fast-import.c 62.5 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

820 821
	pack_fd = odb_mkstemp(tmpfile, sizeof(tmpfile),
			      "pack/tmp_pack_XXXXXX");
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
	qsort(idx, object_count, sizeof(struct object_entry*), oecmp);

	/* Generate the fan-out array. */
	c = idx;
	for (i = 0; i < 256; i++) {
J
Junio C Hamano 已提交
871
		struct object_entry **next = c;
872 873 874 875 876 877 878 879 880
		while (next < last) {
			if ((*next)->sha1[0] != i)
				break;
			next++;
		}
		array[i] = htonl(next - idx);
		c = next;
	}

881 882
	idx_fd = odb_mkstemp(tmpfile, sizeof(tmpfile),
			     "pack/tmp_idx_XXXXXX");
883
	f = sha1fd(idx_fd, tmpfile);
884
	sha1write(f, array, 256 * sizeof(int));
885
	git_SHA1_Init(&ctx);
886
	for (c = idx; c != last; c++) {
887
		uint32_t offset = htonl((*c)->offset);
888 889
		sha1write(f, &offset, 4);
		sha1write(f, (*c)->sha1, sizeof((*c)->sha1));
890
		git_SHA1_Update(&ctx, (*c)->sha1, 20);
891
	}
892
	sha1write(f, pack_data->sha1, sizeof(pack_data->sha1));
893
	sha1close(f, NULL, CSUM_FSYNC);
894
	free(idx);
895
	git_SHA1_Final(pack_data->sha1, &ctx);
896 897 898
	return tmpfile;
}

899
static char *keep_pack(char *curr_index_name)
900 901
{
	static char name[PATH_MAX];
902
	static const char *keep_msg = "fast-import";
903 904 905 906 907
	int keep_fd;

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

908
	keep_fd = odb_pack_keep(name, sizeof(name), pack_data->sha1);
909 910
	if (keep_fd < 0)
		die("cannot create keep file");
911 912 913
	write_or_die(keep_fd, keep_msg, strlen(keep_msg));
	if (close(keep_fd))
		die("failed to write keep file");
914 915 916 917 918 919 920 921 922 923 924 925 926

	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;
}

927
static void unkeep_all_packs(void)
928 929 930 931 932 933 934 935 936 937
{
	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);
	}
938 939
}

940
static void end_packfile(void)
941
{
942 943
	struct packed_git *old_p = pack_data, *new_p;

944
	clear_delta_base_cache();
945
	if (object_count) {
946
		char *idx_name;
947 948 949
		int i;
		struct branch *b;
		struct tag *t;
950

951
		close_pack_windows(pack_data);
952
		fixup_pack_header_footer(pack_data->pack_fd, pack_data->sha1,
953 954
				    pack_data->pack_name, object_count,
				    NULL, 0);
955
		close(pack_data->pack_fd);
956
		idx_name = keep_pack(create_index());
957 958 959 960 961

		/* 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);
962
		all_packs[pack_id] = new_p;
963
		install_packed_git(new_p);
964 965

		/* Print the boundary */
966 967 968 969 970 971 972
		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));
				}
973
			}
974 975 976 977 978 979
			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);
980 981 982
		}

		pack_id++;
983
	}
984 985
	else {
		close(old_p->pack_fd);
986
		unlink(old_p->pack_name);
987
	}
988 989 990
	free(old_p);

	/* We can't carry a delta across packfiles. */
P
Pierre Habouzit 已提交
991
	strbuf_release(&last_blob.data);
992 993
	last_blob.offset = 0;
	last_blob.depth = 0;
994 995
}

996
static void cycle_packfile(void)
997 998 999 1000 1001
{
	end_packfile();
	start_packfile();
}

1002
static size_t encode_header(
1003
	enum object_type type,
1004
	size_t size,
1005
	unsigned char *hdr)
1006 1007 1008 1009
{
	int n = 1;
	unsigned char c;

1010
	if (type < OBJ_COMMIT || type > OBJ_REF_DELTA)
1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024
		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;
}

1025 1026
static int store_object(
	enum object_type type,
1027
	struct strbuf *dat,
1028
	struct last_object *last,
1029
	unsigned char *sha1out,
1030
	uintmax_t mark)
1031 1032
{
	void *out, *delta;
1033 1034 1035
	struct object_entry *e;
	unsigned char hdr[96];
	unsigned char sha1[20];
1036
	unsigned long hdrlen, deltalen;
1037
	git_SHA_CTX c;
1038 1039
	z_stream s;

1040
	hdrlen = sprintf((char*)hdr,"%s %lu", typename(type),
1041
		(unsigned long)dat->len) + 1;
1042 1043 1044 1045
	git_SHA1_Init(&c);
	git_SHA1_Update(&c, hdr, hdrlen);
	git_SHA1_Update(&c, dat->buf, dat->len);
	git_SHA1_Final(sha1, &c);
1046
	if (sha1out)
1047
		hashcpy(sha1out, sha1);
1048 1049

	e = insert_object(sha1);
1050 1051
	if (mark)
		insert_mark(mark, e);
1052
	if (e->offset) {
1053
		duplicate_count_by_type[type]++;
1054
		return 1;
1055 1056 1057 1058 1059 1060
	} 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;
1061
	}
1062

P
Pierre Habouzit 已提交
1063 1064
	if (last && last->data.buf && last->depth < max_depth) {
		delta = diff_delta(last->data.buf, last->data.len,
1065
			dat->buf, dat->len,
1066
			&deltalen, 0);
1067
		if (delta && deltalen >= dat->len) {
1068 1069 1070 1071 1072
			free(delta);
			delta = NULL;
		}
	} else
		delta = NULL;
1073 1074

	memset(&s, 0, sizeof(s));
1075
	deflateInit(&s, pack_compression_level);
1076 1077 1078 1079
	if (delta) {
		s.next_in = delta;
		s.avail_in = deltalen;
	} else {
1080 1081
		s.next_in = (void *)dat->buf;
		s.avail_in = dat->len;
1082 1083 1084 1085 1086 1087 1088 1089
	}
	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. */
1090
	if ((pack_size + 60 + s.total_out) > max_packsize
1091 1092 1093 1094
		|| (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;
1095
		cycle_packfile();
1096 1097 1098 1099 1100

		/* We cannot carry a delta into the new pack. */
		if (delta) {
			free(delta);
			delta = NULL;
1101 1102

			memset(&s, 0, sizeof(s));
1103
			deflateInit(&s, pack_compression_level);
1104 1105
			s.next_in = (void *)dat->buf;
			s.avail_in = dat->len;
1106 1107 1108 1109 1110
			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);
1111 1112 1113 1114 1115 1116 1117 1118
		}
	}

	e->type = type;
	e->pack_id = pack_id;
	e->offset = pack_size;
	object_count++;
	object_count_by_type[type]++;
1119 1120

	if (delta) {
1121 1122 1123
		unsigned long ofs = e->offset - last->offset;
		unsigned pos = sizeof(hdr) - 1;

1124
		delta_count_by_type[type]++;
1125
		e->depth = last->depth + 1;
1126 1127

		hdrlen = encode_header(OBJ_OFS_DELTA, deltalen, hdr);
1128
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1129 1130 1131 1132 1133
		pack_size += hdrlen;

		hdr[pos] = ofs & 127;
		while (ofs >>= 7)
			hdr[--pos] = 128 | (--ofs & 127);
1134
		write_or_die(pack_data->pack_fd, hdr + pos, sizeof(hdr) - pos);
1135
		pack_size += sizeof(hdr) - pos;
1136
	} else {
1137
		e->depth = 0;
1138
		hdrlen = encode_header(type, dat->len, hdr);
1139
		write_or_die(pack_data->pack_fd, hdr, hdrlen);
1140
		pack_size += hdrlen;
1141 1142
	}

1143
	write_or_die(pack_data->pack_fd, out, s.total_out);
1144
	pack_size += s.total_out;
1145 1146

	free(out);
1147
	free(delta);
1148
	if (last) {
P
Pierre Habouzit 已提交
1149 1150 1151
		if (last->no_swap) {
			last->data = *dat;
		} else {
1152
			strbuf_swap(&last->data, dat);
P
Pierre Habouzit 已提交
1153
		}
1154
		last->offset = e->offset;
1155
		last->depth = e->depth;
1156 1157 1158 1159
	}
	return 0;
}

1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177
/* 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.
 */
1178 1179 1180
static void *gfi_unpack_entry(
	struct object_entry *oe,
	unsigned long *sizep)
1181
{
1182
	enum object_type type;
1183
	struct packed_git *p = all_packs[oe->pack_id];
1184
	if (p == pack_data && p->pack_size < (pack_size + 20)) {
1185 1186 1187 1188 1189 1190 1191 1192
		/* 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.
		 */
1193
		close_pack_windows(p);
1194 1195 1196 1197 1198 1199 1200

		/* 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.
		 */
1201
		p->pack_size = pack_size + 20;
1202
	}
1203
	return unpack_entry(p, oe->offset, &type, sizep);
1204 1205
}

1206
static const char *get_mode(const char *str, uint16_t *modep)
1207 1208
{
	unsigned char c;
1209
	uint16_t mode = 0;
1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221

	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)
{
1222
	unsigned char* sha1 = root->versions[1].sha1;
1223 1224 1225 1226 1227 1228 1229
	struct object_entry *myoe;
	struct tree_content *t;
	unsigned long size;
	char *buf;
	const char *c;

	root->tree = t = new_tree_content(8);
1230
	if (is_null_sha1(sha1))
1231 1232
		return;

1233
	myoe = find_object(sha1);
1234
	if (myoe && myoe->pack_id != MAX_PACK_ID) {
1235
		if (myoe->type != OBJ_TREE)
1236
			die("Not a tree: %s", sha1_to_hex(sha1));
1237
		t->delta_depth = myoe->depth;
1238
		buf = gfi_unpack_entry(myoe, &size);
1239 1240
		if (!buf)
			die("Can't load tree %s", sha1_to_hex(sha1));
1241
	} else {
1242 1243 1244
		enum object_type type;
		buf = read_sha1_file(sha1, &type, &size);
		if (!buf || type != OBJ_TREE)
1245
			die("Can't load tree %s", sha1_to_hex(sha1));
1246 1247 1248 1249 1250 1251 1252
	}

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

		if (t->entry_count == t->entry_capacity)
1253
			root->tree = t = grow_tree_content(t, t->entry_count);
1254 1255 1256
		t->entries[t->entry_count++] = e;

		e->tree = NULL;
1257
		c = get_mode(c, &e->versions[1].mode);
1258
		if (!c)
1259 1260
			die("Corrupt mode in %s", sha1_to_hex(sha1));
		e->versions[0].mode = e->versions[1].mode;
1261
		e->name = to_atom(c, strlen(c));
1262
		c += e->name->str_len + 1;
1263 1264
		hashcpy(e->versions[0].sha1, (unsigned char*)c);
		hashcpy(e->versions[1].sha1, (unsigned char*)c);
1265 1266 1267 1268 1269
		c += 20;
	}
	free(buf);
}

1270
static int tecmp0 (const void *_a, const void *_b)
1271 1272 1273 1274
{
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
1275 1276
		a->name->str_dat, a->name->str_len, a->versions[0].mode,
		b->name->str_dat, b->name->str_len, b->versions[0].mode);
1277 1278
}

1279
static int tecmp1 (const void *_a, const void *_b)
1280
{
1281 1282 1283 1284 1285 1286 1287
	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);
}

1288
static void mktree(struct tree_content *t, int v, struct strbuf *b)
1289 1290
{
	size_t maxlen = 0;
1291 1292
	unsigned int i;

1293 1294 1295 1296
	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);
1297 1298

	for (i = 0; i < t->entry_count; i++) {
1299 1300
		if (t->entries[i]->versions[v].mode)
			maxlen += t->entries[i]->name->str_len + 34;
1301 1302
	}

1303 1304
	strbuf_reset(b);
	strbuf_grow(b, maxlen);
1305 1306
	for (i = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
1307 1308
		if (!e->versions[v].mode)
			continue;
1309 1310 1311
		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);
1312
	}
1313 1314 1315 1316 1317 1318
}

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
1319
	struct last_object lo = { STRBUF_INIT, 0, 0, /* no_swap */ 1 };
1320
	struct object_entry *le;
1321 1322 1323 1324 1325 1326 1327 1328 1329

	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]);
	}

1330
	le = find_object(root->versions[0].sha1);
P
Pierre Habouzit 已提交
1331
	if (S_ISDIR(root->versions[0].mode) && le && le->pack_id == pack_id) {
1332
		mktree(t, 0, &old_tree);
P
Pierre Habouzit 已提交
1333
		lo.data = old_tree;
1334
		lo.offset = le->offset;
1335 1336 1337
		lo.depth = t->delta_depth;
	}

1338 1339
	mktree(t, 1, &new_tree);
	store_object(OBJ_TREE, &new_tree, &lo, root->versions[1].sha1, 0);
1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353

	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;
1354 1355 1356 1357 1358 1359
}

static int tree_content_set(
	struct tree_entry *root,
	const char *p,
	const unsigned char *sha1,
1360 1361
	const uint16_t mode,
	struct tree_content *subtree)
1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372
{
	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);
1373 1374
	if (!n)
		die("Empty path component found in input");
1375 1376
	if (!slash1 && !S_ISDIR(mode) && subtree)
		die("Non-directories cannot have subtrees");
1377 1378 1379 1380 1381

	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) {
1382 1383
				if (!S_ISDIR(mode)
						&& e->versions[1].mode == mode
1384
						&& !hashcmp(e->versions[1].sha1, sha1))
1385
					return 0;
1386 1387
				e->versions[1].mode = mode;
				hashcpy(e->versions[1].sha1, sha1);
1388
				if (e->tree)
1389
					release_tree_content_recursive(e->tree);
1390
				e->tree = subtree;
1391
				hashclr(root->versions[1].sha1);
1392 1393
				return 1;
			}
1394
			if (!S_ISDIR(e->versions[1].mode)) {
1395
				e->tree = new_tree_content(8);
1396
				e->versions[1].mode = S_IFDIR;
1397 1398 1399
			}
			if (!e->tree)
				load_tree(e);
1400
			if (tree_content_set(e, slash1 + 1, sha1, mode, subtree)) {
1401
				hashclr(root->versions[1].sha1);
1402 1403 1404 1405 1406 1407 1408
				return 1;
			}
			return 0;
		}
	}

	if (t->entry_count == t->entry_capacity)
1409
		root->tree = t = grow_tree_content(t, t->entry_count);
1410
	e = new_tree_entry();
1411
	e->name = to_atom(p, n);
1412 1413
	e->versions[0].mode = 0;
	hashclr(e->versions[0].sha1);
1414 1415 1416
	t->entries[t->entry_count++] = e;
	if (slash1) {
		e->tree = new_tree_content(8);
1417
		e->versions[1].mode = S_IFDIR;
1418
		tree_content_set(e, slash1 + 1, sha1, mode, subtree);
1419
	} else {
1420
		e->tree = subtree;
1421 1422
		e->versions[1].mode = mode;
		hashcpy(e->versions[1].sha1, sha1);
1423
	}
1424
	hashclr(root->versions[1].sha1);
1425 1426 1427
	return 1;
}

1428 1429 1430 1431
static int tree_content_remove(
	struct tree_entry *root,
	const char *p,
	struct tree_entry *backup_leaf)
1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446
{
	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)) {
1447
			if (!slash1 || !S_ISDIR(e->versions[1].mode))
1448 1449 1450
				goto del_entry;
			if (!e->tree)
				load_tree(e);
1451
			if (tree_content_remove(e, slash1 + 1, backup_leaf)) {
1452 1453 1454 1455 1456 1457
				for (n = 0; n < e->tree->entry_count; n++) {
					if (e->tree->entries[n]->versions[1].mode) {
						hashclr(root->versions[1].sha1);
						return 1;
					}
				}
1458
				backup_leaf = NULL;
1459
				goto del_entry;
1460 1461 1462 1463 1464 1465 1466
			}
			return 0;
		}
	}
	return 0;

del_entry:
1467 1468 1469
	if (backup_leaf)
		memcpy(backup_leaf, e, sizeof(*backup_leaf));
	else if (e->tree)
1470
		release_tree_content_recursive(e->tree);
1471
	e->tree = NULL;
1472 1473 1474
	e->versions[1].mode = 0;
	hashclr(e->versions[1].sha1);
	hashclr(root->versions[1].sha1);
1475
	return 1;
1476 1477
}

1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514
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;
}

1515
static int update_branch(struct branch *b)
1516 1517
{
	static const char *msg = "fast-import";
1518 1519 1520
	struct ref_lock *lock;
	unsigned char old_sha1[20];

1521 1522
	if (is_null_sha1(b->sha1))
		return 0;
1523 1524
	if (read_ref(b->name, old_sha1))
		hashclr(old_sha1);
1525
	lock = lock_any_ref_for_update(b->name, old_sha1, 0);
1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537
	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);
		}

1538
		if (!in_merge_bases(old_cmit, &new_cmit, 1)) {
1539
			unlock_ref(lock);
1540
			warning("Not updating %s"
1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552
				" (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)
{
1553 1554 1555 1556
	unsigned int i;
	struct branch *b;

	for (i = 0; i < branch_table_sz; i++) {
1557 1558
		for (b = branch_table[i]; b; b = b->table_next_branch)
			failure |= update_branch(b);
1559 1560 1561
	}
}

1562
static void dump_tags(void)
1563 1564 1565 1566
{
	static const char *msg = "fast-import";
	struct tag *t;
	struct ref_lock *lock;
1567
	char ref_name[PATH_MAX];
1568 1569

	for (t = first_tag; t; t = t->next_tag) {
1570 1571
		sprintf(ref_name, "tags/%s", t->name);
		lock = lock_ref_sha1(ref_name, NULL);
1572
		if (!lock || write_ref_sha1(lock, t->sha1, msg) < 0)
1573
			failure |= error("Unable to update %s", ref_name);
1574 1575 1576
	}
}

1577
static void dump_marks_helper(FILE *f,
1578
	uintmax_t base,
1579 1580
	struct mark_set *m)
{
1581
	uintmax_t k;
1582 1583 1584 1585 1586 1587 1588 1589 1590
	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])
1591
				fprintf(f, ":%" PRIuMAX " %s\n", base + k,
1592 1593 1594 1595 1596
					sha1_to_hex(m->data.marked[k]->sha1));
		}
	}
}

1597
static void dump_marks(void)
1598
{
1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610
	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;
1611
	}
1612 1613 1614

	f = fdopen(mark_fd, "w");
	if (!f) {
1615
		int saved_errno = errno;
1616 1617
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1618
			mark_file, strerror(saved_errno));
1619
		return;
1620
	}
1621

B
Brandon Casey 已提交
1622
	/*
1623 1624
	 * 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
1625 1626 1627
	 * won't try to close() it.
	 */
	mark_lock.fd = -1;
1628 1629 1630

	dump_marks_helper(f, 0, marks);
	if (ferror(f) || fclose(f)) {
1631
		int saved_errno = errno;
1632 1633
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1634
			mark_file, strerror(saved_errno));
1635 1636 1637 1638
		return;
	}

	if (commit_lock_file(&mark_lock)) {
1639
		int saved_errno = errno;
1640 1641
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to commit marks file %s: %s",
1642
			mark_file, strerror(saved_errno));
1643 1644
		return;
	}
1645 1646
}

1647
static int read_next_command(void)
1648
{
1649 1650 1651 1652 1653 1654 1655
	static int stdin_eof = 0;

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

1656
	do {
1657
		if (unread_command_buf) {
1658
			unread_command_buf = 0;
1659 1660 1661
		} else {
			struct recent_command *rc;

1662
			strbuf_detach(&command_buf, NULL);
1663 1664 1665
			stdin_eof = strbuf_getline(&command_buf, stdin, '\n');
			if (stdin_eof)
				return EOF;
1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683

			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] == '#');
1684 1685

	return 0;
1686 1687
}

J
Junio C Hamano 已提交
1688
static void skip_optional_lf(void)
1689 1690 1691 1692 1693 1694
{
	int term_char = fgetc(stdin);
	if (term_char != '\n' && term_char != EOF)
		ungetc(term_char, stdin);
}

1695
static void parse_mark(void)
1696
{
1697
	if (!prefixcmp(command_buf.buf, "mark :")) {
1698
		next_mark = strtoumax(command_buf.buf + 6, NULL, 10);
1699 1700 1701
		read_next_command();
	}
	else
1702
		next_mark = 0;
1703 1704
}

1705
static void parse_data(struct strbuf *sb)
1706
{
1707
	strbuf_reset(sb);
1708

1709
	if (prefixcmp(command_buf.buf, "data "))
1710 1711
		die("Expected 'data n' command, found: %s", command_buf.buf);

1712
	if (!prefixcmp(command_buf.buf + 5, "<<")) {
1713
		char *term = xstrdup(command_buf.buf + 5 + 2);
1714 1715
		size_t term_len = command_buf.len - 5 - 2;

1716
		strbuf_detach(&command_buf, NULL);
1717
		for (;;) {
1718
			if (strbuf_getline(&command_buf, stdin, '\n') == EOF)
1719 1720 1721 1722
				die("EOF in data (terminator '%s' not found)", term);
			if (term_len == command_buf.len
				&& !strcmp(term, command_buf.buf))
				break;
1723 1724
			strbuf_addbuf(sb, &command_buf);
			strbuf_addch(sb, '\n');
1725 1726 1727 1728
		}
		free(term);
	}
	else {
1729 1730
		size_t n = 0, length;

1731
		length = strtoul(command_buf.buf + 5, NULL, 10);
1732

1733
		while (n < length) {
1734
			size_t s = strbuf_fread(sb, length - n, stdin);
1735
			if (!s && feof(stdin))
1736 1737
				die("EOF in data (%lu bytes remaining)",
					(unsigned long)(length - n));
1738 1739
			n += s;
		}
1740 1741
	}

1742
	skip_optional_lf();
1743 1744
}

1745 1746 1747
static int validate_raw_date(const char *src, char *result, int maxlen)
{
	const char *orig_src = src;
1748
	char *endp;
1749

1750 1751
	errno = 0;

1752
	strtoul(src, &endp, 10);
1753
	if (errno || endp == src || *endp != ' ')
1754 1755 1756 1757 1758 1759
		return -1;

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

1760
	strtoul(src + 1, &endp, 10);
1761
	if (errno || endp == src || *endp || (endp - orig_src) >= maxlen)
1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803
		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;
}

1804
static void parse_new_blob(void)
1805
{
P
Pierre Habouzit 已提交
1806
	static struct strbuf buf = STRBUF_INIT;
1807 1808

	read_next_command();
1809 1810
	parse_mark();
	parse_data(&buf);
P
Pierre Habouzit 已提交
1811
	store_object(OBJ_BLOB, &buf, &last_blob, NULL, next_mark);
1812 1813
}

1814
static void unload_one_branch(void)
1815
{
1816 1817
	while (cur_active_branches
		&& cur_active_branches >= max_active_branches) {
1818
		uintmax_t min_commit = ULONG_MAX;
1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835
		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;
		}
1836
		e->active = 0;
1837 1838
		e->active_next_branch = NULL;
		if (e->branch_tree.tree) {
1839
			release_tree_content_recursive(e->branch_tree.tree);
1840 1841 1842
			e->branch_tree.tree = NULL;
		}
		cur_active_branches--;
1843 1844 1845
	}
}

1846
static void load_branch(struct branch *b)
1847
{
1848
	load_tree(&b->branch_tree);
1849 1850 1851 1852 1853 1854 1855
	if (!b->active) {
		b->active = 1;
		b->active_next_branch = active_branches;
		active_branches = b;
		cur_active_branches++;
		branch_load_count++;
	}
1856 1857
}

1858
static void file_change_m(struct branch *b)
1859
{
1860
	const char *p = command_buf.buf + 2;
1861
	static struct strbuf uq = STRBUF_INIT;
1862
	const char *endp;
1863
	struct object_entry *oe = oe;
1864
	unsigned char sha1[20];
1865
	uint16_t mode, inline_data = 0;
1866

1867 1868 1869 1870
	p = get_mode(p, &mode);
	if (!p)
		die("Corrupt mode: %s", command_buf.buf);
	switch (mode) {
1871 1872 1873
	case 0644:
	case 0755:
		mode |= S_IFREG;
1874 1875
	case S_IFREG | 0644:
	case S_IFREG | 0755:
1876
	case S_IFLNK:
1877
	case S_IFGITLINK:
1878 1879 1880 1881 1882 1883
		/* ok */
		break;
	default:
		die("Corrupt mode: %s", command_buf.buf);
	}

1884 1885
	if (*p == ':') {
		char *x;
1886
		oe = find_mark(strtoumax(p + 1, &x, 10));
1887
		hashcpy(sha1, oe->sha1);
1888
		p = x;
1889
	} else if (!prefixcmp(p, "inline")) {
1890 1891
		inline_data = 1;
		p += 6;
1892 1893 1894 1895 1896 1897
	} else {
		if (get_sha1_hex(p, sha1))
			die("Invalid SHA1: %s", command_buf.buf);
		oe = find_object(sha1);
		p += 40;
	}
1898 1899 1900
	if (*p++ != ' ')
		die("Missing space after SHA1: %s", command_buf.buf);

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

1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921
	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
1922
		static struct strbuf buf = STRBUF_INIT;
1923

1924 1925 1926 1927
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
1928
		read_next_command();
1929
		parse_data(&buf);
P
Pierre Habouzit 已提交
1930
		store_object(OBJ_BLOB, &buf, &last_blob, sha1, 0);
1931
	} else if (oe) {
1932
		if (oe->type != OBJ_BLOB)
1933
			die("Not a blob (actually a %s): %s",
1934
				typename(oe->type), command_buf.buf);
1935
	} else {
1936 1937
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
1938
			die("Blob not found: %s", command_buf.buf);
1939
		if (type != OBJ_BLOB)
1940
			die("Not a blob (actually a %s): %s",
1941
			    typename(type), command_buf.buf);
1942
	}
1943

1944
	tree_content_set(&b->branch_tree, p, sha1, mode, NULL);
1945
}
1946

1947 1948
static void file_change_d(struct branch *b)
{
1949
	const char *p = command_buf.buf + 2;
1950
	static struct strbuf uq = STRBUF_INIT;
1951 1952
	const char *endp;

1953 1954
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
1955 1956
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
1957
		p = uq.buf;
1958
	}
1959
	tree_content_remove(&b->branch_tree, p, NULL);
1960 1961
}

1962
static void file_change_cr(struct branch *b, int rename)
1963 1964
{
	const char *s, *d;
1965 1966
	static struct strbuf s_uq = STRBUF_INIT;
	static struct strbuf d_uq = STRBUF_INIT;
1967 1968 1969 1970
	const char *endp;
	struct tree_entry leaf;

	s = command_buf.buf + 2;
1971 1972
	strbuf_reset(&s_uq);
	if (!unquote_c_style(&s_uq, s, &endp)) {
1973 1974
		if (*endp != ' ')
			die("Missing space after source: %s", command_buf.buf);
1975
	} else {
1976 1977 1978
		endp = strchr(s, ' ');
		if (!endp)
			die("Missing space after source: %s", command_buf.buf);
1979
		strbuf_add(&s_uq, s, endp - s);
1980
	}
1981
	s = s_uq.buf;
1982 1983 1984 1985 1986 1987

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

	d = endp;
1988 1989
	strbuf_reset(&d_uq);
	if (!unquote_c_style(&d_uq, d, &endp)) {
1990 1991
		if (*endp)
			die("Garbage after dest in: %s", command_buf.buf);
1992
		d = d_uq.buf;
1993 1994 1995
	}

	memset(&leaf, 0, sizeof(leaf));
1996 1997 1998 1999
	if (rename)
		tree_content_remove(&b->branch_tree, s, &leaf);
	else
		tree_content_get(&b->branch_tree, s, &leaf);
2000 2001 2002 2003 2004 2005 2006 2007
	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);
}

2008 2009 2010 2011 2012 2013 2014 2015
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);
}

2016
static void parse_from_commit(struct branch *b, char *buf, unsigned long size)
2017 2018 2019 2020 2021 2022 2023 2024 2025 2026
{
	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);
}

2027
static void parse_from_existing(struct branch *b)
2028 2029 2030 2031 2032 2033 2034 2035 2036 2037
{
	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);
2038
		parse_from_commit(b, buf, size);
2039 2040 2041 2042
		free(buf);
	}
}

2043
static int parse_from(struct branch *b)
2044
{
2045
	const char *from;
2046 2047
	struct branch *s;

2048
	if (prefixcmp(command_buf.buf, "from "))
2049
		return 0;
2050

2051 2052 2053 2054
	if (b->branch_tree.tree) {
		release_tree_content_recursive(b->branch_tree.tree);
		b->branch_tree.tree = NULL;
	}
2055 2056 2057 2058 2059 2060

	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) {
2061
		unsigned char *t = s->branch_tree.versions[1].sha1;
2062
		hashcpy(b->sha1, s->sha1);
2063 2064
		hashcpy(b->branch_tree.versions[0].sha1, t);
		hashcpy(b->branch_tree.versions[1].sha1, t);
2065
	} else if (*from == ':') {
2066
		uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2067 2068
		struct object_entry *oe = find_mark(idnum);
		if (oe->type != OBJ_COMMIT)
2069
			die("Mark :%" PRIuMAX " not a commit", idnum);
2070
		hashcpy(b->sha1, oe->sha1);
2071
		if (oe->pack_id != MAX_PACK_ID) {
2072
			unsigned long size;
2073
			char *buf = gfi_unpack_entry(oe, &size);
2074
			parse_from_commit(b, buf, size);
2075
			free(buf);
2076
		} else
2077
			parse_from_existing(b);
2078
	} else if (!get_sha1(from, b->sha1))
2079
		parse_from_existing(b);
2080
	else
2081 2082 2083
		die("Invalid ref name or SHA1 expression: %s", from);

	read_next_command();
2084
	return 1;
2085 2086
}

2087
static struct hash_list *parse_merge(unsigned int *count)
2088
{
2089
	struct hash_list *list = NULL, *n, *e = e;
2090
	const char *from;
2091 2092 2093
	struct branch *s;

	*count = 0;
2094
	while (!prefixcmp(command_buf.buf, "merge ")) {
2095 2096 2097 2098 2099 2100
		from = strchr(command_buf.buf, ' ') + 1;
		n = xmalloc(sizeof(*n));
		s = lookup_branch(from);
		if (s)
			hashcpy(n->sha1, s->sha1);
		else if (*from == ':') {
2101
			uintmax_t idnum = strtoumax(from + 1, NULL, 10);
2102 2103
			struct object_entry *oe = find_mark(idnum);
			if (oe->type != OBJ_COMMIT)
2104
				die("Mark :%" PRIuMAX " not a commit", idnum);
2105
			hashcpy(n->sha1, oe->sha1);
2106 2107 2108
		} 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
2109
				commit_type, &size, n->sha1);
2110 2111 2112 2113
			if (!buf || size < 46)
				die("Not a valid commit: %s", from);
			free(buf);
		} else
2114 2115 2116 2117 2118 2119 2120 2121
			die("Invalid ref name or SHA1 expression: %s", from);

		n->next = NULL;
		if (list)
			e->next = n;
		else
			list = n;
		e = n;
2122
		(*count)++;
2123 2124 2125 2126 2127
		read_next_command();
	}
	return list;
}

2128
static void parse_new_commit(void)
2129
{
2130
	static struct strbuf msg = STRBUF_INIT;
2131 2132 2133 2134
	struct branch *b;
	char *sp;
	char *author = NULL;
	char *committer = NULL;
2135 2136
	struct hash_list *merge_list = NULL;
	unsigned int merge_count;
2137 2138 2139 2140

	/* Obtain the branch name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	b = lookup_branch(sp);
2141
	if (!b)
2142
		b = new_branch(sp);
2143 2144

	read_next_command();
2145
	parse_mark();
2146
	if (!prefixcmp(command_buf.buf, "author ")) {
2147
		author = parse_ident(command_buf.buf + 7);
2148 2149
		read_next_command();
	}
2150
	if (!prefixcmp(command_buf.buf, "committer ")) {
2151
		committer = parse_ident(command_buf.buf + 10);
2152 2153 2154 2155
		read_next_command();
	}
	if (!committer)
		die("Expected committer but didn't get one");
2156
	parse_data(&msg);
2157
	read_next_command();
2158 2159
	parse_from(b);
	merge_list = parse_merge(&merge_count);
2160 2161

	/* ensure the branch is active/loaded */
2162
	if (!b->branch_tree.tree || !max_active_branches) {
2163 2164 2165
		unload_one_branch();
		load_branch(b);
	}
2166

2167
	/* file_change* */
2168
	while (command_buf.len > 0) {
2169
		if (!prefixcmp(command_buf.buf, "M "))
2170
			file_change_m(b);
2171
		else if (!prefixcmp(command_buf.buf, "D "))
2172
			file_change_d(b);
2173
		else if (!prefixcmp(command_buf.buf, "R "))
2174 2175 2176
			file_change_cr(b, 1);
		else if (!prefixcmp(command_buf.buf, "C "))
			file_change_cr(b, 0);
2177 2178
		else if (!strcmp("deleteall", command_buf.buf))
			file_change_deleteall(b);
2179 2180 2181 2182
		else {
			unread_command_buf = 1;
			break;
		}
2183 2184
		if (read_next_command() == EOF)
			break;
2185 2186
	}

2187
	/* build the tree and the commit */
2188
	store_tree(&b->branch_tree);
2189 2190
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
2191 2192 2193

	strbuf_reset(&new_data);
	strbuf_addf(&new_data, "tree %s\n",
2194
		sha1_to_hex(b->branch_tree.versions[1].sha1));
2195
	if (!is_null_sha1(b->sha1))
2196
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(b->sha1));
2197 2198
	while (merge_list) {
		struct hash_list *next = merge_list->next;
2199
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(merge_list->sha1));
2200 2201 2202
		free(merge_list);
		merge_list = next;
	}
2203 2204 2205 2206 2207 2208
	strbuf_addf(&new_data,
		"author %s\n"
		"committer %s\n"
		"\n",
		author ? author : committer, committer);
	strbuf_addbuf(&new_data, &msg);
2209
	free(author);
2210 2211
	free(committer);

2212
	if (!store_object(OBJ_COMMIT, &new_data, NULL, b->sha1, next_mark))
2213
		b->pack_id = pack_id;
2214
	b->last_commit = object_count_by_type[OBJ_COMMIT];
2215 2216
}

2217
static void parse_new_tag(void)
2218
{
2219
	static struct strbuf msg = STRBUF_INIT;
2220 2221 2222 2223 2224
	char *sp;
	const char *from;
	char *tagger;
	struct branch *s;
	struct tag *t;
2225
	uintmax_t from_mark = 0;
2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240
	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 ... */
2241
	if (prefixcmp(command_buf.buf, "from "))
2242 2243 2244 2245
		die("Expected from command, got %s", command_buf.buf);
	from = strchr(command_buf.buf, ' ') + 1;
	s = lookup_branch(from);
	if (s) {
2246
		hashcpy(sha1, s->sha1);
2247
	} else if (*from == ':') {
2248
		struct object_entry *oe;
2249
		from_mark = strtoumax(from + 1, NULL, 10);
2250
		oe = find_mark(from_mark);
2251
		if (oe->type != OBJ_COMMIT)
2252
			die("Mark :%" PRIuMAX " not a commit", from_mark);
2253
		hashcpy(sha1, oe->sha1);
2254 2255 2256 2257 2258
	} else if (!get_sha1(from, sha1)) {
		unsigned long size;
		char *buf;

		buf = read_object_with_reference(sha1,
2259
			commit_type, &size, sha1);
2260 2261 2262 2263 2264 2265 2266 2267
		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 ... */
2268 2269 2270 2271 2272
	if (!prefixcmp(command_buf.buf, "tagger ")) {
		tagger = parse_ident(command_buf.buf + 7);
		read_next_command();
	} else
		tagger = NULL;
2273 2274

	/* tag payload/message */
2275
	parse_data(&msg);
2276 2277

	/* build the tag object */
2278
	strbuf_reset(&new_data);
2279

2280
	strbuf_addf(&new_data,
2281 2282 2283 2284 2285 2286 2287 2288
		    "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');
2289
	strbuf_addbuf(&new_data, &msg);
2290 2291
	free(tagger);

2292
	if (store_object(OBJ_TAG, &new_data, NULL, t->sha1, 0))
2293 2294 2295
		t->pack_id = MAX_PACK_ID;
	else
		t->pack_id = pack_id;
2296 2297
}

2298
static void parse_reset_branch(void)
2299 2300 2301 2302 2303 2304 2305 2306
{
	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) {
2307 2308 2309
		hashclr(b->sha1);
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
2310 2311 2312 2313 2314
		if (b->branch_tree.tree) {
			release_tree_content_recursive(b->branch_tree.tree);
			b->branch_tree.tree = NULL;
		}
	}
2315 2316 2317
	else
		b = new_branch(sp);
	read_next_command();
2318
	parse_from(b);
2319
	if (command_buf.len > 0)
2320
		unread_command_buf = 1;
2321 2322
}

2323
static void parse_checkpoint(void)
2324
{
2325 2326 2327 2328 2329 2330
	if (object_count) {
		cycle_packfile();
		dump_branches();
		dump_tags();
		dump_marks();
	}
2331
	skip_optional_lf();
2332 2333
}

2334
static void parse_progress(void)
2335
{
2336
	fwrite(command_buf.buf, 1, command_buf.len, stdout);
2337 2338 2339 2340 2341
	fputc('\n', stdout);
	fflush(stdout);
	skip_optional_lf();
}

2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369
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;
2370
			e->offset = 1; /* just not zero! */
2371 2372 2373 2374 2375 2376
		}
		insert_mark(mark, e);
	}
	fclose(f);
}

2377
static int git_pack_config(const char *k, const char *v, void *cb)
2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394
{
	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;
	}
2395
	return git_default_config(k, v, cb);
2396 2397
}

2398
static const char fast_import_usage[] =
S
Stephan Beyer 已提交
2399
"git fast-import [--date-format=f] [--max-pack-size=n] [--depth=n] [--active-branches=n] [--export-marks=marks.file]";
2400

2401 2402
int main(int argc, const char **argv)
{
2403
	unsigned int i, show_stats = 1;
2404

2405 2406
	git_extract_argv0_path(argv[0]);

2407
	setup_git_directory();
2408
	git_config(git_pack_config, NULL);
2409 2410 2411
	if (!pack_compression_seen && core_compression_seen)
		pack_compression_level = core_compression_level;

2412
	alloc_objects(object_entry_alloc);
2413
	strbuf_init(&command_buf, 0);
2414 2415 2416 2417
	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));
2418

2419 2420 2421 2422 2423
	for (i = 1; i < argc; i++) {
		const char *a = argv[i];

		if (*a != '-' || !strcmp(a, "--"))
			break;
2424
		else if (!prefixcmp(a, "--date-format=")) {
2425 2426 2427 2428 2429 2430 2431 2432 2433 2434
			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);
		}
2435
		else if (!prefixcmp(a, "--max-pack-size="))
2436
			max_packsize = strtoumax(a + 16, NULL, 0) * 1024 * 1024;
2437
		else if (!prefixcmp(a, "--depth=")) {
2438
			max_depth = strtoul(a + 8, NULL, 0);
2439 2440 2441
			if (max_depth > MAX_DEPTH)
				die("--depth cannot exceed %u", MAX_DEPTH);
		}
2442
		else if (!prefixcmp(a, "--active-branches="))
2443
			max_active_branches = strtoul(a + 18, NULL, 0);
2444 2445
		else if (!prefixcmp(a, "--import-marks="))
			import_marks(a + 15);
2446
		else if (!prefixcmp(a, "--export-marks="))
2447
			mark_file = a + 15;
2448
		else if (!prefixcmp(a, "--export-pack-edges=")) {
2449 2450 2451 2452 2453 2454
			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"))
2455
			force_update = 1;
2456 2457 2458 2459
		else if (!strcmp(a, "--quiet"))
			show_stats = 0;
		else if (!strcmp(a, "--stats"))
			show_stats = 1;
2460 2461 2462
		else
			die("unknown option %s", a);
	}
2463
	if (i != argc)
2464 2465
		usage(fast_import_usage);

2466 2467 2468 2469 2470
	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;

2471
	prepare_packed_git();
2472
	start_packfile();
2473
	set_die_routine(die_nicely);
2474 2475
	while (read_next_command() != EOF) {
		if (!strcmp("blob", command_buf.buf))
2476
			parse_new_blob();
2477
		else if (!prefixcmp(command_buf.buf, "commit "))
2478
			parse_new_commit();
2479
		else if (!prefixcmp(command_buf.buf, "tag "))
2480
			parse_new_tag();
2481
		else if (!prefixcmp(command_buf.buf, "reset "))
2482
			parse_reset_branch();
2483
		else if (!strcmp("checkpoint", command_buf.buf))
2484
			parse_checkpoint();
2485
		else if (!prefixcmp(command_buf.buf, "progress "))
2486
			parse_progress();
2487 2488
		else
			die("Unsupported command: %s", command_buf.buf);
2489
	}
2490
	end_packfile();
2491

2492
	dump_branches();
2493
	dump_tags();
2494
	unkeep_all_packs();
2495
	dump_marks();
2496

2497 2498 2499
	if (pack_edges)
		fclose(pack_edges);

2500 2501 2502 2503 2504 2505 2506 2507 2508
	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");
2509 2510 2511 2512 2513 2514
		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]);
2515
		fprintf(stderr, "Total branches:  %10lu (%10lu loads     )\n", branch_count, branch_load_count);
2516
		fprintf(stderr, "      marks:     %10" PRIuMAX " (%10" PRIuMAX " unique    )\n", (((uintmax_t)1) << marks->shift) * 1024, marks_set_count);
2517
		fprintf(stderr, "      atoms:     %10u\n", atom_cnt);
2518
		fprintf(stderr, "Memory total:    %10" PRIuMAX " KiB\n", (total_allocd + alloc_count*sizeof(struct object_entry))/1024);
2519
		fprintf(stderr, "       pools:    %10lu KiB\n", (unsigned long)(total_allocd/1024));
2520
		fprintf(stderr, "     objects:    %10" PRIuMAX " KiB\n", (alloc_count*sizeof(struct object_entry))/1024);
2521 2522 2523 2524 2525
		fprintf(stderr, "---------------------------------------------------------------------\n");
		pack_report();
		fprintf(stderr, "---------------------------------------------------------------------\n");
		fprintf(stderr, "\n");
	}
2526

2527
	return failure ? 1 : 0;
2528
}