fast-import.c 62.7 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
	char *loc = git_path("fast_import_crash_%"PRIuMAX, (uintmax_t) getpid());
380 381 382
	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

	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");
393 394
	fprintf(rpt, "    fast-import process: %"PRIuMAX"\n", (uintmax_t) getpid());
	fprintf(rpt, "    parent process     : %"PRIuMAX"\n", (uintmax_t) getppid());
395 396 397 398
	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
	fputc('\n', rpt);
	fputs("-------------------\n", rpt);
	fputs("END OF CRASH REPORT\n", rpt);
	fclose(rpt);
}

462 463 464 465
static void end_packfile(void);
static void unkeep_all_packs(void);
static void dump_marks(void);

466 467 468
static NORETURN void die_nicely(const char *err, va_list params)
{
	static int zombie;
469
	char message[2 * PATH_MAX];
470

471
	vsnprintf(message, sizeof(message), err, params);
472
	fputs("fatal: ", stderr);
473
	fputs(message, stderr);
474 475 476 477
	fputc('\n', stderr);

	if (!zombie) {
		zombie = 1;
478
		write_crash_report(message);
479 480 481
		end_packfile();
		unkeep_all_packs();
		dump_marks();
482 483 484
	}
	exit(128);
}
485

486
static void alloc_objects(unsigned int cnt)
487
{
488
	struct object_entry_pool *b;
489

490
	b = xmalloc(sizeof(struct object_entry_pool)
491
		+ cnt * sizeof(struct object_entry));
492
	b->next_pool = blocks;
493 494 495 496 497
	b->next_free = b->entries;
	b->end = b->entries + cnt;
	blocks = b;
	alloc_count += cnt;
}
498

499
static struct object_entry *new_object(unsigned char *sha1)
500
{
501
	struct object_entry *e;
502

503
	if (blocks->next_free == blocks->end)
504
		alloc_objects(object_entry_alloc);
505

506
	e = blocks->next_free++;
507
	hashcpy(e->sha1, sha1);
508
	return e;
509 510
}

511
static struct object_entry *find_object(unsigned char *sha1)
512 513 514 515
{
	unsigned int h = sha1[0] << 8 | sha1[1];
	struct object_entry *e;
	for (e = object_table[h]; e; e = e->next)
516
		if (!hashcmp(sha1, e->sha1))
517 518 519 520
			return e;
	return NULL;
}

521
static struct object_entry *insert_object(unsigned char *sha1)
522 523
{
	unsigned int h = sha1[0] << 8 | sha1[1];
524
	struct object_entry *e = object_table[h];
525
	struct object_entry *p = NULL;
526 527

	while (e) {
528
		if (!hashcmp(sha1, e->sha1))
529 530 531 532 533 534
			return e;
		p = e;
		e = e->next;
	}

	e = new_object(sha1);
535
	e->next = NULL;
536 537 538 539
	e->offset = 0;
	if (p)
		p->next = e;
	else
540
		object_table[h] = e;
541 542
	return e;
}
543

544 545 546 547 548 549 550 551
static unsigned int hc_str(const char *s, size_t len)
{
	unsigned int r = 0;
	while (len-- > 0)
		r = r * 31 + *s++;
	return r;
}

552
static void *pool_alloc(size_t len)
553 554 555 556
{
	struct mem_pool *p;
	void *r;

557 558 559 560
	/* round up to a 'uintmax_t' alignment */
	if (len & (sizeof(uintmax_t) - 1))
		len += sizeof(uintmax_t) - (len & (sizeof(uintmax_t) - 1));

561 562 563 564 565 566 567 568 569 570 571 572
	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;
573
		p->next_free = (char *) p->space;
574 575 576 577 578 579 580 581 582
		p->end = p->next_free + mem_pool_alloc;
		mem_pool = p;
	}

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

583
static void *pool_calloc(size_t count, size_t size)
584 585 586 587 588 589 590
{
	size_t len = count * size;
	void *r = pool_alloc(len);
	memset(r, 0, len);
	return r;
}

591
static char *pool_strdup(const char *s)
592 593 594 595 596 597
{
	char *r = pool_alloc(strlen(s) + 1);
	strcpy(r, s);
	return r;
}

598
static void insert_mark(uintmax_t idnum, struct object_entry *oe)
599 600 601 602 603 604 605 606 607
{
	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) {
608
		uintmax_t i = idnum >> s->shift;
609 610 611 612 613 614 615 616 617 618 619 620
		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;
}

621
static struct object_entry *find_mark(uintmax_t idnum)
622
{
623
	uintmax_t orig_idnum = idnum;
624 625 626 627
	struct mark_set *s = marks;
	struct object_entry *oe = NULL;
	if ((idnum >> s->shift) < 1024) {
		while (s && s->shift) {
628
			uintmax_t i = idnum >> s->shift;
629 630 631 632 633 634 635
			idnum -= i << s->shift;
			s = s->data.sets[i];
		}
		if (s)
			oe = s->data.marked[idnum];
	}
	if (!oe)
636
		die("mark :%" PRIuMAX " not declared", orig_idnum);
637 638 639
	return oe;
}

640
static struct atom_str *to_atom(const char *s, unsigned short len)
641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658
{
	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;
}

659
static struct branch *lookup_branch(const char *name)
660 661 662 663 664 665 666 667 668 669
{
	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;
}

670
static struct branch *new_branch(const char *name)
671 672 673 674 675 676
{
	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);
677
	switch (check_ref_format(name)) {
678 679 680
	case 0: break; /* its valid */
	case CHECK_REF_FORMAT_ONELEVEL:
		break; /* valid, but too few '/', allow anyway */
681
	default:
682
		die("Branch name doesn't conform to GIT standards: %s", name);
683
	}
684 685 686 687

	b = pool_calloc(1, sizeof(struct branch));
	b->name = pool_strdup(name);
	b->table_next_branch = branch_table[hc];
688 689
	b->branch_tree.versions[0].mode = S_IFDIR;
	b->branch_tree.versions[1].mode = S_IFDIR;
690
	b->active = 0;
691
	b->pack_id = MAX_PACK_ID;
692 693 694 695 696 697 698 699 700 701 702
	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;
}

703
static struct tree_content *new_tree_content(unsigned int cnt)
704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725
{
	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;
726
	t->delta_depth = 0;
727 728 729 730 731 732 733 734
	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);
735 736 737 738 739 740
	f->next_avail = avail_tree_table[hc];
	avail_tree_table[hc] = f;
}

static void release_tree_content_recursive(struct tree_content *t)
{
741 742 743
	unsigned int i;
	for (i = 0; i < t->entry_count; i++)
		release_tree_entry(t->entries[i]);
744
	release_tree_content(t);
745 746
}

747
static struct tree_content *grow_tree_content(
748 749 750 751 752
	struct tree_content *t,
	int amt)
{
	struct tree_content *r = new_tree_content(t->entry_count + amt);
	r->entry_count = t->entry_count;
753
	r->delta_depth = t->delta_depth;
754 755 756 757 758
	memcpy(r->entries,t->entries,t->entry_count*sizeof(t->entries[0]));
	release_tree_content(t);
	return r;
}

759
static struct tree_entry *new_tree_entry(void)
760 761 762 763 764
{
	struct tree_entry *e;

	if (!avail_tree_entry) {
		unsigned int n = tree_entry_alloc;
765
		total_allocd += n * sizeof(struct tree_entry);
766
		avail_tree_entry = e = xmalloc(n * sizeof(struct tree_entry));
767
		while (n-- > 1) {
768 769 770
			*((void**)e) = e + 1;
			e++;
		}
771
		*((void**)e) = NULL;
772 773 774 775 776 777 778 779 780 781
	}

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

static void release_tree_entry(struct tree_entry *e)
{
	if (e->tree)
782
		release_tree_content_recursive(e->tree);
783 784 785 786
	*((void**)e) = avail_tree_entry;
	avail_tree_entry = e;
}

787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811
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;
}

812
static void start_packfile(void)
813
{
814
	static char tmpfile[PATH_MAX];
815
	struct packed_git *p;
816
	struct pack_header hdr;
817
	int pack_fd;
818

819
	snprintf(tmpfile, sizeof(tmpfile),
820
		"%s/pack/tmp_pack_XXXXXX", get_object_directory());
821
	pack_fd = xmkstemp(tmpfile);
822 823
	p = xcalloc(1, sizeof(*p) + strlen(tmpfile) + 2);
	strcpy(p->pack_name, tmpfile);
824
	p->pack_fd = pack_fd;
825 826 827 828

	hdr.hdr_signature = htonl(PACK_SIGNATURE);
	hdr.hdr_version = htonl(2);
	hdr.hdr_entries = 0;
829
	write_or_die(p->pack_fd, &hdr, sizeof(hdr));
830 831

	pack_data = p;
832 833
	pack_size = sizeof(hdr);
	object_count = 0;
834 835 836

	all_packs = xrealloc(all_packs, sizeof(*all_packs) * (pack_id + 1));
	all_packs[pack_id] = p;
837 838 839 840 841 842 843 844 845
}

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);
}

846
static char *create_index(void)
847
{
848
	static char tmpfile[PATH_MAX];
849
	git_SHA_CTX ctx;
850 851 852
	struct sha1file *f;
	struct object_entry **idx, **c, **last, *e;
	struct object_entry_pool *o;
853
	uint32_t array[256];
854
	int i, idx_fd;
855 856 857 858 859

	/* 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)
860 861 862
		for (e = o->next_free; e-- != o->entries;)
			if (pack_id == e->pack_id)
				*c++ = e;
863
	last = idx + object_count;
864 865
	if (c != last)
		die("internal consistency error creating the index");
866 867 868 869 870 871 872 873 874 875 876 877 878 879 880
	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;
	}

881
	snprintf(tmpfile, sizeof(tmpfile),
882
		"%s/pack/tmp_idx_XXXXXX", get_object_directory());
883
	idx_fd = xmkstemp(tmpfile);
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";
904 905 906 907 908 909 910 911 912 913
	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");
914 915 916
	write_or_die(keep_fd, keep_msg, strlen(keep_msg));
	if (close(keep_fd))
		die("failed to write keep file");
917 918 919 920 921 922 923 924 925 926 927 928 929

	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;
}

930
static void unkeep_all_packs(void)
931 932 933 934 935 936 937 938 939 940
{
	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);
	}
941 942
}

943
static void end_packfile(void)
944
{
945 946
	struct packed_git *old_p = pack_data, *new_p;

947
	if (object_count) {
948
		char *idx_name;
949 950 951
		int i;
		struct branch *b;
		struct tag *t;
952

953
		close_pack_windows(pack_data);
954
		fixup_pack_header_footer(pack_data->pack_fd, pack_data->sha1,
955 956
				    pack_data->pack_name, object_count,
				    NULL, 0);
957
		close(pack_data->pack_fd);
958
		idx_name = keep_pack(create_index());
959 960 961 962 963

		/* 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);
964
		all_packs[pack_id] = new_p;
965
		install_packed_git(new_p);
966 967

		/* Print the boundary */
968 969 970 971 972 973 974
		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));
				}
975
			}
976 977 978 979 980 981
			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);
982 983 984
		}

		pack_id++;
985
	}
986 987
	else {
		close(old_p->pack_fd);
988
		unlink(old_p->pack_name);
989
	}
990 991 992
	free(old_p);

	/* We can't carry a delta across packfiles. */
P
Pierre Habouzit 已提交
993
	strbuf_release(&last_blob.data);
994 995
	last_blob.offset = 0;
	last_blob.depth = 0;
996 997
}

998
static void cycle_packfile(void)
999 1000 1001 1002 1003
{
	end_packfile();
	start_packfile();
}

1004
static size_t encode_header(
1005
	enum object_type type,
1006
	size_t size,
1007
	unsigned char *hdr)
1008 1009 1010 1011
{
	int n = 1;
	unsigned char c;

1012
	if (type < OBJ_COMMIT || type > OBJ_REF_DELTA)
1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026
		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;
}

1027 1028
static int store_object(
	enum object_type type,
1029
	struct strbuf *dat,
1030
	struct last_object *last,
1031
	unsigned char *sha1out,
1032
	uintmax_t mark)
1033 1034
{
	void *out, *delta;
1035 1036 1037
	struct object_entry *e;
	unsigned char hdr[96];
	unsigned char sha1[20];
1038
	unsigned long hdrlen, deltalen;
1039
	git_SHA_CTX c;
1040 1041
	z_stream s;

1042
	hdrlen = sprintf((char*)hdr,"%s %lu", typename(type),
1043
		(unsigned long)dat->len) + 1;
1044 1045 1046 1047
	git_SHA1_Init(&c);
	git_SHA1_Update(&c, hdr, hdrlen);
	git_SHA1_Update(&c, dat->buf, dat->len);
	git_SHA1_Final(sha1, &c);
1048
	if (sha1out)
1049
		hashcpy(sha1out, sha1);
1050 1051

	e = insert_object(sha1);
1052 1053
	if (mark)
		insert_mark(mark, e);
1054
	if (e->offset) {
1055
		duplicate_count_by_type[type]++;
1056
		return 1;
1057 1058 1059 1060 1061 1062
	} 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;
1063
	}
1064

P
Pierre Habouzit 已提交
1065 1066
	if (last && last->data.buf && last->depth < max_depth) {
		delta = diff_delta(last->data.buf, last->data.len,
1067
			dat->buf, dat->len,
1068
			&deltalen, 0);
1069
		if (delta && deltalen >= dat->len) {
1070 1071 1072 1073 1074
			free(delta);
			delta = NULL;
		}
	} else
		delta = NULL;
1075 1076

	memset(&s, 0, sizeof(s));
1077
	deflateInit(&s, pack_compression_level);
1078 1079 1080 1081
	if (delta) {
		s.next_in = delta;
		s.avail_in = deltalen;
	} else {
1082 1083
		s.next_in = (void *)dat->buf;
		s.avail_in = dat->len;
1084 1085 1086 1087 1088 1089 1090 1091
	}
	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. */
1092
	if ((pack_size + 60 + s.total_out) > max_packsize
1093 1094 1095 1096
		|| (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;
1097
		cycle_packfile();
1098 1099 1100 1101 1102

		/* We cannot carry a delta into the new pack. */
		if (delta) {
			free(delta);
			delta = NULL;
1103 1104

			memset(&s, 0, sizeof(s));
1105
			deflateInit(&s, pack_compression_level);
1106 1107
			s.next_in = (void *)dat->buf;
			s.avail_in = dat->len;
1108 1109 1110 1111 1112
			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);
1113 1114 1115 1116 1117 1118 1119 1120
		}
	}

	e->type = type;
	e->pack_id = pack_id;
	e->offset = pack_size;
	object_count++;
	object_count_by_type[type]++;
1121 1122

	if (delta) {
1123 1124 1125
		unsigned long ofs = e->offset - last->offset;
		unsigned pos = sizeof(hdr) - 1;

1126
		delta_count_by_type[type]++;
1127
		e->depth = last->depth + 1;
1128 1129

		hdrlen = encode_header(OBJ_OFS_DELTA, deltalen, hdr);
1130
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1131 1132 1133 1134 1135
		pack_size += hdrlen;

		hdr[pos] = ofs & 127;
		while (ofs >>= 7)
			hdr[--pos] = 128 | (--ofs & 127);
1136
		write_or_die(pack_data->pack_fd, hdr + pos, sizeof(hdr) - pos);
1137
		pack_size += sizeof(hdr) - pos;
1138
	} else {
1139
		e->depth = 0;
1140
		hdrlen = encode_header(type, dat->len, hdr);
1141
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1142
		pack_size += hdrlen;
1143 1144
	}

1145
	write_or_die(pack_data->pack_fd, out, s.total_out);
1146
	pack_size += s.total_out;
1147 1148

	free(out);
1149
	free(delta);
1150
	if (last) {
P
Pierre Habouzit 已提交
1151 1152 1153
		if (last->no_swap) {
			last->data = *dat;
		} else {
1154
			strbuf_swap(&last->data, dat);
P
Pierre Habouzit 已提交
1155
		}
1156
		last->offset = e->offset;
1157
		last->depth = e->depth;
1158 1159 1160 1161
	}
	return 0;
}

1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179
/* 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.
 */
1180 1181 1182
static void *gfi_unpack_entry(
	struct object_entry *oe,
	unsigned long *sizep)
1183
{
1184
	enum object_type type;
1185
	struct packed_git *p = all_packs[oe->pack_id];
1186
	if (p == pack_data && p->pack_size < (pack_size + 20)) {
1187 1188 1189 1190 1191 1192 1193 1194
		/* 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.
		 */
1195
		close_pack_windows(p);
1196 1197 1198 1199 1200 1201 1202

		/* 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.
		 */
1203
		p->pack_size = pack_size + 20;
1204
	}
1205
	return unpack_entry(p, oe->offset, &type, sizep);
1206 1207
}

1208
static const char *get_mode(const char *str, uint16_t *modep)
1209 1210
{
	unsigned char c;
1211
	uint16_t mode = 0;
1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223

	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)
{
1224
	unsigned char* sha1 = root->versions[1].sha1;
1225 1226 1227 1228 1229 1230 1231
	struct object_entry *myoe;
	struct tree_content *t;
	unsigned long size;
	char *buf;
	const char *c;

	root->tree = t = new_tree_content(8);
1232
	if (is_null_sha1(sha1))
1233 1234
		return;

1235
	myoe = find_object(sha1);
1236
	if (myoe && myoe->pack_id != MAX_PACK_ID) {
1237
		if (myoe->type != OBJ_TREE)
1238
			die("Not a tree: %s", sha1_to_hex(sha1));
1239
		t->delta_depth = myoe->depth;
1240
		buf = gfi_unpack_entry(myoe, &size);
1241 1242
		if (!buf)
			die("Can't load tree %s", sha1_to_hex(sha1));
1243
	} else {
1244 1245 1246
		enum object_type type;
		buf = read_sha1_file(sha1, &type, &size);
		if (!buf || type != OBJ_TREE)
1247
			die("Can't load tree %s", sha1_to_hex(sha1));
1248 1249 1250 1251 1252 1253 1254
	}

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

		if (t->entry_count == t->entry_capacity)
1255
			root->tree = t = grow_tree_content(t, t->entry_count);
1256 1257 1258
		t->entries[t->entry_count++] = e;

		e->tree = NULL;
1259
		c = get_mode(c, &e->versions[1].mode);
1260
		if (!c)
1261 1262
			die("Corrupt mode in %s", sha1_to_hex(sha1));
		e->versions[0].mode = e->versions[1].mode;
1263
		e->name = to_atom(c, strlen(c));
1264
		c += e->name->str_len + 1;
1265 1266
		hashcpy(e->versions[0].sha1, (unsigned char*)c);
		hashcpy(e->versions[1].sha1, (unsigned char*)c);
1267 1268 1269 1270 1271
		c += 20;
	}
	free(buf);
}

1272
static int tecmp0 (const void *_a, const void *_b)
1273 1274 1275 1276
{
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
1277 1278
		a->name->str_dat, a->name->str_len, a->versions[0].mode,
		b->name->str_dat, b->name->str_len, b->versions[0].mode);
1279 1280
}

1281
static int tecmp1 (const void *_a, const void *_b)
1282
{
1283 1284 1285 1286 1287 1288 1289
	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);
}

1290
static void mktree(struct tree_content *t, int v, struct strbuf *b)
1291 1292
{
	size_t maxlen = 0;
1293 1294
	unsigned int i;

1295 1296 1297 1298
	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);
1299 1300

	for (i = 0; i < t->entry_count; i++) {
1301 1302
		if (t->entries[i]->versions[v].mode)
			maxlen += t->entries[i]->name->str_len + 34;
1303 1304
	}

1305 1306
	strbuf_reset(b);
	strbuf_grow(b, maxlen);
1307 1308
	for (i = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
1309 1310
		if (!e->versions[v].mode)
			continue;
1311 1312 1313
		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);
1314
	}
1315 1316 1317 1318 1319 1320
}

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
1321
	struct last_object lo = { STRBUF_INIT, 0, 0, /* no_swap */ 1 };
1322
	struct object_entry *le;
1323 1324 1325 1326 1327 1328 1329 1330 1331

	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]);
	}

1332
	le = find_object(root->versions[0].sha1);
P
Pierre Habouzit 已提交
1333
	if (S_ISDIR(root->versions[0].mode) && le && le->pack_id == pack_id) {
1334
		mktree(t, 0, &old_tree);
P
Pierre Habouzit 已提交
1335
		lo.data = old_tree;
1336
		lo.offset = le->offset;
1337 1338 1339
		lo.depth = t->delta_depth;
	}

1340 1341
	mktree(t, 1, &new_tree);
	store_object(OBJ_TREE, &new_tree, &lo, root->versions[1].sha1, 0);
1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355

	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;
1356 1357 1358 1359 1360 1361
}

static int tree_content_set(
	struct tree_entry *root,
	const char *p,
	const unsigned char *sha1,
1362 1363
	const uint16_t mode,
	struct tree_content *subtree)
1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374
{
	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);
1375 1376
	if (!n)
		die("Empty path component found in input");
1377 1378
	if (!slash1 && !S_ISDIR(mode) && subtree)
		die("Non-directories cannot have subtrees");
1379 1380 1381 1382 1383

	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) {
1384 1385
				if (!S_ISDIR(mode)
						&& e->versions[1].mode == mode
1386
						&& !hashcmp(e->versions[1].sha1, sha1))
1387
					return 0;
1388 1389
				e->versions[1].mode = mode;
				hashcpy(e->versions[1].sha1, sha1);
1390
				if (e->tree)
1391
					release_tree_content_recursive(e->tree);
1392
				e->tree = subtree;
1393
				hashclr(root->versions[1].sha1);
1394 1395
				return 1;
			}
1396
			if (!S_ISDIR(e->versions[1].mode)) {
1397
				e->tree = new_tree_content(8);
1398
				e->versions[1].mode = S_IFDIR;
1399 1400 1401
			}
			if (!e->tree)
				load_tree(e);
1402
			if (tree_content_set(e, slash1 + 1, sha1, mode, subtree)) {
1403
				hashclr(root->versions[1].sha1);
1404 1405 1406 1407 1408 1409 1410
				return 1;
			}
			return 0;
		}
	}

	if (t->entry_count == t->entry_capacity)
1411
		root->tree = t = grow_tree_content(t, t->entry_count);
1412
	e = new_tree_entry();
1413
	e->name = to_atom(p, n);
1414 1415
	e->versions[0].mode = 0;
	hashclr(e->versions[0].sha1);
1416 1417 1418
	t->entries[t->entry_count++] = e;
	if (slash1) {
		e->tree = new_tree_content(8);
1419
		e->versions[1].mode = S_IFDIR;
1420
		tree_content_set(e, slash1 + 1, sha1, mode, subtree);
1421
	} else {
1422
		e->tree = subtree;
1423 1424
		e->versions[1].mode = mode;
		hashcpy(e->versions[1].sha1, sha1);
1425
	}
1426
	hashclr(root->versions[1].sha1);
1427 1428 1429
	return 1;
}

1430 1431 1432 1433
static int tree_content_remove(
	struct tree_entry *root,
	const char *p,
	struct tree_entry *backup_leaf)
1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448
{
	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)) {
1449
			if (!slash1 || !S_ISDIR(e->versions[1].mode))
1450 1451 1452
				goto del_entry;
			if (!e->tree)
				load_tree(e);
1453
			if (tree_content_remove(e, slash1 + 1, backup_leaf)) {
1454 1455 1456 1457 1458 1459
				for (n = 0; n < e->tree->entry_count; n++) {
					if (e->tree->entries[n]->versions[1].mode) {
						hashclr(root->versions[1].sha1);
						return 1;
					}
				}
1460
				backup_leaf = NULL;
1461
				goto del_entry;
1462 1463 1464 1465 1466 1467 1468
			}
			return 0;
		}
	}
	return 0;

del_entry:
1469 1470 1471
	if (backup_leaf)
		memcpy(backup_leaf, e, sizeof(*backup_leaf));
	else if (e->tree)
1472
		release_tree_content_recursive(e->tree);
1473
	e->tree = NULL;
1474 1475 1476
	e->versions[1].mode = 0;
	hashclr(e->versions[1].sha1);
	hashclr(root->versions[1].sha1);
1477
	return 1;
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 1513 1514 1515 1516
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;
}

1517
static int update_branch(struct branch *b)
1518 1519
{
	static const char *msg = "fast-import";
1520 1521 1522
	struct ref_lock *lock;
	unsigned char old_sha1[20];

1523 1524
	if (is_null_sha1(b->sha1))
		return 0;
1525 1526
	if (read_ref(b->name, old_sha1))
		hashclr(old_sha1);
1527
	lock = lock_any_ref_for_update(b->name, old_sha1, 0);
1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539
	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);
		}

1540
		if (!in_merge_bases(old_cmit, &new_cmit, 1)) {
1541
			unlock_ref(lock);
1542
			warning("Not updating %s"
1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554
				" (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)
{
1555 1556 1557 1558
	unsigned int i;
	struct branch *b;

	for (i = 0; i < branch_table_sz; i++) {
1559 1560
		for (b = branch_table[i]; b; b = b->table_next_branch)
			failure |= update_branch(b);
1561 1562 1563
	}
}

1564
static void dump_tags(void)
1565 1566 1567 1568
{
	static const char *msg = "fast-import";
	struct tag *t;
	struct ref_lock *lock;
1569
	char ref_name[PATH_MAX];
1570 1571

	for (t = first_tag; t; t = t->next_tag) {
1572 1573
		sprintf(ref_name, "tags/%s", t->name);
		lock = lock_ref_sha1(ref_name, NULL);
1574
		if (!lock || write_ref_sha1(lock, t->sha1, msg) < 0)
1575
			failure |= error("Unable to update %s", ref_name);
1576 1577 1578
	}
}

1579
static void dump_marks_helper(FILE *f,
1580
	uintmax_t base,
1581 1582
	struct mark_set *m)
{
1583
	uintmax_t k;
1584 1585 1586 1587 1588 1589 1590 1591 1592
	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])
1593
				fprintf(f, ":%" PRIuMAX " %s\n", base + k,
1594 1595 1596 1597 1598
					sha1_to_hex(m->data.marked[k]->sha1));
		}
	}
}

1599
static void dump_marks(void)
1600
{
1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612
	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;
1613
	}
1614 1615 1616

	f = fdopen(mark_fd, "w");
	if (!f) {
1617
		int saved_errno = errno;
1618 1619
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1620
			mark_file, strerror(saved_errno));
1621
		return;
1622
	}
1623

B
Brandon Casey 已提交
1624
	/*
1625 1626
	 * 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
1627 1628 1629
	 * won't try to close() it.
	 */
	mark_lock.fd = -1;
1630 1631 1632

	dump_marks_helper(f, 0, marks);
	if (ferror(f) || fclose(f)) {
1633
		int saved_errno = errno;
1634 1635
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1636
			mark_file, strerror(saved_errno));
1637 1638 1639 1640
		return;
	}

	if (commit_lock_file(&mark_lock)) {
1641
		int saved_errno = errno;
1642 1643
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to commit marks file %s: %s",
1644
			mark_file, strerror(saved_errno));
1645 1646
		return;
	}
1647 1648
}

1649
static int read_next_command(void)
1650
{
1651 1652 1653 1654 1655 1656 1657
	static int stdin_eof = 0;

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

1658
	do {
1659
		if (unread_command_buf) {
1660
			unread_command_buf = 0;
1661 1662 1663
		} else {
			struct recent_command *rc;

1664
			strbuf_detach(&command_buf, NULL);
1665 1666 1667
			stdin_eof = strbuf_getline(&command_buf, stdin, '\n');
			if (stdin_eof)
				return EOF;
1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685

			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] == '#');
1686 1687

	return 0;
1688 1689
}

J
Junio C Hamano 已提交
1690
static void skip_optional_lf(void)
1691 1692 1693 1694 1695 1696
{
	int term_char = fgetc(stdin);
	if (term_char != '\n' && term_char != EOF)
		ungetc(term_char, stdin);
}

1697
static void parse_mark(void)
1698
{
1699
	if (!prefixcmp(command_buf.buf, "mark :")) {
1700
		next_mark = strtoumax(command_buf.buf + 6, NULL, 10);
1701 1702 1703
		read_next_command();
	}
	else
1704
		next_mark = 0;
1705 1706
}

1707
static void parse_data(struct strbuf *sb)
1708
{
1709
	strbuf_reset(sb);
1710

1711
	if (prefixcmp(command_buf.buf, "data "))
1712 1713
		die("Expected 'data n' command, found: %s", command_buf.buf);

1714
	if (!prefixcmp(command_buf.buf + 5, "<<")) {
1715
		char *term = xstrdup(command_buf.buf + 5 + 2);
1716 1717
		size_t term_len = command_buf.len - 5 - 2;

1718
		strbuf_detach(&command_buf, NULL);
1719
		for (;;) {
1720
			if (strbuf_getline(&command_buf, stdin, '\n') == EOF)
1721 1722 1723 1724
				die("EOF in data (terminator '%s' not found)", term);
			if (term_len == command_buf.len
				&& !strcmp(term, command_buf.buf))
				break;
1725 1726
			strbuf_addbuf(sb, &command_buf);
			strbuf_addch(sb, '\n');
1727 1728 1729 1730
		}
		free(term);
	}
	else {
1731 1732
		size_t n = 0, length;

1733
		length = strtoul(command_buf.buf + 5, NULL, 10);
1734

1735
		while (n < length) {
1736
			size_t s = strbuf_fread(sb, length - n, stdin);
1737
			if (!s && feof(stdin))
1738 1739
				die("EOF in data (%lu bytes remaining)",
					(unsigned long)(length - n));
1740 1741
			n += s;
		}
1742 1743
	}

1744
	skip_optional_lf();
1745 1746
}

1747 1748 1749 1750
static int validate_raw_date(const char *src, char *result, int maxlen)
{
	const char *orig_src = src;
	char *endp, sign;
1751
	unsigned long date;
1752

1753 1754 1755 1756
	errno = 0;

	date = strtoul(src, &endp, 10);
	if (errno || endp == src || *endp != ' ')
1757 1758 1759 1760 1761 1762 1763
		return -1;

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

1764 1765
	date = strtoul(src + 1, &endp, 10);
	if (errno || endp == src || *endp || (endp - orig_src) >= maxlen)
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 1802 1803 1804 1805 1806 1807
		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;
}

1808
static void parse_new_blob(void)
1809
{
P
Pierre Habouzit 已提交
1810
	static struct strbuf buf = STRBUF_INIT;
1811 1812

	read_next_command();
1813 1814
	parse_mark();
	parse_data(&buf);
P
Pierre Habouzit 已提交
1815
	store_object(OBJ_BLOB, &buf, &last_blob, NULL, next_mark);
1816 1817
}

1818
static void unload_one_branch(void)
1819
{
1820 1821
	while (cur_active_branches
		&& cur_active_branches >= max_active_branches) {
1822
		uintmax_t min_commit = ULONG_MAX;
1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839
		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;
		}
1840
		e->active = 0;
1841 1842
		e->active_next_branch = NULL;
		if (e->branch_tree.tree) {
1843
			release_tree_content_recursive(e->branch_tree.tree);
1844 1845 1846
			e->branch_tree.tree = NULL;
		}
		cur_active_branches--;
1847 1848 1849
	}
}

1850
static void load_branch(struct branch *b)
1851
{
1852
	load_tree(&b->branch_tree);
1853 1854 1855 1856 1857 1858 1859
	if (!b->active) {
		b->active = 1;
		b->active_next_branch = active_branches;
		active_branches = b;
		cur_active_branches++;
		branch_load_count++;
	}
1860 1861
}

1862
static void file_change_m(struct branch *b)
1863
{
1864
	const char *p = command_buf.buf + 2;
1865
	static struct strbuf uq = STRBUF_INIT;
1866
	const char *endp;
1867
	struct object_entry *oe = oe;
1868
	unsigned char sha1[20];
1869
	uint16_t mode, inline_data = 0;
1870

1871 1872 1873 1874
	p = get_mode(p, &mode);
	if (!p)
		die("Corrupt mode: %s", command_buf.buf);
	switch (mode) {
1875 1876 1877
	case 0644:
	case 0755:
		mode |= S_IFREG;
1878 1879
	case S_IFREG | 0644:
	case S_IFREG | 0755:
1880
	case S_IFLNK:
1881
	case S_IFGITLINK:
1882 1883 1884 1885 1886 1887
		/* ok */
		break;
	default:
		die("Corrupt mode: %s", command_buf.buf);
	}

1888 1889
	if (*p == ':') {
		char *x;
1890
		oe = find_mark(strtoumax(p + 1, &x, 10));
1891
		hashcpy(sha1, oe->sha1);
1892
		p = x;
1893
	} else if (!prefixcmp(p, "inline")) {
1894 1895
		inline_data = 1;
		p += 6;
1896 1897 1898 1899 1900 1901
	} else {
		if (get_sha1_hex(p, sha1))
			die("Invalid SHA1: %s", command_buf.buf);
		oe = find_object(sha1);
		p += 40;
	}
1902 1903 1904
	if (*p++ != ' ')
		die("Missing space after SHA1: %s", command_buf.buf);

1905 1906
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1907 1908
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1909
		p = uq.buf;
1910
	}
1911

1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925
	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
1926
		static struct strbuf buf = STRBUF_INIT;
1927

1928 1929 1930 1931
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
1932
		read_next_command();
1933
		parse_data(&buf);
P
Pierre Habouzit 已提交
1934
		store_object(OBJ_BLOB, &buf, &last_blob, sha1, 0);
1935
	} else if (oe) {
1936
		if (oe->type != OBJ_BLOB)
1937
			die("Not a blob (actually a %s): %s",
1938
				typename(oe->type), command_buf.buf);
1939
	} else {
1940 1941
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
1942
			die("Blob not found: %s", command_buf.buf);
1943
		if (type != OBJ_BLOB)
1944
			die("Not a blob (actually a %s): %s",
1945
			    typename(type), command_buf.buf);
1946
	}
1947

1948
	tree_content_set(&b->branch_tree, p, sha1, mode, NULL);
1949
}
1950

1951 1952
static void file_change_d(struct branch *b)
{
1953
	const char *p = command_buf.buf + 2;
1954
	static struct strbuf uq = STRBUF_INIT;
1955 1956
	const char *endp;

1957 1958
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1959 1960
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1961
		p = uq.buf;
1962
	}
1963
	tree_content_remove(&b->branch_tree, p, NULL);
1964 1965
}

1966
static void file_change_cr(struct branch *b, int rename)
1967 1968
{
	const char *s, *d;
1969 1970
	static struct strbuf s_uq = STRBUF_INIT;
	static struct strbuf d_uq = STRBUF_INIT;
1971 1972 1973 1974
	const char *endp;
	struct tree_entry leaf;

	s = command_buf.buf + 2;
1975 1976
	strbuf_reset(&s_uq);
	if (!unquote_c_style(&s_uq, s, &endp)) {
1977 1978
		if (*endp != ' ')
			die("Missing space after source: %s", command_buf.buf);
1979
	} else {
1980 1981 1982
		endp = strchr(s, ' ');
		if (!endp)
			die("Missing space after source: %s", command_buf.buf);
1983
		strbuf_add(&s_uq, s, endp - s);
1984
	}
1985
	s = s_uq.buf;
1986 1987 1988 1989 1990 1991

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

	d = endp;
1992 1993
	strbuf_reset(&d_uq);
	if (!unquote_c_style(&d_uq, d, &endp)) {
1994 1995
		if (*endp)
			die("Garbage after dest in: %s", command_buf.buf);
1996
		d = d_uq.buf;
1997 1998 1999
	}

	memset(&leaf, 0, sizeof(leaf));
2000 2001 2002 2003
	if (rename)
		tree_content_remove(&b->branch_tree, s, &leaf);
	else
		tree_content_get(&b->branch_tree, s, &leaf);
2004 2005 2006 2007 2008 2009 2010 2011
	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);
}

2012 2013 2014 2015 2016 2017 2018 2019
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);
}

2020
static void parse_from_commit(struct branch *b, char *buf, unsigned long size)
2021 2022 2023 2024 2025 2026 2027 2028 2029 2030
{
	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);
}

2031
static void parse_from_existing(struct branch *b)
2032 2033 2034 2035 2036 2037 2038 2039 2040 2041
{
	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);
2042
		parse_from_commit(b, buf, size);
2043 2044 2045 2046
		free(buf);
	}
}

2047
static int parse_from(struct branch *b)
2048
{
2049
	const char *from;
2050 2051
	struct branch *s;

2052
	if (prefixcmp(command_buf.buf, "from "))
2053
		return 0;
2054

2055 2056 2057 2058
	if (b->branch_tree.tree) {
		release_tree_content_recursive(b->branch_tree.tree);
		b->branch_tree.tree = NULL;
	}
2059 2060 2061 2062 2063 2064

	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) {
2065
		unsigned char *t = s->branch_tree.versions[1].sha1;
2066
		hashcpy(b->sha1, s->sha1);
2067 2068
		hashcpy(b->branch_tree.versions[0].sha1, t);
		hashcpy(b->branch_tree.versions[1].sha1, t);
2069
	} else if (*from == ':') {
2070
		uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2071 2072
		struct object_entry *oe = find_mark(idnum);
		if (oe->type != OBJ_COMMIT)
2073
			die("Mark :%" PRIuMAX " not a commit", idnum);
2074
		hashcpy(b->sha1, oe->sha1);
2075
		if (oe->pack_id != MAX_PACK_ID) {
2076
			unsigned long size;
2077
			char *buf = gfi_unpack_entry(oe, &size);
2078
			parse_from_commit(b, buf, size);
2079
			free(buf);
2080
		} else
2081
			parse_from_existing(b);
2082
	} else if (!get_sha1(from, b->sha1))
2083
		parse_from_existing(b);
2084
	else
2085 2086 2087
		die("Invalid ref name or SHA1 expression: %s", from);

	read_next_command();
2088
	return 1;
2089 2090
}

2091
static struct hash_list *parse_merge(unsigned int *count)
2092
{
2093
	struct hash_list *list = NULL, *n, *e = e;
2094
	const char *from;
2095 2096 2097
	struct branch *s;

	*count = 0;
2098
	while (!prefixcmp(command_buf.buf, "merge ")) {
2099 2100 2101 2102 2103 2104
		from = strchr(command_buf.buf, ' ') + 1;
		n = xmalloc(sizeof(*n));
		s = lookup_branch(from);
		if (s)
			hashcpy(n->sha1, s->sha1);
		else if (*from == ':') {
2105
			uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2106 2107
			struct object_entry *oe = find_mark(idnum);
			if (oe->type != OBJ_COMMIT)
2108
				die("Mark :%" PRIuMAX " not a commit", idnum);
2109
			hashcpy(n->sha1, oe->sha1);
2110 2111 2112
		} 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
2113
				commit_type, &size, n->sha1);
2114 2115 2116 2117
			if (!buf || size < 46)
				die("Not a valid commit: %s", from);
			free(buf);
		} else
2118 2119 2120 2121 2122 2123 2124 2125
			die("Invalid ref name or SHA1 expression: %s", from);

		n->next = NULL;
		if (list)
			e->next = n;
		else
			list = n;
		e = n;
2126
		(*count)++;
2127 2128 2129 2130 2131
		read_next_command();
	}
	return list;
}

2132
static void parse_new_commit(void)
2133
{
2134
	static struct strbuf msg = STRBUF_INIT;
2135 2136 2137 2138
	struct branch *b;
	char *sp;
	char *author = NULL;
	char *committer = NULL;
2139 2140
	struct hash_list *merge_list = NULL;
	unsigned int merge_count;
2141 2142 2143 2144

	/* Obtain the branch name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	b = lookup_branch(sp);
2145
	if (!b)
2146
		b = new_branch(sp);
2147 2148

	read_next_command();
2149
	parse_mark();
2150
	if (!prefixcmp(command_buf.buf, "author ")) {
2151
		author = parse_ident(command_buf.buf + 7);
2152 2153
		read_next_command();
	}
2154
	if (!prefixcmp(command_buf.buf, "committer ")) {
2155
		committer = parse_ident(command_buf.buf + 10);
2156 2157 2158 2159
		read_next_command();
	}
	if (!committer)
		die("Expected committer but didn't get one");
2160
	parse_data(&msg);
2161
	read_next_command();
2162 2163
	parse_from(b);
	merge_list = parse_merge(&merge_count);
2164 2165

	/* ensure the branch is active/loaded */
2166
	if (!b->branch_tree.tree || !max_active_branches) {
2167 2168 2169
		unload_one_branch();
		load_branch(b);
	}
2170

2171
	/* file_change* */
2172
	while (command_buf.len > 0) {
2173
		if (!prefixcmp(command_buf.buf, "M "))
2174
			file_change_m(b);
2175
		else if (!prefixcmp(command_buf.buf, "D "))
2176
			file_change_d(b);
2177
		else if (!prefixcmp(command_buf.buf, "R "))
2178 2179 2180
			file_change_cr(b, 1);
		else if (!prefixcmp(command_buf.buf, "C "))
			file_change_cr(b, 0);
2181 2182
		else if (!strcmp("deleteall", command_buf.buf))
			file_change_deleteall(b);
2183 2184 2185 2186
		else {
			unread_command_buf = 1;
			break;
		}
2187 2188
		if (read_next_command() == EOF)
			break;
2189 2190
	}

2191
	/* build the tree and the commit */
2192
	store_tree(&b->branch_tree);
2193 2194
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
2195 2196 2197

	strbuf_reset(&new_data);
	strbuf_addf(&new_data, "tree %s\n",
2198
		sha1_to_hex(b->branch_tree.versions[1].sha1));
2199
	if (!is_null_sha1(b->sha1))
2200
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(b->sha1));
2201 2202
	while (merge_list) {
		struct hash_list *next = merge_list->next;
2203
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(merge_list->sha1));
2204 2205 2206
		free(merge_list);
		merge_list = next;
	}
2207 2208 2209 2210 2211 2212
	strbuf_addf(&new_data,
		"author %s\n"
		"committer %s\n"
		"\n",
		author ? author : committer, committer);
	strbuf_addbuf(&new_data, &msg);
2213
	free(author);
2214 2215
	free(committer);

2216
	if (!store_object(OBJ_COMMIT, &new_data, NULL, b->sha1, next_mark))
2217
		b->pack_id = pack_id;
2218
	b->last_commit = object_count_by_type[OBJ_COMMIT];
2219 2220
}

2221
static void parse_new_tag(void)
2222
{
2223
	static struct strbuf msg = STRBUF_INIT;
2224 2225 2226 2227 2228
	char *sp;
	const char *from;
	char *tagger;
	struct branch *s;
	struct tag *t;
2229
	uintmax_t from_mark = 0;
2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244
	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 ... */
2245
	if (prefixcmp(command_buf.buf, "from "))
2246 2247 2248 2249
		die("Expected from command, got %s", command_buf.buf);
	from = strchr(command_buf.buf, ' ') + 1;
	s = lookup_branch(from);
	if (s) {
2250
		hashcpy(sha1, s->sha1);
2251
	} else if (*from == ':') {
2252
		struct object_entry *oe;
2253
		from_mark = strtoumax(from + 1, NULL, 10);
2254
		oe = find_mark(from_mark);
2255
		if (oe->type != OBJ_COMMIT)
2256
			die("Mark :%" PRIuMAX " not a commit", from_mark);
2257
		hashcpy(sha1, oe->sha1);
2258 2259 2260 2261 2262
	} else if (!get_sha1(from, sha1)) {
		unsigned long size;
		char *buf;

		buf = read_object_with_reference(sha1,
2263
			commit_type, &size, sha1);
2264 2265 2266 2267 2268 2269 2270 2271
		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 ... */
2272 2273 2274 2275 2276
	if (!prefixcmp(command_buf.buf, "tagger ")) {
		tagger = parse_ident(command_buf.buf + 7);
		read_next_command();
	} else
		tagger = NULL;
2277 2278

	/* tag payload/message */
2279
	parse_data(&msg);
2280 2281

	/* build the tag object */
2282
	strbuf_reset(&new_data);
2283

2284
	strbuf_addf(&new_data,
2285 2286 2287 2288 2289 2290 2291 2292
		    "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');
2293
	strbuf_addbuf(&new_data, &msg);
2294 2295
	free(tagger);

2296
	if (store_object(OBJ_TAG, &new_data, NULL, t->sha1, 0))
2297 2298 2299
		t->pack_id = MAX_PACK_ID;
	else
		t->pack_id = pack_id;
2300 2301
}

2302
static void parse_reset_branch(void)
2303 2304 2305 2306 2307 2308 2309 2310
{
	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) {
2311 2312 2313
		hashclr(b->sha1);
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
2314 2315 2316 2317 2318
		if (b->branch_tree.tree) {
			release_tree_content_recursive(b->branch_tree.tree);
			b->branch_tree.tree = NULL;
		}
	}
2319 2320 2321
	else
		b = new_branch(sp);
	read_next_command();
2322
	parse_from(b);
2323
	if (command_buf.len > 0)
2324
		unread_command_buf = 1;
2325 2326
}

2327
static void parse_checkpoint(void)
2328
{
2329 2330 2331 2332 2333 2334
	if (object_count) {
		cycle_packfile();
		dump_branches();
		dump_tags();
		dump_marks();
	}
2335
	skip_optional_lf();
2336 2337
}

2338
static void parse_progress(void)
2339
{
2340
	fwrite(command_buf.buf, 1, command_buf.len, stdout);
2341 2342 2343 2344 2345
	fputc('\n', stdout);
	fflush(stdout);
	skip_optional_lf();
}

2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373
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;
2374
			e->offset = 1; /* just not zero! */
2375 2376 2377 2378 2379 2380
		}
		insert_mark(mark, e);
	}
	fclose(f);
}

2381
static int git_pack_config(const char *k, const char *v, void *cb)
2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398
{
	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;
	}
2399
	return git_default_config(k, v, cb);
2400 2401
}

2402
static const char fast_import_usage[] =
S
Stephan Beyer 已提交
2403
"git fast-import [--date-format=f] [--max-pack-size=n] [--depth=n] [--active-branches=n] [--export-marks=marks.file]";
2404

2405 2406
int main(int argc, const char **argv)
{
2407
	unsigned int i, show_stats = 1;
2408

2409
	setup_git_directory();
2410
	git_config(git_pack_config, NULL);
2411 2412 2413
	if (!pack_compression_seen && core_compression_seen)
		pack_compression_level = core_compression_level;

2414
	alloc_objects(object_entry_alloc);
2415
	strbuf_init(&command_buf, 0);
2416 2417 2418 2419
	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));
2420

2421 2422 2423 2424 2425
	for (i = 1; i < argc; i++) {
		const char *a = argv[i];

		if (*a != '-' || !strcmp(a, "--"))
			break;
2426
		else if (!prefixcmp(a, "--date-format=")) {
2427 2428 2429 2430 2431 2432 2433 2434 2435 2436
			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);
		}
2437
		else if (!prefixcmp(a, "--max-pack-size="))
2438
			max_packsize = strtoumax(a + 16, NULL, 0) * 1024 * 1024;
2439
		else if (!prefixcmp(a, "--depth=")) {
2440
			max_depth = strtoul(a + 8, NULL, 0);
2441 2442 2443
			if (max_depth > MAX_DEPTH)
				die("--depth cannot exceed %u", MAX_DEPTH);
		}
2444
		else if (!prefixcmp(a, "--active-branches="))
2445
			max_active_branches = strtoul(a + 18, NULL, 0);
2446 2447
		else if (!prefixcmp(a, "--import-marks="))
			import_marks(a + 15);
2448
		else if (!prefixcmp(a, "--export-marks="))
2449
			mark_file = a + 15;
2450
		else if (!prefixcmp(a, "--export-pack-edges=")) {
2451 2452 2453 2454 2455 2456
			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"))
2457
			force_update = 1;
2458 2459 2460 2461
		else if (!strcmp(a, "--quiet"))
			show_stats = 0;
		else if (!strcmp(a, "--stats"))
			show_stats = 1;
2462 2463 2464
		else
			die("unknown option %s", a);
	}
2465
	if (i != argc)
2466 2467
		usage(fast_import_usage);

2468 2469 2470 2471 2472
	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;

2473
	prepare_packed_git();
2474
	start_packfile();
2475
	set_die_routine(die_nicely);
2476 2477
	while (read_next_command() != EOF) {
		if (!strcmp("blob", command_buf.buf))
2478
			parse_new_blob();
2479
		else if (!prefixcmp(command_buf.buf, "commit "))
2480
			parse_new_commit();
2481
		else if (!prefixcmp(command_buf.buf, "tag "))
2482
			parse_new_tag();
2483
		else if (!prefixcmp(command_buf.buf, "reset "))
2484
			parse_reset_branch();
2485
		else if (!strcmp("checkpoint", command_buf.buf))
2486
			parse_checkpoint();
2487
		else if (!prefixcmp(command_buf.buf, "progress "))
2488
			parse_progress();
2489 2490
		else
			die("Unsupported command: %s", command_buf.buf);
2491
	}
2492
	end_packfile();
2493

2494
	dump_branches();
2495
	dump_tags();
2496
	unkeep_all_packs();
2497
	dump_marks();
2498

2499 2500 2501
	if (pack_edges)
		fclose(pack_edges);

2502 2503 2504 2505 2506 2507 2508 2509 2510
	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");
2511 2512 2513 2514 2515 2516
		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]);
2517
		fprintf(stderr, "Total branches:  %10lu (%10lu loads     )\n", branch_count, branch_load_count);
2518
		fprintf(stderr, "      marks:     %10" PRIuMAX " (%10" PRIuMAX " unique    )\n", (((uintmax_t)1) << marks->shift) * 1024, marks_set_count);
2519
		fprintf(stderr, "      atoms:     %10u\n", atom_cnt);
2520
		fprintf(stderr, "Memory total:    %10" PRIuMAX " KiB\n", (total_allocd + alloc_count*sizeof(struct object_entry))/1024);
2521
		fprintf(stderr, "       pools:    %10lu KiB\n", (unsigned long)(total_allocd/1024));
2522
		fprintf(stderr, "     objects:    %10" PRIuMAX " KiB\n", (alloc_count*sizeof(struct object_entry))/1024);
2523 2524 2525 2526 2527
		fprintf(stderr, "---------------------------------------------------------------------\n");
		pack_report();
		fprintf(stderr, "---------------------------------------------------------------------\n");
		fprintf(stderr, "\n");
	}
2528

2529
	return failure ? 1 : 0;
2530
}