fast-import.c 62.6 KB
Newer Older
1
/*
2
(See Documentation/git-fast-import.txt for maintained documentation.)
3 4 5 6 7
Format of STDIN stream:

  stream ::= cmd*;

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

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

20
  new_commit ::= 'commit' sp ref_str lf
21
    mark?
22 23
    ('author' sp name sp '<' email '>' sp when lf)?
    'committer' sp name sp '<' email '>' sp when lf
24
    commit_msg
25
    ('from' sp (ref_str | hexsha1 | sha1exp_str | idnum) lf)?
26
    ('merge' sp (ref_str | hexsha1 | sha1exp_str | idnum) lf)*
27
    file_change*
28
    lf?;
29
  commit_msg ::= data;
30

31 32 33 34 35 36
  file_change ::= file_clr
    | file_del
    | file_rnm
    | file_cpy
    | file_obm
    | file_inm;
37
  file_clr ::= 'deleteall' lf;
38
  file_del ::= 'D' sp path_str lf;
39
  file_rnm ::= 'R' sp path_str sp path_str lf;
40
  file_cpy ::= 'C' sp path_str sp path_str lf;
41 42 43
  file_obm ::= 'M' sp mode sp (hexsha1 | idnum) sp path_str lf;
  file_inm ::= 'M' sp mode sp 'inline' sp path_str lf
    data;
44 45 46

  new_tag ::= 'tag' sp tag_str lf
    'from' sp (ref_str | hexsha1 | sha1exp_str | idnum) lf
47
    ('tagger' sp name sp '<' email '>' sp when lf)?
48 49 50
    tag_msg;
  tag_msg ::= data;

51 52
  reset_branch ::= 'reset' sp ref_str lf
    ('from' sp (ref_str | hexsha1 | sha1exp_str | idnum) lf)?
53
    lf?;
54

55
  checkpoint ::= 'checkpoint' lf
56
    lf?;
57

58 59 60
  progress ::= 'progress' sp not_lf* lf
    lf?;

61 62 63
     # 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
64
     # idnum can be updated in the future to a new object by issuing
65
     # a new mark directive with the old idnum.
66
     #
67
  mark ::= 'mark' sp idnum lf;
68
  data ::= (delimited_data | exact_data)
69
    lf?;
70 71 72 73 74 75

    # 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)*
76
    delim lf;
77 78

     # note: declen indicates the length of binary_data in bytes.
M
Mike Ralphson 已提交
79
     # declen does not include the lf preceding the binary data.
80
     #
81 82
  exact_data ::= 'data' sp declen lf
    binary_data;
83 84 85

     # note: quoted strings are C-style quoting supporting \c for
     # common escapes of 'c' (e..g \n, \t, \\, \") or \nnn where nnn
86
     # is the signed byte value in octal.  Note that the only
87 88
     # characters which must actually be escaped to protect the
     # stream formatting is: \, " and LF.  Otherwise these values
89
     # are UTF8.
90
     #
91 92 93
  ref_str     ::= ref;
  sha1exp_str ::= sha1exp;
  tag_str     ::= tag;
94
  path_str    ::= path    | '"' quoted(path)    '"' ;
95 96
  mode        ::= '100644' | '644'
                | '100755' | '755'
J
Junio C Hamano 已提交
97
                | '120000'
98
                ;
99 100

  declen ::= # unsigned 32 bit value, ascii base10 notation;
101
  bigint ::= # unsigned integer value, ascii base10 notation;
102
  binary_data ::= # file content, not interpreted;
103

104 105 106 107
  when         ::= raw_when | rfc2822_when;
  raw_when     ::= ts sp tz;
  rfc2822_when ::= # Valid RFC 2822 date and time;

108 109
  sp ::= # ASCII space character;
  lf ::= # ASCII newline (LF) character;
110 111

     # note: a colon (':') must precede the numerical value assigned to
112
     # an idnum.  This is to distinguish it from a ref or tag name as
113
     # GIT does not permit ':' in ref or tag strings.
114
     #
115
  idnum   ::= ':' bigint;
116 117 118
  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";
119 120
  sha1exp ::= # Any valid GIT SHA1 expression;
  hexsha1 ::= # SHA1 in hexadecimal format;
121 122

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

     # 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
136
     # must be the first character on that line (an lf
M
Mike Ralphson 已提交
137
     # preceded it).
138 139 140
     #
  comment ::= '#' not_lf* lf;
  not_lf  ::= # Any byte that is not ASCII newline (LF);
141 142
*/

143 144 145 146
#include "builtin.h"
#include "cache.h"
#include "object.h"
#include "blob.h"
147
#include "tree.h"
148
#include "commit.h"
149 150
#include "delta.h"
#include "pack.h"
151
#include "refs.h"
152
#include "csum-file.h"
153
#include "quote.h"
154
#include "exec_cmd.h"
155

156 157
#define PACK_ID_BITS 16
#define MAX_PACK_ID ((1<<PACK_ID_BITS)-1)
158 159
#define DEPTH_BITS 13
#define MAX_DEPTH ((1<<DEPTH_BITS)-1)
160

161 162 163
struct object_entry
{
	struct object_entry *next;
164
	uint32_t offset;
165 166 167
	uint32_t type : TYPE_BITS,
		pack_id : PACK_ID_BITS,
		depth : DEPTH_BITS;
168 169 170
	unsigned char sha1[20];
};

171
struct object_entry_pool
172
{
173
	struct object_entry_pool *next_pool;
174 175
	struct object_entry *next_free;
	struct object_entry *end;
176
	struct object_entry entries[FLEX_ARRAY]; /* more */
177 178
};

179 180 181 182 183 184
struct mark_set
{
	union {
		struct object_entry *marked[1024];
		struct mark_set *sets[1024];
	} data;
185
	unsigned int shift;
186 187
};

188 189
struct last_object
{
P
Pierre Habouzit 已提交
190
	struct strbuf data;
191
	uint32_t offset;
192
	unsigned int depth;
P
Pierre Habouzit 已提交
193
	unsigned no_swap : 1;
194 195
};

196 197 198 199 200
struct mem_pool
{
	struct mem_pool *next_pool;
	char *next_free;
	char *end;
201
	uintmax_t space[FLEX_ARRAY]; /* more */
202 203 204 205 206
};

struct atom_str
{
	struct atom_str *next_atom;
207
	unsigned short str_len;
208 209 210 211
	char str_dat[FLEX_ARRAY]; /* more */
};

struct tree_content;
212 213
struct tree_entry
{
214
	struct tree_content *tree;
215
	struct atom_str *name;
216 217
	struct tree_entry_ms
	{
218
		uint16_t mode;
219 220
		unsigned char sha1[20];
	} versions[2];
221 222
};

223
struct tree_content
224
{
225 226
	unsigned int entry_capacity; /* must match avail_tree_content */
	unsigned int entry_count;
227
	unsigned int delta_depth;
228 229 230 231 232 233 234
	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;
235 236 237 238
};

struct branch
{
239 240
	struct branch *table_next_branch;
	struct branch *active_next_branch;
241
	const char *name;
242
	struct tree_entry branch_tree;
243
	uintmax_t last_commit;
244 245
	unsigned active : 1;
	unsigned pack_id : PACK_ID_BITS;
246
	unsigned char sha1[20];
247 248
};

249 250 251 252
struct tag
{
	struct tag *next_tag;
	const char *name;
253
	unsigned int pack_id;
254 255 256
	unsigned char sha1[20];
};

257 258 259 260 261
struct hash_list
{
	struct hash_list *next;
	unsigned char sha1[20];
};
262

263 264 265 266 267 268
typedef enum {
	WHENSPEC_RAW = 1,
	WHENSPEC_RFC2822,
	WHENSPEC_NOW,
} whenspec_type;

269 270 271 272 273 274 275
struct recent_command
{
	struct recent_command *prev;
	struct recent_command *next;
	char *buf;
};

276
/* Configured limits on output */
277
static unsigned long max_depth = 10;
278
static off_t max_packsize = (1LL << 32) - 1;
279
static int force_update;
280 281
static int pack_compression_level = Z_DEFAULT_COMPRESSION;
static int pack_compression_seen;
282 283 284 285 286 287 288

/* 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];
289
static unsigned long object_count;
290
static unsigned long branch_count;
291
static unsigned long branch_load_count;
292
static int failure;
293
static FILE *pack_edges;
294

295 296 297 298 299
/* 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;

300
/* Atom management */
301 302 303 304 305
static unsigned int atom_table_sz = 4451;
static unsigned int atom_cnt;
static struct atom_str **atom_table;

/* The .pack file being generated */
306
static unsigned int pack_id;
307
static struct packed_git *pack_data;
308
static struct packed_git **all_packs;
309
static unsigned long pack_size;
310 311

/* Table of objects we've written. */
312
static unsigned int object_entry_alloc = 5000;
313 314
static struct object_entry_pool *blocks;
static struct object_entry *object_table[1 << 16];
315
static struct mark_set *marks;
316
static const char *mark_file;
317 318

/* Our last blob */
P
Pierre Habouzit 已提交
319
static struct last_object last_blob = { STRBUF_INIT, 0, 0, 0 };
320 321 322 323 324 325

/* 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;
326 327
static struct strbuf old_tree = STRBUF_INIT;
static struct strbuf new_tree = STRBUF_INIT;
328

329
/* Branch data */
330 331 332
static unsigned long max_active_branches = 5;
static unsigned long cur_active_branches;
static unsigned long branch_table_sz = 1039;
333 334 335
static struct branch **branch_table;
static struct branch *active_branches;

336 337 338 339
/* Tag data */
static struct tag *first_tag;
static struct tag *last_tag;

340
/* Input stream parsing */
341
static whenspec_type whenspec = WHENSPEC_RAW;
342
static struct strbuf command_buf = STRBUF_INIT;
343
static int unread_command_buf;
344 345 346 347
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;
348
static uintmax_t next_mark;
349
static struct strbuf new_data = STRBUF_INIT;
350

351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376
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);
}

377 378
static void dump_marks_helper(FILE *, uintmax_t, struct mark_set *);

379
static void write_crash_report(const char *err)
380
{
381
	char *loc = git_path("fast_import_crash_%"PRIuMAX, (uintmax_t) getpid());
382 383 384
	FILE *rpt = fopen(loc, "w");
	struct branch *b;
	unsigned long lu;
385
	struct recent_command *rc;
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");
395 396
	fprintf(rpt, "    fast-import process: %"PRIuMAX"\n", (uintmax_t) getpid());
	fprintf(rpt, "    parent process     : %"PRIuMAX"\n", (uintmax_t) getppid());
397 398 399 400
	fprintf(rpt, "    at %s\n", show_date(time(NULL), 0, DATE_LOCAL));
	fputc('\n', rpt);

	fputs("fatal: ", rpt);
401
	fputs(err, rpt);
402 403
	fputc('\n', rpt);

404 405 406 407 408 409 410 411 412 413 414 415
	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);
	}

416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436
	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);
	}

437 438 439 440 441 442 443 444 445 446 447 448 449
	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);
		}
	}

450 451 452 453 454 455 456 457
	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);

458 459 460 461 462 463
	fputc('\n', rpt);
	fputs("-------------------\n", rpt);
	fputs("END OF CRASH REPORT\n", rpt);
	fclose(rpt);
}

464 465 466 467
static void end_packfile(void);
static void unkeep_all_packs(void);
static void dump_marks(void);

468 469 470
static NORETURN void die_nicely(const char *err, va_list params)
{
	static int zombie;
471
	char message[2 * PATH_MAX];
472

473
	vsnprintf(message, sizeof(message), err, params);
474
	fputs("fatal: ", stderr);
475
	fputs(message, stderr);
476 477 478 479
	fputc('\n', stderr);

	if (!zombie) {
		zombie = 1;
480
		write_crash_report(message);
481 482 483
		end_packfile();
		unkeep_all_packs();
		dump_marks();
484 485 486
	}
	exit(128);
}
487

488
static void alloc_objects(unsigned int cnt)
489
{
490
	struct object_entry_pool *b;
491

492
	b = xmalloc(sizeof(struct object_entry_pool)
493
		+ cnt * sizeof(struct object_entry));
494
	b->next_pool = blocks;
495 496 497 498 499
	b->next_free = b->entries;
	b->end = b->entries + cnt;
	blocks = b;
	alloc_count += cnt;
}
500

501
static struct object_entry *new_object(unsigned char *sha1)
502
{
503
	struct object_entry *e;
504

505
	if (blocks->next_free == blocks->end)
506
		alloc_objects(object_entry_alloc);
507

508
	e = blocks->next_free++;
509
	hashcpy(e->sha1, sha1);
510
	return e;
511 512
}

513
static struct object_entry *find_object(unsigned char *sha1)
514 515 516 517
{
	unsigned int h = sha1[0] << 8 | sha1[1];
	struct object_entry *e;
	for (e = object_table[h]; e; e = e->next)
518
		if (!hashcmp(sha1, e->sha1))
519 520 521 522
			return e;
	return NULL;
}

523
static struct object_entry *insert_object(unsigned char *sha1)
524 525
{
	unsigned int h = sha1[0] << 8 | sha1[1];
526
	struct object_entry *e = object_table[h];
527
	struct object_entry *p = NULL;
528 529

	while (e) {
530
		if (!hashcmp(sha1, e->sha1))
531 532 533 534 535 536
			return e;
		p = e;
		e = e->next;
	}

	e = new_object(sha1);
537
	e->next = NULL;
538 539 540 541
	e->offset = 0;
	if (p)
		p->next = e;
	else
542
		object_table[h] = e;
543 544
	return e;
}
545

546 547 548 549 550 551 552 553
static unsigned int hc_str(const char *s, size_t len)
{
	unsigned int r = 0;
	while (len-- > 0)
		r = r * 31 + *s++;
	return r;
}

554
static void *pool_alloc(size_t len)
555 556 557 558
{
	struct mem_pool *p;
	void *r;

559 560 561 562
	/* round up to a 'uintmax_t' alignment */
	if (len & (sizeof(uintmax_t) - 1))
		len += sizeof(uintmax_t) - (len & (sizeof(uintmax_t) - 1));

563 564 565 566 567 568 569 570 571 572 573 574
	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;
575
		p->next_free = (char *) p->space;
576 577 578 579 580 581 582 583 584
		p->end = p->next_free + mem_pool_alloc;
		mem_pool = p;
	}

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

585
static void *pool_calloc(size_t count, size_t size)
586 587 588 589 590 591 592
{
	size_t len = count * size;
	void *r = pool_alloc(len);
	memset(r, 0, len);
	return r;
}

593
static char *pool_strdup(const char *s)
594 595 596 597 598 599
{
	char *r = pool_alloc(strlen(s) + 1);
	strcpy(r, s);
	return r;
}

600
static void insert_mark(uintmax_t idnum, struct object_entry *oe)
601 602 603 604 605 606 607 608 609
{
	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) {
610
		uintmax_t i = idnum >> s->shift;
611 612 613 614 615 616 617 618 619 620 621 622
		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;
}

623
static struct object_entry *find_mark(uintmax_t idnum)
624
{
625
	uintmax_t orig_idnum = idnum;
626 627 628 629
	struct mark_set *s = marks;
	struct object_entry *oe = NULL;
	if ((idnum >> s->shift) < 1024) {
		while (s && s->shift) {
630
			uintmax_t i = idnum >> s->shift;
631 632 633 634 635 636 637
			idnum -= i << s->shift;
			s = s->data.sets[i];
		}
		if (s)
			oe = s->data.marked[idnum];
	}
	if (!oe)
638
		die("mark :%" PRIuMAX " not declared", orig_idnum);
639 640 641
	return oe;
}

642
static struct atom_str *to_atom(const char *s, unsigned short len)
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
{
	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;
}

661
static struct branch *lookup_branch(const char *name)
662 663 664 665 666 667 668 669 670 671
{
	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;
}

672
static struct branch *new_branch(const char *name)
673 674
{
	unsigned int hc = hc_str(name, strlen(name)) % branch_table_sz;
675
	struct branch *b = lookup_branch(name);
676 677 678

	if (b)
		die("Invalid attempt to create duplicate branch: %s", name);
679
	switch (check_ref_format(name)) {
680 681 682
	case 0: break; /* its valid */
	case CHECK_REF_FORMAT_ONELEVEL:
		break; /* valid, but too few '/', allow anyway */
683
	default:
684
		die("Branch name doesn't conform to GIT standards: %s", name);
685
	}
686 687 688 689

	b = pool_calloc(1, sizeof(struct branch));
	b->name = pool_strdup(name);
	b->table_next_branch = branch_table[hc];
690 691
	b->branch_tree.versions[0].mode = S_IFDIR;
	b->branch_tree.versions[1].mode = S_IFDIR;
692
	b->active = 0;
693
	b->pack_id = MAX_PACK_ID;
694 695 696 697 698 699 700 701 702 703 704
	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;
}

705
static struct tree_content *new_tree_content(unsigned int cnt)
706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727
{
	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;
728
	t->delta_depth = 0;
729 730 731 732 733 734 735 736
	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);
737 738 739 740 741 742
	f->next_avail = avail_tree_table[hc];
	avail_tree_table[hc] = f;
}

static void release_tree_content_recursive(struct tree_content *t)
{
743 744 745
	unsigned int i;
	for (i = 0; i < t->entry_count; i++)
		release_tree_entry(t->entries[i]);
746
	release_tree_content(t);
747 748
}

749
static struct tree_content *grow_tree_content(
750 751 752 753 754
	struct tree_content *t,
	int amt)
{
	struct tree_content *r = new_tree_content(t->entry_count + amt);
	r->entry_count = t->entry_count;
755
	r->delta_depth = t->delta_depth;
756 757 758 759 760
	memcpy(r->entries,t->entries,t->entry_count*sizeof(t->entries[0]));
	release_tree_content(t);
	return r;
}

761
static struct tree_entry *new_tree_entry(void)
762 763 764 765 766
{
	struct tree_entry *e;

	if (!avail_tree_entry) {
		unsigned int n = tree_entry_alloc;
767
		total_allocd += n * sizeof(struct tree_entry);
768
		avail_tree_entry = e = xmalloc(n * sizeof(struct tree_entry));
769
		while (n-- > 1) {
770 771 772
			*((void**)e) = e + 1;
			e++;
		}
773
		*((void**)e) = NULL;
774 775 776 777 778 779 780 781 782 783
	}

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

static void release_tree_entry(struct tree_entry *e)
{
	if (e->tree)
784
		release_tree_content_recursive(e->tree);
785 786 787 788
	*((void**)e) = avail_tree_entry;
	avail_tree_entry = e;
}

789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813
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;
}

814
static void start_packfile(void)
815
{
816
	static char tmpfile[PATH_MAX];
817
	struct packed_git *p;
818
	struct pack_header hdr;
819
	int pack_fd;
820

821 822
	pack_fd = odb_mkstemp(tmpfile, sizeof(tmpfile),
			      "pack/tmp_pack_XXXXXX");
823 824
	p = xcalloc(1, sizeof(*p) + strlen(tmpfile) + 2);
	strcpy(p->pack_name, tmpfile);
825
	p->pack_fd = pack_fd;
826 827 828 829

	hdr.hdr_signature = htonl(PACK_SIGNATURE);
	hdr.hdr_version = htonl(2);
	hdr.hdr_entries = 0;
830
	write_or_die(p->pack_fd, &hdr, sizeof(hdr));
831 832

	pack_data = p;
833 834
	pack_size = sizeof(hdr);
	object_count = 0;
835 836 837

	all_packs = xrealloc(all_packs, sizeof(*all_packs) * (pack_id + 1));
	all_packs[pack_id] = p;
838 839 840 841 842 843 844 845 846
}

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);
}

847
static char *create_index(void)
848
{
849
	static char tmpfile[PATH_MAX];
850
	git_SHA_CTX ctx;
851 852 853
	struct sha1file *f;
	struct object_entry **idx, **c, **last, *e;
	struct object_entry_pool *o;
854
	uint32_t array[256];
855
	int i, idx_fd;
856 857 858 859 860

	/* 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)
861 862 863
		for (e = o->next_free; e-- != o->entries;)
			if (pack_id == e->pack_id)
				*c++ = e;
864
	last = idx + object_count;
865 866
	if (c != last)
		die("internal consistency error creating the index");
867 868 869 870 871
	qsort(idx, object_count, sizeof(struct object_entry*), oecmp);

	/* Generate the fan-out array. */
	c = idx;
	for (i = 0; i < 256; i++) {
J
Junio C Hamano 已提交
872
		struct object_entry **next = c;
873 874 875 876 877 878 879 880 881
		while (next < last) {
			if ((*next)->sha1[0] != i)
				break;
			next++;
		}
		array[i] = htonl(next - idx);
		c = next;
	}

882 883
	idx_fd = odb_mkstemp(tmpfile, sizeof(tmpfile),
			     "pack/tmp_idx_XXXXXX");
884
	f = sha1fd(idx_fd, tmpfile);
885
	sha1write(f, array, 256 * sizeof(int));
886
	git_SHA1_Init(&ctx);
887
	for (c = idx; c != last; c++) {
888
		uint32_t offset = htonl((*c)->offset);
889 890
		sha1write(f, &offset, 4);
		sha1write(f, (*c)->sha1, sizeof((*c)->sha1));
891
		git_SHA1_Update(&ctx, (*c)->sha1, 20);
892
	}
893
	sha1write(f, pack_data->sha1, sizeof(pack_data->sha1));
894
	sha1close(f, NULL, CSUM_FSYNC);
895
	free(idx);
896
	git_SHA1_Final(pack_data->sha1, &ctx);
897 898 899
	return tmpfile;
}

900
static char *keep_pack(char *curr_index_name)
901 902
{
	static char name[PATH_MAX];
903
	static const char *keep_msg = "fast-import";
904 905
	int keep_fd;

906
	keep_fd = odb_pack_keep(name, sizeof(name), pack_data->sha1);
907
	if (keep_fd < 0)
908
		die_errno("cannot create keep file");
909 910
	write_or_die(keep_fd, keep_msg, strlen(keep_msg));
	if (close(keep_fd))
911
		die_errno("failed to write keep file");
912 913 914 915 916 917 918 919 920 921 922 923 924

	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;
}

925
static void unkeep_all_packs(void)
926 927 928 929 930 931 932 933
{
	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));
934
		unlink_or_warn(name);
935
	}
936 937
}

938
static void end_packfile(void)
939
{
940 941
	struct packed_git *old_p = pack_data, *new_p;

942
	clear_delta_base_cache();
943
	if (object_count) {
944
		char *idx_name;
945 946 947
		int i;
		struct branch *b;
		struct tag *t;
948

949
		close_pack_windows(pack_data);
950
		fixup_pack_header_footer(pack_data->pack_fd, pack_data->sha1,
951 952
				    pack_data->pack_name, object_count,
				    NULL, 0);
953
		close(pack_data->pack_fd);
954
		idx_name = keep_pack(create_index());
955

M
Mike Ralphson 已提交
956
		/* Register the packfile with core git's machinery. */
957 958 959
		new_p = add_packed_git(idx_name, strlen(idx_name), 1);
		if (!new_p)
			die("core git rejected index %s", idx_name);
960
		all_packs[pack_id] = new_p;
961
		install_packed_git(new_p);
962 963

		/* Print the boundary */
964 965 966 967 968 969 970
		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));
				}
971
			}
972 973 974 975 976 977
			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);
978 979 980
		}

		pack_id++;
981
	}
982 983
	else {
		close(old_p->pack_fd);
984
		unlink_or_warn(old_p->pack_name);
985
	}
986 987 988
	free(old_p);

	/* We can't carry a delta across packfiles. */
P
Pierre Habouzit 已提交
989
	strbuf_release(&last_blob.data);
990 991
	last_blob.offset = 0;
	last_blob.depth = 0;
992 993
}

994
static void cycle_packfile(void)
995 996 997 998 999
{
	end_packfile();
	start_packfile();
}

1000
static size_t encode_header(
1001
	enum object_type type,
1002
	size_t size,
1003
	unsigned char *hdr)
1004 1005 1006 1007
{
	int n = 1;
	unsigned char c;

1008
	if (type < OBJ_COMMIT || type > OBJ_REF_DELTA)
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022
		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;
}

1023 1024
static int store_object(
	enum object_type type,
1025
	struct strbuf *dat,
1026
	struct last_object *last,
1027
	unsigned char *sha1out,
1028
	uintmax_t mark)
1029 1030
{
	void *out, *delta;
1031 1032 1033
	struct object_entry *e;
	unsigned char hdr[96];
	unsigned char sha1[20];
1034
	unsigned long hdrlen, deltalen;
1035
	git_SHA_CTX c;
1036 1037
	z_stream s;

1038
	hdrlen = sprintf((char *)hdr,"%s %lu", typename(type),
1039
		(unsigned long)dat->len) + 1;
1040 1041 1042 1043
	git_SHA1_Init(&c);
	git_SHA1_Update(&c, hdr, hdrlen);
	git_SHA1_Update(&c, dat->buf, dat->len);
	git_SHA1_Final(sha1, &c);
1044
	if (sha1out)
1045
		hashcpy(sha1out, sha1);
1046 1047

	e = insert_object(sha1);
1048 1049
	if (mark)
		insert_mark(mark, e);
1050
	if (e->offset) {
1051
		duplicate_count_by_type[type]++;
1052
		return 1;
1053 1054 1055 1056 1057 1058
	} 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;
1059
	}
1060

P
Pierre Habouzit 已提交
1061 1062
	if (last && last->data.buf && last->depth < max_depth) {
		delta = diff_delta(last->data.buf, last->data.len,
1063
			dat->buf, dat->len,
1064
			&deltalen, 0);
1065
		if (delta && deltalen >= dat->len) {
1066 1067 1068 1069 1070
			free(delta);
			delta = NULL;
		}
	} else
		delta = NULL;
1071 1072

	memset(&s, 0, sizeof(s));
1073
	deflateInit(&s, pack_compression_level);
1074 1075 1076 1077
	if (delta) {
		s.next_in = delta;
		s.avail_in = deltalen;
	} else {
1078 1079
		s.next_in = (void *)dat->buf;
		s.avail_in = dat->len;
1080 1081 1082 1083 1084 1085 1086 1087
	}
	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. */
1088
	if ((pack_size + 60 + s.total_out) > max_packsize
1089 1090 1091 1092
		|| (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;
1093
		cycle_packfile();
1094 1095 1096 1097 1098

		/* We cannot carry a delta into the new pack. */
		if (delta) {
			free(delta);
			delta = NULL;
1099 1100

			memset(&s, 0, sizeof(s));
1101
			deflateInit(&s, pack_compression_level);
1102 1103
			s.next_in = (void *)dat->buf;
			s.avail_in = dat->len;
1104 1105 1106 1107 1108
			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);
1109 1110 1111 1112 1113 1114 1115 1116
		}
	}

	e->type = type;
	e->pack_id = pack_id;
	e->offset = pack_size;
	object_count++;
	object_count_by_type[type]++;
1117 1118

	if (delta) {
1119 1120 1121
		unsigned long ofs = e->offset - last->offset;
		unsigned pos = sizeof(hdr) - 1;

1122
		delta_count_by_type[type]++;
1123
		e->depth = last->depth + 1;
1124 1125

		hdrlen = encode_header(OBJ_OFS_DELTA, deltalen, hdr);
1126
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1127 1128 1129 1130 1131
		pack_size += hdrlen;

		hdr[pos] = ofs & 127;
		while (ofs >>= 7)
			hdr[--pos] = 128 | (--ofs & 127);
1132
		write_or_die(pack_data->pack_fd, hdr + pos, sizeof(hdr) - pos);
1133
		pack_size += sizeof(hdr) - pos;
1134
	} else {
1135
		e->depth = 0;
1136
		hdrlen = encode_header(type, dat->len, hdr);
1137
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1138
		pack_size += hdrlen;
1139 1140
	}

1141
	write_or_die(pack_data->pack_fd, out, s.total_out);
1142
	pack_size += s.total_out;
1143 1144

	free(out);
1145
	free(delta);
1146
	if (last) {
P
Pierre Habouzit 已提交
1147 1148 1149
		if (last->no_swap) {
			last->data = *dat;
		} else {
1150
			strbuf_swap(&last->data, dat);
P
Pierre Habouzit 已提交
1151
		}
1152
		last->offset = e->offset;
1153
		last->depth = e->depth;
1154 1155 1156 1157
	}
	return 0;
}

1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175
/* 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.
 */
1176 1177 1178
static void *gfi_unpack_entry(
	struct object_entry *oe,
	unsigned long *sizep)
1179
{
1180
	enum object_type type;
1181
	struct packed_git *p = all_packs[oe->pack_id];
1182
	if (p == pack_data && p->pack_size < (pack_size + 20)) {
1183 1184 1185 1186 1187 1188 1189 1190
		/* 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.
		 */
1191
		close_pack_windows(p);
1192 1193 1194 1195 1196 1197 1198

		/* 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.
		 */
1199
		p->pack_size = pack_size + 20;
1200
	}
1201
	return unpack_entry(p, oe->offset, &type, sizep);
1202 1203
}

1204
static const char *get_mode(const char *str, uint16_t *modep)
1205 1206
{
	unsigned char c;
1207
	uint16_t mode = 0;
1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219

	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)
{
1220
	unsigned char *sha1 = root->versions[1].sha1;
1221 1222 1223 1224 1225 1226 1227
	struct object_entry *myoe;
	struct tree_content *t;
	unsigned long size;
	char *buf;
	const char *c;

	root->tree = t = new_tree_content(8);
1228
	if (is_null_sha1(sha1))
1229 1230
		return;

1231
	myoe = find_object(sha1);
1232
	if (myoe && myoe->pack_id != MAX_PACK_ID) {
1233
		if (myoe->type != OBJ_TREE)
1234
			die("Not a tree: %s", sha1_to_hex(sha1));
1235
		t->delta_depth = myoe->depth;
1236
		buf = gfi_unpack_entry(myoe, &size);
1237 1238
		if (!buf)
			die("Can't load tree %s", sha1_to_hex(sha1));
1239
	} else {
1240 1241 1242
		enum object_type type;
		buf = read_sha1_file(sha1, &type, &size);
		if (!buf || type != OBJ_TREE)
1243
			die("Can't load tree %s", sha1_to_hex(sha1));
1244 1245 1246 1247 1248 1249 1250
	}

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

		if (t->entry_count == t->entry_capacity)
1251
			root->tree = t = grow_tree_content(t, t->entry_count);
1252 1253 1254
		t->entries[t->entry_count++] = e;

		e->tree = NULL;
1255
		c = get_mode(c, &e->versions[1].mode);
1256
		if (!c)
1257 1258
			die("Corrupt mode in %s", sha1_to_hex(sha1));
		e->versions[0].mode = e->versions[1].mode;
1259
		e->name = to_atom(c, strlen(c));
1260
		c += e->name->str_len + 1;
1261 1262
		hashcpy(e->versions[0].sha1, (unsigned char *)c);
		hashcpy(e->versions[1].sha1, (unsigned char *)c);
1263 1264 1265 1266 1267
		c += 20;
	}
	free(buf);
}

1268
static int tecmp0 (const void *_a, const void *_b)
1269 1270 1271 1272
{
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
1273 1274
		a->name->str_dat, a->name->str_len, a->versions[0].mode,
		b->name->str_dat, b->name->str_len, b->versions[0].mode);
1275 1276
}

1277
static int tecmp1 (const void *_a, const void *_b)
1278
{
1279 1280 1281 1282 1283 1284 1285
	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);
}

1286
static void mktree(struct tree_content *t, int v, struct strbuf *b)
1287 1288
{
	size_t maxlen = 0;
1289 1290
	unsigned int i;

1291 1292 1293 1294
	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);
1295 1296

	for (i = 0; i < t->entry_count; i++) {
1297 1298
		if (t->entries[i]->versions[v].mode)
			maxlen += t->entries[i]->name->str_len + 34;
1299 1300
	}

1301 1302
	strbuf_reset(b);
	strbuf_grow(b, maxlen);
1303 1304
	for (i = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
1305 1306
		if (!e->versions[v].mode)
			continue;
1307 1308 1309
		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);
1310
	}
1311 1312 1313 1314 1315 1316
}

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
1317
	struct last_object lo = { STRBUF_INIT, 0, 0, /* no_swap */ 1 };
1318
	struct object_entry *le;
1319 1320 1321 1322 1323 1324 1325 1326 1327

	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]);
	}

1328
	le = find_object(root->versions[0].sha1);
P
Pierre Habouzit 已提交
1329
	if (S_ISDIR(root->versions[0].mode) && le && le->pack_id == pack_id) {
1330
		mktree(t, 0, &old_tree);
P
Pierre Habouzit 已提交
1331
		lo.data = old_tree;
1332
		lo.offset = le->offset;
1333 1334 1335
		lo.depth = t->delta_depth;
	}

1336 1337
	mktree(t, 1, &new_tree);
	store_object(OBJ_TREE, &new_tree, &lo, root->versions[1].sha1, 0);
1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351

	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;
1352 1353 1354 1355 1356 1357
}

static int tree_content_set(
	struct tree_entry *root,
	const char *p,
	const unsigned char *sha1,
1358 1359
	const uint16_t mode,
	struct tree_content *subtree)
1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370
{
	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);
1371 1372
	if (!n)
		die("Empty path component found in input");
1373 1374
	if (!slash1 && !S_ISDIR(mode) && subtree)
		die("Non-directories cannot have subtrees");
1375 1376 1377 1378 1379

	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) {
1380 1381
				if (!S_ISDIR(mode)
						&& e->versions[1].mode == mode
1382
						&& !hashcmp(e->versions[1].sha1, sha1))
1383
					return 0;
1384 1385
				e->versions[1].mode = mode;
				hashcpy(e->versions[1].sha1, sha1);
1386
				if (e->tree)
1387
					release_tree_content_recursive(e->tree);
1388
				e->tree = subtree;
1389
				hashclr(root->versions[1].sha1);
1390 1391
				return 1;
			}
1392
			if (!S_ISDIR(e->versions[1].mode)) {
1393
				e->tree = new_tree_content(8);
1394
				e->versions[1].mode = S_IFDIR;
1395 1396 1397
			}
			if (!e->tree)
				load_tree(e);
1398
			if (tree_content_set(e, slash1 + 1, sha1, mode, subtree)) {
1399
				hashclr(root->versions[1].sha1);
1400 1401 1402 1403 1404 1405 1406
				return 1;
			}
			return 0;
		}
	}

	if (t->entry_count == t->entry_capacity)
1407
		root->tree = t = grow_tree_content(t, t->entry_count);
1408
	e = new_tree_entry();
1409
	e->name = to_atom(p, n);
1410 1411
	e->versions[0].mode = 0;
	hashclr(e->versions[0].sha1);
1412 1413 1414
	t->entries[t->entry_count++] = e;
	if (slash1) {
		e->tree = new_tree_content(8);
1415
		e->versions[1].mode = S_IFDIR;
1416
		tree_content_set(e, slash1 + 1, sha1, mode, subtree);
1417
	} else {
1418
		e->tree = subtree;
1419 1420
		e->versions[1].mode = mode;
		hashcpy(e->versions[1].sha1, sha1);
1421
	}
1422
	hashclr(root->versions[1].sha1);
1423 1424 1425
	return 1;
}

1426 1427 1428 1429
static int tree_content_remove(
	struct tree_entry *root,
	const char *p,
	struct tree_entry *backup_leaf)
1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444
{
	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)) {
1445
			if (!slash1 || !S_ISDIR(e->versions[1].mode))
1446 1447 1448
				goto del_entry;
			if (!e->tree)
				load_tree(e);
1449
			if (tree_content_remove(e, slash1 + 1, backup_leaf)) {
1450 1451 1452 1453 1454 1455
				for (n = 0; n < e->tree->entry_count; n++) {
					if (e->tree->entries[n]->versions[1].mode) {
						hashclr(root->versions[1].sha1);
						return 1;
					}
				}
1456
				backup_leaf = NULL;
1457
				goto del_entry;
1458 1459 1460 1461 1462 1463 1464
			}
			return 0;
		}
	}
	return 0;

del_entry:
1465 1466 1467
	if (backup_leaf)
		memcpy(backup_leaf, e, sizeof(*backup_leaf));
	else if (e->tree)
1468
		release_tree_content_recursive(e->tree);
1469
	e->tree = NULL;
1470 1471 1472
	e->versions[1].mode = 0;
	hashclr(e->versions[1].sha1);
	hashclr(root->versions[1].sha1);
1473
	return 1;
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 1506 1507 1508 1509 1510 1511 1512
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;
}

1513
static int update_branch(struct branch *b)
1514 1515
{
	static const char *msg = "fast-import";
1516 1517 1518
	struct ref_lock *lock;
	unsigned char old_sha1[20];

1519 1520
	if (is_null_sha1(b->sha1))
		return 0;
1521 1522
	if (read_ref(b->name, old_sha1))
		hashclr(old_sha1);
1523
	lock = lock_any_ref_for_update(b->name, old_sha1, 0);
1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535
	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);
		}

1536
		if (!in_merge_bases(old_cmit, &new_cmit, 1)) {
1537
			unlock_ref(lock);
1538
			warning("Not updating %s"
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550
				" (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)
{
1551 1552 1553 1554
	unsigned int i;
	struct branch *b;

	for (i = 0; i < branch_table_sz; i++) {
1555 1556
		for (b = branch_table[i]; b; b = b->table_next_branch)
			failure |= update_branch(b);
1557 1558 1559
	}
}

1560
static void dump_tags(void)
1561 1562 1563 1564
{
	static const char *msg = "fast-import";
	struct tag *t;
	struct ref_lock *lock;
1565
	char ref_name[PATH_MAX];
1566 1567

	for (t = first_tag; t; t = t->next_tag) {
1568 1569
		sprintf(ref_name, "tags/%s", t->name);
		lock = lock_ref_sha1(ref_name, NULL);
1570
		if (!lock || write_ref_sha1(lock, t->sha1, msg) < 0)
1571
			failure |= error("Unable to update %s", ref_name);
1572 1573 1574
	}
}

1575
static void dump_marks_helper(FILE *f,
1576
	uintmax_t base,
1577 1578
	struct mark_set *m)
{
1579
	uintmax_t k;
1580 1581 1582 1583 1584 1585 1586 1587 1588
	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])
1589
				fprintf(f, ":%" PRIuMAX " %s\n", base + k,
1590 1591 1592 1593 1594
					sha1_to_hex(m->data.marked[k]->sha1));
		}
	}
}

1595
static void dump_marks(void)
1596
{
1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608
	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;
1609
	}
1610 1611 1612

	f = fdopen(mark_fd, "w");
	if (!f) {
1613
		int saved_errno = errno;
1614 1615
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1616
			mark_file, strerror(saved_errno));
1617
		return;
1618
	}
1619

B
Brandon Casey 已提交
1620
	/*
1621 1622
	 * 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
1623 1624 1625
	 * won't try to close() it.
	 */
	mark_lock.fd = -1;
1626 1627 1628

	dump_marks_helper(f, 0, marks);
	if (ferror(f) || fclose(f)) {
1629
		int saved_errno = errno;
1630 1631
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1632
			mark_file, strerror(saved_errno));
1633 1634 1635 1636
		return;
	}

	if (commit_lock_file(&mark_lock)) {
1637
		int saved_errno = errno;
1638 1639
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to commit marks file %s: %s",
1640
			mark_file, strerror(saved_errno));
1641 1642
		return;
	}
1643 1644
}

1645
static int read_next_command(void)
1646
{
1647 1648 1649 1650 1651 1652 1653
	static int stdin_eof = 0;

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

1654
	do {
1655
		if (unread_command_buf) {
1656
			unread_command_buf = 0;
1657 1658 1659
		} else {
			struct recent_command *rc;

1660
			strbuf_detach(&command_buf, NULL);
1661 1662 1663
			stdin_eof = strbuf_getline(&command_buf, stdin, '\n');
			if (stdin_eof)
				return EOF;
1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681

			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] == '#');
1682 1683

	return 0;
1684 1685
}

J
Junio C Hamano 已提交
1686
static void skip_optional_lf(void)
1687 1688 1689 1690 1691 1692
{
	int term_char = fgetc(stdin);
	if (term_char != '\n' && term_char != EOF)
		ungetc(term_char, stdin);
}

1693
static void parse_mark(void)
1694
{
1695
	if (!prefixcmp(command_buf.buf, "mark :")) {
1696
		next_mark = strtoumax(command_buf.buf + 6, NULL, 10);
1697 1698 1699
		read_next_command();
	}
	else
1700
		next_mark = 0;
1701 1702
}

1703
static void parse_data(struct strbuf *sb)
1704
{
1705
	strbuf_reset(sb);
1706

1707
	if (prefixcmp(command_buf.buf, "data "))
1708 1709
		die("Expected 'data n' command, found: %s", command_buf.buf);

1710
	if (!prefixcmp(command_buf.buf + 5, "<<")) {
1711
		char *term = xstrdup(command_buf.buf + 5 + 2);
1712 1713
		size_t term_len = command_buf.len - 5 - 2;

1714
		strbuf_detach(&command_buf, NULL);
1715
		for (;;) {
1716
			if (strbuf_getline(&command_buf, stdin, '\n') == EOF)
1717 1718 1719 1720
				die("EOF in data (terminator '%s' not found)", term);
			if (term_len == command_buf.len
				&& !strcmp(term, command_buf.buf))
				break;
1721 1722
			strbuf_addbuf(sb, &command_buf);
			strbuf_addch(sb, '\n');
1723 1724 1725 1726
		}
		free(term);
	}
	else {
1727 1728
		size_t n = 0, length;

1729
		length = strtoul(command_buf.buf + 5, NULL, 10);
1730

1731
		while (n < length) {
1732
			size_t s = strbuf_fread(sb, length - n, stdin);
1733
			if (!s && feof(stdin))
1734 1735
				die("EOF in data (%lu bytes remaining)",
					(unsigned long)(length - n));
1736 1737
			n += s;
		}
1738 1739
	}

1740
	skip_optional_lf();
1741 1742
}

1743 1744 1745
static int validate_raw_date(const char *src, char *result, int maxlen)
{
	const char *orig_src = src;
1746
	char *endp;
1747

1748 1749
	errno = 0;

1750
	strtoul(src, &endp, 10);
1751
	if (errno || endp == src || *endp != ' ')
1752 1753 1754 1755 1756 1757
		return -1;

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

1758
	strtoul(src + 1, &endp, 10);
1759
	if (errno || endp == src || *endp || (endp - orig_src) >= maxlen)
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 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801
		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;
}

1802
static void parse_new_blob(void)
1803
{
P
Pierre Habouzit 已提交
1804
	static struct strbuf buf = STRBUF_INIT;
1805 1806

	read_next_command();
1807 1808
	parse_mark();
	parse_data(&buf);
P
Pierre Habouzit 已提交
1809
	store_object(OBJ_BLOB, &buf, &last_blob, NULL, next_mark);
1810 1811
}

1812
static void unload_one_branch(void)
1813
{
1814 1815
	while (cur_active_branches
		&& cur_active_branches >= max_active_branches) {
1816
		uintmax_t min_commit = ULONG_MAX;
1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833
		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;
		}
1834
		e->active = 0;
1835 1836
		e->active_next_branch = NULL;
		if (e->branch_tree.tree) {
1837
			release_tree_content_recursive(e->branch_tree.tree);
1838 1839 1840
			e->branch_tree.tree = NULL;
		}
		cur_active_branches--;
1841 1842 1843
	}
}

1844
static void load_branch(struct branch *b)
1845
{
1846
	load_tree(&b->branch_tree);
1847 1848 1849 1850 1851 1852 1853
	if (!b->active) {
		b->active = 1;
		b->active_next_branch = active_branches;
		active_branches = b;
		cur_active_branches++;
		branch_load_count++;
	}
1854 1855
}

1856
static void file_change_m(struct branch *b)
1857
{
1858
	const char *p = command_buf.buf + 2;
1859
	static struct strbuf uq = STRBUF_INIT;
1860
	const char *endp;
1861
	struct object_entry *oe = oe;
1862
	unsigned char sha1[20];
1863
	uint16_t mode, inline_data = 0;
1864

1865 1866 1867 1868
	p = get_mode(p, &mode);
	if (!p)
		die("Corrupt mode: %s", command_buf.buf);
	switch (mode) {
1869 1870 1871
	case 0644:
	case 0755:
		mode |= S_IFREG;
1872 1873
	case S_IFREG | 0644:
	case S_IFREG | 0755:
1874
	case S_IFLNK:
1875
	case S_IFGITLINK:
1876 1877 1878 1879 1880 1881
		/* ok */
		break;
	default:
		die("Corrupt mode: %s", command_buf.buf);
	}

1882 1883
	if (*p == ':') {
		char *x;
1884
		oe = find_mark(strtoumax(p + 1, &x, 10));
1885
		hashcpy(sha1, oe->sha1);
1886
		p = x;
1887
	} else if (!prefixcmp(p, "inline")) {
1888 1889
		inline_data = 1;
		p += 6;
1890 1891 1892 1893 1894 1895
	} else {
		if (get_sha1_hex(p, sha1))
			die("Invalid SHA1: %s", command_buf.buf);
		oe = find_object(sha1);
		p += 40;
	}
1896 1897 1898
	if (*p++ != ' ')
		die("Missing space after SHA1: %s", command_buf.buf);

1899 1900
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1901 1902
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1903
		p = uq.buf;
1904
	}
1905

1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919
	if (S_ISGITLINK(mode)) {
		if (inline_data)
			die("Git links cannot be specified 'inline': %s",
				command_buf.buf);
		else if (oe) {
			if (oe->type != OBJ_COMMIT)
				die("Not a commit (actually a %s): %s",
					typename(oe->type), command_buf.buf);
		}
		/*
		 * Accept the sha1 without checking; it expected to be in
		 * another repository.
		 */
	} else if (inline_data) {
P
Pierre Habouzit 已提交
1920
		static struct strbuf buf = STRBUF_INIT;
1921

1922 1923 1924 1925
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
1926
		read_next_command();
1927
		parse_data(&buf);
P
Pierre Habouzit 已提交
1928
		store_object(OBJ_BLOB, &buf, &last_blob, sha1, 0);
1929
	} else if (oe) {
1930
		if (oe->type != OBJ_BLOB)
1931
			die("Not a blob (actually a %s): %s",
1932
				typename(oe->type), command_buf.buf);
1933
	} else {
1934 1935
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
1936
			die("Blob not found: %s", command_buf.buf);
1937
		if (type != OBJ_BLOB)
1938
			die("Not a blob (actually a %s): %s",
1939
			    typename(type), command_buf.buf);
1940
	}
1941

1942
	tree_content_set(&b->branch_tree, p, sha1, mode, NULL);
1943
}
1944

1945 1946
static void file_change_d(struct branch *b)
{
1947
	const char *p = command_buf.buf + 2;
1948
	static struct strbuf uq = STRBUF_INIT;
1949 1950
	const char *endp;

1951 1952
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1953 1954
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1955
		p = uq.buf;
1956
	}
1957
	tree_content_remove(&b->branch_tree, p, NULL);
1958 1959
}

1960
static void file_change_cr(struct branch *b, int rename)
1961 1962
{
	const char *s, *d;
1963 1964
	static struct strbuf s_uq = STRBUF_INIT;
	static struct strbuf d_uq = STRBUF_INIT;
1965 1966 1967 1968
	const char *endp;
	struct tree_entry leaf;

	s = command_buf.buf + 2;
1969 1970
	strbuf_reset(&s_uq);
	if (!unquote_c_style(&s_uq, s, &endp)) {
1971 1972
		if (*endp != ' ')
			die("Missing space after source: %s", command_buf.buf);
1973
	} else {
1974 1975 1976
		endp = strchr(s, ' ');
		if (!endp)
			die("Missing space after source: %s", command_buf.buf);
1977
		strbuf_add(&s_uq, s, endp - s);
1978
	}
1979
	s = s_uq.buf;
1980 1981 1982 1983 1984 1985

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

	d = endp;
1986 1987
	strbuf_reset(&d_uq);
	if (!unquote_c_style(&d_uq, d, &endp)) {
1988 1989
		if (*endp)
			die("Garbage after dest in: %s", command_buf.buf);
1990
		d = d_uq.buf;
1991 1992 1993
	}

	memset(&leaf, 0, sizeof(leaf));
1994 1995 1996 1997
	if (rename)
		tree_content_remove(&b->branch_tree, s, &leaf);
	else
		tree_content_get(&b->branch_tree, s, &leaf);
1998 1999 2000 2001 2002 2003 2004 2005
	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);
}

2006 2007 2008 2009 2010 2011 2012 2013
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);
}

2014
static void parse_from_commit(struct branch *b, char *buf, unsigned long size)
2015 2016 2017 2018 2019 2020 2021 2022 2023 2024
{
	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);
}

2025
static void parse_from_existing(struct branch *b)
2026 2027 2028 2029 2030 2031 2032 2033 2034 2035
{
	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);
2036
		parse_from_commit(b, buf, size);
2037 2038 2039 2040
		free(buf);
	}
}

2041
static int parse_from(struct branch *b)
2042
{
2043
	const char *from;
2044 2045
	struct branch *s;

2046
	if (prefixcmp(command_buf.buf, "from "))
2047
		return 0;
2048

2049 2050 2051 2052
	if (b->branch_tree.tree) {
		release_tree_content_recursive(b->branch_tree.tree);
		b->branch_tree.tree = NULL;
	}
2053 2054 2055 2056 2057 2058

	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) {
2059
		unsigned char *t = s->branch_tree.versions[1].sha1;
2060
		hashcpy(b->sha1, s->sha1);
2061 2062
		hashcpy(b->branch_tree.versions[0].sha1, t);
		hashcpy(b->branch_tree.versions[1].sha1, t);
2063
	} else if (*from == ':') {
2064
		uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2065 2066
		struct object_entry *oe = find_mark(idnum);
		if (oe->type != OBJ_COMMIT)
2067
			die("Mark :%" PRIuMAX " not a commit", idnum);
2068
		hashcpy(b->sha1, oe->sha1);
2069
		if (oe->pack_id != MAX_PACK_ID) {
2070
			unsigned long size;
2071
			char *buf = gfi_unpack_entry(oe, &size);
2072
			parse_from_commit(b, buf, size);
2073
			free(buf);
2074
		} else
2075
			parse_from_existing(b);
2076
	} else if (!get_sha1(from, b->sha1))
2077
		parse_from_existing(b);
2078
	else
2079 2080 2081
		die("Invalid ref name or SHA1 expression: %s", from);

	read_next_command();
2082
	return 1;
2083 2084
}

2085
static struct hash_list *parse_merge(unsigned int *count)
2086
{
2087
	struct hash_list *list = NULL, *n, *e = e;
2088
	const char *from;
2089 2090 2091
	struct branch *s;

	*count = 0;
2092
	while (!prefixcmp(command_buf.buf, "merge ")) {
2093 2094 2095 2096 2097 2098
		from = strchr(command_buf.buf, ' ') + 1;
		n = xmalloc(sizeof(*n));
		s = lookup_branch(from);
		if (s)
			hashcpy(n->sha1, s->sha1);
		else if (*from == ':') {
2099
			uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2100 2101
			struct object_entry *oe = find_mark(idnum);
			if (oe->type != OBJ_COMMIT)
2102
				die("Mark :%" PRIuMAX " not a commit", idnum);
2103
			hashcpy(n->sha1, oe->sha1);
2104 2105 2106
		} 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
2107
				commit_type, &size, n->sha1);
2108 2109 2110 2111
			if (!buf || size < 46)
				die("Not a valid commit: %s", from);
			free(buf);
		} else
2112 2113 2114 2115 2116 2117 2118 2119
			die("Invalid ref name or SHA1 expression: %s", from);

		n->next = NULL;
		if (list)
			e->next = n;
		else
			list = n;
		e = n;
2120
		(*count)++;
2121 2122 2123 2124 2125
		read_next_command();
	}
	return list;
}

2126
static void parse_new_commit(void)
2127
{
2128
	static struct strbuf msg = STRBUF_INIT;
2129 2130 2131 2132
	struct branch *b;
	char *sp;
	char *author = NULL;
	char *committer = NULL;
2133 2134
	struct hash_list *merge_list = NULL;
	unsigned int merge_count;
2135 2136 2137 2138

	/* Obtain the branch name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	b = lookup_branch(sp);
2139
	if (!b)
2140
		b = new_branch(sp);
2141 2142

	read_next_command();
2143
	parse_mark();
2144
	if (!prefixcmp(command_buf.buf, "author ")) {
2145
		author = parse_ident(command_buf.buf + 7);
2146 2147
		read_next_command();
	}
2148
	if (!prefixcmp(command_buf.buf, "committer ")) {
2149
		committer = parse_ident(command_buf.buf + 10);
2150 2151 2152 2153
		read_next_command();
	}
	if (!committer)
		die("Expected committer but didn't get one");
2154
	parse_data(&msg);
2155
	read_next_command();
2156 2157
	parse_from(b);
	merge_list = parse_merge(&merge_count);
2158 2159

	/* ensure the branch is active/loaded */
2160
	if (!b->branch_tree.tree || !max_active_branches) {
2161 2162 2163
		unload_one_branch();
		load_branch(b);
	}
2164

2165
	/* file_change* */
2166
	while (command_buf.len > 0) {
2167
		if (!prefixcmp(command_buf.buf, "M "))
2168
			file_change_m(b);
2169
		else if (!prefixcmp(command_buf.buf, "D "))
2170
			file_change_d(b);
2171
		else if (!prefixcmp(command_buf.buf, "R "))
2172 2173 2174
			file_change_cr(b, 1);
		else if (!prefixcmp(command_buf.buf, "C "))
			file_change_cr(b, 0);
2175 2176
		else if (!strcmp("deleteall", command_buf.buf))
			file_change_deleteall(b);
2177 2178 2179 2180
		else {
			unread_command_buf = 1;
			break;
		}
2181 2182
		if (read_next_command() == EOF)
			break;
2183 2184
	}

2185
	/* build the tree and the commit */
2186
	store_tree(&b->branch_tree);
2187 2188
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
2189 2190 2191

	strbuf_reset(&new_data);
	strbuf_addf(&new_data, "tree %s\n",
2192
		sha1_to_hex(b->branch_tree.versions[1].sha1));
2193
	if (!is_null_sha1(b->sha1))
2194
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(b->sha1));
2195 2196
	while (merge_list) {
		struct hash_list *next = merge_list->next;
2197
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(merge_list->sha1));
2198 2199 2200
		free(merge_list);
		merge_list = next;
	}
2201 2202 2203 2204 2205 2206
	strbuf_addf(&new_data,
		"author %s\n"
		"committer %s\n"
		"\n",
		author ? author : committer, committer);
	strbuf_addbuf(&new_data, &msg);
2207
	free(author);
2208 2209
	free(committer);

2210
	if (!store_object(OBJ_COMMIT, &new_data, NULL, b->sha1, next_mark))
2211
		b->pack_id = pack_id;
2212
	b->last_commit = object_count_by_type[OBJ_COMMIT];
2213 2214
}

2215
static void parse_new_tag(void)
2216
{
2217
	static struct strbuf msg = STRBUF_INIT;
2218 2219 2220 2221 2222
	char *sp;
	const char *from;
	char *tagger;
	struct branch *s;
	struct tag *t;
2223
	uintmax_t from_mark = 0;
2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238
	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 ... */
2239
	if (prefixcmp(command_buf.buf, "from "))
2240 2241 2242 2243
		die("Expected from command, got %s", command_buf.buf);
	from = strchr(command_buf.buf, ' ') + 1;
	s = lookup_branch(from);
	if (s) {
2244
		hashcpy(sha1, s->sha1);
2245
	} else if (*from == ':') {
2246
		struct object_entry *oe;
2247
		from_mark = strtoumax(from + 1, NULL, 10);
2248
		oe = find_mark(from_mark);
2249
		if (oe->type != OBJ_COMMIT)
2250
			die("Mark :%" PRIuMAX " not a commit", from_mark);
2251
		hashcpy(sha1, oe->sha1);
2252 2253 2254 2255 2256
	} else if (!get_sha1(from, sha1)) {
		unsigned long size;
		char *buf;

		buf = read_object_with_reference(sha1,
2257
			commit_type, &size, sha1);
2258 2259 2260 2261 2262 2263 2264 2265
		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 ... */
2266 2267 2268 2269 2270
	if (!prefixcmp(command_buf.buf, "tagger ")) {
		tagger = parse_ident(command_buf.buf + 7);
		read_next_command();
	} else
		tagger = NULL;
2271 2272

	/* tag payload/message */
2273
	parse_data(&msg);
2274 2275

	/* build the tag object */
2276
	strbuf_reset(&new_data);
2277

2278
	strbuf_addf(&new_data,
2279 2280 2281 2282 2283 2284 2285 2286
		    "object %s\n"
		    "type %s\n"
		    "tag %s\n",
		    sha1_to_hex(sha1), commit_type, t->name);
	if (tagger)
		strbuf_addf(&new_data,
			    "tagger %s\n", tagger);
	strbuf_addch(&new_data, '\n');
2287
	strbuf_addbuf(&new_data, &msg);
2288 2289
	free(tagger);

2290
	if (store_object(OBJ_TAG, &new_data, NULL, t->sha1, 0))
2291 2292 2293
		t->pack_id = MAX_PACK_ID;
	else
		t->pack_id = pack_id;
2294 2295
}

2296
static void parse_reset_branch(void)
2297 2298 2299 2300 2301 2302 2303 2304
{
	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) {
2305 2306 2307
		hashclr(b->sha1);
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
2308 2309 2310 2311 2312
		if (b->branch_tree.tree) {
			release_tree_content_recursive(b->branch_tree.tree);
			b->branch_tree.tree = NULL;
		}
	}
2313 2314 2315
	else
		b = new_branch(sp);
	read_next_command();
2316
	parse_from(b);
2317
	if (command_buf.len > 0)
2318
		unread_command_buf = 1;
2319 2320
}

2321
static void parse_checkpoint(void)
2322
{
2323 2324 2325 2326 2327 2328
	if (object_count) {
		cycle_packfile();
		dump_branches();
		dump_tags();
		dump_marks();
	}
2329
	skip_optional_lf();
2330 2331
}

2332
static void parse_progress(void)
2333
{
2334
	fwrite(command_buf.buf, 1, command_buf.len, stdout);
2335 2336 2337 2338 2339
	fputc('\n', stdout);
	fflush(stdout);
	skip_optional_lf();
}

2340 2341 2342 2343 2344
static void import_marks(const char *input_file)
{
	char line[512];
	FILE *f = fopen(input_file, "r");
	if (!f)
2345
		die_errno("cannot read '%s'", input_file);
2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367
	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;
2368
			e->offset = 1; /* just not zero! */
2369 2370 2371 2372 2373 2374
		}
		insert_mark(mark, e);
	}
	fclose(f);
}

2375
static int git_pack_config(const char *k, const char *v, void *cb)
2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392
{
	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;
	}
2393
	return git_default_config(k, v, cb);
2394 2395
}

2396
static const char fast_import_usage[] =
S
Stephan Beyer 已提交
2397
"git fast-import [--date-format=f] [--max-pack-size=n] [--depth=n] [--active-branches=n] [--export-marks=marks.file]";
2398

2399 2400
int main(int argc, const char **argv)
{
2401
	unsigned int i, show_stats = 1;
2402

2403 2404
	git_extract_argv0_path(argv[0]);

2405
	setup_git_directory();
2406
	git_config(git_pack_config, NULL);
2407 2408 2409
	if (!pack_compression_seen && core_compression_seen)
		pack_compression_level = core_compression_level;

2410
	alloc_objects(object_entry_alloc);
2411
	strbuf_init(&command_buf, 0);
2412 2413 2414 2415
	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));
2416

2417 2418 2419 2420 2421
	for (i = 1; i < argc; i++) {
		const char *a = argv[i];

		if (*a != '-' || !strcmp(a, "--"))
			break;
2422
		else if (!prefixcmp(a, "--date-format=")) {
2423 2424 2425 2426 2427 2428 2429 2430 2431 2432
			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);
		}
2433
		else if (!prefixcmp(a, "--max-pack-size="))
2434
			max_packsize = strtoumax(a + 16, NULL, 0) * 1024 * 1024;
2435
		else if (!prefixcmp(a, "--depth=")) {
2436
			max_depth = strtoul(a + 8, NULL, 0);
2437 2438 2439
			if (max_depth > MAX_DEPTH)
				die("--depth cannot exceed %u", MAX_DEPTH);
		}
2440
		else if (!prefixcmp(a, "--active-branches="))
2441
			max_active_branches = strtoul(a + 18, NULL, 0);
2442 2443
		else if (!prefixcmp(a, "--import-marks="))
			import_marks(a + 15);
2444
		else if (!prefixcmp(a, "--export-marks="))
2445
			mark_file = a + 15;
2446
		else if (!prefixcmp(a, "--export-pack-edges=")) {
2447 2448 2449 2450
			if (pack_edges)
				fclose(pack_edges);
			pack_edges = fopen(a + 20, "a");
			if (!pack_edges)
2451
				die_errno("Cannot open '%s'", a + 20);
2452
		} else if (!strcmp(a, "--force"))
2453
			force_update = 1;
2454 2455 2456 2457
		else if (!strcmp(a, "--quiet"))
			show_stats = 0;
		else if (!strcmp(a, "--stats"))
			show_stats = 1;
2458 2459 2460
		else
			die("unknown option %s", a);
	}
2461
	if (i != argc)
2462 2463
		usage(fast_import_usage);

2464 2465 2466 2467 2468
	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;

2469
	prepare_packed_git();
2470
	start_packfile();
2471
	set_die_routine(die_nicely);
2472 2473
	while (read_next_command() != EOF) {
		if (!strcmp("blob", command_buf.buf))
2474
			parse_new_blob();
2475
		else if (!prefixcmp(command_buf.buf, "commit "))
2476
			parse_new_commit();
2477
		else if (!prefixcmp(command_buf.buf, "tag "))
2478
			parse_new_tag();
2479
		else if (!prefixcmp(command_buf.buf, "reset "))
2480
			parse_reset_branch();
2481
		else if (!strcmp("checkpoint", command_buf.buf))
2482
			parse_checkpoint();
2483
		else if (!prefixcmp(command_buf.buf, "progress "))
2484
			parse_progress();
2485 2486
		else
			die("Unsupported command: %s", command_buf.buf);
2487
	}
2488
	end_packfile();
2489

2490
	dump_branches();
2491
	dump_tags();
2492
	unkeep_all_packs();
2493
	dump_marks();
2494

2495 2496 2497
	if (pack_edges)
		fclose(pack_edges);

2498 2499 2500 2501 2502 2503 2504 2505 2506
	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");
2507 2508 2509 2510 2511 2512
		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]);
2513
		fprintf(stderr, "Total branches:  %10lu (%10lu loads     )\n", branch_count, branch_load_count);
2514
		fprintf(stderr, "      marks:     %10" PRIuMAX " (%10" PRIuMAX " unique    )\n", (((uintmax_t)1) << marks->shift) * 1024, marks_set_count);
2515
		fprintf(stderr, "      atoms:     %10u\n", atom_cnt);
2516
		fprintf(stderr, "Memory total:    %10" PRIuMAX " KiB\n", (total_allocd + alloc_count*sizeof(struct object_entry))/1024);
2517
		fprintf(stderr, "       pools:    %10lu KiB\n", (unsigned long)(total_allocd/1024));
2518
		fprintf(stderr, "     objects:    %10" PRIuMAX " KiB\n", (alloc_count*sizeof(struct object_entry))/1024);
2519 2520 2521 2522 2523
		fprintf(stderr, "---------------------------------------------------------------------\n");
		pack_report();
		fprintf(stderr, "---------------------------------------------------------------------\n");
		fprintf(stderr, "\n");
	}
2524

2525
	return failure ? 1 : 0;
2526
}