
	b->active = 0;
683
	b->pack_id = MAX_PACK_ID;
684 685 686 687 688 689 690 691 692 693 694
	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;
}

695
static struct tree_content *new_tree_content(unsigned int cnt)
696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717
{
	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;
718
	t->delta_depth = 0;
719 720 721 722 723 724 725 726
	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);
727 728 729 730 731 732
	f->next_avail = avail_tree_table[hc];
	avail_tree_table[hc] = f;
}

static void release_tree_content_recursive(struct tree_content *t)
{
733 734 735
	unsigned int i;
	for (i = 0; i < t->entry_count; i++)
		release_tree_entry(t->entries[i]);
736
	release_tree_content(t);
737 738
}

739
static struct tree_content *grow_tree_content(
740 741 742 743 744
	struct tree_content *t,
	int amt)
{
	struct tree_content *r = new_tree_content(t->entry_count + amt);
	r->entry_count = t->entry_count;
745
	r->delta_depth = t->delta_depth;
746 747 748 749 750
	memcpy(r->entries,t->entries,t->entry_count*sizeof(t->entries[0]));
	release_tree_content(t);
	return r;
}

751
static struct tree_entry *new_tree_entry(void)
752 753 754 755 756
{
	struct tree_entry *e;

	if (!avail_tree_entry) {
		unsigned int n = tree_entry_alloc;
757
		total_allocd += n * sizeof(struct tree_entry);
758
		avail_tree_entry = e = xmalloc(n * sizeof(struct tree_entry));
759
		while (n-- > 1) {
760 761 762
			*((void**)e) = e + 1;
			e++;
		}
763
		*((void**)e) = NULL;
764 765 766 767 768 769 770 771 772 773
	}

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

static void release_tree_entry(struct tree_entry *e)
{
	if (e->tree)
774
		release_tree_content_recursive(e->tree);
775 776 777 778
	*((void**)e) = avail_tree_entry;
	avail_tree_entry = e;
}

779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803
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;
}

804
static void start_packfile(void)
805
{
806
	static char tmpfile[PATH_MAX];
807
	struct packed_git *p;
808
	struct pack_header hdr;
809
	int pack_fd;
810

811
	snprintf(tmpfile, sizeof(tmpfile),
812
		"%s/tmp_pack_XXXXXX", get_object_directory());
813
	pack_fd = xmkstemp(tmpfile);
814 815
	p = xcalloc(1, sizeof(*p) + strlen(tmpfile) + 2);
	strcpy(p->pack_name, tmpfile);
816
	p->pack_fd = pack_fd;
817 818 819 820

	hdr.hdr_signature = htonl(PACK_SIGNATURE);
	hdr.hdr_version = htonl(2);
	hdr.hdr_entries = 0;
821
	write_or_die(p->pack_fd, &hdr, sizeof(hdr));
822 823

	pack_data = p;
824 825
	pack_size = sizeof(hdr);
	object_count = 0;
826 827 828

	all_packs = xrealloc(all_packs, sizeof(*all_packs) * (pack_id + 1));
	all_packs[pack_id] = p;
829 830 831 832 833 834 835 836 837
}

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);
}

838
static char *create_index(void)
839
{
840 841
	static char tmpfile[PATH_MAX];
	SHA_CTX ctx;
842 843 844
	struct sha1file *f;
	struct object_entry **idx, **c, **last, *e;
	struct object_entry_pool *o;
845
	uint32_t array[256];
846
	int i, idx_fd;
847 848 849 850 851

	/* 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)
852 853 854
		for (e = o->next_free; e-- != o->entries;)
			if (pack_id == e->pack_id)
				*c++ = e;
855
	last = idx + object_count;
856 857
	if (c != last)
		die("internal consistency error creating the index");
858 859 860 861 862 863 864 865 866 867 868 869 870 871 872
	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;
	}

873
	snprintf(tmpfile, sizeof(tmpfile),
874
		"%s/tmp_idx_XXXXXX", get_object_directory());
875
	idx_fd = xmkstemp(tmpfile);
876
	f = sha1fd(idx_fd, tmpfile);
877
	sha1write(f, array, 256 * sizeof(int));
878
	SHA1_Init(&ctx);
879
	for (c = idx; c != last; c++) {
880
		uint32_t offset = htonl((*c)->offset);
881 882
		sha1write(f, &offset, 4);
		sha1write(f, (*c)->sha1, sizeof((*c)->sha1));
883
		SHA1_Update(&ctx, (*c)->sha1, 20);
884
	}
885
	sha1write(f, pack_data->sha1, sizeof(pack_data->sha1));
886
	sha1close(f, NULL, 1);
887
	free(idx);
888 889 890 891
	SHA1_Final(pack_data->sha1, &ctx);
	return tmpfile;
}

892
static char *keep_pack(char *curr_index_name)
893 894
{
	static char name[PATH_MAX];
895
	static const char *keep_msg = "fast-import";
896 897 898 899 900 901 902 903 904 905
	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");
906 907 908
	write_or_die(keep_fd, keep_msg, strlen(keep_msg));
	if (close(keep_fd))
		die("failed to write keep file");
909 910 911 912 913 914 915 916 917 918 919 920 921

	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;
}

922
static void unkeep_all_packs(void)
923 924 925 926 927 928 929 930 931 932
{
	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);
	}
933 934
}

935
static void end_packfile(void)
936
{
937 938
	struct packed_git *old_p = pack_data, *new_p;

939
	if (object_count) {
940
		char *idx_name;
941 942 943
		int i;
		struct branch *b;
		struct tag *t;
944

945
		close_pack_windows(pack_data);
946 947 948
		fixup_pack_header_footer(pack_data->pack_fd, pack_data->sha1,
				    pack_data->pack_name, object_count);
		close(pack_data->pack_fd);
949
		idx_name = keep_pack(create_index());
950 951 952 953 954

		/* 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);
955
		all_packs[pack_id] = new_p;
956
		install_packed_git(new_p);
957 958

		/* Print the boundary */
959 960 961 962 963 964 965
		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));
				}
966
			}
967 968 969 970 971 972
			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);
973 974 975
		}

		pack_id++;
976
	}
977
	else
978
		unlink(old_p->pack_name);
979 980 981
	free(old_p);

	/* We can't carry a delta across packfiles. */
P
Pierre Habouzit 已提交
982
	strbuf_release(&last_blob.data);
983 984
	last_blob.offset = 0;
	last_blob.depth = 0;
985 986
}

987
static void cycle_packfile(void)
988 989 990 991 992
{
	end_packfile();
	start_packfile();
}

993
static size_t encode_header(
994
	enum object_type type,
995
	size_t size,
996
	unsigned char *hdr)
997 998 999 1000
{
	int n = 1;
	unsigned char c;

1001
	if (type < OBJ_COMMIT || type > OBJ_REF_DELTA)
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
		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;
}

1016 1017
static int store_object(
	enum object_type type,
1018
	struct strbuf *dat,
1019
	struct last_object *last,
1020
	unsigned char *sha1out,
1021
	uintmax_t mark)
1022 1023
{
	void *out, *delta;
1024 1025 1026
	struct object_entry *e;
	unsigned char hdr[96];
	unsigned char sha1[20];
1027
	unsigned long hdrlen, deltalen;
1028 1029 1030
	SHA_CTX c;
	z_stream s;

1031
	hdrlen = sprintf((char*)hdr,"%s %lu", typename(type),
1032
		(unsigned long)dat->len) + 1;
1033 1034
	SHA1_Init(&c);
	SHA1_Update(&c, hdr, hdrlen);
1035
	SHA1_Update(&c, dat->buf, dat->len);
1036
	SHA1_Final(sha1, &c);
1037
	if (sha1out)
1038
		hashcpy(sha1out, sha1);
1039 1040

	e = insert_object(sha1);
1041 1042
	if (mark)
		insert_mark(mark, e);
1043
	if (e->offset) {
1044
		duplicate_count_by_type[type]++;
1045
		return 1;
1046 1047 1048 1049 1050 1051
	} 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;
1052
	}
1053

P
Pierre Habouzit 已提交
1054 1055
	if (last && last->data.buf && last->depth < max_depth) {
		delta = diff_delta(last->data.buf, last->data.len,
1056
			dat->buf, dat->len,
1057
			&deltalen, 0);
1058
		if (delta && deltalen >= dat->len) {
1059 1060 1061 1062 1063
			free(delta);
			delta = NULL;
		}
	} else
		delta = NULL;
1064 1065

	memset(&s, 0, sizeof(s));
1066
	deflateInit(&s, pack_compression_level);
1067 1068 1069 1070
	if (delta) {
		s.next_in = delta;
		s.avail_in = deltalen;
	} else {
1071 1072
		s.next_in = (void *)dat->buf;
		s.avail_in = dat->len;
1073 1074 1075 1076 1077 1078 1079 1080
	}
	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. */
1081
	if ((pack_size + 60 + s.total_out) > max_packsize
1082 1083 1084 1085
		|| (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;
1086
		cycle_packfile();
1087 1088 1089 1090 1091

		/* We cannot carry a delta into the new pack. */
		if (delta) {
			free(delta);
			delta = NULL;
1092 1093

			memset(&s, 0, sizeof(s));
1094
			deflateInit(&s, pack_compression_level);
1095 1096
			s.next_in = (void *)dat->buf;
			s.avail_in = dat->len;
1097 1098 1099 1100 1101
			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);
1102 1103 1104 1105 1106 1107 1108 1109
		}
	}

	e->type = type;
	e->pack_id = pack_id;
	e->offset = pack_size;
	object_count++;
	object_count_by_type[type]++;
1110 1111

	if (delta) {
1112 1113 1114
		unsigned long ofs = e->offset - last->offset;
		unsigned pos = sizeof(hdr) - 1;

1115
		delta_count_by_type[type]++;
1116
		e->depth = last->depth + 1;
1117 1118

		hdrlen = encode_header(OBJ_OFS_DELTA, deltalen, hdr);
1119
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1120 1121 1122 1123 1124
		pack_size += hdrlen;

		hdr[pos] = ofs & 127;
		while (ofs >>= 7)
			hdr[--pos] = 128 | (--ofs & 127);
1125
		write_or_die(pack_data->pack_fd, hdr + pos, sizeof(hdr) - pos);
1126
		pack_size += sizeof(hdr) - pos;
1127
	} else {
1128
		e->depth = 0;
1129
		hdrlen = encode_header(type, dat->len, hdr);
1130
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1131
		pack_size += hdrlen;
1132 1133
	}

1134
	write_or_die(pack_data->pack_fd, out, s.total_out);
1135
	pack_size += s.total_out;
1136 1137

	free(out);
1138
	free(delta);
1139
	if (last) {
P
Pierre Habouzit 已提交
1140 1141 1142
		if (last->no_swap) {
			last->data = *dat;
		} else {
1143
			strbuf_swap(&last->data, dat);
P
Pierre Habouzit 已提交
1144
		}
1145
		last->offset = e->offset;
1146
		last->depth = e->depth;
1147 1148 1149 1150
	}
	return 0;
}

1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168
/* 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.
 */
1169 1170 1171
static void *gfi_unpack_entry(
	struct object_entry *oe,
	unsigned long *sizep)
1172
{
1173
	enum object_type type;
1174
	struct packed_git *p = all_packs[oe->pack_id];
1175
	if (p == pack_data && p->pack_size < (pack_size + 20)) {
1176 1177 1178 1179 1180 1181 1182 1183
		/* 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.
		 */
1184
		close_pack_windows(p);
1185 1186 1187 1188 1189 1190 1191

		/* 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.
		 */
1192
		p->pack_size = pack_size + 20;
1193
	}
1194
	return unpack_entry(p, oe->offset, &type, sizep);
1195 1196
}

1197
static const char *get_mode(const char *str, uint16_t *modep)
1198 1199
{
	unsigned char c;
1200
	uint16_t mode = 0;
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212

	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)
{
1213
	unsigned char* sha1 = root->versions[1].sha1;
1214 1215 1216 1217 1218 1219 1220
	struct object_entry *myoe;
	struct tree_content *t;
	unsigned long size;
	char *buf;
	const char *c;

	root->tree = t = new_tree_content(8);
1221
	if (is_null_sha1(sha1))
1222 1223
		return;

1224
	myoe = find_object(sha1);
1225
	if (myoe && myoe->pack_id != MAX_PACK_ID) {
1226
		if (myoe->type != OBJ_TREE)
1227
			die("Not a tree: %s", sha1_to_hex(sha1));
1228
		t->delta_depth = myoe->depth;
1229
		buf = gfi_unpack_entry(myoe, &size);
1230 1231
		if (!buf)
			die("Can't load tree %s", sha1_to_hex(sha1));
1232
	} else {
1233 1234 1235
		enum object_type type;
		buf = read_sha1_file(sha1, &type, &size);
		if (!buf || type != OBJ_TREE)
1236
			die("Can't load tree %s", sha1_to_hex(sha1));
1237 1238 1239 1240 1241 1242 1243
	}

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

		if (t->entry_count == t->entry_capacity)
1244
			root->tree = t = grow_tree_content(t, t->entry_count);
1245 1246 1247
		t->entries[t->entry_count++] = e;

		e->tree = NULL;
1248
		c = get_mode(c, &e->versions[1].mode);
1249
		if (!c)
1250 1251
			die("Corrupt mode in %s", sha1_to_hex(sha1));
		e->versions[0].mode = e->versions[1].mode;
1252
		e->name = to_atom(c, strlen(c));
1253
		c += e->name->str_len + 1;
1254 1255
		hashcpy(e->versions[0].sha1, (unsigned char*)c);
		hashcpy(e->versions[1].sha1, (unsigned char*)c);
1256 1257 1258 1259 1260
		c += 20;
	}
	free(buf);
}

1261
static int tecmp0 (const void *_a, const void *_b)
1262 1263 1264 1265
{
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
1266 1267
		a->name->str_dat, a->name->str_len, a->versions[0].mode,
		b->name->str_dat, b->name->str_len, b->versions[0].mode);
1268 1269
}

1270
static int tecmp1 (const void *_a, const void *_b)
1271
{
1272 1273 1274 1275 1276 1277 1278
	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);
}

1279
static void mktree(struct tree_content *t, int v, struct strbuf *b)
1280 1281
{
	size_t maxlen = 0;
1282 1283
	unsigned int i;

1284 1285 1286 1287
	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);
1288 1289

	for (i = 0; i < t->entry_count; i++) {
1290 1291
		if (t->entries[i]->versions[v].mode)
			maxlen += t->entries[i]->name->str_len + 34;
1292 1293
	}

1294 1295
	strbuf_reset(b);
	strbuf_grow(b, maxlen);
1296 1297
	for (i = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
1298 1299
		if (!e->versions[v].mode)
			continue;
1300 1301 1302
		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);
1303
	}
1304 1305 1306 1307 1308 1309
}

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
1310
	struct last_object lo = { STRBUF_INIT, 0, 0, /* no_swap */ 1 };
1311
	struct object_entry *le;
1312 1313 1314 1315 1316 1317 1318 1319 1320

	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]);
	}

1321
	le = find_object(root->versions[0].sha1);
P
Pierre Habouzit 已提交
1322
	if (S_ISDIR(root->versions[0].mode) && le && le->pack_id == pack_id) {
1323
		mktree(t, 0, &old_tree);
P
Pierre Habouzit 已提交
1324
		lo.data = old_tree;
1325
		lo.offset = le->offset;
1326 1327 1328
		lo.depth = t->delta_depth;
	}

1329 1330
	mktree(t, 1, &new_tree);
	store_object(OBJ_TREE, &new_tree, &lo, root->versions[1].sha1, 0);
1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344

	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;
1345 1346 1347 1348 1349 1350
}

static int tree_content_set(
	struct tree_entry *root,
	const char *p,
	const unsigned char *sha1,
1351 1352
	const uint16_t mode,
	struct tree_content *subtree)
1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363
{
	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);
1364 1365
	if (!n)
		die("Empty path component found in input");
1366 1367
	if (!slash1 && !S_ISDIR(mode) && subtree)
		die("Non-directories cannot have subtrees");
1368 1369 1370 1371 1372

	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) {
1373 1374
				if (!S_ISDIR(mode)
						&& e->versions[1].mode == mode
1375
						&& !hashcmp(e->versions[1].sha1, sha1))
1376
					return 0;
1377 1378
				e->versions[1].mode = mode;
				hashcpy(e->versions[1].sha1, sha1);
1379
				if (e->tree)
1380
					release_tree_content_recursive(e->tree);
1381
				e->tree = subtree;
1382
				hashclr(root->versions[1].sha1);
1383 1384
				return 1;
			}
1385
			if (!S_ISDIR(e->versions[1].mode)) {
1386
				e->tree = new_tree_content(8);
1387
				e->versions[1].mode = S_IFDIR;
1388 1389 1390
			}
			if (!e->tree)
				load_tree(e);
1391
			if (tree_content_set(e, slash1 + 1, sha1, mode, subtree)) {
1392
				hashclr(root->versions[1].sha1);
1393 1394 1395 1396 1397 1398 1399
				return 1;
			}
			return 0;
		}
	}

	if (t->entry_count == t->entry_capacity)
1400
		root->tree = t = grow_tree_content(t, t->entry_count);
1401
	e = new_tree_entry();
1402
	e->name = to_atom(p, n);
1403 1404
	e->versions[0].mode = 0;
	hashclr(e->versions[0].sha1);
1405 1406 1407
	t->entries[t->entry_count++] = e;
	if (slash1) {
		e->tree = new_tree_content(8);
1408
		e->versions[1].mode = S_IFDIR;
1409
		tree_content_set(e, slash1 + 1, sha1, mode, subtree);
1410
	} else {
1411
		e->tree = subtree;
1412 1413
		e->versions[1].mode = mode;
		hashcpy(e->versions[1].sha1, sha1);
1414
	}
1415
	hashclr(root->versions[1].sha1);
1416 1417 1418
	return 1;
}

1419 1420 1421 1422
static int tree_content_remove(
	struct tree_entry *root,
	const char *p,
	struct tree_entry *backup_leaf)
1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437
{
	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)) {
1438
			if (!slash1 || !S_ISDIR(e->versions[1].mode))
1439 1440 1441
				goto del_entry;
			if (!e->tree)
				load_tree(e);
1442
			if (tree_content_remove(e, slash1 + 1, backup_leaf)) {
1443 1444 1445 1446 1447 1448
				for (n = 0; n < e->tree->entry_count; n++) {
					if (e->tree->entries[n]->versions[1].mode) {
						hashclr(root->versions[1].sha1);
						return 1;
					}
				}
1449
				backup_leaf = NULL;
1450
				goto del_entry;
1451 1452 1453 1454 1455 1456 1457
			}
			return 0;
		}
	}
	return 0;

del_entry:
1458 1459 1460
	if (backup_leaf)
		memcpy(backup_leaf, e, sizeof(*backup_leaf));
	else if (e->tree)
1461
		release_tree_content_recursive(e->tree);
1462
	e->tree = NULL;
1463 1464 1465
	e->versions[1].mode = 0;
	hashclr(e->versions[1].sha1);
	hashclr(root->versions[1].sha1);
1466
	return 1;
1467 1468
}

1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505
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;
}

1506
static int update_branch(struct branch *b)
1507 1508
{
	static const char *msg = "fast-import";
1509 1510 1511 1512 1513
	struct ref_lock *lock;
	unsigned char old_sha1[20];

	if (read_ref(b->name, old_sha1))
		hashclr(old_sha1);
1514
	lock = lock_any_ref_for_update(b->name, old_sha1, 0);
1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526
	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);
		}

1527
		if (!in_merge_bases(old_cmit, &new_cmit, 1)) {
1528
			unlock_ref(lock);
1529
			warning("Not updating %s"
1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541
				" (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)
{
1542 1543 1544 1545
	unsigned int i;
	struct branch *b;

	for (i = 0; i < branch_table_sz; i++) {
1546 1547
		for (b = branch_table[i]; b; b = b->table_next_branch)
			failure |= update_branch(b);
1548 1549 1550
	}
}

1551
static void dump_tags(void)
1552 1553 1554 1555
{
	static const char *msg = "fast-import";
	struct tag *t;
	struct ref_lock *lock;
1556
	char ref_name[PATH_MAX];
1557 1558

	for (t = first_tag; t; t = t->next_tag) {
1559 1560
		sprintf(ref_name, "tags/%s", t->name);
		lock = lock_ref_sha1(ref_name, NULL);
1561
		if (!lock || write_ref_sha1(lock, t->sha1, msg) < 0)
1562
			failure |= error("Unable to update %s", ref_name);
1563 1564 1565
	}
}

1566
static void dump_marks_helper(FILE *f,
1567
	uintmax_t base,
1568 1569
	struct mark_set *m)
{
1570
	uintmax_t k;
1571 1572 1573 1574 1575 1576 1577 1578 1579
	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])
1580
				fprintf(f, ":%" PRIuMAX " %s\n", base + k,
1581 1582 1583 1584 1585
					sha1_to_hex(m->data.marked[k]->sha1));
		}
	}
}

1586
static void dump_marks(void)
1587
{
1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599
	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;
1600
	}
1601 1602 1603

	f = fdopen(mark_fd, "w");
	if (!f) {
1604
		int saved_errno = errno;
1605 1606
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1607
			mark_file, strerror(saved_errno));
1608
		return;
1609
	}
1610

B
Brandon Casey 已提交
1611
	/*
1612 1613
	 * 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
1614 1615 1616
	 * won't try to close() it.
	 */
	mark_lock.fd = -1;
1617 1618 1619

	dump_marks_helper(f, 0, marks);
	if (ferror(f) || fclose(f)) {
1620
		int saved_errno = errno;
1621 1622
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1623
			mark_file, strerror(saved_errno));
1624 1625 1626 1627
		return;
	}

	if (commit_lock_file(&mark_lock)) {
1628
		int saved_errno = errno;
1629 1630
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to commit marks file %s: %s",
1631
			mark_file, strerror(saved_errno));
1632 1633
		return;
	}
1634 1635
}

1636
static int read_next_command(void)
1637
{
1638 1639 1640 1641 1642 1643 1644
	static int stdin_eof = 0;

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

1645
	do {
1646
		if (unread_command_buf) {
1647
			unread_command_buf = 0;
1648 1649 1650
		} else {
			struct recent_command *rc;

1651
			strbuf_detach(&command_buf, NULL);
1652 1653 1654
			stdin_eof = strbuf_getline(&command_buf, stdin, '\n');
			if (stdin_eof)
				return EOF;
1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672

			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] == '#');
1673 1674

	return 0;
1675 1676
}

J
Junio C Hamano 已提交
1677
static void skip_optional_lf(void)
1678 1679 1680 1681 1682 1683
{
	int term_char = fgetc(stdin);
	if (term_char != '\n' && term_char != EOF)
		ungetc(term_char, stdin);
}

1684
static void cmd_mark(void)
1685
{
1686
	if (!prefixcmp(command_buf.buf, "mark :")) {
1687
		next_mark = strtoumax(command_buf.buf + 6, NULL, 10);
1688 1689 1690
		read_next_command();
	}
	else
1691
		next_mark = 0;
1692 1693
}

1694
static void cmd_data(struct strbuf *sb)
1695
{
1696
	strbuf_reset(sb);
1697

1698
	if (prefixcmp(command_buf.buf, "data "))
1699 1700
		die("Expected 'data n' command, found: %s", command_buf.buf);

1701
	if (!prefixcmp(command_buf.buf + 5, "<<")) {
1702
		char *term = xstrdup(command_buf.buf + 5 + 2);
1703 1704
		size_t term_len = command_buf.len - 5 - 2;

1705
		strbuf_detach(&command_buf, NULL);
1706
		for (;;) {
1707
			if (strbuf_getline(&command_buf, stdin, '\n') == EOF)
1708 1709 1710 1711
				die("EOF in data (terminator '%s' not found)", term);
			if (term_len == command_buf.len
				&& !strcmp(term, command_buf.buf))
				break;
1712 1713
			strbuf_addbuf(sb, &command_buf);
			strbuf_addch(sb, '\n');
1714 1715 1716 1717
		}
		free(term);
	}
	else {
1718 1719
		size_t n = 0, length;

1720
		length = strtoul(command_buf.buf + 5, NULL, 10);
1721

1722
		while (n < length) {
1723
			size_t s = strbuf_fread(sb, length - n, stdin);
1724
			if (!s && feof(stdin))
1725 1726
				die("EOF in data (%lu bytes remaining)",
					(unsigned long)(length - n));
1727 1728
			n += s;
		}
1729 1730
	}

1731
	skip_optional_lf();
1732 1733
}

1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791
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;
}

1792
static void cmd_new_blob(void)
1793
{
P
Pierre Habouzit 已提交
1794
	static struct strbuf buf = STRBUF_INIT;
1795 1796 1797

	read_next_command();
	cmd_mark();
1798
	cmd_data(&buf);
P
Pierre Habouzit 已提交
1799
	store_object(OBJ_BLOB, &buf, &last_blob, NULL, next_mark);
1800 1801
}

1802
static void unload_one_branch(void)
1803
{
1804 1805
	while (cur_active_branches
		&& cur_active_branches >= max_active_branches) {
1806
		uintmax_t min_commit = ULONG_MAX;
1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823
		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;
		}
1824
		e->active = 0;
1825 1826
		e->active_next_branch = NULL;
		if (e->branch_tree.tree) {
1827
			release_tree_content_recursive(e->branch_tree.tree);
1828 1829 1830
			e->branch_tree.tree = NULL;
		}
		cur_active_branches--;
1831 1832 1833
	}
}

1834
static void load_branch(struct branch *b)
1835
{
1836
	load_tree(&b->branch_tree);
1837 1838 1839 1840 1841 1842 1843
	if (!b->active) {
		b->active = 1;
		b->active_next_branch = active_branches;
		active_branches = b;
		cur_active_branches++;
		branch_load_count++;
	}
1844 1845
}

1846
static void file_change_m(struct branch *b)
1847
{
1848
	const char *p = command_buf.buf + 2;
1849
	static struct strbuf uq = STRBUF_INIT;
1850
	const char *endp;
1851
	struct object_entry *oe = oe;
1852
	unsigned char sha1[20];
1853
	uint16_t mode, inline_data = 0;
1854

1855 1856 1857 1858 1859 1860
	p = get_mode(p, &mode);
	if (!p)
		die("Corrupt mode: %s", command_buf.buf);
	switch (mode) {
	case S_IFREG | 0644:
	case S_IFREG | 0755:
1861
	case S_IFLNK:
1862 1863 1864 1865 1866 1867 1868 1869
	case 0644:
	case 0755:
		/* ok */
		break;
	default:
		die("Corrupt mode: %s", command_buf.buf);
	}

1870 1871
	if (*p == ':') {
		char *x;
1872
		oe = find_mark(strtoumax(p + 1, &x, 10));
1873
		hashcpy(sha1, oe->sha1);
1874
		p = x;
1875
	} else if (!prefixcmp(p, "inline")) {
1876 1877
		inline_data = 1;
		p += 6;
1878 1879 1880 1881 1882 1883
	} else {
		if (get_sha1_hex(p, sha1))
			die("Invalid SHA1: %s", command_buf.buf);
		oe = find_object(sha1);
		p += 40;
	}
1884 1885 1886
	if (*p++ != ' ')
		die("Missing space after SHA1: %s", command_buf.buf);

1887 1888
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1889 1890
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1891
		p = uq.buf;
1892
	}
1893

1894
	if (inline_data) {
P
Pierre Habouzit 已提交
1895
		static struct strbuf buf = STRBUF_INIT;
1896

1897 1898 1899 1900
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
1901
		read_next_command();
1902
		cmd_data(&buf);
P
Pierre Habouzit 已提交
1903
		store_object(OBJ_BLOB, &buf, &last_blob, sha1, 0);
1904
	} else if (oe) {
1905
		if (oe->type != OBJ_BLOB)
1906
			die("Not a blob (actually a %s): %s",
1907
				typename(oe->type), command_buf.buf);
1908
	} else {
1909 1910
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
1911
			die("Blob not found: %s", command_buf.buf);
1912
		if (type != OBJ_BLOB)
1913
			die("Not a blob (actually a %s): %s",
1914
			    typename(type), command_buf.buf);
1915
	}
1916

1917
	tree_content_set(&b->branch_tree, p, sha1, S_IFREG | mode, NULL);
1918
}
1919

1920 1921
static void file_change_d(struct branch *b)
{
1922
	const char *p = command_buf.buf + 2;
1923
	static struct strbuf uq = STRBUF_INIT;
1924 1925
	const char *endp;

1926 1927
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1928 1929
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1930
		p = uq.buf;
1931
	}
1932
	tree_content_remove(&b->branch_tree, p, NULL);
1933 1934
}

1935
static void file_change_cr(struct branch *b, int rename)
1936 1937
{
	const char *s, *d;
1938 1939
	static struct strbuf s_uq = STRBUF_INIT;
	static struct strbuf d_uq = STRBUF_INIT;
1940 1941 1942 1943
	const char *endp;
	struct tree_entry leaf;

	s = command_buf.buf + 2;
1944 1945
	strbuf_reset(&s_uq);
	if (!unquote_c_style(&s_uq, s, &endp)) {
1946 1947
		if (*endp != ' ')
			die("Missing space after source: %s", command_buf.buf);
1948
	} else {
1949 1950 1951
		endp = strchr(s, ' ');
		if (!endp)
			die("Missing space after source: %s", command_buf.buf);
1952
		strbuf_add(&s_uq, s, endp - s);
1953
	}
1954
	s = s_uq.buf;
1955 1956 1957 1958 1959 1960

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

	d = endp;
1961 1962
	strbuf_reset(&d_uq);
	if (!unquote_c_style(&d_uq, d, &endp)) {
1963 1964
		if (*endp)
			die("Garbage after dest in: %s", command_buf.buf);
1965
		d = d_uq.buf;
1966 1967 1968
	}

	memset(&leaf, 0, sizeof(leaf));
1969 1970 1971 1972
	if (rename)
		tree_content_remove(&b->branch_tree, s, &leaf);
	else
		tree_content_get(&b->branch_tree, s, &leaf);
1973 1974 1975 1976 1977 1978 1979 1980
	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);
}

1981 1982 1983 1984 1985 1986 1987 1988
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);
}

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
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);
	}
}

2016
static int cmd_from(struct branch *b)
2017
{
2018
	const char *from;
2019 2020
	struct branch *s;

2021
	if (prefixcmp(command_buf.buf, "from "))
2022
		return 0;
2023

2024 2025 2026 2027
	if (b->branch_tree.tree) {
		release_tree_content_recursive(b->branch_tree.tree);
		b->branch_tree.tree = NULL;
	}
2028 2029 2030 2031 2032 2033

	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) {
2034
		unsigned char *t = s->branch_tree.versions[1].sha1;
2035
		hashcpy(b->sha1, s->sha1);
2036 2037
		hashcpy(b->branch_tree.versions[0].sha1, t);
		hashcpy(b->branch_tree.versions[1].sha1, t);
2038
	} else if (*from == ':') {
2039
		uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2040 2041
		struct object_entry *oe = find_mark(idnum);
		if (oe->type != OBJ_COMMIT)
2042
			die("Mark :%" PRIuMAX " not a commit", idnum);
2043
		hashcpy(b->sha1, oe->sha1);
2044
		if (oe->pack_id != MAX_PACK_ID) {
2045
			unsigned long size;
2046 2047
			char *buf = gfi_unpack_entry(oe, &size);
			cmd_from_commit(b, buf, size);
2048
			free(buf);
2049 2050
		} else
			cmd_from_existing(b);
2051 2052 2053
	} else if (!get_sha1(from, b->sha1))
		cmd_from_existing(b);
	else
2054 2055 2056
		die("Invalid ref name or SHA1 expression: %s", from);

	read_next_command();
2057
	return 1;
2058 2059
}

2060
static struct hash_list *cmd_merge(unsigned int *count)
2061
{
2062
	struct hash_list *list = NULL, *n, *e = e;
2063
	const char *from;
2064 2065 2066
	struct branch *s;

	*count = 0;
2067
	while (!prefixcmp(command_buf.buf, "merge ")) {
2068 2069 2070 2071 2072 2073
		from = strchr(command_buf.buf, ' ') + 1;
		n = xmalloc(sizeof(*n));
		s = lookup_branch(from);
		if (s)
			hashcpy(n->sha1, s->sha1);
		else if (*from == ':') {
2074
			uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2075 2076
			struct object_entry *oe = find_mark(idnum);
			if (oe->type != OBJ_COMMIT)
2077
				die("Mark :%" PRIuMAX " not a commit", idnum);
2078
			hashcpy(n->sha1, oe->sha1);
2079 2080 2081
		} 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
2082
				commit_type, &size, n->sha1);
2083 2084 2085 2086
			if (!buf || size < 46)
				die("Not a valid commit: %s", from);
			free(buf);
		} else
2087 2088 2089 2090 2091 2092 2093 2094
			die("Invalid ref name or SHA1 expression: %s", from);

		n->next = NULL;
		if (list)
			e->next = n;
		else
			list = n;
		e = n;
2095
		(*count)++;
2096 2097 2098 2099 2100
		read_next_command();
	}
	return list;
}

2101
static void cmd_new_commit(void)
2102
{
2103
	static struct strbuf msg = STRBUF_INIT;
2104 2105 2106 2107
	struct branch *b;
	char *sp;
	char *author = NULL;
	char *committer = NULL;
2108 2109
	struct hash_list *merge_list = NULL;
	unsigned int merge_count;
2110 2111 2112 2113

	/* Obtain the branch name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	b = lookup_branch(sp);
2114
	if (!b)
2115
		b = new_branch(sp);
2116 2117 2118

	read_next_command();
	cmd_mark();
2119
	if (!prefixcmp(command_buf.buf, "author ")) {
2120
		author = parse_ident(command_buf.buf + 7);
2121 2122
		read_next_command();
	}
2123
	if (!prefixcmp(command_buf.buf, "committer ")) {
2124
		committer = parse_ident(command_buf.buf + 10);
2125 2126 2127 2128
		read_next_command();
	}
	if (!committer)
		die("Expected committer but didn't get one");
2129
	cmd_data(&msg);
2130 2131
	read_next_command();
	cmd_from(b);
2132
	merge_list = cmd_merge(&merge_count);
2133 2134

	/* ensure the branch is active/loaded */
2135
	if (!b->branch_tree.tree || !max_active_branches) {
2136 2137 2138
		unload_one_branch();
		load_branch(b);
	}
2139

2140
	/* file_change* */
2141
	while (command_buf.len > 0) {
2142
		if (!prefixcmp(command_buf.buf, "M "))
2143
			file_change_m(b);
2144
		else if (!prefixcmp(command_buf.buf, "D "))
2145
			file_change_d(b);
2146
		else if (!prefixcmp(command_buf.buf, "R "))
2147 2148 2149
			file_change_cr(b, 1);
		else if (!prefixcmp(command_buf.buf, "C "))
			file_change_cr(b, 0);
2150 2151
		else if (!strcmp("deleteall", command_buf.buf))
			file_change_deleteall(b);
2152 2153 2154 2155
		else {
			unread_command_buf = 1;
			break;
		}
2156 2157
		if (read_next_command() == EOF)
			break;
2158 2159
	}

2160
	/* build the tree and the commit */
2161
	store_tree(&b->branch_tree);
2162 2163
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
2164 2165 2166

	strbuf_reset(&new_data);
	strbuf_addf(&new_data, "tree %s\n",
2167
		sha1_to_hex(b->branch_tree.versions[1].sha1));
2168
	if (!is_null_sha1(b->sha1))
2169
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(b->sha1));
2170 2171
	while (merge_list) {
		struct hash_list *next = merge_list->next;
2172
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(merge_list->sha1));
2173 2174 2175
		free(merge_list);
		merge_list = next;
	}
2176 2177 2178 2179 2180 2181
	strbuf_addf(&new_data,
		"author %s\n"
		"committer %s\n"
		"\n",
		author ? author : committer, committer);
	strbuf_addbuf(&new_data, &msg);
2182
	free(author);
2183 2184
	free(committer);

2185
	if (!store_object(OBJ_COMMIT, &new_data, NULL, b->sha1, next_mark))
2186
		b->pack_id = pack_id;
2187
	b->last_commit = object_count_by_type[OBJ_COMMIT];
2188 2189
}

2190
static void cmd_new_tag(void)
2191
{
2192
	static struct strbuf msg = STRBUF_INIT;
2193 2194 2195 2196 2197
	char *sp;
	const char *from;
	char *tagger;
	struct branch *s;
	struct tag *t;
2198
	uintmax_t from_mark = 0;
2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213
	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 ... */
2214
	if (prefixcmp(command_buf.buf, "from "))
2215 2216 2217 2218
		die("Expected from command, got %s", command_buf.buf);
	from = strchr(command_buf.buf, ' ') + 1;
	s = lookup_branch(from);
	if (s) {
2219
		hashcpy(sha1, s->sha1);
2220
	} else if (*from == ':') {
2221
		struct object_entry *oe;
2222
		from_mark = strtoumax(from + 1, NULL, 10);
2223
		oe = find_mark(from_mark);
2224
		if (oe->type != OBJ_COMMIT)
2225
			die("Mark :%" PRIuMAX " not a commit", from_mark);
2226
		hashcpy(sha1, oe->sha1);
2227 2228 2229 2230 2231
	} else if (!get_sha1(from, sha1)) {
		unsigned long size;
		char *buf;

		buf = read_object_with_reference(sha1,
2232
			commit_type, &size, sha1);
2233 2234 2235 2236 2237 2238 2239 2240
		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 ... */
2241
	if (prefixcmp(command_buf.buf, "tagger "))
2242
		die("Expected tagger command, got %s", command_buf.buf);
2243
	tagger = parse_ident(command_buf.buf + 7);
2244 2245 2246

	/* tag payload/message */
	read_next_command();
2247
	cmd_data(&msg);
2248 2249

	/* build the tag object */
2250 2251 2252 2253 2254 2255 2256 2257 2258
	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);
2259 2260
	free(tagger);

2261
	if (store_object(OBJ_TAG, &new_data, NULL, t->sha1, 0))
2262 2263 2264
		t->pack_id = MAX_PACK_ID;
	else
		t->pack_id = pack_id;
2265 2266
}

2267
static void cmd_reset_branch(void)
2268 2269 2270 2271 2272 2273 2274 2275
{
	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) {
2276 2277 2278
		hashclr(b->sha1);
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
2279 2280 2281 2282 2283
		if (b->branch_tree.tree) {
			release_tree_content_recursive(b->branch_tree.tree);
			b->branch_tree.tree = NULL;
		}
	}
2284 2285 2286
	else
		b = new_branch(sp);
	read_next_command();
2287
	if (!cmd_from(b) && command_buf.len > 0)
2288
		unread_command_buf = 1;
2289 2290
}

2291
static void cmd_checkpoint(void)
2292
{
2293 2294 2295 2296 2297 2298
	if (object_count) {
		cycle_packfile();
		dump_branches();
		dump_tags();
		dump_marks();
	}
2299
	skip_optional_lf();
2300 2301
}

2302 2303
static void cmd_progress(void)
{
2304
	fwrite(command_buf.buf, 1, command_buf.len, stdout);
2305 2306 2307 2308 2309
	fputc('\n', stdout);
	fflush(stdout);
	skip_optional_lf();
}

2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337
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;
2338
			e->offset = 1; /* just not zero! */
2339 2340 2341 2342 2343 2344
		}
		insert_mark(mark, e);
	}
	fclose(f);
}

2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365
static int git_pack_config(const char *k, const char *v)
{
	if (!strcmp(k, "pack.depth")) {
		max_depth = git_config_int(k, v);
		if (max_depth > MAX_DEPTH)
			max_depth = MAX_DEPTH;
		return 0;
	}
	if (!strcmp(k, "pack.compression")) {
		int level = git_config_int(k, v);
		if (level == -1)
			level = Z_DEFAULT_COMPRESSION;
		else if (level < 0 || level > Z_BEST_COMPRESSION)
			die("bad pack compression level %d", level);
		pack_compression_level = level;
		pack_compression_seen = 1;
		return 0;
	}
	return git_default_config(k, v);
}

2366
static const char fast_import_usage[] =
2367
"git-fast-import [--date-format=f] [--max-pack-size=n] [--depth=n] [--active-branches=n] [--export-marks=marks.file]";
2368

2369 2370
int main(int argc, const char **argv)
{
2371
	unsigned int i, show_stats = 1;
2372

2373 2374 2375 2376
	git_config(git_pack_config);
	if (!pack_compression_seen && core_compression_seen)
		pack_compression_level = core_compression_level;

2377
	alloc_objects(object_entry_alloc);
2378
	strbuf_init(&command_buf, 0);
2379 2380 2381 2382
	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));
2383

2384 2385 2386 2387 2388
	for (i = 1; i < argc; i++) {
		const char *a = argv[i];

		if (*a != '-' || !strcmp(a, "--"))
			break;
2389
		else if (!prefixcmp(a, "--date-format=")) {
2390 2391 2392 2393 2394 2395 2396 2397 2398 2399
			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);
		}
2400
		else if (!prefixcmp(a, "--max-pack-size="))
2401
			max_packsize = strtoumax(a + 16, NULL, 0) * 1024 * 1024;
2402
		else if (!prefixcmp(a, "--depth=")) {
2403
			max_depth = strtoul(a + 8, NULL, 0);
2404 2405 2406
			if (max_depth > MAX_DEPTH)
				die("--depth cannot exceed %u", MAX_DEPTH);
		}
2407
		else if (!prefixcmp(a, "--active-branches="))
2408
			max_active_branches = strtoul(a + 18, NULL, 0);
2409 2410
		else if (!prefixcmp(a, "--import-marks="))
			import_marks(a + 15);
2411
		else if (!prefixcmp(a, "--export-marks="))
2412
			mark_file = a + 15;
2413
		else if (!prefixcmp(a, "--export-pack-edges=")) {
2414 2415 2416 2417 2418 2419
			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"))
2420
			force_update = 1;
2421 2422 2423 2424
		else if (!strcmp(a, "--quiet"))
			show_stats = 0;
		else if (!strcmp(a, "--stats"))
			show_stats = 1;
2425 2426 2427
		else
			die("unknown option %s", a);
	}
2428
	if (i != argc)
2429 2430
		usage(fast_import_usage);

2431 2432 2433 2434 2435
	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;

2436
	prepare_packed_git();
2437
	start_packfile();
2438
	set_die_routine(die_nicely);
2439 2440
	while (read_next_command() != EOF) {
		if (!strcmp("blob", command_buf.buf))
2441
			cmd_new_blob();
2442
		else if (!prefixcmp(command_buf.buf, "commit "))
2443
			cmd_new_commit();
2444
		else if (!prefixcmp(command_buf.buf, "tag "))
2445
			cmd_new_tag();
2446
		else if (!prefixcmp(command_buf.buf, "reset "))
2447
			cmd_reset_branch();
2448 2449
		else if (!strcmp("checkpoint", command_buf.buf))
			cmd_checkpoint();
2450 2451
		else if (!prefixcmp(command_buf.buf, "progress "))
			cmd_progress();
2452 2453
		else
			die("Unsupported command: %s", command_buf.buf);
2454
	}
2455
	end_packfile();
2456

2457
	dump_branches();
2458
	dump_tags();
2459
	unkeep_all_packs();
2460
	dump_marks();
2461

2462 2463 2464
	if (pack_edges)
		fclose(pack_edges);

2465 2466 2467 2468 2469 2470 2471 2472 2473
	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");
2474 2475 2476 2477 2478 2479
		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]);
2480
		fprintf(stderr, "Total branches:  %10lu (%10lu loads     )\n", branch_count, branch_load_count);
2481
		fprintf(stderr, "      marks:     %10" PRIuMAX " (%10" PRIuMAX " unique    )\n", (((uintmax_t)1) << marks->shift) * 1024, marks_set_count);
2482
		fprintf(stderr, "      atoms:     %10u\n", atom_cnt);
2483
		fprintf(stderr, "Memory total:    %10" PRIuMAX " KiB\n", (total_allocd + alloc_count*sizeof(struct object_entry))/1024);
2484
		fprintf(stderr, "       pools:    %10lu KiB\n", (unsigned long)(total_allocd/1024));
2485
		fprintf(stderr, "     objects:    %10" PRIuMAX " KiB\n", (alloc_count*sizeof(struct object_entry))/1024);
2486 2487 2488 2489 2490
		fprintf(stderr, "---------------------------------------------------------------------\n");
		pack_report();
		fprintf(stderr, "---------------------------------------------------------------------\n");
		fprintf(stderr, "\n");
	}
2491

2492
	return failure ? 1 : 0;
2493
}