fast-import.c 62.8 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
#include "exec_cmd.h"
154

155 156
#define PACK_ID_BITS 16
#define MAX_PACK_ID ((1<<PACK_ID_BITS)-1)
157 158
#define DEPTH_BITS 13
#define MAX_DEPTH ((1<<DEPTH_BITS)-1)
159

160 161 162
struct object_entry
{
	struct object_entry *next;
163
	uint32_t offset;
164 165 166
	uint32_t type : TYPE_BITS,
		pack_id : PACK_ID_BITS,
		depth : DEPTH_BITS;
167 168 169
	unsigned char sha1[20];
};

170
struct object_entry_pool
171
{
172
	struct object_entry_pool *next_pool;
173 174
	struct object_entry *next_free;
	struct object_entry *end;
175
	struct object_entry entries[FLEX_ARRAY]; /* more */
176 177
};

178 179 180 181 182 183
struct mark_set
{
	union {
		struct object_entry *marked[1024];
		struct mark_set *sets[1024];
	} data;
184
	unsigned int shift;
185 186
};

187 188
struct last_object
{
P
Pierre Habouzit 已提交
189
	struct strbuf data;
190
	uint32_t offset;
191
	unsigned int depth;
P
Pierre Habouzit 已提交
192
	unsigned no_swap : 1;
193 194
};

195 196 197 198 199
struct mem_pool
{
	struct mem_pool *next_pool;
	char *next_free;
	char *end;
200
	uintmax_t space[FLEX_ARRAY]; /* more */
201 202 203 204 205
};

struct atom_str
{
	struct atom_str *next_atom;
206
	unsigned short str_len;
207 208 209 210
	char str_dat[FLEX_ARRAY]; /* more */
};

struct tree_content;
211 212
struct tree_entry
{
213 214
	struct tree_content *tree;
	struct atom_str* name;
215 216
	struct tree_entry_ms
	{
217
		uint16_t mode;
218 219
		unsigned char sha1[20];
	} versions[2];
220 221
};

222
struct tree_content
223
{
224 225
	unsigned int entry_capacity; /* must match avail_tree_content */
	unsigned int entry_count;
226
	unsigned int delta_depth;
227 228 229 230 231 232 233
	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;
234 235 236 237
};

struct branch
{
238 239
	struct branch *table_next_branch;
	struct branch *active_next_branch;
240
	const char *name;
241
	struct tree_entry branch_tree;
242
	uintmax_t last_commit;
243 244
	unsigned active : 1;
	unsigned pack_id : PACK_ID_BITS;
245
	unsigned char sha1[20];
246 247
};

248 249 250 251
struct tag
{
	struct tag *next_tag;
	const char *name;
252
	unsigned int pack_id;
253 254 255
	unsigned char sha1[20];
};

256 257 258 259 260
struct hash_list
{
	struct hash_list *next;
	unsigned char sha1[20];
};
261

262 263 264 265 266 267
typedef enum {
	WHENSPEC_RAW = 1,
	WHENSPEC_RFC2822,
	WHENSPEC_NOW,
} whenspec_type;

268 269 270 271 272 273 274
struct recent_command
{
	struct recent_command *prev;
	struct recent_command *next;
	char *buf;
};

275
/* Configured limits on output */
276
static unsigned long max_depth = 10;
277
static off_t max_packsize = (1LL << 32) - 1;
278
static int force_update;
279 280
static int pack_compression_level = Z_DEFAULT_COMPRESSION;
static int pack_compression_seen;
281 282 283 284 285 286 287

/* 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];
288
static unsigned long object_count;
289
static unsigned long branch_count;
290
static unsigned long branch_load_count;
291
static int failure;
292
static FILE *pack_edges;
293

294 295 296 297 298
/* 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;

299
/* Atom management */
300 301 302 303 304
static unsigned int atom_table_sz = 4451;
static unsigned int atom_cnt;
static struct atom_str **atom_table;

/* The .pack file being generated */
305
static unsigned int pack_id;
306
static struct packed_git *pack_data;
307
static struct packed_git **all_packs;
308
static unsigned long pack_size;
309 310

/* Table of objects we've written. */
311
static unsigned int object_entry_alloc = 5000;
312 313
static struct object_entry_pool *blocks;
static struct object_entry *object_table[1 << 16];
314
static struct mark_set *marks;
315
static const char* mark_file;
316 317

/* Our last blob */
P
Pierre Habouzit 已提交
318
static struct last_object last_blob = { STRBUF_INIT, 0, 0, 0 };
319 320 321 322 323 324

/* 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;
325 326
static struct strbuf old_tree = STRBUF_INIT;
static struct strbuf new_tree = STRBUF_INIT;
327

328
/* Branch data */
329 330 331
static unsigned long max_active_branches = 5;
static unsigned long cur_active_branches;
static unsigned long branch_table_sz = 1039;
332 333 334
static struct branch **branch_table;
static struct branch *active_branches;

335 336 337 338
/* Tag data */
static struct tag *first_tag;
static struct tag *last_tag;

339
/* Input stream parsing */
340
static whenspec_type whenspec = WHENSPEC_RAW;
341
static struct strbuf command_buf = STRBUF_INIT;
342
static int unread_command_buf;
343 344 345 346
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;
347
static uintmax_t next_mark;
348
static struct strbuf new_data = STRBUF_INIT;
349

350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375
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);
}

376 377
static void dump_marks_helper(FILE *, uintmax_t, struct mark_set *);

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

	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");
394 395
	fprintf(rpt, "    fast-import process: %"PRIuMAX"\n", (uintmax_t) getpid());
	fprintf(rpt, "    parent process     : %"PRIuMAX"\n", (uintmax_t) getppid());
396 397 398 399
	fprintf(rpt, "    at %s\n", show_date(time(NULL), 0, DATE_LOCAL));
	fputc('\n', rpt);

	fputs("fatal: ", rpt);
400
	fputs(err, rpt);
401 402
	fputc('\n', rpt);

403 404 405 406 407 408 409 410 411 412 413 414
	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);
	}

415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435
	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);
	}

436 437 438 439 440 441 442 443 444 445 446 447 448
	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);
		}
	}

449 450 451 452 453 454 455 456
	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);

457 458 459 460 461 462
	fputc('\n', rpt);
	fputs("-------------------\n", rpt);
	fputs("END OF CRASH REPORT\n", rpt);
	fclose(rpt);
}

463 464 465 466
static void end_packfile(void);
static void unkeep_all_packs(void);
static void dump_marks(void);

467 468 469
static NORETURN void die_nicely(const char *err, va_list params)
{
	static int zombie;
470
	char message[2 * PATH_MAX];
471

472
	vsnprintf(message, sizeof(message), err, params);
473
	fputs("fatal: ", stderr);
474
	fputs(message, stderr);
475 476 477 478
	fputc('\n', stderr);

	if (!zombie) {
		zombie = 1;
479
		write_crash_report(message);
480 481 482
		end_packfile();
		unkeep_all_packs();
		dump_marks();
483 484 485
	}
	exit(128);
}
486

487
static void alloc_objects(unsigned int cnt)
488
{
489
	struct object_entry_pool *b;
490

491
	b = xmalloc(sizeof(struct object_entry_pool)
492
		+ cnt * sizeof(struct object_entry));
493
	b->next_pool = blocks;
494 495 496 497 498
	b->next_free = b->entries;
	b->end = b->entries + cnt;
	blocks = b;
	alloc_count += cnt;
}
499

500
static struct object_entry *new_object(unsigned char *sha1)
501
{
502
	struct object_entry *e;
503

504
	if (blocks->next_free == blocks->end)
505
		alloc_objects(object_entry_alloc);
506

507
	e = blocks->next_free++;
508
	hashcpy(e->sha1, sha1);
509
	return e;
510 511
}

512
static struct object_entry *find_object(unsigned char *sha1)
513 514 515 516
{
	unsigned int h = sha1[0] << 8 | sha1[1];
	struct object_entry *e;
	for (e = object_table[h]; e; e = e->next)
517
		if (!hashcmp(sha1, e->sha1))
518 519 520 521
			return e;
	return NULL;
}

522
static struct object_entry *insert_object(unsigned char *sha1)
523 524
{
	unsigned int h = sha1[0] << 8 | sha1[1];
525
	struct object_entry *e = object_table[h];
526
	struct object_entry *p = NULL;
527 528

	while (e) {
529
		if (!hashcmp(sha1, e->sha1))
530 531 532 533 534 535
			return e;
		p = e;
		e = e->next;
	}

	e = new_object(sha1);
536
	e->next = NULL;
537 538 539 540
	e->offset = 0;
	if (p)
		p->next = e;
	else
541
		object_table[h] = e;
542 543
	return e;
}
544

545 546 547 548 549 550 551 552
static unsigned int hc_str(const char *s, size_t len)
{
	unsigned int r = 0;
	while (len-- > 0)
		r = r * 31 + *s++;
	return r;
}

553
static void *pool_alloc(size_t len)
554 555 556 557
{
	struct mem_pool *p;
	void *r;

558 559 560 561
	/* round up to a 'uintmax_t' alignment */
	if (len & (sizeof(uintmax_t) - 1))
		len += sizeof(uintmax_t) - (len & (sizeof(uintmax_t) - 1));

562 563 564 565 566 567 568 569 570 571 572 573
	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;
574
		p->next_free = (char *) p->space;
575 576 577 578 579 580 581 582 583
		p->end = p->next_free + mem_pool_alloc;
		mem_pool = p;
	}

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

584
static void *pool_calloc(size_t count, size_t size)
585 586 587 588 589 590 591
{
	size_t len = count * size;
	void *r = pool_alloc(len);
	memset(r, 0, len);
	return r;
}

592
static char *pool_strdup(const char *s)
593 594 595 596 597 598
{
	char *r = pool_alloc(strlen(s) + 1);
	strcpy(r, s);
	return r;
}

599
static void insert_mark(uintmax_t idnum, struct object_entry *oe)
600 601 602 603 604 605 606 607 608
{
	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) {
609
		uintmax_t i = idnum >> s->shift;
610 611 612 613 614 615 616 617 618 619 620 621
		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;
}

622
static struct object_entry *find_mark(uintmax_t idnum)
623
{
624
	uintmax_t orig_idnum = idnum;
625 626 627 628
	struct mark_set *s = marks;
	struct object_entry *oe = NULL;
	if ((idnum >> s->shift) < 1024) {
		while (s && s->shift) {
629
			uintmax_t i = idnum >> s->shift;
630 631 632 633 634 635 636
			idnum -= i << s->shift;
			s = s->data.sets[i];
		}
		if (s)
			oe = s->data.marked[idnum];
	}
	if (!oe)
637
		die("mark :%" PRIuMAX " not declared", orig_idnum);
638 639 640
	return oe;
}

641
static struct atom_str *to_atom(const char *s, unsigned short len)
642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659
{
	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;
}

660
static struct branch *lookup_branch(const char *name)
661 662 663 664 665 666 667 668 669 670
{
	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;
}

671
static struct branch *new_branch(const char *name)
672 673 674 675 676 677
{
	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);
678
	switch (check_ref_format(name)) {
679 680 681
	case 0: break; /* its valid */
	case CHECK_REF_FORMAT_ONELEVEL:
		break; /* valid, but too few '/', allow anyway */
682
	default:
683
		die("Branch name doesn't conform to GIT standards: %s", name);
684
	}
685 686 687 688

	b = pool_calloc(1, sizeof(struct branch));
	b->name = pool_strdup(name);
	b->table_next_branch = branch_table[hc];
689 690
	b->branch_tree.versions[0].mode = S_IFDIR;
	b->branch_tree.versions[1].mode = S_IFDIR;
691
	b->active = 0;
692
	b->pack_id = MAX_PACK_ID;
693 694 695 696 697 698 699 700 701 702 703
	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;
}

704
static struct tree_content *new_tree_content(unsigned int cnt)
705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726
{
	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;
727
	t->delta_depth = 0;
728 729 730 731 732 733 734 735
	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);
736 737 738 739 740 741
	f->next_avail = avail_tree_table[hc];
	avail_tree_table[hc] = f;
}

static void release_tree_content_recursive(struct tree_content *t)
{
742 743 744
	unsigned int i;
	for (i = 0; i < t->entry_count; i++)
		release_tree_entry(t->entries[i]);
745
	release_tree_content(t);
746 747
}

748
static struct tree_content *grow_tree_content(
749 750 751 752 753
	struct tree_content *t,
	int amt)
{
	struct tree_content *r = new_tree_content(t->entry_count + amt);
	r->entry_count = t->entry_count;
754
	r->delta_depth = t->delta_depth;
755 756 757 758 759
	memcpy(r->entries,t->entries,t->entry_count*sizeof(t->entries[0]));
	release_tree_content(t);
	return r;
}

760
static struct tree_entry *new_tree_entry(void)
761 762 763 764 765
{
	struct tree_entry *e;

	if (!avail_tree_entry) {
		unsigned int n = tree_entry_alloc;
766
		total_allocd += n * sizeof(struct tree_entry);
767
		avail_tree_entry = e = xmalloc(n * sizeof(struct tree_entry));
768
		while (n-- > 1) {
769 770 771
			*((void**)e) = e + 1;
			e++;
		}
772
		*((void**)e) = NULL;
773 774 775 776 777 778 779 780 781 782
	}

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

static void release_tree_entry(struct tree_entry *e)
{
	if (e->tree)
783
		release_tree_content_recursive(e->tree);
784 785 786 787
	*((void**)e) = avail_tree_entry;
	avail_tree_entry = e;
}

788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812
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;
}

813
static void start_packfile(void)
814
{
815
	static char tmpfile[PATH_MAX];
816
	struct packed_git *p;
817
	struct pack_header hdr;
818
	int pack_fd;
819

820
	snprintf(tmpfile, sizeof(tmpfile),
821
		"%s/pack/tmp_pack_XXXXXX", get_object_directory());
822
	pack_fd = xmkstemp(tmpfile);
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");
867 868 869 870 871 872 873 874 875 876 877 878 879 880 881
	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;
	}

882
	snprintf(tmpfile, sizeof(tmpfile),
883
		"%s/pack/tmp_idx_XXXXXX", get_object_directory());
884
	idx_fd = xmkstemp(tmpfile);
885
	f = sha1fd(idx_fd, tmpfile);
886
	sha1write(f, array, 256 * sizeof(int));
887
	git_SHA1_Init(&ctx);
888
	for (c = idx; c != last; c++) {
889
		uint32_t offset = htonl((*c)->offset);
890 891
		sha1write(f, &offset, 4);
		sha1write(f, (*c)->sha1, sizeof((*c)->sha1));
892
		git_SHA1_Update(&ctx, (*c)->sha1, 20);
893
	}
894
	sha1write(f, pack_data->sha1, sizeof(pack_data->sha1));
895
	sha1close(f, NULL, CSUM_FSYNC);
896
	free(idx);
897
	git_SHA1_Final(pack_data->sha1, &ctx);
898 899 900
	return tmpfile;
}

901
static char *keep_pack(char *curr_index_name)
902 903
{
	static char name[PATH_MAX];
904
	static const char *keep_msg = "fast-import";
905 906 907 908 909 910 911 912 913 914
	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");
915 916 917
	write_or_die(keep_fd, keep_msg, strlen(keep_msg));
	if (close(keep_fd))
		die("failed to write keep file");
918 919 920 921 922 923 924 925 926 927 928 929 930

	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;
}

931
static void unkeep_all_packs(void)
932 933 934 935 936 937 938 939 940 941
{
	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);
	}
942 943
}

944
static void end_packfile(void)
945
{
946 947
	struct packed_git *old_p = pack_data, *new_p;

948
	clear_delta_base_cache();
949
	if (object_count) {
950
		char *idx_name;
951 952 953
		int i;
		struct branch *b;
		struct tag *t;
954

955
		close_pack_windows(pack_data);
956
		fixup_pack_header_footer(pack_data->pack_fd, pack_data->sha1,
957 958
				    pack_data->pack_name, object_count,
				    NULL, 0);
959
		close(pack_data->pack_fd);
960
		idx_name = keep_pack(create_index());
961 962 963 964 965

		/* 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);
966
		all_packs[pack_id] = new_p;
967
		install_packed_git(new_p);
968 969

		/* Print the boundary */
970 971 972 973 974 975 976
		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));
				}
977
			}
978 979 980 981 982 983
			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);
984 985 986
		}

		pack_id++;
987
	}
988 989
	else {
		close(old_p->pack_fd);
990
		unlink(old_p->pack_name);
991
	}
992 993 994
	free(old_p);

	/* We can't carry a delta across packfiles. */
P
Pierre Habouzit 已提交
995
	strbuf_release(&last_blob.data);
996 997
	last_blob.offset = 0;
	last_blob.depth = 0;
998 999
}

1000
static void cycle_packfile(void)
1001 1002 1003 1004 1005
{
	end_packfile();
	start_packfile();
}

1006
static size_t encode_header(
1007
	enum object_type type,
1008
	size_t size,
1009
	unsigned char *hdr)
1010 1011 1012 1013
{
	int n = 1;
	unsigned char c;

1014
	if (type < OBJ_COMMIT || type > OBJ_REF_DELTA)
1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028
		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;
}

1029 1030
static int store_object(
	enum object_type type,
1031
	struct strbuf *dat,
1032
	struct last_object *last,
1033
	unsigned char *sha1out,
1034
	uintmax_t mark)
1035 1036
{
	void *out, *delta;
1037 1038 1039
	struct object_entry *e;
	unsigned char hdr[96];
	unsigned char sha1[20];
1040
	unsigned long hdrlen, deltalen;
1041
	git_SHA_CTX c;
1042 1043
	z_stream s;

1044
	hdrlen = sprintf((char*)hdr,"%s %lu", typename(type),
1045
		(unsigned long)dat->len) + 1;
1046 1047 1048 1049
	git_SHA1_Init(&c);
	git_SHA1_Update(&c, hdr, hdrlen);
	git_SHA1_Update(&c, dat->buf, dat->len);
	git_SHA1_Final(sha1, &c);
1050
	if (sha1out)
1051
		hashcpy(sha1out, sha1);
1052 1053

	e = insert_object(sha1);
1054 1055
	if (mark)
		insert_mark(mark, e);
1056
	if (e->offset) {
1057
		duplicate_count_by_type[type]++;
1058
		return 1;
1059 1060 1061 1062 1063 1064
	} 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;
1065
	}
1066

P
Pierre Habouzit 已提交
1067 1068
	if (last && last->data.buf && last->depth < max_depth) {
		delta = diff_delta(last->data.buf, last->data.len,
1069
			dat->buf, dat->len,
1070
			&deltalen, 0);
1071
		if (delta && deltalen >= dat->len) {
1072 1073 1074 1075 1076
			free(delta);
			delta = NULL;
		}
	} else
		delta = NULL;
1077 1078

	memset(&s, 0, sizeof(s));
1079
	deflateInit(&s, pack_compression_level);
1080 1081 1082 1083
	if (delta) {
		s.next_in = delta;
		s.avail_in = deltalen;
	} else {
1084 1085
		s.next_in = (void *)dat->buf;
		s.avail_in = dat->len;
1086 1087 1088 1089 1090 1091 1092 1093
	}
	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. */
1094
	if ((pack_size + 60 + s.total_out) > max_packsize
1095 1096 1097 1098
		|| (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;
1099
		cycle_packfile();
1100 1101 1102 1103 1104

		/* We cannot carry a delta into the new pack. */
		if (delta) {
			free(delta);
			delta = NULL;
1105 1106

			memset(&s, 0, sizeof(s));
1107
			deflateInit(&s, pack_compression_level);
1108 1109
			s.next_in = (void *)dat->buf;
			s.avail_in = dat->len;
1110 1111 1112 1113 1114
			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);
1115 1116 1117 1118 1119 1120 1121 1122
		}
	}

	e->type = type;
	e->pack_id = pack_id;
	e->offset = pack_size;
	object_count++;
	object_count_by_type[type]++;
1123 1124

	if (delta) {
1125 1126 1127
		unsigned long ofs = e->offset - last->offset;
		unsigned pos = sizeof(hdr) - 1;

1128
		delta_count_by_type[type]++;
1129
		e->depth = last->depth + 1;
1130 1131

		hdrlen = encode_header(OBJ_OFS_DELTA, deltalen, hdr);
1132
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1133 1134 1135 1136 1137
		pack_size += hdrlen;

		hdr[pos] = ofs & 127;
		while (ofs >>= 7)
			hdr[--pos] = 128 | (--ofs & 127);
1138
		write_or_die(pack_data->pack_fd, hdr + pos, sizeof(hdr) - pos);
1139
		pack_size += sizeof(hdr) - pos;
1140
	} else {
1141
		e->depth = 0;
1142
		hdrlen = encode_header(type, dat->len, hdr);
1143
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1144
		pack_size += hdrlen;
1145 1146
	}

1147
	write_or_die(pack_data->pack_fd, out, s.total_out);
1148
	pack_size += s.total_out;
1149 1150

	free(out);
1151
	free(delta);
1152
	if (last) {
P
Pierre Habouzit 已提交
1153 1154 1155
		if (last->no_swap) {
			last->data = *dat;
		} else {
1156
			strbuf_swap(&last->data, dat);
P
Pierre Habouzit 已提交
1157
		}
1158
		last->offset = e->offset;
1159
		last->depth = e->depth;
1160 1161 1162 1163
	}
	return 0;
}

1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181
/* 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.
 */
1182 1183 1184
static void *gfi_unpack_entry(
	struct object_entry *oe,
	unsigned long *sizep)
1185
{
1186
	enum object_type type;
1187
	struct packed_git *p = all_packs[oe->pack_id];
1188
	if (p == pack_data && p->pack_size < (pack_size + 20)) {
1189 1190 1191 1192 1193 1194 1195 1196
		/* 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.
		 */
1197
		close_pack_windows(p);
1198 1199 1200 1201 1202 1203 1204

		/* 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.
		 */
1205
		p->pack_size = pack_size + 20;
1206
	}
1207
	return unpack_entry(p, oe->offset, &type, sizep);
1208 1209
}

1210
static const char *get_mode(const char *str, uint16_t *modep)
1211 1212
{
	unsigned char c;
1213
	uint16_t mode = 0;
1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225

	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)
{
1226
	unsigned char* sha1 = root->versions[1].sha1;
1227 1228 1229 1230 1231 1232 1233
	struct object_entry *myoe;
	struct tree_content *t;
	unsigned long size;
	char *buf;
	const char *c;

	root->tree = t = new_tree_content(8);
1234
	if (is_null_sha1(sha1))
1235 1236
		return;

1237
	myoe = find_object(sha1);
1238
	if (myoe && myoe->pack_id != MAX_PACK_ID) {
1239
		if (myoe->type != OBJ_TREE)
1240
			die("Not a tree: %s", sha1_to_hex(sha1));
1241
		t->delta_depth = myoe->depth;
1242
		buf = gfi_unpack_entry(myoe, &size);
1243 1244
		if (!buf)
			die("Can't load tree %s", sha1_to_hex(sha1));
1245
	} else {
1246 1247 1248
		enum object_type type;
		buf = read_sha1_file(sha1, &type, &size);
		if (!buf || type != OBJ_TREE)
1249
			die("Can't load tree %s", sha1_to_hex(sha1));
1250 1251 1252 1253 1254 1255 1256
	}

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

		if (t->entry_count == t->entry_capacity)
1257
			root->tree = t = grow_tree_content(t, t->entry_count);
1258 1259 1260
		t->entries[t->entry_count++] = e;

		e->tree = NULL;
1261
		c = get_mode(c, &e->versions[1].mode);
1262
		if (!c)
1263 1264
			die("Corrupt mode in %s", sha1_to_hex(sha1));
		e->versions[0].mode = e->versions[1].mode;
1265
		e->name = to_atom(c, strlen(c));
1266
		c += e->name->str_len + 1;
1267 1268
		hashcpy(e->versions[0].sha1, (unsigned char*)c);
		hashcpy(e->versions[1].sha1, (unsigned char*)c);
1269 1270 1271 1272 1273
		c += 20;
	}
	free(buf);
}

1274
static int tecmp0 (const void *_a, const void *_b)
1275 1276 1277 1278
{
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
1279 1280
		a->name->str_dat, a->name->str_len, a->versions[0].mode,
		b->name->str_dat, b->name->str_len, b->versions[0].mode);
1281 1282
}

1283
static int tecmp1 (const void *_a, const void *_b)
1284
{
1285 1286 1287 1288 1289 1290 1291
	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);
}

1292
static void mktree(struct tree_content *t, int v, struct strbuf *b)
1293 1294
{
	size_t maxlen = 0;
1295 1296
	unsigned int i;

1297 1298 1299 1300
	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);
1301 1302

	for (i = 0; i < t->entry_count; i++) {
1303 1304
		if (t->entries[i]->versions[v].mode)
			maxlen += t->entries[i]->name->str_len + 34;
1305 1306
	}

1307 1308
	strbuf_reset(b);
	strbuf_grow(b, maxlen);
1309 1310
	for (i = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
1311 1312
		if (!e->versions[v].mode)
			continue;
1313 1314 1315
		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);
1316
	}
1317 1318 1319 1320 1321 1322
}

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
1323
	struct last_object lo = { STRBUF_INIT, 0, 0, /* no_swap */ 1 };
1324
	struct object_entry *le;
1325 1326 1327 1328 1329 1330 1331 1332 1333

	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]);
	}

1334
	le = find_object(root->versions[0].sha1);
P
Pierre Habouzit 已提交
1335
	if (S_ISDIR(root->versions[0].mode) && le && le->pack_id == pack_id) {
1336
		mktree(t, 0, &old_tree);
P
Pierre Habouzit 已提交
1337
		lo.data = old_tree;
1338
		lo.offset = le->offset;
1339 1340 1341
		lo.depth = t->delta_depth;
	}

1342 1343
	mktree(t, 1, &new_tree);
	store_object(OBJ_TREE, &new_tree, &lo, root->versions[1].sha1, 0);
1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357

	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;
1358 1359 1360 1361 1362 1363
}

static int tree_content_set(
	struct tree_entry *root,
	const char *p,
	const unsigned char *sha1,
1364 1365
	const uint16_t mode,
	struct tree_content *subtree)
1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
{
	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);
1377 1378
	if (!n)
		die("Empty path component found in input");
1379 1380
	if (!slash1 && !S_ISDIR(mode) && subtree)
		die("Non-directories cannot have subtrees");
1381 1382 1383 1384 1385

	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) {
1386 1387
				if (!S_ISDIR(mode)
						&& e->versions[1].mode == mode
1388
						&& !hashcmp(e->versions[1].sha1, sha1))
1389
					return 0;
1390 1391
				e->versions[1].mode = mode;
				hashcpy(e->versions[1].sha1, sha1);
1392
				if (e->tree)
1393
					release_tree_content_recursive(e->tree);
1394
				e->tree = subtree;
1395
				hashclr(root->versions[1].sha1);
1396 1397
				return 1;
			}
1398
			if (!S_ISDIR(e->versions[1].mode)) {
1399
				e->tree = new_tree_content(8);
1400
				e->versions[1].mode = S_IFDIR;
1401 1402 1403
			}
			if (!e->tree)
				load_tree(e);
1404
			if (tree_content_set(e, slash1 + 1, sha1, mode, subtree)) {
1405
				hashclr(root->versions[1].sha1);
1406 1407 1408 1409 1410 1411 1412
				return 1;
			}
			return 0;
		}
	}

	if (t->entry_count == t->entry_capacity)
1413
		root->tree = t = grow_tree_content(t, t->entry_count);
1414
	e = new_tree_entry();
1415
	e->name = to_atom(p, n);
1416 1417
	e->versions[0].mode = 0;
	hashclr(e->versions[0].sha1);
1418 1419 1420
	t->entries[t->entry_count++] = e;
	if (slash1) {
		e->tree = new_tree_content(8);
1421
		e->versions[1].mode = S_IFDIR;
1422
		tree_content_set(e, slash1 + 1, sha1, mode, subtree);
1423
	} else {
1424
		e->tree = subtree;
1425 1426
		e->versions[1].mode = mode;
		hashcpy(e->versions[1].sha1, sha1);
1427
	}
1428
	hashclr(root->versions[1].sha1);
1429 1430 1431
	return 1;
}

1432 1433 1434 1435
static int tree_content_remove(
	struct tree_entry *root,
	const char *p,
	struct tree_entry *backup_leaf)
1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450
{
	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)) {
1451
			if (!slash1 || !S_ISDIR(e->versions[1].mode))
1452 1453 1454
				goto del_entry;
			if (!e->tree)
				load_tree(e);
1455
			if (tree_content_remove(e, slash1 + 1, backup_leaf)) {
1456 1457 1458 1459 1460 1461
				for (n = 0; n < e->tree->entry_count; n++) {
					if (e->tree->entries[n]->versions[1].mode) {
						hashclr(root->versions[1].sha1);
						return 1;
					}
				}
1462
				backup_leaf = NULL;
1463
				goto del_entry;
1464 1465 1466 1467 1468 1469 1470
			}
			return 0;
		}
	}
	return 0;

del_entry:
1471 1472 1473
	if (backup_leaf)
		memcpy(backup_leaf, e, sizeof(*backup_leaf));
	else if (e->tree)
1474
		release_tree_content_recursive(e->tree);
1475
	e->tree = NULL;
1476 1477 1478
	e->versions[1].mode = 0;
	hashclr(e->versions[1].sha1);
	hashclr(root->versions[1].sha1);
1479
	return 1;
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 1517 1518
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;
}

1519
static int update_branch(struct branch *b)
1520 1521
{
	static const char *msg = "fast-import";
1522 1523 1524
	struct ref_lock *lock;
	unsigned char old_sha1[20];

1525 1526
	if (is_null_sha1(b->sha1))
		return 0;
1527 1528
	if (read_ref(b->name, old_sha1))
		hashclr(old_sha1);
1529
	lock = lock_any_ref_for_update(b->name, old_sha1, 0);
1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541
	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);
		}

1542
		if (!in_merge_bases(old_cmit, &new_cmit, 1)) {
1543
			unlock_ref(lock);
1544
			warning("Not updating %s"
1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556
				" (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)
{
1557 1558 1559 1560
	unsigned int i;
	struct branch *b;

	for (i = 0; i < branch_table_sz; i++) {
1561 1562
		for (b = branch_table[i]; b; b = b->table_next_branch)
			failure |= update_branch(b);
1563 1564 1565
	}
}

1566
static void dump_tags(void)
1567 1568 1569 1570
{
	static const char *msg = "fast-import";
	struct tag *t;
	struct ref_lock *lock;
1571
	char ref_name[PATH_MAX];
1572 1573

	for (t = first_tag; t; t = t->next_tag) {
1574 1575
		sprintf(ref_name, "tags/%s", t->name);
		lock = lock_ref_sha1(ref_name, NULL);
1576
		if (!lock || write_ref_sha1(lock, t->sha1, msg) < 0)
1577
			failure |= error("Unable to update %s", ref_name);
1578 1579 1580
	}
}

1581
static void dump_marks_helper(FILE *f,
1582
	uintmax_t base,
1583 1584
	struct mark_set *m)
{
1585
	uintmax_t k;
1586 1587 1588 1589 1590 1591 1592 1593 1594
	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])
1595
				fprintf(f, ":%" PRIuMAX " %s\n", base + k,
1596 1597 1598 1599 1600
					sha1_to_hex(m->data.marked[k]->sha1));
		}
	}
}

1601
static void dump_marks(void)
1602
{
1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614
	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;
1615
	}
1616 1617 1618

	f = fdopen(mark_fd, "w");
	if (!f) {
1619
		int saved_errno = errno;
1620 1621
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1622
			mark_file, strerror(saved_errno));
1623
		return;
1624
	}
1625

B
Brandon Casey 已提交
1626
	/*
1627 1628
	 * 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
1629 1630 1631
	 * won't try to close() it.
	 */
	mark_lock.fd = -1;
1632 1633 1634

	dump_marks_helper(f, 0, marks);
	if (ferror(f) || fclose(f)) {
1635
		int saved_errno = errno;
1636 1637
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1638
			mark_file, strerror(saved_errno));
1639 1640 1641 1642
		return;
	}

	if (commit_lock_file(&mark_lock)) {
1643
		int saved_errno = errno;
1644 1645
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to commit marks file %s: %s",
1646
			mark_file, strerror(saved_errno));
1647 1648
		return;
	}
1649 1650
}

1651
static int read_next_command(void)
1652
{
1653 1654 1655 1656 1657 1658 1659
	static int stdin_eof = 0;

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

1660
	do {
1661
		if (unread_command_buf) {
1662
			unread_command_buf = 0;
1663 1664 1665
		} else {
			struct recent_command *rc;

1666
			strbuf_detach(&command_buf, NULL);
1667 1668 1669
			stdin_eof = strbuf_getline(&command_buf, stdin, '\n');
			if (stdin_eof)
				return EOF;
1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687

			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] == '#');
1688 1689

	return 0;
1690 1691
}

J
Junio C Hamano 已提交
1692
static void skip_optional_lf(void)
1693 1694 1695 1696 1697 1698
{
	int term_char = fgetc(stdin);
	if (term_char != '\n' && term_char != EOF)
		ungetc(term_char, stdin);
}

1699
static void parse_mark(void)
1700
{
1701
	if (!prefixcmp(command_buf.buf, "mark :")) {
1702
		next_mark = strtoumax(command_buf.buf + 6, NULL, 10);
1703 1704 1705
		read_next_command();
	}
	else
1706
		next_mark = 0;
1707 1708
}

1709
static void parse_data(struct strbuf *sb)
1710
{
1711
	strbuf_reset(sb);
1712

1713
	if (prefixcmp(command_buf.buf, "data "))
1714 1715
		die("Expected 'data n' command, found: %s", command_buf.buf);

1716
	if (!prefixcmp(command_buf.buf + 5, "<<")) {
1717
		char *term = xstrdup(command_buf.buf + 5 + 2);
1718 1719
		size_t term_len = command_buf.len - 5 - 2;

1720
		strbuf_detach(&command_buf, NULL);
1721
		for (;;) {
1722
			if (strbuf_getline(&command_buf, stdin, '\n') == EOF)
1723 1724 1725 1726
				die("EOF in data (terminator '%s' not found)", term);
			if (term_len == command_buf.len
				&& !strcmp(term, command_buf.buf))
				break;
1727 1728
			strbuf_addbuf(sb, &command_buf);
			strbuf_addch(sb, '\n');
1729 1730 1731 1732
		}
		free(term);
	}
	else {
1733 1734
		size_t n = 0, length;

1735
		length = strtoul(command_buf.buf + 5, NULL, 10);
1736

1737
		while (n < length) {
1738
			size_t s = strbuf_fread(sb, length - n, stdin);
1739
			if (!s && feof(stdin))
1740 1741
				die("EOF in data (%lu bytes remaining)",
					(unsigned long)(length - n));
1742 1743
			n += s;
		}
1744 1745
	}

1746
	skip_optional_lf();
1747 1748
}

1749 1750 1751 1752
static int validate_raw_date(const char *src, char *result, int maxlen)
{
	const char *orig_src = src;
	char *endp, sign;
1753
	unsigned long date;
1754

1755 1756 1757 1758
	errno = 0;

	date = strtoul(src, &endp, 10);
	if (errno || endp == src || *endp != ' ')
1759 1760 1761 1762 1763 1764 1765
		return -1;

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

1766 1767
	date = strtoul(src + 1, &endp, 10);
	if (errno || endp == src || *endp || (endp - orig_src) >= maxlen)
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 1808 1809
		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;
}

1810
static void parse_new_blob(void)
1811
{
P
Pierre Habouzit 已提交
1812
	static struct strbuf buf = STRBUF_INIT;
1813 1814

	read_next_command();
1815 1816
	parse_mark();
	parse_data(&buf);
P
Pierre Habouzit 已提交
1817
	store_object(OBJ_BLOB, &buf, &last_blob, NULL, next_mark);
1818 1819
}

1820
static void unload_one_branch(void)
1821
{
1822 1823
	while (cur_active_branches
		&& cur_active_branches >= max_active_branches) {
1824
		uintmax_t min_commit = ULONG_MAX;
1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841
		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;
		}
1842
		e->active = 0;
1843 1844
		e->active_next_branch = NULL;
		if (e->branch_tree.tree) {
1845
			release_tree_content_recursive(e->branch_tree.tree);
1846 1847 1848
			e->branch_tree.tree = NULL;
		}
		cur_active_branches--;
1849 1850 1851
	}
}

1852
static void load_branch(struct branch *b)
1853
{
1854
	load_tree(&b->branch_tree);
1855 1856 1857 1858 1859 1860 1861
	if (!b->active) {
		b->active = 1;
		b->active_next_branch = active_branches;
		active_branches = b;
		cur_active_branches++;
		branch_load_count++;
	}
1862 1863
}

1864
static void file_change_m(struct branch *b)
1865
{
1866
	const char *p = command_buf.buf + 2;
1867
	static struct strbuf uq = STRBUF_INIT;
1868
	const char *endp;
1869
	struct object_entry *oe = oe;
1870
	unsigned char sha1[20];
1871
	uint16_t mode, inline_data = 0;
1872

1873 1874 1875 1876
	p = get_mode(p, &mode);
	if (!p)
		die("Corrupt mode: %s", command_buf.buf);
	switch (mode) {
1877 1878 1879
	case 0644:
	case 0755:
		mode |= S_IFREG;
1880 1881
	case S_IFREG | 0644:
	case S_IFREG | 0755:
1882
	case S_IFLNK:
1883
	case S_IFGITLINK:
1884 1885 1886 1887 1888 1889
		/* ok */
		break;
	default:
		die("Corrupt mode: %s", command_buf.buf);
	}

1890 1891
	if (*p == ':') {
		char *x;
1892
		oe = find_mark(strtoumax(p + 1, &x, 10));
1893
		hashcpy(sha1, oe->sha1);
1894
		p = x;
1895
	} else if (!prefixcmp(p, "inline")) {
1896 1897
		inline_data = 1;
		p += 6;
1898 1899 1900 1901 1902 1903
	} else {
		if (get_sha1_hex(p, sha1))
			die("Invalid SHA1: %s", command_buf.buf);
		oe = find_object(sha1);
		p += 40;
	}
1904 1905 1906
	if (*p++ != ' ')
		die("Missing space after SHA1: %s", command_buf.buf);

1907 1908
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1909 1910
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1911
		p = uq.buf;
1912
	}
1913

1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927
	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
1928
		static struct strbuf buf = STRBUF_INIT;
1929

1930 1931 1932 1933
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
1934
		read_next_command();
1935
		parse_data(&buf);
P
Pierre Habouzit 已提交
1936
		store_object(OBJ_BLOB, &buf, &last_blob, sha1, 0);
1937
	} else if (oe) {
1938
		if (oe->type != OBJ_BLOB)
1939
			die("Not a blob (actually a %s): %s",
1940
				typename(oe->type), command_buf.buf);
1941
	} else {
1942 1943
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
1944
			die("Blob not found: %s", command_buf.buf);
1945
		if (type != OBJ_BLOB)
1946
			die("Not a blob (actually a %s): %s",
1947
			    typename(type), command_buf.buf);
1948
	}
1949

1950
	tree_content_set(&b->branch_tree, p, sha1, mode, NULL);
1951
}
1952

1953 1954
static void file_change_d(struct branch *b)
{
1955
	const char *p = command_buf.buf + 2;
1956
	static struct strbuf uq = STRBUF_INIT;
1957 1958
	const char *endp;

1959 1960
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1961 1962
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1963
		p = uq.buf;
1964
	}
1965
	tree_content_remove(&b->branch_tree, p, NULL);
1966 1967
}

1968
static void file_change_cr(struct branch *b, int rename)
1969 1970
{
	const char *s, *d;
1971 1972
	static struct strbuf s_uq = STRBUF_INIT;
	static struct strbuf d_uq = STRBUF_INIT;
1973 1974 1975 1976
	const char *endp;
	struct tree_entry leaf;

	s = command_buf.buf + 2;
1977 1978
	strbuf_reset(&s_uq);
	if (!unquote_c_style(&s_uq, s, &endp)) {
1979 1980
		if (*endp != ' ')
			die("Missing space after source: %s", command_buf.buf);
1981
	} else {
1982 1983 1984
		endp = strchr(s, ' ');
		if (!endp)
			die("Missing space after source: %s", command_buf.buf);
1985
		strbuf_add(&s_uq, s, endp - s);
1986
	}
1987
	s = s_uq.buf;
1988 1989 1990 1991 1992 1993

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

	d = endp;
1994 1995
	strbuf_reset(&d_uq);
	if (!unquote_c_style(&d_uq, d, &endp)) {
1996 1997
		if (*endp)
			die("Garbage after dest in: %s", command_buf.buf);
1998
		d = d_uq.buf;
1999 2000 2001
	}

	memset(&leaf, 0, sizeof(leaf));
2002 2003 2004 2005
	if (rename)
		tree_content_remove(&b->branch_tree, s, &leaf);
	else
		tree_content_get(&b->branch_tree, s, &leaf);
2006 2007 2008 2009 2010 2011 2012 2013
	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);
}

2014 2015 2016 2017 2018 2019 2020 2021
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);
}

2022
static void parse_from_commit(struct branch *b, char *buf, unsigned long size)
2023 2024 2025 2026 2027 2028 2029 2030 2031 2032
{
	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);
}

2033
static void parse_from_existing(struct branch *b)
2034 2035 2036 2037 2038 2039 2040 2041 2042 2043
{
	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);
2044
		parse_from_commit(b, buf, size);
2045 2046 2047 2048
		free(buf);
	}
}

2049
static int parse_from(struct branch *b)
2050
{
2051
	const char *from;
2052 2053
	struct branch *s;

2054
	if (prefixcmp(command_buf.buf, "from "))
2055
		return 0;
2056

2057 2058 2059 2060
	if (b->branch_tree.tree) {
		release_tree_content_recursive(b->branch_tree.tree);
		b->branch_tree.tree = NULL;
	}
2061 2062 2063 2064 2065 2066

	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) {
2067
		unsigned char *t = s->branch_tree.versions[1].sha1;
2068
		hashcpy(b->sha1, s->sha1);
2069 2070
		hashcpy(b->branch_tree.versions[0].sha1, t);
		hashcpy(b->branch_tree.versions[1].sha1, t);
2071
	} else if (*from == ':') {
2072
		uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2073 2074
		struct object_entry *oe = find_mark(idnum);
		if (oe->type != OBJ_COMMIT)
2075
			die("Mark :%" PRIuMAX " not a commit", idnum);
2076
		hashcpy(b->sha1, oe->sha1);
2077
		if (oe->pack_id != MAX_PACK_ID) {
2078
			unsigned long size;
2079
			char *buf = gfi_unpack_entry(oe, &size);
2080
			parse_from_commit(b, buf, size);
2081
			free(buf);
2082
		} else
2083
			parse_from_existing(b);
2084
	} else if (!get_sha1(from, b->sha1))
2085
		parse_from_existing(b);
2086
	else
2087 2088 2089
		die("Invalid ref name or SHA1 expression: %s", from);

	read_next_command();
2090
	return 1;
2091 2092
}

2093
static struct hash_list *parse_merge(unsigned int *count)
2094
{
2095
	struct hash_list *list = NULL, *n, *e = e;
2096
	const char *from;
2097 2098 2099
	struct branch *s;

	*count = 0;
2100
	while (!prefixcmp(command_buf.buf, "merge ")) {
2101 2102 2103 2104 2105 2106
		from = strchr(command_buf.buf, ' ') + 1;
		n = xmalloc(sizeof(*n));
		s = lookup_branch(from);
		if (s)
			hashcpy(n->sha1, s->sha1);
		else if (*from == ':') {
2107
			uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2108 2109
			struct object_entry *oe = find_mark(idnum);
			if (oe->type != OBJ_COMMIT)
2110
				die("Mark :%" PRIuMAX " not a commit", idnum);
2111
			hashcpy(n->sha1, oe->sha1);
2112 2113 2114
		} 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
2115
				commit_type, &size, n->sha1);
2116 2117 2118 2119
			if (!buf || size < 46)
				die("Not a valid commit: %s", from);
			free(buf);
		} else
2120 2121 2122 2123 2124 2125 2126 2127
			die("Invalid ref name or SHA1 expression: %s", from);

		n->next = NULL;
		if (list)
			e->next = n;
		else
			list = n;
		e = n;
2128
		(*count)++;
2129 2130 2131 2132 2133
		read_next_command();
	}
	return list;
}

2134
static void parse_new_commit(void)
2135
{
2136
	static struct strbuf msg = STRBUF_INIT;
2137 2138 2139 2140
	struct branch *b;
	char *sp;
	char *author = NULL;
	char *committer = NULL;
2141 2142
	struct hash_list *merge_list = NULL;
	unsigned int merge_count;
2143 2144 2145 2146

	/* Obtain the branch name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	b = lookup_branch(sp);
2147
	if (!b)
2148
		b = new_branch(sp);
2149 2150

	read_next_command();
2151
	parse_mark();
2152
	if (!prefixcmp(command_buf.buf, "author ")) {
2153
		author = parse_ident(command_buf.buf + 7);
2154 2155
		read_next_command();
	}
2156
	if (!prefixcmp(command_buf.buf, "committer ")) {
2157
		committer = parse_ident(command_buf.buf + 10);
2158 2159 2160 2161
		read_next_command();
	}
	if (!committer)
		die("Expected committer but didn't get one");
2162
	parse_data(&msg);
2163
	read_next_command();
2164 2165
	parse_from(b);
	merge_list = parse_merge(&merge_count);
2166 2167

	/* ensure the branch is active/loaded */
2168
	if (!b->branch_tree.tree || !max_active_branches) {
2169 2170 2171
		unload_one_branch();
		load_branch(b);
	}
2172

2173
	/* file_change* */
2174
	while (command_buf.len > 0) {
2175
		if (!prefixcmp(command_buf.buf, "M "))
2176
			file_change_m(b);
2177
		else if (!prefixcmp(command_buf.buf, "D "))
2178
			file_change_d(b);
2179
		else if (!prefixcmp(command_buf.buf, "R "))
2180 2181 2182
			file_change_cr(b, 1);
		else if (!prefixcmp(command_buf.buf, "C "))
			file_change_cr(b, 0);
2183 2184
		else if (!strcmp("deleteall", command_buf.buf))
			file_change_deleteall(b);
2185 2186 2187 2188
		else {
			unread_command_buf = 1;
			break;
		}
2189 2190
		if (read_next_command() == EOF)
			break;
2191 2192
	}

2193
	/* build the tree and the commit */
2194
	store_tree(&b->branch_tree);
2195 2196
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
2197 2198 2199

	strbuf_reset(&new_data);
	strbuf_addf(&new_data, "tree %s\n",
2200
		sha1_to_hex(b->branch_tree.versions[1].sha1));
2201
	if (!is_null_sha1(b->sha1))
2202
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(b->sha1));
2203 2204
	while (merge_list) {
		struct hash_list *next = merge_list->next;
2205
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(merge_list->sha1));
2206 2207 2208
		free(merge_list);
		merge_list = next;
	}
2209 2210 2211 2212 2213 2214
	strbuf_addf(&new_data,
		"author %s\n"
		"committer %s\n"
		"\n",
		author ? author : committer, committer);
	strbuf_addbuf(&new_data, &msg);
2215
	free(author);
2216 2217
	free(committer);

2218
	if (!store_object(OBJ_COMMIT, &new_data, NULL, b->sha1, next_mark))
2219
		b->pack_id = pack_id;
2220
	b->last_commit = object_count_by_type[OBJ_COMMIT];
2221 2222
}

2223
static void parse_new_tag(void)
2224
{
2225
	static struct strbuf msg = STRBUF_INIT;
2226 2227 2228 2229 2230
	char *sp;
	const char *from;
	char *tagger;
	struct branch *s;
	struct tag *t;
2231
	uintmax_t from_mark = 0;
2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246
	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 ... */
2247
	if (prefixcmp(command_buf.buf, "from "))
2248 2249 2250 2251
		die("Expected from command, got %s", command_buf.buf);
	from = strchr(command_buf.buf, ' ') + 1;
	s = lookup_branch(from);
	if (s) {
2252
		hashcpy(sha1, s->sha1);
2253
	} else if (*from == ':') {
2254
		struct object_entry *oe;
2255
		from_mark = strtoumax(from + 1, NULL, 10);
2256
		oe = find_mark(from_mark);
2257
		if (oe->type != OBJ_COMMIT)
2258
			die("Mark :%" PRIuMAX " not a commit", from_mark);
2259
		hashcpy(sha1, oe->sha1);
2260 2261 2262 2263 2264
	} else if (!get_sha1(from, sha1)) {
		unsigned long size;
		char *buf;

		buf = read_object_with_reference(sha1,
2265
			commit_type, &size, sha1);
2266 2267 2268 2269 2270 2271 2272 2273
		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 ... */
2274 2275 2276 2277 2278
	if (!prefixcmp(command_buf.buf, "tagger ")) {
		tagger = parse_ident(command_buf.buf + 7);
		read_next_command();
	} else
		tagger = NULL;
2279 2280

	/* tag payload/message */
2281
	parse_data(&msg);
2282 2283

	/* build the tag object */
2284
	strbuf_reset(&new_data);
2285

2286
	strbuf_addf(&new_data,
2287 2288 2289 2290 2291 2292 2293 2294
		    "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');
2295
	strbuf_addbuf(&new_data, &msg);
2296 2297
	free(tagger);

2298
	if (store_object(OBJ_TAG, &new_data, NULL, t->sha1, 0))
2299 2300 2301
		t->pack_id = MAX_PACK_ID;
	else
		t->pack_id = pack_id;
2302 2303
}

2304
static void parse_reset_branch(void)
2305 2306 2307 2308 2309 2310 2311 2312
{
	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) {
2313 2314 2315
		hashclr(b->sha1);
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
2316 2317 2318 2319 2320
		if (b->branch_tree.tree) {
			release_tree_content_recursive(b->branch_tree.tree);
			b->branch_tree.tree = NULL;
		}
	}
2321 2322 2323
	else
		b = new_branch(sp);
	read_next_command();
2324
	parse_from(b);
2325
	if (command_buf.len > 0)
2326
		unread_command_buf = 1;
2327 2328
}

2329
static void parse_checkpoint(void)
2330
{
2331 2332 2333 2334 2335 2336
	if (object_count) {
		cycle_packfile();
		dump_branches();
		dump_tags();
		dump_marks();
	}
2337
	skip_optional_lf();
2338 2339
}

2340
static void parse_progress(void)
2341
{
2342
	fwrite(command_buf.buf, 1, command_buf.len, stdout);
2343 2344 2345 2346 2347
	fputc('\n', stdout);
	fflush(stdout);
	skip_optional_lf();
}

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 2374 2375
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;
2376
			e->offset = 1; /* just not zero! */
2377 2378 2379 2380 2381 2382
		}
		insert_mark(mark, e);
	}
	fclose(f);
}

2383
static int git_pack_config(const char *k, const char *v, void *cb)
2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400
{
	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;
	}
2401
	return git_default_config(k, v, cb);
2402 2403
}

2404
static const char fast_import_usage[] =
S
Stephan Beyer 已提交
2405
"git fast-import [--date-format=f] [--max-pack-size=n] [--depth=n] [--active-branches=n] [--export-marks=marks.file]";
2406

2407 2408
int main(int argc, const char **argv)
{
2409
	unsigned int i, show_stats = 1;
2410

2411 2412
	git_extract_argv0_path(argv[0]);

2413
	setup_git_directory();
2414
	git_config(git_pack_config, NULL);
2415 2416 2417
	if (!pack_compression_seen && core_compression_seen)
		pack_compression_level = core_compression_level;

2418
	alloc_objects(object_entry_alloc);
2419
	strbuf_init(&command_buf, 0);
2420 2421 2422 2423
	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));
2424

2425 2426 2427 2428 2429
	for (i = 1; i < argc; i++) {
		const char *a = argv[i];

		if (*a != '-' || !strcmp(a, "--"))
			break;
2430
		else if (!prefixcmp(a, "--date-format=")) {
2431 2432 2433 2434 2435 2436 2437 2438 2439 2440
			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);
		}
2441
		else if (!prefixcmp(a, "--max-pack-size="))
2442
			max_packsize = strtoumax(a + 16, NULL, 0) * 1024 * 1024;
2443
		else if (!prefixcmp(a, "--depth=")) {
2444
			max_depth = strtoul(a + 8, NULL, 0);
2445 2446 2447
			if (max_depth > MAX_DEPTH)
				die("--depth cannot exceed %u", MAX_DEPTH);
		}
2448
		else if (!prefixcmp(a, "--active-branches="))
2449
			max_active_branches = strtoul(a + 18, NULL, 0);
2450 2451
		else if (!prefixcmp(a, "--import-marks="))
			import_marks(a + 15);
2452
		else if (!prefixcmp(a, "--export-marks="))
2453
			mark_file = a + 15;
2454
		else if (!prefixcmp(a, "--export-pack-edges=")) {
2455 2456 2457 2458 2459 2460
			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"))
2461
			force_update = 1;
2462 2463 2464 2465
		else if (!strcmp(a, "--quiet"))
			show_stats = 0;
		else if (!strcmp(a, "--stats"))
			show_stats = 1;
2466 2467 2468
		else
			die("unknown option %s", a);
	}
2469
	if (i != argc)
2470 2471
		usage(fast_import_usage);

2472 2473 2474 2475 2476
	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;

2477
	prepare_packed_git();
2478
	start_packfile();
2479
	set_die_routine(die_nicely);
2480 2481
	while (read_next_command() != EOF) {
		if (!strcmp("blob", command_buf.buf))
2482
			parse_new_blob();
2483
		else if (!prefixcmp(command_buf.buf, "commit "))
2484
			parse_new_commit();
2485
		else if (!prefixcmp(command_buf.buf, "tag "))
2486
			parse_new_tag();
2487
		else if (!prefixcmp(command_buf.buf, "reset "))
2488
			parse_reset_branch();
2489
		else if (!strcmp("checkpoint", command_buf.buf))
2490
			parse_checkpoint();
2491
		else if (!prefixcmp(command_buf.buf, "progress "))
2492
			parse_progress();
2493 2494
		else
			die("Unsupported command: %s", command_buf.buf);
2495
	}
2496
	end_packfile();
2497

2498
	dump_branches();
2499
	dump_tags();
2500
	unkeep_all_packs();
2501
	dump_marks();
2502

2503 2504 2505
	if (pack_edges)
		fclose(pack_edges);

2506 2507 2508 2509 2510 2511 2512 2513 2514
	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");
2515 2516 2517 2518 2519 2520
		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]);
2521
		fprintf(stderr, "Total branches:  %10lu (%10lu loads     )\n", branch_count, branch_load_count);
2522
		fprintf(stderr, "      marks:     %10" PRIuMAX " (%10" PRIuMAX " unique    )\n", (((uintmax_t)1) << marks->shift) * 1024, marks_set_count);
2523
		fprintf(stderr, "      atoms:     %10u\n", atom_cnt);
2524
		fprintf(stderr, "Memory total:    %10" PRIuMAX " KiB\n", (total_allocd + alloc_count*sizeof(struct object_entry))/1024);
2525
		fprintf(stderr, "       pools:    %10lu KiB\n", (unsigned long)(total_allocd/1024));
2526
		fprintf(stderr, "     objects:    %10" PRIuMAX " KiB\n", (alloc_count*sizeof(struct object_entry))/1024);
2527 2528 2529 2530 2531
		fprintf(stderr, "---------------------------------------------------------------------\n");
		pack_report();
		fprintf(stderr, "---------------------------------------------------------------------\n");
		fprintf(stderr, "\n");
	}
2532

2533
	return failure ? 1 : 0;
2534
}