fast-import.c 87.7 KB
Newer Older
1
/*
2
(See Documentation/git-fast-import.txt for maintained documentation.)
3 4 5 6 7
Format of STDIN stream:

  stream ::= cmd*;

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

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

20
  new_commit ::= 'commit' sp ref_str lf
21
    mark?
22 23
    ('author' (sp name)? sp '<' email '>' sp when lf)?
    'committer' (sp name)? sp '<' email '>' sp when lf
24
    commit_msg
25 26
    ('from' sp commit-ish lf)?
    ('merge' sp commit-ish lf)*
D
David Barr 已提交
27
    (file_change | ls)*
28
    lf?;
29
  commit_msg ::= data;
30

D
David Barr 已提交
31 32
  ls ::= 'ls' sp '"' quoted(path) '"' lf;

33 34 35 36 37 38
  file_change ::= file_clr
    | file_del
    | file_rnm
    | file_cpy
    | file_obm
    | file_inm;
39
  file_clr ::= 'deleteall' lf;
40
  file_del ::= 'D' sp path_str lf;
41
  file_rnm ::= 'R' sp path_str sp path_str lf;
42
  file_cpy ::= 'C' sp path_str sp path_str lf;
43 44 45
  file_obm ::= 'M' sp mode sp (hexsha1 | idnum) sp path_str lf;
  file_inm ::= 'M' sp mode sp 'inline' sp path_str lf
    data;
46 47
  note_obm ::= 'N' sp (hexsha1 | idnum) sp commit-ish lf;
  note_inm ::= 'N' sp 'inline' sp commit-ish lf
48
    data;
49 50

  new_tag ::= 'tag' sp tag_str lf
51
    'from' sp commit-ish lf
52
    ('tagger' (sp name)? sp '<' email '>' sp when lf)?
53 54 55
    tag_msg;
  tag_msg ::= data;

56
  reset_branch ::= 'reset' sp ref_str lf
57
    ('from' sp commit-ish lf)?
58
    lf?;
59

60
  checkpoint ::= 'checkpoint' lf
61
    lf?;
62

63 64 65
  progress ::= 'progress' sp not_lf* lf
    lf?;

66 67 68
     # 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
69
     # idnum can be updated in the future to a new object by issuing
70
     # a new mark directive with the old idnum.
71
     #
72
  mark ::= 'mark' sp idnum lf;
73
  data ::= (delimited_data | exact_data)
74
    lf?;
75 76 77 78 79 80

    # 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)*
81
    delim lf;
82 83

     # note: declen indicates the length of binary_data in bytes.
M
Mike Ralphson 已提交
84
     # declen does not include the lf preceding the binary data.
85
     #
86 87
  exact_data ::= 'data' sp declen lf
    binary_data;
88 89 90

     # note: quoted strings are C-style quoting supporting \c for
     # common escapes of 'c' (e..g \n, \t, \\, \") or \nnn where nnn
91
     # is the signed byte value in octal.  Note that the only
92 93
     # characters which must actually be escaped to protect the
     # stream formatting is: \, " and LF.  Otherwise these values
94
     # are UTF8.
95
     #
96
  commit-ish  ::= (ref_str | hexsha1 | sha1exp_str | idnum);
97 98 99
  ref_str     ::= ref;
  sha1exp_str ::= sha1exp;
  tag_str     ::= tag;
100
  path_str    ::= path    | '"' quoted(path)    '"' ;
101 102
  mode        ::= '100644' | '644'
                | '100755' | '755'
J
Junio C Hamano 已提交
103
                | '120000'
104
                ;
105 106

  declen ::= # unsigned 32 bit value, ascii base10 notation;
107
  bigint ::= # unsigned integer value, ascii base10 notation;
108
  binary_data ::= # file content, not interpreted;
109

110 111 112 113
  when         ::= raw_when | rfc2822_when;
  raw_when     ::= ts sp tz;
  rfc2822_when ::= # Valid RFC 2822 date and time;

114 115
  sp ::= # ASCII space character;
  lf ::= # ASCII newline (LF) character;
116 117

     # note: a colon (':') must precede the numerical value assigned to
118
     # an idnum.  This is to distinguish it from a ref or tag name as
119
     # GIT does not permit ':' in ref or tag strings.
120
     #
121
  idnum   ::= ':' bigint;
122 123 124
  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";
125 126
  sha1exp ::= # Any valid GIT SHA1 expression;
  hexsha1 ::= # SHA1 in hexadecimal format;
127 128

     # note: name and email are UTF8 strings, however name must not
129
     # contain '<' or lf and email must not contain any of the
130
     # following: '<', '>', lf.
131
     #
132
  name  ::= # valid GIT author/committer name;
133
  email ::= # valid GIT author/committer email;
134 135
  ts    ::= # time since the epoch in seconds, ascii base10 notation;
  tz    ::= # GIT style timezone;
136

D
David Barr 已提交
137
     # note: comments, ls and cat requests may appear anywhere
138 139 140
     # in the input, except within a data command.  Any form
     # of the data command always escapes the related input
     # from comment processing.
141 142
     #
     # In case it is not clear, the '#' that starts the comment
143
     # must be the first character on that line (an lf
M
Mike Ralphson 已提交
144
     # preceded it).
145
     #
D
David Barr 已提交
146

147
  cat_blob ::= 'cat-blob' sp (hexsha1 | idnum) lf;
D
David Barr 已提交
148
  ls_tree  ::= 'ls' sp (hexsha1 | idnum) sp path_str lf;
149

150 151
  comment ::= '#' not_lf* lf;
  not_lf  ::= # Any byte that is not ASCII newline (LF);
152 153
*/

154 155 156 157
#include "builtin.h"
#include "cache.h"
#include "object.h"
#include "blob.h"
158
#include "tree.h"
159
#include "commit.h"
160 161
#include "delta.h"
#include "pack.h"
162
#include "refs.h"
163
#include "csum-file.h"
164
#include "quote.h"
165
#include "exec_cmd.h"
166
#include "dir.h"
167

168 169
#define PACK_ID_BITS 16
#define MAX_PACK_ID ((1<<PACK_ID_BITS)-1)
170 171
#define DEPTH_BITS 13
#define MAX_DEPTH ((1<<DEPTH_BITS)-1)
172

173 174 175 176 177
/*
 * We abuse the setuid bit on directories to mean "do not delta".
 */
#define NO_DELTA S_ISUID

178
struct object_entry {
179
	struct pack_idx_entry idx;
180
	struct object_entry *next;
181 182 183
	uint32_t type : TYPE_BITS,
		pack_id : PACK_ID_BITS,
		depth : DEPTH_BITS;
184 185
};

186
struct object_entry_pool {
187
	struct object_entry_pool *next_pool;
188 189
	struct object_entry *next_free;
	struct object_entry *end;
190
	struct object_entry entries[FLEX_ARRAY]; /* more */
191 192
};

193
struct mark_set {
194 195 196 197
	union {
		struct object_entry *marked[1024];
		struct mark_set *sets[1024];
	} data;
198
	unsigned int shift;
199 200
};

201
struct last_object {
P
Pierre Habouzit 已提交
202
	struct strbuf data;
203
	off_t offset;
204
	unsigned int depth;
P
Pierre Habouzit 已提交
205
	unsigned no_swap : 1;
206 207
};

208
struct mem_pool {
209 210 211
	struct mem_pool *next_pool;
	char *next_free;
	char *end;
212
	uintmax_t space[FLEX_ARRAY]; /* more */
213 214
};

215
struct atom_str {
216
	struct atom_str *next_atom;
217
	unsigned short str_len;
218 219 220 221
	char str_dat[FLEX_ARRAY]; /* more */
};

struct tree_content;
222
struct tree_entry {
223
	struct tree_content *tree;
224
	struct atom_str *name;
225
	struct tree_entry_ms {
226
		uint16_t mode;
227 228
		unsigned char sha1[20];
	} versions[2];
229 230
};

231
struct tree_content {
232 233
	unsigned int entry_capacity; /* must match avail_tree_content */
	unsigned int entry_count;
234
	unsigned int delta_depth;
235 236 237
	struct tree_entry *entries[FLEX_ARRAY]; /* more */
};

238
struct avail_tree_content {
239 240
	unsigned int entry_capacity; /* must match tree_content */
	struct avail_tree_content *next_avail;
241 242
};

243
struct branch {
244 245
	struct branch *table_next_branch;
	struct branch *active_next_branch;
246
	const char *name;
247
	struct tree_entry branch_tree;
248
	uintmax_t last_commit;
249
	uintmax_t num_notes;
250
	unsigned active : 1;
251
	unsigned delete : 1;
252
	unsigned pack_id : PACK_ID_BITS;
253
	unsigned char sha1[20];
254 255
};

256
struct tag {
257 258
	struct tag *next_tag;
	const char *name;
259
	unsigned int pack_id;
260 261 262
	unsigned char sha1[20];
};

263
struct hash_list {
264 265 266
	struct hash_list *next;
	unsigned char sha1[20];
};
267

268 269 270
typedef enum {
	WHENSPEC_RAW = 1,
	WHENSPEC_RFC2822,
271
	WHENSPEC_NOW
272 273
} whenspec_type;

274
struct recent_command {
275 276 277 278 279
	struct recent_command *prev;
	struct recent_command *next;
	char *buf;
};

280
/* Configured limits on output */
281
static unsigned long max_depth = 10;
282
static off_t max_packsize;
283
static int force_update;
284 285
static int pack_compression_level = Z_DEFAULT_COMPRESSION;
static int pack_compression_seen;
286 287 288 289 290 291 292

/* 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];
293
static uintmax_t delta_count_attempts_by_type[1 << TYPE_BITS];
294
static unsigned long object_count;
295
static unsigned long branch_count;
296
static unsigned long branch_load_count;
297
static int failure;
298
static FILE *pack_edges;
299
static unsigned int show_stats = 1;
S
Sverre Rabbelier 已提交
300
static int global_argc;
301
static char **global_argv;
302

303 304 305 306 307
/* 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;

308
/* Atom management */
309 310 311 312 313
static unsigned int atom_table_sz = 4451;
static unsigned int atom_cnt;
static struct atom_str **atom_table;

/* The .pack file being generated */
314
static struct pack_idx_option pack_idx_opts;
315
static unsigned int pack_id;
316
static struct sha1file *pack_file;
317
static struct packed_git *pack_data;
318
static struct packed_git **all_packs;
319
static off_t pack_size;
320 321

/* Table of objects we've written. */
322
static unsigned int object_entry_alloc = 5000;
323 324
static struct object_entry_pool *blocks;
static struct object_entry *object_table[1 << 16];
325
static struct mark_set *marks;
326 327
static const char *export_marks_file;
static const char *import_marks_file;
328
static int import_marks_file_from_stream;
329
static int import_marks_file_ignore_missing;
330
static int relative_marks_paths;
331 332

/* Our last blob */
P
Pierre Habouzit 已提交
333
static struct last_object last_blob = { STRBUF_INIT, 0, 0, 0 };
334 335 336 337 338 339

/* 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;
340 341
static struct strbuf old_tree = STRBUF_INIT;
static struct strbuf new_tree = STRBUF_INIT;
342

343
/* Branch data */
344 345 346
static unsigned long max_active_branches = 5;
static unsigned long cur_active_branches;
static unsigned long branch_table_sz = 1039;
347 348 349
static struct branch **branch_table;
static struct branch *active_branches;

350 351 352 353
/* Tag data */
static struct tag *first_tag;
static struct tag *last_tag;

354
/* Input stream parsing */
355
static whenspec_type whenspec = WHENSPEC_RAW;
356
static struct strbuf command_buf = STRBUF_INIT;
357
static int unread_command_buf;
358 359 360 361
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;
362
static uintmax_t next_mark;
363
static struct strbuf new_data = STRBUF_INIT;
364
static int seen_data_command;
365
static int require_explicit_termination;
366

367 368 369
/* Signal handling */
static volatile sig_atomic_t checkpoint_requested;

370 371 372
/* Where to write output of cat-blob commands */
static int cat_blob_fd = STDOUT_FILENO;

S
Sverre Rabbelier 已提交
373
static void parse_argv(void);
374 375
static void parse_cat_blob(const char *p);
static void parse_ls(const char *p, struct branch *b);
376

377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402
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);
}

403 404
static void dump_marks_helper(FILE *, uintmax_t, struct mark_set *);

405
static void write_crash_report(const char *err)
406
{
407
	char *loc = git_path("fast_import_crash_%"PRIuMAX, (uintmax_t) getpid());
408 409 410
	FILE *rpt = fopen(loc, "w");
	struct branch *b;
	unsigned long lu;
411
	struct recent_command *rc;
412 413 414 415 416 417 418 419 420

	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");
421 422
	fprintf(rpt, "    fast-import process: %"PRIuMAX"\n", (uintmax_t) getpid());
	fprintf(rpt, "    parent process     : %"PRIuMAX"\n", (uintmax_t) getppid());
423 424 425 426
	fprintf(rpt, "    at %s\n", show_date(time(NULL), 0, DATE_LOCAL));
	fputc('\n', rpt);

	fputs("fatal: ", rpt);
427
	fputs(err, rpt);
428 429
	fputc('\n', rpt);

430 431 432 433 434 435 436 437 438 439 440 441
	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);
	}

442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462
	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);
	}

463 464 465 466 467 468 469 470 471 472 473 474 475
	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);
		}
	}

476 477 478
	fputc('\n', rpt);
	fputs("Marks\n", rpt);
	fputs("-----\n", rpt);
479 480
	if (export_marks_file)
		fprintf(rpt, "  exported to %s\n", export_marks_file);
481 482 483
	else
		dump_marks_helper(rpt, 0, marks);

484 485 486 487 488 489
	fputc('\n', rpt);
	fputs("-------------------\n", rpt);
	fputs("END OF CRASH REPORT\n", rpt);
	fclose(rpt);
}

490 491 492 493
static void end_packfile(void);
static void unkeep_all_packs(void);
static void dump_marks(void);

494 495 496
static NORETURN void die_nicely(const char *err, va_list params)
{
	static int zombie;
497
	char message[2 * PATH_MAX];
498

499
	vsnprintf(message, sizeof(message), err, params);
500
	fputs("fatal: ", stderr);
501
	fputs(message, stderr);
502 503 504 505
	fputc('\n', stderr);

	if (!zombie) {
		zombie = 1;
506
		write_crash_report(message);
507 508 509
		end_packfile();
		unkeep_all_packs();
		dump_marks();
510 511 512
	}
	exit(128);
}
513

514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539
#ifndef SIGUSR1	/* Windows, for example */

static void set_checkpoint_signal(void)
{
}

#else

static void checkpoint_signal(int signo)
{
	checkpoint_requested = 1;
}

static void set_checkpoint_signal(void)
{
	struct sigaction sa;

	memset(&sa, 0, sizeof(sa));
	sa.sa_handler = checkpoint_signal;
	sigemptyset(&sa.sa_mask);
	sa.sa_flags = SA_RESTART;
	sigaction(SIGUSR1, &sa, NULL);
}

#endif

540
static void alloc_objects(unsigned int cnt)
541
{
542
	struct object_entry_pool *b;
543

544
	b = xmalloc(sizeof(struct object_entry_pool)
545
		+ cnt * sizeof(struct object_entry));
546
	b->next_pool = blocks;
547 548 549 550 551
	b->next_free = b->entries;
	b->end = b->entries + cnt;
	blocks = b;
	alloc_count += cnt;
}
552

553
static struct object_entry *new_object(unsigned char *sha1)
554
{
555
	struct object_entry *e;
556

557
	if (blocks->next_free == blocks->end)
558
		alloc_objects(object_entry_alloc);
559

560
	e = blocks->next_free++;
561
	hashcpy(e->idx.sha1, sha1);
562
	return e;
563 564
}

565
static struct object_entry *find_object(unsigned char *sha1)
566 567 568 569
{
	unsigned int h = sha1[0] << 8 | sha1[1];
	struct object_entry *e;
	for (e = object_table[h]; e; e = e->next)
570
		if (!hashcmp(sha1, e->idx.sha1))
571 572 573 574
			return e;
	return NULL;
}

575
static struct object_entry *insert_object(unsigned char *sha1)
576 577
{
	unsigned int h = sha1[0] << 8 | sha1[1];
578
	struct object_entry *e = object_table[h];
579 580

	while (e) {
581
		if (!hashcmp(sha1, e->idx.sha1))
582 583 584 585 586
			return e;
		e = e->next;
	}

	e = new_object(sha1);
587
	e->next = object_table[h];
588
	e->idx.offset = 0;
589
	object_table[h] = e;
590 591
	return e;
}
592

593 594 595 596 597 598 599 600
static unsigned int hc_str(const char *s, size_t len)
{
	unsigned int r = 0;
	while (len-- > 0)
		r = r * 31 + *s++;
	return r;
}

601
static void *pool_alloc(size_t len)
602 603 604 605
{
	struct mem_pool *p;
	void *r;

606 607 608 609
	/* round up to a 'uintmax_t' alignment */
	if (len & (sizeof(uintmax_t) - 1))
		len += sizeof(uintmax_t) - (len & (sizeof(uintmax_t) - 1));

610 611 612 613 614 615 616 617 618 619 620 621
	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;
622
		p->next_free = (char *) p->space;
623 624 625 626 627 628 629 630 631
		p->end = p->next_free + mem_pool_alloc;
		mem_pool = p;
	}

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

632
static void *pool_calloc(size_t count, size_t size)
633 634 635 636 637 638 639
{
	size_t len = count * size;
	void *r = pool_alloc(len);
	memset(r, 0, len);
	return r;
}

640
static char *pool_strdup(const char *s)
641 642 643 644 645 646
{
	char *r = pool_alloc(strlen(s) + 1);
	strcpy(r, s);
	return r;
}

647
static void insert_mark(uintmax_t idnum, struct object_entry *oe)
648 649 650 651 652 653 654 655 656
{
	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) {
657
		uintmax_t i = idnum >> s->shift;
658 659 660 661 662 663 664 665 666 667 668 669
		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;
}

670
static struct object_entry *find_mark(uintmax_t idnum)
671
{
672
	uintmax_t orig_idnum = idnum;
673 674 675 676
	struct mark_set *s = marks;
	struct object_entry *oe = NULL;
	if ((idnum >> s->shift) < 1024) {
		while (s && s->shift) {
677
			uintmax_t i = idnum >> s->shift;
678 679 680 681 682 683 684
			idnum -= i << s->shift;
			s = s->data.sets[i];
		}
		if (s)
			oe = s->data.marked[idnum];
	}
	if (!oe)
685
		die("mark :%" PRIuMAX " not declared", orig_idnum);
686 687 688
	return oe;
}

689
static struct atom_str *to_atom(const char *s, unsigned short len)
690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707
{
	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;
}

708
static struct branch *lookup_branch(const char *name)
709 710 711 712 713 714 715 716 717 718
{
	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;
}

719
static struct branch *new_branch(const char *name)
720 721
{
	unsigned int hc = hc_str(name, strlen(name)) % branch_table_sz;
722
	struct branch *b = lookup_branch(name);
723 724 725

	if (b)
		die("Invalid attempt to create duplicate branch: %s", name);
726
	if (check_refname_format(name, REFNAME_ALLOW_ONELEVEL))
727
		die("Branch name doesn't conform to GIT standards: %s", name);
728 729 730 731

	b = pool_calloc(1, sizeof(struct branch));
	b->name = pool_strdup(name);
	b->table_next_branch = branch_table[hc];
732 733
	b->branch_tree.versions[0].mode = S_IFDIR;
	b->branch_tree.versions[1].mode = S_IFDIR;
734
	b->num_notes = 0;
735
	b->active = 0;
736
	b->pack_id = MAX_PACK_ID;
737 738 739 740 741 742 743 744 745 746 747
	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;
}

748
static struct tree_content *new_tree_content(unsigned int cnt)
749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770
{
	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;
771
	t->delta_depth = 0;
772 773 774 775 776 777 778 779
	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);
780 781 782 783 784 785
	f->next_avail = avail_tree_table[hc];
	avail_tree_table[hc] = f;
}

static void release_tree_content_recursive(struct tree_content *t)
{
786 787 788
	unsigned int i;
	for (i = 0; i < t->entry_count; i++)
		release_tree_entry(t->entries[i]);
789
	release_tree_content(t);
790 791
}

792
static struct tree_content *grow_tree_content(
793 794 795 796 797
	struct tree_content *t,
	int amt)
{
	struct tree_content *r = new_tree_content(t->entry_count + amt);
	r->entry_count = t->entry_count;
798
	r->delta_depth = t->delta_depth;
799 800 801 802 803
	memcpy(r->entries,t->entries,t->entry_count*sizeof(t->entries[0]));
	release_tree_content(t);
	return r;
}

804
static struct tree_entry *new_tree_entry(void)
805 806 807 808 809
{
	struct tree_entry *e;

	if (!avail_tree_entry) {
		unsigned int n = tree_entry_alloc;
810
		total_allocd += n * sizeof(struct tree_entry);
811
		avail_tree_entry = e = xmalloc(n * sizeof(struct tree_entry));
812
		while (n-- > 1) {
813 814 815
			*((void**)e) = e + 1;
			e++;
		}
816
		*((void**)e) = NULL;
817 818 819 820 821 822 823 824 825 826
	}

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

static void release_tree_entry(struct tree_entry *e)
{
	if (e->tree)
827
		release_tree_content_recursive(e->tree);
828 829 830 831
	*((void**)e) = avail_tree_entry;
	avail_tree_entry = e;
}

832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856
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;
}

857
static void start_packfile(void)
858
{
859
	static char tmp_file[PATH_MAX];
860
	struct packed_git *p;
861
	struct pack_header hdr;
862
	int pack_fd;
863

864
	pack_fd = odb_mkstemp(tmp_file, sizeof(tmp_file),
865
			      "pack/tmp_pack_XXXXXX");
866 867
	p = xcalloc(1, sizeof(*p) + strlen(tmp_file) + 2);
	strcpy(p->pack_name, tmp_file);
868
	p->pack_fd = pack_fd;
869
	p->do_not_close = 1;
870
	pack_file = sha1fd(pack_fd, p->pack_name);
871 872 873 874

	hdr.hdr_signature = htonl(PACK_SIGNATURE);
	hdr.hdr_version = htonl(2);
	hdr.hdr_entries = 0;
875
	sha1write(pack_file, &hdr, sizeof(hdr));
876 877

	pack_data = p;
878 879
	pack_size = sizeof(hdr);
	object_count = 0;
880

881
	REALLOC_ARRAY(all_packs, pack_id + 1);
882
	all_packs[pack_id] = p;
883 884
}

885
static const char *create_index(void)
886
{
887 888 889
	const char *tmpfile;
	struct pack_idx_entry **idx, **c, **last;
	struct object_entry *e;
890 891
	struct object_entry_pool *o;

892 893
	/* Build the table of object IDs. */
	idx = xmalloc(object_count * sizeof(*idx));
894 895
	c = idx;
	for (o = blocks; o; o = o->next_pool)
896 897
		for (e = o->next_free; e-- != o->entries;)
			if (pack_id == e->pack_id)
898
				*c++ = &e->idx;
899
	last = idx + object_count;
900 901
	if (c != last)
		die("internal consistency error creating the index");
902

903
	tmpfile = write_idx_file(NULL, idx, object_count, &pack_idx_opts, pack_data->sha1);
904
	free(idx);
905 906 907
	return tmpfile;
}

908
static char *keep_pack(const char *curr_index_name)
909 910
{
	static char name[PATH_MAX];
911
	static const char *keep_msg = "fast-import";
912 913
	int keep_fd;

914
	keep_fd = odb_pack_keep(name, sizeof(name), pack_data->sha1);
915
	if (keep_fd < 0)
916
		die_errno("cannot create keep file");
917 918
	write_or_die(keep_fd, keep_msg, strlen(keep_msg));
	if (close(keep_fd))
919
		die_errno("failed to write keep file");
920 921 922 923 924 925 926 927 928 929

	snprintf(name, sizeof(name), "%s/pack/pack-%s.pack",
		 get_object_directory(), sha1_to_hex(pack_data->sha1));
	if (move_temp_to_file(pack_data->pack_name, name))
		die("cannot store pack file");

	snprintf(name, sizeof(name), "%s/pack/pack-%s.idx",
		 get_object_directory(), sha1_to_hex(pack_data->sha1));
	if (move_temp_to_file(curr_index_name, name))
		die("cannot store index file");
930
	free((void *)curr_index_name);
931 932 933
	return name;
}

934
static void unkeep_all_packs(void)
935 936 937 938 939 940 941 942
{
	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));
943
		unlink_or_warn(name);
944
	}
945 946
}

947
static void end_packfile(void)
948
{
949 950
	if (!pack_data)
		return;
951

952
	clear_delta_base_cache();
953
	if (object_count) {
954
		struct packed_git *new_p;
955
		unsigned char cur_pack_sha1[20];
956
		char *idx_name;
957 958 959
		int i;
		struct branch *b;
		struct tag *t;
960

961
		close_pack_windows(pack_data);
962
		sha1close(pack_file, cur_pack_sha1, 0);
963
		fixup_pack_header_footer(pack_data->pack_fd, pack_data->sha1,
964
				    pack_data->pack_name, object_count,
965
				    cur_pack_sha1, pack_size);
966
		close(pack_data->pack_fd);
967
		idx_name = keep_pack(create_index());
968

M
Mike Ralphson 已提交
969
		/* Register the packfile with core git's machinery. */
970 971 972
		new_p = add_packed_git(idx_name, strlen(idx_name), 1);
		if (!new_p)
			die("core git rejected index %s", idx_name);
973
		all_packs[pack_id] = new_p;
974
		install_packed_git(new_p);
975 976

		/* Print the boundary */
977 978 979 980 981 982 983
		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));
				}
984
			}
985 986 987 988 989 990
			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);
991 992 993
		}

		pack_id++;
994
	}
995
	else {
996 997
		close(pack_data->pack_fd);
		unlink_or_warn(pack_data->pack_name);
998
	}
999 1000
	free(pack_data);
	pack_data = NULL;
1001 1002

	/* We can't carry a delta across packfiles. */
P
Pierre Habouzit 已提交
1003
	strbuf_release(&last_blob.data);
1004 1005
	last_blob.offset = 0;
	last_blob.depth = 0;
1006 1007
}

1008
static void cycle_packfile(void)
1009 1010 1011 1012 1013
{
	end_packfile();
	start_packfile();
}

1014 1015
static int store_object(
	enum object_type type,
1016
	struct strbuf *dat,
1017
	struct last_object *last,
1018
	unsigned char *sha1out,
1019
	uintmax_t mark)
1020 1021
{
	void *out, *delta;
1022 1023 1024
	struct object_entry *e;
	unsigned char hdr[96];
	unsigned char sha1[20];
1025
	unsigned long hdrlen, deltalen;
1026
	git_SHA_CTX c;
1027
	git_zstream s;
1028

1029
	hdrlen = sprintf((char *)hdr,"%s %lu", typename(type),
1030
		(unsigned long)dat->len) + 1;
1031 1032 1033 1034
	git_SHA1_Init(&c);
	git_SHA1_Update(&c, hdr, hdrlen);
	git_SHA1_Update(&c, dat->buf, dat->len);
	git_SHA1_Final(sha1, &c);
1035
	if (sha1out)
1036
		hashcpy(sha1out, sha1);
1037 1038

	e = insert_object(sha1);
1039 1040
	if (mark)
		insert_mark(mark, e);
1041
	if (e->idx.offset) {
1042
		duplicate_count_by_type[type]++;
1043
		return 1;
1044 1045 1046
	} else if (find_sha1_pack(sha1, packed_git)) {
		e->type = type;
		e->pack_id = MAX_PACK_ID;
1047
		e->idx.offset = 1; /* just not zero! */
1048 1049
		duplicate_count_by_type[type]++;
		return 1;
1050
	}
1051

1052
	if (last && last->data.buf && last->depth < max_depth && dat->len > 20) {
1053
		delta_count_attempts_by_type[type]++;
P
Pierre Habouzit 已提交
1054
		delta = diff_delta(last->data.buf, last->data.len,
1055
			dat->buf, dat->len,
1056
			&deltalen, dat->len - 20);
1057 1058
	} else
		delta = NULL;
1059 1060

	memset(&s, 0, sizeof(s));
1061
	git_deflate_init(&s, pack_compression_level);
1062 1063 1064 1065
	if (delta) {
		s.next_in = delta;
		s.avail_in = deltalen;
	} else {
1066 1067
		s.next_in = (void *)dat->buf;
		s.avail_in = dat->len;
1068
	}
J
Junio C Hamano 已提交
1069
	s.avail_out = git_deflate_bound(&s, s.avail_in);
1070
	s.next_out = out = xmalloc(s.avail_out);
1071 1072 1073
	while (git_deflate(&s, Z_FINISH) == Z_OK)
		; /* nothing */
	git_deflate_end(&s);
1074 1075

	/* Determine if we should auto-checkpoint. */
1076
	if ((max_packsize && (pack_size + 60 + s.total_out) > max_packsize)
1077 1078 1079 1080
		|| (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;
1081
		cycle_packfile();
1082 1083 1084 1085 1086

		/* We cannot carry a delta into the new pack. */
		if (delta) {
			free(delta);
			delta = NULL;
1087 1088

			memset(&s, 0, sizeof(s));
1089
			git_deflate_init(&s, pack_compression_level);
1090 1091
			s.next_in = (void *)dat->buf;
			s.avail_in = dat->len;
J
Junio C Hamano 已提交
1092
			s.avail_out = git_deflate_bound(&s, s.avail_in);
1093
			s.next_out = out = xrealloc(out, s.avail_out);
1094 1095 1096
			while (git_deflate(&s, Z_FINISH) == Z_OK)
				; /* nothing */
			git_deflate_end(&s);
1097 1098 1099 1100 1101
		}
	}

	e->type = type;
	e->pack_id = pack_id;
1102
	e->idx.offset = pack_size;
1103 1104
	object_count++;
	object_count_by_type[type]++;
1105

1106 1107
	crc32_begin(pack_file);

1108
	if (delta) {
1109
		off_t ofs = e->idx.offset - last->offset;
1110 1111
		unsigned pos = sizeof(hdr) - 1;

1112
		delta_count_by_type[type]++;
1113
		e->depth = last->depth + 1;
1114

1115
		hdrlen = encode_in_pack_object_header(OBJ_OFS_DELTA, deltalen, hdr);
1116
		sha1write(pack_file, hdr, hdrlen);
1117 1118 1119 1120 1121
		pack_size += hdrlen;

		hdr[pos] = ofs & 127;
		while (ofs >>= 7)
			hdr[--pos] = 128 | (--ofs & 127);
1122
		sha1write(pack_file, hdr + pos, sizeof(hdr) - pos);
1123
		pack_size += sizeof(hdr) - pos;
1124
	} else {
1125
		e->depth = 0;
1126
		hdrlen = encode_in_pack_object_header(type, dat->len, hdr);
1127
		sha1write(pack_file, hdr, hdrlen);
1128
		pack_size += hdrlen;
1129 1130
	}

1131
	sha1write(pack_file, out, s.total_out);
1132
	pack_size += s.total_out;
1133

1134 1135
	e->idx.crc32 = crc32_end(pack_file);

1136
	free(out);
1137
	free(delta);
1138
	if (last) {
P
Pierre Habouzit 已提交
1139 1140 1141
		if (last->no_swap) {
			last->data = *dat;
		} else {
1142
			strbuf_swap(&last->data, dat);
P
Pierre Habouzit 已提交
1143
		}
1144
		last->offset = e->idx.offset;
1145
		last->depth = e->depth;
1146 1147 1148 1149
	}
	return 0;
}

1150
static void truncate_pack(struct sha1file_checkpoint *checkpoint)
1151
{
1152
	if (sha1file_truncate(pack_file, checkpoint))
1153
		die_errno("cannot truncate pack to skip duplicate");
1154
	pack_size = checkpoint->offset;
1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166
}

static void stream_blob(uintmax_t len, unsigned char *sha1out, uintmax_t mark)
{
	size_t in_sz = 64 * 1024, out_sz = 64 * 1024;
	unsigned char *in_buf = xmalloc(in_sz);
	unsigned char *out_buf = xmalloc(out_sz);
	struct object_entry *e;
	unsigned char sha1[20];
	unsigned long hdrlen;
	off_t offset;
	git_SHA_CTX c;
1167
	git_zstream s;
1168
	struct sha1file_checkpoint checkpoint;
1169 1170 1171
	int status = Z_OK;

	/* Determine if we should auto-checkpoint. */
1172
	if ((max_packsize && (pack_size + 60 + len) > max_packsize)
1173 1174 1175
		|| (pack_size + 60 + len) < pack_size)
		cycle_packfile();

1176 1177
	sha1file_checkpoint(pack_file, &checkpoint);
	offset = checkpoint.offset;
1178

1179 1180 1181 1182 1183 1184 1185
	hdrlen = snprintf((char *)out_buf, out_sz, "blob %" PRIuMAX, len) + 1;
	if (out_sz <= hdrlen)
		die("impossibly large object header");

	git_SHA1_Init(&c);
	git_SHA1_Update(&c, out_buf, hdrlen);

1186 1187
	crc32_begin(pack_file);

1188
	memset(&s, 0, sizeof(s));
1189
	git_deflate_init(&s, pack_compression_level);
1190

1191
	hdrlen = encode_in_pack_object_header(OBJ_BLOB, len, out_buf);
1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210
	if (out_sz <= hdrlen)
		die("impossibly large object header");

	s.next_out = out_buf + hdrlen;
	s.avail_out = out_sz - hdrlen;

	while (status != Z_STREAM_END) {
		if (0 < len && !s.avail_in) {
			size_t cnt = in_sz < len ? in_sz : (size_t)len;
			size_t n = fread(in_buf, 1, cnt, stdin);
			if (!n && feof(stdin))
				die("EOF in data (%" PRIuMAX " bytes remaining)", len);

			git_SHA1_Update(&c, in_buf, n);
			s.next_in = in_buf;
			s.avail_in = n;
			len -= n;
		}

1211
		status = git_deflate(&s, len ? 0 : Z_FINISH);
1212 1213 1214

		if (!s.avail_out || status == Z_STREAM_END) {
			size_t n = s.next_out - out_buf;
1215
			sha1write(pack_file, out_buf, n);
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229
			pack_size += n;
			s.next_out = out_buf;
			s.avail_out = out_sz;
		}

		switch (status) {
		case Z_OK:
		case Z_BUF_ERROR:
		case Z_STREAM_END:
			continue;
		default:
			die("unexpected deflate failure: %d", status);
		}
	}
1230
	git_deflate_end(&s);
1231 1232 1233 1234 1235 1236 1237 1238 1239 1240
	git_SHA1_Final(sha1, &c);

	if (sha1out)
		hashcpy(sha1out, sha1);

	e = insert_object(sha1);

	if (mark)
		insert_mark(mark, e);

1241
	if (e->idx.offset) {
1242
		duplicate_count_by_type[OBJ_BLOB]++;
1243
		truncate_pack(&checkpoint);
1244 1245 1246 1247

	} else if (find_sha1_pack(sha1, packed_git)) {
		e->type = OBJ_BLOB;
		e->pack_id = MAX_PACK_ID;
1248
		e->idx.offset = 1; /* just not zero! */
1249
		duplicate_count_by_type[OBJ_BLOB]++;
1250
		truncate_pack(&checkpoint);
1251 1252 1253 1254 1255

	} else {
		e->depth = 0;
		e->type = OBJ_BLOB;
		e->pack_id = pack_id;
1256
		e->idx.offset = offset;
1257
		e->idx.crc32 = crc32_end(pack_file);
1258 1259 1260 1261 1262 1263 1264 1265
		object_count++;
		object_count_by_type[OBJ_BLOB]++;
	}

	free(in_buf);
	free(out_buf);
}

1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283
/* 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.
 */
1284 1285 1286
static void *gfi_unpack_entry(
	struct object_entry *oe,
	unsigned long *sizep)
1287
{
1288
	enum object_type type;
1289
	struct packed_git *p = all_packs[oe->pack_id];
1290
	if (p == pack_data && p->pack_size < (pack_size + 20)) {
1291 1292 1293 1294 1295 1296 1297 1298
		/* 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.
		 */
1299
		close_pack_windows(p);
1300
		sha1flush(pack_file);
1301 1302 1303 1304 1305 1306 1307

		/* 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.
		 */
1308
		p->pack_size = pack_size + 20;
1309
	}
1310
	return unpack_entry(p, oe->idx.offset, &type, sizep);
1311 1312
}

1313
static const char *get_mode(const char *str, uint16_t *modep)
1314 1315
{
	unsigned char c;
1316
	uint16_t mode = 0;
1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328

	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)
{
1329
	unsigned char *sha1 = root->versions[1].sha1;
1330 1331 1332 1333 1334 1335 1336
	struct object_entry *myoe;
	struct tree_content *t;
	unsigned long size;
	char *buf;
	const char *c;

	root->tree = t = new_tree_content(8);
1337
	if (is_null_sha1(sha1))
1338 1339
		return;

1340
	myoe = find_object(sha1);
1341
	if (myoe && myoe->pack_id != MAX_PACK_ID) {
1342
		if (myoe->type != OBJ_TREE)
1343
			die("Not a tree: %s", sha1_to_hex(sha1));
1344
		t->delta_depth = myoe->depth;
1345
		buf = gfi_unpack_entry(myoe, &size);
1346 1347
		if (!buf)
			die("Can't load tree %s", sha1_to_hex(sha1));
1348
	} else {
1349 1350 1351
		enum object_type type;
		buf = read_sha1_file(sha1, &type, &size);
		if (!buf || type != OBJ_TREE)
1352
			die("Can't load tree %s", sha1_to_hex(sha1));
1353 1354 1355 1356 1357 1358 1359
	}

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

		if (t->entry_count == t->entry_capacity)
1360
			root->tree = t = grow_tree_content(t, t->entry_count);
1361 1362 1363
		t->entries[t->entry_count++] = e;

		e->tree = NULL;
1364
		c = get_mode(c, &e->versions[1].mode);
1365
		if (!c)
1366 1367
			die("Corrupt mode in %s", sha1_to_hex(sha1));
		e->versions[0].mode = e->versions[1].mode;
1368
		e->name = to_atom(c, strlen(c));
1369
		c += e->name->str_len + 1;
1370 1371
		hashcpy(e->versions[0].sha1, (unsigned char *)c);
		hashcpy(e->versions[1].sha1, (unsigned char *)c);
1372 1373 1374 1375 1376
		c += 20;
	}
	free(buf);
}

1377
static int tecmp0 (const void *_a, const void *_b)
1378 1379 1380 1381
{
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
1382 1383
		a->name->str_dat, a->name->str_len, a->versions[0].mode,
		b->name->str_dat, b->name->str_len, b->versions[0].mode);
1384 1385
}

1386
static int tecmp1 (const void *_a, const void *_b)
1387
{
1388 1389 1390 1391 1392 1393 1394
	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);
}

1395
static void mktree(struct tree_content *t, int v, struct strbuf *b)
1396 1397
{
	size_t maxlen = 0;
1398 1399
	unsigned int i;

1400 1401 1402 1403
	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);
1404 1405

	for (i = 0; i < t->entry_count; i++) {
1406 1407
		if (t->entries[i]->versions[v].mode)
			maxlen += t->entries[i]->name->str_len + 34;
1408 1409
	}

1410 1411
	strbuf_reset(b);
	strbuf_grow(b, maxlen);
1412 1413
	for (i = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
1414 1415
		if (!e->versions[v].mode)
			continue;
1416 1417 1418
		strbuf_addf(b, "%o %s%c",
			(unsigned int)(e->versions[v].mode & ~NO_DELTA),
			e->name->str_dat, '\0');
1419
		strbuf_add(b, e->versions[v].sha1, 20);
1420
	}
1421 1422 1423 1424
}

static void store_tree(struct tree_entry *root)
{
1425
	struct tree_content *t;
1426
	unsigned int i, j, del;
P
Pierre Habouzit 已提交
1427
	struct last_object lo = { STRBUF_INIT, 0, 0, /* no_swap */ 1 };
1428
	struct object_entry *le = NULL;
1429 1430 1431 1432

	if (!is_null_sha1(root->versions[1].sha1))
		return;

1433 1434 1435 1436
	if (!root->tree)
		load_tree(root);
	t = root->tree;

1437 1438 1439 1440 1441
	for (i = 0; i < t->entry_count; i++) {
		if (t->entries[i]->tree)
			store_tree(t->entries[i]);
	}

1442 1443
	if (!(root->versions[0].mode & NO_DELTA))
		le = find_object(root->versions[0].sha1);
P
Pierre Habouzit 已提交
1444
	if (S_ISDIR(root->versions[0].mode) && le && le->pack_id == pack_id) {
1445
		mktree(t, 0, &old_tree);
P
Pierre Habouzit 已提交
1446
		lo.data = old_tree;
1447
		lo.offset = le->idx.offset;
1448 1449 1450
		lo.depth = t->delta_depth;
	}

1451 1452
	mktree(t, 1, &new_tree);
	store_object(OBJ_TREE, &new_tree, &lo, root->versions[1].sha1, 0);
1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466

	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;
1467 1468
}

1469 1470 1471 1472 1473 1474 1475 1476
static void tree_content_replace(
	struct tree_entry *root,
	const unsigned char *sha1,
	const uint16_t mode,
	struct tree_content *newtree)
{
	if (!S_ISDIR(mode))
		die("Root cannot be a non-directory");
1477
	hashclr(root->versions[0].sha1);
1478 1479 1480 1481 1482 1483
	hashcpy(root->versions[1].sha1, sha1);
	if (root->tree)
		release_tree_content_recursive(root->tree);
	root->tree = newtree;
}

1484 1485 1486 1487
static int tree_content_set(
	struct tree_entry *root,
	const char *p,
	const unsigned char *sha1,
1488 1489
	const uint16_t mode,
	struct tree_content *subtree)
1490
{
1491
	struct tree_content *t;
1492 1493 1494 1495
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

1496 1497
	slash1 = strchrnul(p, '/');
	n = slash1 - p;
1498 1499
	if (!n)
		die("Empty path component found in input");
1500
	if (!*slash1 && !S_ISDIR(mode) && subtree)
1501
		die("Non-directories cannot have subtrees");
1502

1503 1504 1505
	if (!root->tree)
		load_tree(root);
	t = root->tree;
1506 1507
	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
1508
		if (e->name->str_len == n && !strncmp_icase(p, e->name->str_dat, n)) {
1509
			if (!*slash1) {
1510 1511
				if (!S_ISDIR(mode)
						&& e->versions[1].mode == mode
1512
						&& !hashcmp(e->versions[1].sha1, sha1))
1513
					return 0;
1514 1515
				e->versions[1].mode = mode;
				hashcpy(e->versions[1].sha1, sha1);
1516
				if (e->tree)
1517
					release_tree_content_recursive(e->tree);
1518
				e->tree = subtree;
1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535

				/*
				 * We need to leave e->versions[0].sha1 alone
				 * to avoid modifying the preimage tree used
				 * when writing out the parent directory.
				 * But after replacing the subdir with a
				 * completely different one, it's not a good
				 * delta base any more, and besides, we've
				 * thrown away the tree entries needed to
				 * make a delta against it.
				 *
				 * So let's just explicitly disable deltas
				 * for the subtree.
				 */
				if (S_ISDIR(e->versions[0].mode))
					e->versions[0].mode |= NO_DELTA;

1536
				hashclr(root->versions[1].sha1);
1537 1538
				return 1;
			}
1539
			if (!S_ISDIR(e->versions[1].mode)) {
1540
				e->tree = new_tree_content(8);
1541
				e->versions[1].mode = S_IFDIR;
1542 1543 1544
			}
			if (!e->tree)
				load_tree(e);
1545
			if (tree_content_set(e, slash1 + 1, sha1, mode, subtree)) {
1546
				hashclr(root->versions[1].sha1);
1547 1548 1549 1550 1551 1552 1553
				return 1;
			}
			return 0;
		}
	}

	if (t->entry_count == t->entry_capacity)
1554
		root->tree = t = grow_tree_content(t, t->entry_count);
1555
	e = new_tree_entry();
1556
	e->name = to_atom(p, n);
1557 1558
	e->versions[0].mode = 0;
	hashclr(e->versions[0].sha1);
1559
	t->entries[t->entry_count++] = e;
1560
	if (*slash1) {
1561
		e->tree = new_tree_content(8);
1562
		e->versions[1].mode = S_IFDIR;
1563
		tree_content_set(e, slash1 + 1, sha1, mode, subtree);
1564
	} else {
1565
		e->tree = subtree;
1566 1567
		e->versions[1].mode = mode;
		hashcpy(e->versions[1].sha1, sha1);
1568
	}
1569
	hashclr(root->versions[1].sha1);
1570 1571 1572
	return 1;
}

1573 1574 1575
static int tree_content_remove(
	struct tree_entry *root,
	const char *p,
1576 1577
	struct tree_entry *backup_leaf,
	int allow_root)
1578
{
1579
	struct tree_content *t;
1580 1581 1582 1583
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

1584 1585
	slash1 = strchrnul(p, '/');
	n = slash1 - p;
1586

1587 1588
	if (!root->tree)
		load_tree(root);
1589 1590 1591 1592 1593 1594

	if (!*p && allow_root) {
		e = root;
		goto del_entry;
	}

1595
	t = root->tree;
1596 1597
	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
1598
		if (e->name->str_len == n && !strncmp_icase(p, e->name->str_dat, n)) {
1599
			if (*slash1 && !S_ISDIR(e->versions[1].mode))
1600 1601 1602 1603 1604 1605 1606
				/*
				 * If p names a file in some subdirectory, and a
				 * file or symlink matching the name of the
				 * parent directory of p exists, then p cannot
				 * exist and need not be deleted.
				 */
				return 1;
1607
			if (!*slash1 || !S_ISDIR(e->versions[1].mode))
1608 1609 1610
				goto del_entry;
			if (!e->tree)
				load_tree(e);
1611
			if (tree_content_remove(e, slash1 + 1, backup_leaf, 0)) {
1612 1613 1614 1615 1616 1617
				for (n = 0; n < e->tree->entry_count; n++) {
					if (e->tree->entries[n]->versions[1].mode) {
						hashclr(root->versions[1].sha1);
						return 1;
					}
				}
1618
				backup_leaf = NULL;
1619
				goto del_entry;
1620 1621 1622 1623 1624 1625 1626
			}
			return 0;
		}
	}
	return 0;

del_entry:
1627 1628 1629
	if (backup_leaf)
		memcpy(backup_leaf, e, sizeof(*backup_leaf));
	else if (e->tree)
1630
		release_tree_content_recursive(e->tree);
1631
	e->tree = NULL;
1632 1633 1634
	e->versions[1].mode = 0;
	hashclr(e->versions[1].sha1);
	hashclr(root->versions[1].sha1);
1635
	return 1;
1636 1637
}

1638 1639 1640
static int tree_content_get(
	struct tree_entry *root,
	const char *p,
1641 1642
	struct tree_entry *leaf,
	int allow_root)
1643
{
1644
	struct tree_content *t;
1645 1646 1647 1648
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

1649 1650
	slash1 = strchrnul(p, '/');
	n = slash1 - p;
1651
	if (!n && !allow_root)
1652
		die("Empty path component found in input");
1653

1654 1655
	if (!root->tree)
		load_tree(root);
1656 1657 1658 1659 1660 1661

	if (!n) {
		e = root;
		goto found_entry;
	}

1662
	t = root->tree;
1663 1664
	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
1665
		if (e->name->str_len == n && !strncmp_icase(p, e->name->str_dat, n)) {
1666
			if (!*slash1)
1667
				goto found_entry;
1668 1669 1670 1671
			if (!S_ISDIR(e->versions[1].mode))
				return 0;
			if (!e->tree)
				load_tree(e);
1672
			return tree_content_get(e, slash1 + 1, leaf, 0);
1673 1674 1675
		}
	}
	return 0;
1676 1677 1678 1679 1680 1681 1682 1683

found_entry:
	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;
1684 1685
}

1686
static int update_branch(struct branch *b)
1687 1688
{
	static const char *msg = "fast-import";
1689
	struct ref_transaction *transaction;
1690
	unsigned char old_sha1[20];
1691
	struct strbuf err = STRBUF_INIT;
1692 1693 1694

	if (read_ref(b->name, old_sha1))
		hashclr(old_sha1);
1695 1696 1697 1698 1699
	if (is_null_sha1(b->sha1)) {
		if (b->delete)
			delete_ref(b->name, old_sha1, 0);
		return 0;
	}
1700 1701 1702 1703 1704
	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);
1705
		if (!old_cmit || !new_cmit)
1706 1707
			return error("Branch %s is missing commits.", b->name);

1708
		if (!in_merge_bases(old_cmit, new_cmit)) {
1709
			warning("Not updating %s"
1710 1711 1712 1713 1714
				" (new tip %s does not contain %s)",
				b->name, sha1_to_hex(b->sha1), sha1_to_hex(old_sha1));
			return -1;
		}
	}
1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726
	transaction = ref_transaction_begin(&err);
	if (!transaction ||
	    ref_transaction_update(transaction, b->name, b->sha1, old_sha1,
				   0, 1, &err) ||
	    ref_transaction_commit(transaction, msg, &err)) {
		ref_transaction_free(transaction);
		error("%s", err.buf);
		strbuf_release(&err);
		return -1;
	}
	ref_transaction_free(transaction);
	strbuf_release(&err);
1727 1728 1729 1730 1731
	return 0;
}

static void dump_branches(void)
{
1732 1733 1734 1735
	unsigned int i;
	struct branch *b;

	for (i = 0; i < branch_table_sz; i++) {
1736 1737
		for (b = branch_table[i]; b; b = b->table_next_branch)
			failure |= update_branch(b);
1738 1739 1740
	}
}

1741
static void dump_tags(void)
1742 1743 1744
{
	static const char *msg = "fast-import";
	struct tag *t;
1745 1746 1747
	struct strbuf ref_name = STRBUF_INIT;
	struct strbuf err = STRBUF_INIT;
	struct ref_transaction *transaction;
1748

1749 1750 1751 1752 1753
	transaction = ref_transaction_begin(&err);
	if (!transaction) {
		failure |= error("%s", err.buf);
		goto cleanup;
	}
1754
	for (t = first_tag; t; t = t->next_tag) {
1755 1756 1757 1758 1759 1760 1761 1762
		strbuf_reset(&ref_name);
		strbuf_addf(&ref_name, "refs/tags/%s", t->name);

		if (ref_transaction_update(transaction, ref_name.buf, t->sha1,
					   NULL, 0, 0, &err)) {
			failure |= error("%s", err.buf);
			goto cleanup;
		}
1763
	}
1764 1765 1766 1767 1768 1769 1770
	if (ref_transaction_commit(transaction, msg, &err))
		failure |= error("%s", err.buf);

 cleanup:
	ref_transaction_free(transaction);
	strbuf_release(&ref_name);
	strbuf_release(&err);
1771 1772
}

1773
static void dump_marks_helper(FILE *f,
1774
	uintmax_t base,
1775 1776
	struct mark_set *m)
{
1777
	uintmax_t k;
1778 1779 1780
	if (m->shift) {
		for (k = 0; k < 1024; k++) {
			if (m->data.sets[k])
1781
				dump_marks_helper(f, base + (k << m->shift),
1782 1783 1784 1785 1786
					m->data.sets[k]);
		}
	} else {
		for (k = 0; k < 1024; k++) {
			if (m->data.marked[k])
1787
				fprintf(f, ":%" PRIuMAX " %s\n", base + k,
1788
					sha1_to_hex(m->data.marked[k]->idx.sha1));
1789 1790 1791 1792
		}
	}
}

1793
static void dump_marks(void)
1794
{
1795 1796 1797 1798
	static struct lock_file mark_lock;
	int mark_fd;
	FILE *f;

1799
	if (!export_marks_file)
1800 1801
		return;

1802
	mark_fd = hold_lock_file_for_update(&mark_lock, export_marks_file, 0);
1803 1804
	if (mark_fd < 0) {
		failure |= error("Unable to write marks file %s: %s",
1805
			export_marks_file, strerror(errno));
1806
		return;
1807
	}
1808 1809 1810

	f = fdopen(mark_fd, "w");
	if (!f) {
1811
		int saved_errno = errno;
1812 1813
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1814
			export_marks_file, strerror(saved_errno));
1815
		return;
1816
	}
1817

B
Brandon Casey 已提交
1818
	/*
1819 1820
	 * 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
1821 1822 1823
	 * won't try to close() it.
	 */
	mark_lock.fd = -1;
1824 1825 1826

	dump_marks_helper(f, 0, marks);
	if (ferror(f) || fclose(f)) {
1827
		int saved_errno = errno;
1828 1829
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1830
			export_marks_file, strerror(saved_errno));
1831 1832 1833 1834 1835
		return;
	}

	if (commit_lock_file(&mark_lock)) {
		failure |= error("Unable to commit marks file %s: %s",
1836
			export_marks_file, strerror(errno));
1837 1838
		return;
	}
1839 1840
}

1841 1842 1843 1844
static void read_marks(void)
{
	char line[512];
	FILE *f = fopen(import_marks_file, "r");
1845 1846 1847 1848 1849
	if (f)
		;
	else if (import_marks_file_ignore_missing && errno == ENOENT)
		return; /* Marks file does not exist */
	else
1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862
		die_errno("cannot read '%s'", import_marks_file);
	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
1863
			|| *end != ' ' || get_sha1_hex(end + 1, sha1))
1864 1865 1866 1867 1868 1869 1870 1871 1872
			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;
1873
			e->idx.offset = 1; /* just not zero! */
1874 1875 1876 1877 1878 1879 1880
		}
		insert_mark(mark, e);
	}
	fclose(f);
}


1881
static int read_next_command(void)
1882
{
1883 1884 1885 1886 1887 1888 1889
	static int stdin_eof = 0;

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

1890
	for (;;) {
1891 1892
		const char *p;

1893
		if (unread_command_buf) {
1894
			unread_command_buf = 0;
1895 1896 1897
		} else {
			struct recent_command *rc;

1898
			strbuf_detach(&command_buf, NULL);
1899 1900 1901
			stdin_eof = strbuf_getline(&command_buf, stdin, '\n');
			if (stdin_eof)
				return EOF;
1902

1903
			if (!seen_data_command
1904 1905
				&& !starts_with(command_buf.buf, "feature ")
				&& !starts_with(command_buf.buf, "option ")) {
S
Sverre Rabbelier 已提交
1906
				parse_argv();
1907 1908
			}

1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924
			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;
		}
1925 1926
		if (skip_prefix(command_buf.buf, "cat-blob ", &p)) {
			parse_cat_blob(p);
1927 1928 1929 1930 1931 1932
			continue;
		}
		if (command_buf.buf[0] == '#')
			continue;
		return 0;
	}
1933 1934
}

J
Junio C Hamano 已提交
1935
static void skip_optional_lf(void)
1936 1937 1938 1939 1940 1941
{
	int term_char = fgetc(stdin);
	if (term_char != '\n' && term_char != EOF)
		ungetc(term_char, stdin);
}

1942
static void parse_mark(void)
1943
{
1944 1945 1946
	const char *v;
	if (skip_prefix(command_buf.buf, "mark :", &v)) {
		next_mark = strtoumax(v, NULL, 10);
1947 1948 1949
		read_next_command();
	}
	else
1950
		next_mark = 0;
1951 1952
}

1953
static int parse_data(struct strbuf *sb, uintmax_t limit, uintmax_t *len_res)
1954
{
1955
	const char *data;
1956
	strbuf_reset(sb);
1957

1958
	if (!skip_prefix(command_buf.buf, "data ", &data))
1959 1960
		die("Expected 'data n' command, found: %s", command_buf.buf);

1961 1962 1963
	if (skip_prefix(data, "<<", &data)) {
		char *term = xstrdup(data);
		size_t term_len = command_buf.len - (data - command_buf.buf);
1964

1965
		strbuf_detach(&command_buf, NULL);
1966
		for (;;) {
1967
			if (strbuf_getline(&command_buf, stdin, '\n') == EOF)
1968 1969 1970 1971
				die("EOF in data (terminator '%s' not found)", term);
			if (term_len == command_buf.len
				&& !strcmp(term, command_buf.buf))
				break;
1972 1973
			strbuf_addbuf(sb, &command_buf);
			strbuf_addch(sb, '\n');
1974 1975 1976 1977
		}
		free(term);
	}
	else {
1978
		uintmax_t len = strtoumax(data, NULL, 10);
1979
		size_t n = 0, length = (size_t)len;
1980

1981 1982 1983 1984 1985 1986
		if (limit && limit < len) {
			*len_res = len;
			return 0;
		}
		if (length < len)
			die("data is too large to use in this context");
1987

1988
		while (n < length) {
1989
			size_t s = strbuf_fread(sb, length - n, stdin);
1990
			if (!s && feof(stdin))
1991 1992
				die("EOF in data (%lu bytes remaining)",
					(unsigned long)(length - n));
1993 1994
			n += s;
		}
1995 1996
	}

1997
	skip_optional_lf();
1998
	return 1;
1999 2000
}

2001
static int validate_raw_date(const char *src, struct strbuf *result)
2002 2003
{
	const char *orig_src = src;
2004
	char *endp;
2005
	unsigned long num;
2006

2007 2008
	errno = 0;

2009 2010
	num = strtoul(src, &endp, 10);
	/* NEEDSWORK: perhaps check for reasonable values? */
2011
	if (errno || endp == src || *endp != ' ')
2012 2013 2014 2015 2016 2017
		return -1;

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

2018
	num = strtoul(src + 1, &endp, 10);
2019
	if (errno || endp == src + 1 || *endp || 1400 < num)
2020 2021
		return -1;

2022
	strbuf_addstr(result, orig_src);
2023 2024 2025 2026 2027
	return 0;
}

static char *parse_ident(const char *buf)
{
2028
	const char *ltgt;
2029
	size_t name_len;
2030
	struct strbuf ident = STRBUF_INIT;
2031

2032 2033 2034 2035
	/* ensure there is a space delimiter even if there is no name */
	if (*buf == '<')
		--buf;

2036 2037 2038 2039 2040 2041 2042
	ltgt = buf + strcspn(buf, "<>");
	if (*ltgt != '<')
		die("Missing < in ident string: %s", buf);
	if (ltgt != buf && ltgt[-1] != ' ')
		die("Missing space before < in ident string: %s", buf);
	ltgt = ltgt + 1 + strcspn(ltgt + 1, "<>");
	if (*ltgt != '>')
2043
		die("Missing > in ident string: %s", buf);
2044 2045
	ltgt++;
	if (*ltgt != ' ')
2046
		die("Missing space after > in ident string: %s", buf);
2047 2048
	ltgt++;
	name_len = ltgt - buf;
2049
	strbuf_add(&ident, buf, name_len);
2050 2051 2052

	switch (whenspec) {
	case WHENSPEC_RAW:
2053
		if (validate_raw_date(ltgt, &ident) < 0)
2054
			die("Invalid raw date \"%s\" in ident: %s", ltgt, buf);
2055 2056
		break;
	case WHENSPEC_RFC2822:
2057
		if (parse_date(ltgt, &ident) < 0)
2058
			die("Invalid rfc2822 date \"%s\" in ident: %s", ltgt, buf);
2059 2060
		break;
	case WHENSPEC_NOW:
2061
		if (strcmp("now", ltgt))
2062
			die("Date in ident must be 'now': %s", buf);
2063
		datestamp(&ident);
2064 2065 2066
		break;
	}

2067
	return strbuf_detach(&ident, NULL);
2068 2069
}

2070 2071 2072 2073
static void parse_and_store_blob(
	struct last_object *last,
	unsigned char *sha1out,
	uintmax_t mark)
2074
{
P
Pierre Habouzit 已提交
2075
	static struct strbuf buf = STRBUF_INIT;
2076
	uintmax_t len;
2077

2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092
	if (parse_data(&buf, big_file_threshold, &len))
		store_object(OBJ_BLOB, &buf, last, sha1out, mark);
	else {
		if (last) {
			strbuf_release(&last->data);
			last->offset = 0;
			last->depth = 0;
		}
		stream_blob(len, sha1out, mark);
		skip_optional_lf();
	}
}

static void parse_new_blob(void)
{
2093
	read_next_command();
2094
	parse_mark();
2095
	parse_and_store_blob(&last_blob, NULL, next_mark);
2096 2097
}

2098
static void unload_one_branch(void)
2099
{
2100 2101
	while (cur_active_branches
		&& cur_active_branches >= max_active_branches) {
2102
		uintmax_t min_commit = ULONG_MAX;
2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119
		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;
		}
2120
		e->active = 0;
2121 2122
		e->active_next_branch = NULL;
		if (e->branch_tree.tree) {
2123
			release_tree_content_recursive(e->branch_tree.tree);
2124 2125 2126
			e->branch_tree.tree = NULL;
		}
		cur_active_branches--;
2127 2128 2129
	}
}

2130
static void load_branch(struct branch *b)
2131
{
2132
	load_tree(&b->branch_tree);
2133 2134 2135 2136 2137 2138 2139
	if (!b->active) {
		b->active = 1;
		b->active_next_branch = active_branches;
		active_branches = b;
		cur_active_branches++;
		branch_load_count++;
	}
2140 2141
}

2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208
static unsigned char convert_num_notes_to_fanout(uintmax_t num_notes)
{
	unsigned char fanout = 0;
	while ((num_notes >>= 8))
		fanout++;
	return fanout;
}

static void construct_path_with_fanout(const char *hex_sha1,
		unsigned char fanout, char *path)
{
	unsigned int i = 0, j = 0;
	if (fanout >= 20)
		die("Too large fanout (%u)", fanout);
	while (fanout) {
		path[i++] = hex_sha1[j++];
		path[i++] = hex_sha1[j++];
		path[i++] = '/';
		fanout--;
	}
	memcpy(path + i, hex_sha1 + j, 40 - j);
	path[i + 40 - j] = '\0';
}

static uintmax_t do_change_note_fanout(
		struct tree_entry *orig_root, struct tree_entry *root,
		char *hex_sha1, unsigned int hex_sha1_len,
		char *fullpath, unsigned int fullpath_len,
		unsigned char fanout)
{
	struct tree_content *t = root->tree;
	struct tree_entry *e, leaf;
	unsigned int i, tmp_hex_sha1_len, tmp_fullpath_len;
	uintmax_t num_notes = 0;
	unsigned char sha1[20];
	char realpath[60];

	for (i = 0; t && i < t->entry_count; i++) {
		e = t->entries[i];
		tmp_hex_sha1_len = hex_sha1_len + e->name->str_len;
		tmp_fullpath_len = fullpath_len;

		/*
		 * We're interested in EITHER existing note entries (entries
		 * with exactly 40 hex chars in path, not including directory
		 * separators), OR directory entries that may contain note
		 * entries (with < 40 hex chars in path).
		 * Also, each path component in a note entry must be a multiple
		 * of 2 chars.
		 */
		if (!e->versions[1].mode ||
		    tmp_hex_sha1_len > 40 ||
		    e->name->str_len % 2)
			continue;

		/* This _may_ be a note entry, or a subdir containing notes */
		memcpy(hex_sha1 + hex_sha1_len, e->name->str_dat,
		       e->name->str_len);
		if (tmp_fullpath_len)
			fullpath[tmp_fullpath_len++] = '/';
		memcpy(fullpath + tmp_fullpath_len, e->name->str_dat,
		       e->name->str_len);
		tmp_fullpath_len += e->name->str_len;
		fullpath[tmp_fullpath_len] = '\0';

		if (tmp_hex_sha1_len == 40 && !get_sha1_hex(hex_sha1, sha1)) {
			/* This is a note entry */
2209 2210 2211 2212 2213
			if (fanout == 0xff) {
				/* Counting mode, no rename */
				num_notes++;
				continue;
			}
2214 2215 2216 2217 2218 2219 2220 2221
			construct_path_with_fanout(hex_sha1, fanout, realpath);
			if (!strcmp(fullpath, realpath)) {
				/* Note entry is in correct location */
				num_notes++;
				continue;
			}

			/* Rename fullpath to realpath */
2222
			if (!tree_content_remove(orig_root, fullpath, &leaf, 0))
2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249
				die("Failed to remove path %s", fullpath);
			tree_content_set(orig_root, realpath,
				leaf.versions[1].sha1,
				leaf.versions[1].mode,
				leaf.tree);
		} else if (S_ISDIR(e->versions[1].mode)) {
			/* This is a subdir that may contain note entries */
			if (!e->tree)
				load_tree(e);
			num_notes += do_change_note_fanout(orig_root, e,
				hex_sha1, tmp_hex_sha1_len,
				fullpath, tmp_fullpath_len, fanout);
		}

		/* The above may have reallocated the current tree_content */
		t = root->tree;
	}
	return num_notes;
}

static uintmax_t change_note_fanout(struct tree_entry *root,
		unsigned char fanout)
{
	char hex_sha1[40], path[60];
	return do_change_note_fanout(root, root, hex_sha1, 0, path, 0, fanout);
}

2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296
/*
 * Given a pointer into a string, parse a mark reference:
 *
 *   idnum ::= ':' bigint;
 *
 * Return the first character after the value in *endptr.
 *
 * Complain if the following character is not what is expected,
 * either a space or end of the string.
 */
static uintmax_t parse_mark_ref(const char *p, char **endptr)
{
	uintmax_t mark;

	assert(*p == ':');
	p++;
	mark = strtoumax(p, endptr, 10);
	if (*endptr == p)
		die("No value after ':' in mark: %s", command_buf.buf);
	return mark;
}

/*
 * Parse the mark reference, and complain if this is not the end of
 * the string.
 */
static uintmax_t parse_mark_ref_eol(const char *p)
{
	char *end;
	uintmax_t mark;

	mark = parse_mark_ref(p, &end);
	if (*end != '\0')
		die("Garbage after mark: %s", command_buf.buf);
	return mark;
}

/*
 * Parse the mark reference, demanding a trailing space.  Return a
 * pointer to the space.
 */
static uintmax_t parse_mark_ref_space(const char **p)
{
	uintmax_t mark;
	char *end;

	mark = parse_mark_ref(*p, &end);
2297
	if (*end++ != ' ')
2298 2299 2300 2301 2302
		die("Missing space after mark: %s", command_buf.buf);
	*p = end;
	return mark;
}

2303
static void file_change_m(const char *p, struct branch *b)
2304
{
2305
	static struct strbuf uq = STRBUF_INIT;
2306
	const char *endp;
2307
	struct object_entry *oe;
2308
	unsigned char sha1[20];
2309
	uint16_t mode, inline_data = 0;
2310

2311 2312 2313 2314
	p = get_mode(p, &mode);
	if (!p)
		die("Corrupt mode: %s", command_buf.buf);
	switch (mode) {
2315 2316 2317
	case 0644:
	case 0755:
		mode |= S_IFREG;
2318 2319
	case S_IFREG | 0644:
	case S_IFREG | 0755:
2320
	case S_IFLNK:
2321
	case S_IFDIR:
2322
	case S_IFGITLINK:
2323 2324 2325 2326 2327 2328
		/* ok */
		break;
	default:
		die("Corrupt mode: %s", command_buf.buf);
	}

2329
	if (*p == ':') {
2330
		oe = find_mark(parse_mark_ref_space(&p));
2331
		hashcpy(sha1, oe->idx.sha1);
2332
	} else if (skip_prefix(p, "inline ", &p)) {
2333
		inline_data = 1;
2334
		oe = NULL; /* not used with inline_data, but makes gcc happy */
2335 2336
	} else {
		if (get_sha1_hex(p, sha1))
2337
			die("Invalid dataref: %s", command_buf.buf);
2338 2339
		oe = find_object(sha1);
		p += 40;
2340
		if (*p++ != ' ')
2341
			die("Missing space after SHA1: %s", command_buf.buf);
2342
	}
2343

2344 2345
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
2346 2347
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
2348
		p = uq.buf;
2349
	}
2350

2351
	/* Git does not track empty, non-toplevel directories. */
2352
	if (S_ISDIR(mode) && !hashcmp(sha1, EMPTY_TREE_SHA1_BIN) && *p) {
2353
		tree_content_remove(&b->branch_tree, p, NULL, 0);
2354 2355 2356
		return;
	}

2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370
	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) {
2371 2372 2373
		if (S_ISDIR(mode))
			die("Directories cannot be specified 'inline': %s",
				command_buf.buf);
2374 2375 2376 2377
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
2378
		read_next_command();
2379
		parse_and_store_blob(&last_blob, sha1, 0);
2380
	} else {
2381 2382 2383 2384
		enum object_type expected = S_ISDIR(mode) ?
						OBJ_TREE: OBJ_BLOB;
		enum object_type type = oe ? oe->type :
					sha1_object_info(sha1, NULL);
2385
		if (type < 0)
2386 2387 2388 2389 2390 2391 2392
			die("%s not found: %s",
					S_ISDIR(mode) ?  "Tree" : "Blob",
					command_buf.buf);
		if (type != expected)
			die("Not a %s (actually a %s): %s",
				typename(expected), typename(type),
				command_buf.buf);
2393
	}
2394

2395 2396 2397 2398
	if (!*p) {
		tree_content_replace(&b->branch_tree, sha1, mode, NULL);
		return;
	}
2399
	tree_content_set(&b->branch_tree, p, sha1, mode, NULL);
2400
}
2401

2402
static void file_change_d(const char *p, struct branch *b)
2403
{
2404
	static struct strbuf uq = STRBUF_INIT;
2405 2406
	const char *endp;

2407 2408
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
2409 2410
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
2411
		p = uq.buf;
2412
	}
2413
	tree_content_remove(&b->branch_tree, p, NULL, 1);
2414 2415
}

2416
static void file_change_cr(const char *s, struct branch *b, int rename)
2417
{
2418
	const char *d;
2419 2420
	static struct strbuf s_uq = STRBUF_INIT;
	static struct strbuf d_uq = STRBUF_INIT;
2421 2422 2423
	const char *endp;
	struct tree_entry leaf;

2424 2425
	strbuf_reset(&s_uq);
	if (!unquote_c_style(&s_uq, s, &endp)) {
2426 2427
		if (*endp != ' ')
			die("Missing space after source: %s", command_buf.buf);
2428
	} else {
2429 2430 2431
		endp = strchr(s, ' ');
		if (!endp)
			die("Missing space after source: %s", command_buf.buf);
2432
		strbuf_add(&s_uq, s, endp - s);
2433
	}
2434
	s = s_uq.buf;
2435 2436 2437 2438 2439 2440

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

	d = endp;
2441 2442
	strbuf_reset(&d_uq);
	if (!unquote_c_style(&d_uq, d, &endp)) {
2443 2444
		if (*endp)
			die("Garbage after dest in: %s", command_buf.buf);
2445
		d = d_uq.buf;
2446 2447 2448
	}

	memset(&leaf, 0, sizeof(leaf));
2449
	if (rename)
2450
		tree_content_remove(&b->branch_tree, s, &leaf, 1);
2451
	else
2452
		tree_content_get(&b->branch_tree, s, &leaf, 1);
2453 2454
	if (!leaf.versions[1].mode)
		die("Path %s not in branch", s);
2455 2456 2457 2458 2459 2460 2461
	if (!*d) {	/* C "path/to/subdir" "" */
		tree_content_replace(&b->branch_tree,
			leaf.versions[1].sha1,
			leaf.versions[1].mode,
			leaf.tree);
		return;
	}
2462 2463 2464 2465 2466 2467
	tree_content_set(&b->branch_tree, d,
		leaf.versions[1].sha1,
		leaf.versions[1].mode,
		leaf.tree);
}

2468
static void note_change_n(const char *p, struct branch *b, unsigned char *old_fanout)
2469 2470
{
	static struct strbuf uq = STRBUF_INIT;
2471
	struct object_entry *oe;
2472 2473
	struct branch *s;
	unsigned char sha1[20], commit_sha1[20];
2474
	char path[60];
2475
	uint16_t inline_data = 0;
2476
	unsigned char new_fanout;
2477

2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494
	/*
	 * When loading a branch, we don't traverse its tree to count the real
	 * number of notes (too expensive to do this for all non-note refs).
	 * This means that recently loaded notes refs might incorrectly have
	 * b->num_notes == 0, and consequently, old_fanout might be wrong.
	 *
	 * Fix this by traversing the tree and counting the number of notes
	 * when b->num_notes == 0. If the notes tree is truly empty, the
	 * calculation should not take long.
	 */
	if (b->num_notes == 0 && *old_fanout == 0) {
		/* Invoke change_note_fanout() in "counting mode". */
		b->num_notes = change_note_fanout(&b->branch_tree, 0xff);
		*old_fanout = convert_num_notes_to_fanout(b->num_notes);
	}

	/* Now parse the notemodify command. */
2495 2496
	/* <dataref> or 'inline' */
	if (*p == ':') {
2497
		oe = find_mark(parse_mark_ref_space(&p));
2498
		hashcpy(sha1, oe->idx.sha1);
2499
	} else if (skip_prefix(p, "inline ", &p)) {
2500
		inline_data = 1;
2501
		oe = NULL; /* not used with inline_data, but makes gcc happy */
2502 2503
	} else {
		if (get_sha1_hex(p, sha1))
2504
			die("Invalid dataref: %s", command_buf.buf);
2505 2506
		oe = find_object(sha1);
		p += 40;
2507
		if (*p++ != ' ')
2508
			die("Missing space after SHA1: %s", command_buf.buf);
2509 2510
	}

2511
	/* <commit-ish> */
2512 2513
	s = lookup_branch(p);
	if (s) {
2514 2515
		if (is_null_sha1(s->sha1))
			die("Can't add a note on empty branch.");
2516 2517
		hashcpy(commit_sha1, s->sha1);
	} else if (*p == ':') {
2518
		uintmax_t commit_mark = parse_mark_ref_eol(p);
2519 2520 2521
		struct object_entry *commit_oe = find_mark(commit_mark);
		if (commit_oe->type != OBJ_COMMIT)
			die("Mark :%" PRIuMAX " not a commit", commit_mark);
2522
		hashcpy(commit_sha1, commit_oe->idx.sha1);
2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538
	} else if (!get_sha1(p, commit_sha1)) {
		unsigned long size;
		char *buf = read_object_with_reference(commit_sha1,
			commit_type, &size, commit_sha1);
		if (!buf || size < 46)
			die("Not a valid commit: %s", p);
		free(buf);
	} else
		die("Invalid ref name or SHA1 expression: %s", p);

	if (inline_data) {
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
		read_next_command();
2539
		parse_and_store_blob(&last_blob, sha1, 0);
2540 2541 2542 2543
	} else if (oe) {
		if (oe->type != OBJ_BLOB)
			die("Not a blob (actually a %s): %s",
				typename(oe->type), command_buf.buf);
2544
	} else if (!is_null_sha1(sha1)) {
2545 2546 2547 2548 2549 2550 2551 2552
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
			die("Blob not found: %s", command_buf.buf);
		if (type != OBJ_BLOB)
			die("Not a blob (actually a %s): %s",
			    typename(type), command_buf.buf);
	}

2553
	construct_path_with_fanout(sha1_to_hex(commit_sha1), *old_fanout, path);
2554
	if (tree_content_remove(&b->branch_tree, path, NULL, 0))
2555 2556 2557 2558 2559 2560 2561 2562 2563
		b->num_notes--;

	if (is_null_sha1(sha1))
		return; /* nothing to insert */

	b->num_notes++;
	new_fanout = convert_num_notes_to_fanout(b->num_notes);
	construct_path_with_fanout(sha1_to_hex(commit_sha1), new_fanout, path);
	tree_content_set(&b->branch_tree, path, sha1, S_IFREG | 0644, NULL);
2564 2565
}

2566 2567 2568 2569 2570 2571
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);
2572
	b->num_notes = 0;
2573 2574
}

2575
static void parse_from_commit(struct branch *b, char *buf, unsigned long size)
2576 2577 2578 2579 2580 2581 2582 2583 2584 2585
{
	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);
}

2586
static void parse_from_existing(struct branch *b)
2587 2588 2589 2590 2591 2592 2593 2594 2595 2596
{
	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);
2597
		parse_from_commit(b, buf, size);
2598 2599 2600 2601
		free(buf);
	}
}

2602
static int parse_from(struct branch *b)
2603
{
2604
	const char *from;
2605 2606
	struct branch *s;

2607
	if (!skip_prefix(command_buf.buf, "from ", &from))
2608
		return 0;
2609

2610 2611 2612 2613
	if (b->branch_tree.tree) {
		release_tree_content_recursive(b->branch_tree.tree);
		b->branch_tree.tree = NULL;
	}
2614 2615 2616 2617 2618

	s = lookup_branch(from);
	if (b == s)
		die("Can't create a branch from itself: %s", b->name);
	else if (s) {
2619
		unsigned char *t = s->branch_tree.versions[1].sha1;
2620
		hashcpy(b->sha1, s->sha1);
2621 2622
		hashcpy(b->branch_tree.versions[0].sha1, t);
		hashcpy(b->branch_tree.versions[1].sha1, t);
2623
	} else if (*from == ':') {
2624
		uintmax_t idnum = parse_mark_ref_eol(from);
2625 2626
		struct object_entry *oe = find_mark(idnum);
		if (oe->type != OBJ_COMMIT)
2627
			die("Mark :%" PRIuMAX " not a commit", idnum);
2628
		hashcpy(b->sha1, oe->idx.sha1);
2629
		if (oe->pack_id != MAX_PACK_ID) {
2630
			unsigned long size;
2631
			char *buf = gfi_unpack_entry(oe, &size);
2632
			parse_from_commit(b, buf, size);
2633
			free(buf);
2634
		} else
2635
			parse_from_existing(b);
2636
	} else if (!get_sha1(from, b->sha1)) {
2637
		parse_from_existing(b);
2638 2639 2640
		if (is_null_sha1(b->sha1))
			b->delete = 1;
	}
2641
	else
2642 2643 2644
		die("Invalid ref name or SHA1 expression: %s", from);

	read_next_command();
2645
	return 1;
2646 2647
}

2648
static struct hash_list *parse_merge(unsigned int *count)
2649
{
2650
	struct hash_list *list = NULL, **tail = &list, *n;
2651
	const char *from;
2652 2653 2654
	struct branch *s;

	*count = 0;
2655
	while (skip_prefix(command_buf.buf, "merge ", &from)) {
2656 2657 2658 2659 2660
		n = xmalloc(sizeof(*n));
		s = lookup_branch(from);
		if (s)
			hashcpy(n->sha1, s->sha1);
		else if (*from == ':') {
2661
			uintmax_t idnum = parse_mark_ref_eol(from);
2662 2663
			struct object_entry *oe = find_mark(idnum);
			if (oe->type != OBJ_COMMIT)
2664
				die("Mark :%" PRIuMAX " not a commit", idnum);
2665
			hashcpy(n->sha1, oe->idx.sha1);
2666 2667 2668
		} 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
2669
				commit_type, &size, n->sha1);
2670 2671 2672 2673
			if (!buf || size < 46)
				die("Not a valid commit: %s", from);
			free(buf);
		} else
2674 2675 2676
			die("Invalid ref name or SHA1 expression: %s", from);

		n->next = NULL;
2677 2678 2679
		*tail = n;
		tail = &n->next;

2680
		(*count)++;
2681 2682 2683 2684 2685
		read_next_command();
	}
	return list;
}

2686
static void parse_new_commit(const char *arg)
2687
{
2688
	static struct strbuf msg = STRBUF_INIT;
2689 2690 2691
	struct branch *b;
	char *author = NULL;
	char *committer = NULL;
2692 2693
	struct hash_list *merge_list = NULL;
	unsigned int merge_count;
2694
	unsigned char prev_fanout, new_fanout;
2695
	const char *v;
2696

2697
	b = lookup_branch(arg);
2698
	if (!b)
2699
		b = new_branch(arg);
2700 2701

	read_next_command();
2702
	parse_mark();
2703 2704
	if (skip_prefix(command_buf.buf, "author ", &v)) {
		author = parse_ident(v);
2705 2706
		read_next_command();
	}
2707 2708
	if (skip_prefix(command_buf.buf, "committer ", &v)) {
		committer = parse_ident(v);
2709 2710 2711 2712
		read_next_command();
	}
	if (!committer)
		die("Expected committer but didn't get one");
2713
	parse_data(&msg, 0, NULL);
2714
	read_next_command();
2715 2716
	parse_from(b);
	merge_list = parse_merge(&merge_count);
2717 2718

	/* ensure the branch is active/loaded */
2719
	if (!b->branch_tree.tree || !max_active_branches) {
2720 2721 2722
		unload_one_branch();
		load_branch(b);
	}
2723

2724 2725
	prev_fanout = convert_num_notes_to_fanout(b->num_notes);

2726
	/* file_change* */
2727
	while (command_buf.len > 0) {
2728 2729 2730 2731 2732 2733 2734 2735 2736 2737
		if (skip_prefix(command_buf.buf, "M ", &v))
			file_change_m(v, b);
		else if (skip_prefix(command_buf.buf, "D ", &v))
			file_change_d(v, b);
		else if (skip_prefix(command_buf.buf, "R ", &v))
			file_change_cr(v, b, 1);
		else if (skip_prefix(command_buf.buf, "C ", &v))
			file_change_cr(v, b, 0);
		else if (skip_prefix(command_buf.buf, "N ", &v))
			note_change_n(v, b, &prev_fanout);
2738 2739
		else if (!strcmp("deleteall", command_buf.buf))
			file_change_deleteall(b);
2740 2741
		else if (skip_prefix(command_buf.buf, "ls ", &v))
			parse_ls(v, b);
2742 2743 2744 2745
		else {
			unread_command_buf = 1;
			break;
		}
2746 2747
		if (read_next_command() == EOF)
			break;
2748 2749
	}

2750 2751 2752 2753
	new_fanout = convert_num_notes_to_fanout(b->num_notes);
	if (new_fanout != prev_fanout)
		b->num_notes = change_note_fanout(&b->branch_tree, new_fanout);

2754
	/* build the tree and the commit */
2755
	store_tree(&b->branch_tree);
2756 2757
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
2758 2759 2760

	strbuf_reset(&new_data);
	strbuf_addf(&new_data, "tree %s\n",
2761
		sha1_to_hex(b->branch_tree.versions[1].sha1));
2762
	if (!is_null_sha1(b->sha1))
2763
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(b->sha1));
2764 2765
	while (merge_list) {
		struct hash_list *next = merge_list->next;
2766
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(merge_list->sha1));
2767 2768 2769
		free(merge_list);
		merge_list = next;
	}
2770 2771 2772 2773 2774 2775
	strbuf_addf(&new_data,
		"author %s\n"
		"committer %s\n"
		"\n",
		author ? author : committer, committer);
	strbuf_addbuf(&new_data, &msg);
2776
	free(author);
2777 2778
	free(committer);

2779
	if (!store_object(OBJ_COMMIT, &new_data, NULL, b->sha1, next_mark))
2780
		b->pack_id = pack_id;
2781
	b->last_commit = object_count_by_type[OBJ_COMMIT];
2782 2783
}

2784
static void parse_new_tag(const char *arg)
2785
{
2786
	static struct strbuf msg = STRBUF_INIT;
2787 2788 2789 2790
	const char *from;
	char *tagger;
	struct branch *s;
	struct tag *t;
2791
	uintmax_t from_mark = 0;
2792
	unsigned char sha1[20];
2793
	enum object_type type;
2794
	const char *v;
2795 2796

	t = pool_alloc(sizeof(struct tag));
2797
	memset(t, 0, sizeof(struct tag));
2798
	t->name = pool_strdup(arg);
2799 2800 2801 2802 2803 2804 2805 2806
	if (last_tag)
		last_tag->next_tag = t;
	else
		first_tag = t;
	last_tag = t;
	read_next_command();

	/* from ... */
2807
	if (!skip_prefix(command_buf.buf, "from ", &from))
2808 2809 2810
		die("Expected from command, got %s", command_buf.buf);
	s = lookup_branch(from);
	if (s) {
2811 2812
		if (is_null_sha1(s->sha1))
			die("Can't tag an empty branch.");
2813
		hashcpy(sha1, s->sha1);
2814
		type = OBJ_COMMIT;
2815
	} else if (*from == ':') {
2816
		struct object_entry *oe;
2817
		from_mark = parse_mark_ref_eol(from);
2818
		oe = find_mark(from_mark);
2819
		type = oe->type;
2820
		hashcpy(sha1, oe->idx.sha1);
2821
	} else if (!get_sha1(from, sha1)) {
2822 2823 2824 2825 2826 2827 2828
		struct object_entry *oe = find_object(sha1);
		if (!oe) {
			type = sha1_object_info(sha1, NULL);
			if (type < 0)
				die("Not a valid object: %s", from);
		} else
			type = oe->type;
2829 2830 2831 2832 2833
	} else
		die("Invalid ref name or SHA1 expression: %s", from);
	read_next_command();

	/* tagger ... */
2834 2835
	if (skip_prefix(command_buf.buf, "tagger ", &v)) {
		tagger = parse_ident(v);
2836 2837 2838
		read_next_command();
	} else
		tagger = NULL;
2839 2840

	/* tag payload/message */
2841
	parse_data(&msg, 0, NULL);
2842 2843

	/* build the tag object */
2844
	strbuf_reset(&new_data);
2845

2846
	strbuf_addf(&new_data,
2847 2848 2849
		    "object %s\n"
		    "type %s\n"
		    "tag %s\n",
2850
		    sha1_to_hex(sha1), typename(type), t->name);
2851 2852 2853 2854
	if (tagger)
		strbuf_addf(&new_data,
			    "tagger %s\n", tagger);
	strbuf_addch(&new_data, '\n');
2855
	strbuf_addbuf(&new_data, &msg);
2856 2857
	free(tagger);

2858
	if (store_object(OBJ_TAG, &new_data, NULL, t->sha1, 0))
2859 2860 2861
		t->pack_id = MAX_PACK_ID;
	else
		t->pack_id = pack_id;
2862 2863
}

2864
static void parse_reset_branch(const char *arg)
2865 2866 2867
{
	struct branch *b;

2868
	b = lookup_branch(arg);
2869
	if (b) {
2870 2871 2872
		hashclr(b->sha1);
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
2873 2874 2875 2876 2877
		if (b->branch_tree.tree) {
			release_tree_content_recursive(b->branch_tree.tree);
			b->branch_tree.tree = NULL;
		}
	}
2878
	else
2879
		b = new_branch(arg);
2880
	read_next_command();
2881
	parse_from(b);
2882
	if (command_buf.len > 0)
2883
		unread_command_buf = 1;
2884 2885
}

2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928
static void cat_blob_write(const char *buf, unsigned long size)
{
	if (write_in_full(cat_blob_fd, buf, size) != size)
		die_errno("Write to frontend failed");
}

static void cat_blob(struct object_entry *oe, unsigned char sha1[20])
{
	struct strbuf line = STRBUF_INIT;
	unsigned long size;
	enum object_type type = 0;
	char *buf;

	if (!oe || oe->pack_id == MAX_PACK_ID) {
		buf = read_sha1_file(sha1, &type, &size);
	} else {
		type = oe->type;
		buf = gfi_unpack_entry(oe, &size);
	}

	/*
	 * Output based on batch_one_object() from cat-file.c.
	 */
	if (type <= 0) {
		strbuf_reset(&line);
		strbuf_addf(&line, "%s missing\n", sha1_to_hex(sha1));
		cat_blob_write(line.buf, line.len);
		strbuf_release(&line);
		free(buf);
		return;
	}
	if (!buf)
		die("Can't read object %s", sha1_to_hex(sha1));
	if (type != OBJ_BLOB)
		die("Object %s is a %s but a blob was expected.",
		    sha1_to_hex(sha1), typename(type));
	strbuf_reset(&line);
	strbuf_addf(&line, "%s %s %lu\n", sha1_to_hex(sha1),
						typename(type), size);
	cat_blob_write(line.buf, line.len);
	strbuf_release(&line);
	cat_blob_write(buf, size);
	cat_blob_write("\n", 1);
2929 2930 2931 2932 2933 2934
	if (oe && oe->pack_id == pack_id) {
		last_blob.offset = oe->idx.offset;
		strbuf_attach(&last_blob.data, buf, size, size);
		last_blob.depth = oe->depth;
	} else
		free(buf);
2935 2936
}

2937
static void parse_cat_blob(const char *p)
2938 2939 2940 2941 2942 2943
{
	struct object_entry *oe = oe;
	unsigned char sha1[20];

	/* cat-blob SP <object> LF */
	if (*p == ':') {
2944
		oe = find_mark(parse_mark_ref_eol(p));
2945 2946 2947 2948 2949
		if (!oe)
			die("Unknown mark: %s", command_buf.buf);
		hashcpy(sha1, oe->idx.sha1);
	} else {
		if (get_sha1_hex(p, sha1))
2950
			die("Invalid dataref: %s", command_buf.buf);
2951 2952 2953 2954 2955 2956 2957 2958
		if (p[40])
			die("Garbage after SHA1: %s", command_buf.buf);
		oe = find_object(sha1);
	}

	cat_blob(oe, sha1);
}

D
David Barr 已提交
2959 2960 2961 2962
static struct object_entry *dereference(struct object_entry *oe,
					unsigned char sha1[20])
{
	unsigned long size;
2963
	char *buf = NULL;
D
David Barr 已提交
2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980
	if (!oe) {
		enum object_type type = sha1_object_info(sha1, NULL);
		if (type < 0)
			die("object not found: %s", sha1_to_hex(sha1));
		/* cache it! */
		oe = insert_object(sha1);
		oe->type = type;
		oe->pack_id = MAX_PACK_ID;
		oe->idx.offset = 1;
	}
	switch (oe->type) {
	case OBJ_TREE:	/* easy case. */
		return oe;
	case OBJ_COMMIT:
	case OBJ_TAG:
		break;
	default:
2981
		die("Not a tree-ish: %s", command_buf.buf);
D
David Barr 已提交
2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015
	}

	if (oe->pack_id != MAX_PACK_ID) {	/* in a pack being written */
		buf = gfi_unpack_entry(oe, &size);
	} else {
		enum object_type unused;
		buf = read_sha1_file(sha1, &unused, &size);
	}
	if (!buf)
		die("Can't load object %s", sha1_to_hex(sha1));

	/* Peel one layer. */
	switch (oe->type) {
	case OBJ_TAG:
		if (size < 40 + strlen("object ") ||
		    get_sha1_hex(buf + strlen("object "), sha1))
			die("Invalid SHA1 in tag: %s", command_buf.buf);
		break;
	case OBJ_COMMIT:
		if (size < 40 + strlen("tree ") ||
		    get_sha1_hex(buf + strlen("tree "), sha1))
			die("Invalid SHA1 in commit: %s", command_buf.buf);
	}

	free(buf);
	return find_object(sha1);
}

static struct object_entry *parse_treeish_dataref(const char **p)
{
	unsigned char sha1[20];
	struct object_entry *e;

	if (**p == ':') {	/* <mark> */
3016
		e = find_mark(parse_mark_ref_space(p));
D
David Barr 已提交
3017 3018 3019 3020 3021
		if (!e)
			die("Unknown mark: %s", command_buf.buf);
		hashcpy(sha1, e->idx.sha1);
	} else {	/* <sha1> */
		if (get_sha1_hex(*p, sha1))
3022
			die("Invalid dataref: %s", command_buf.buf);
D
David Barr 已提交
3023 3024
		e = find_object(sha1);
		*p += 40;
3025 3026
		if (*(*p)++ != ' ')
			die("Missing space after tree-ish: %s", command_buf.buf);
D
David Barr 已提交
3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053
	}

	while (!e || e->type != OBJ_TREE)
		e = dereference(e, sha1);
	return e;
}

static void print_ls(int mode, const unsigned char *sha1, const char *path)
{
	static struct strbuf line = STRBUF_INIT;

	/* See show_tree(). */
	const char *type =
		S_ISGITLINK(mode) ? commit_type :
		S_ISDIR(mode) ? tree_type :
		blob_type;

	if (!mode) {
		/* missing SP path LF */
		strbuf_reset(&line);
		strbuf_addstr(&line, "missing ");
		quote_c_style(path, &line, NULL, 0);
		strbuf_addch(&line, '\n');
	} else {
		/* mode SP type SP object_name TAB path LF */
		strbuf_reset(&line);
		strbuf_addf(&line, "%06o %s %s\t",
3054
				mode & ~NO_DELTA, type, sha1_to_hex(sha1));
D
David Barr 已提交
3055 3056 3057 3058 3059 3060
		quote_c_style(path, &line, NULL, 0);
		strbuf_addch(&line, '\n');
	}
	cat_blob_write(line.buf, line.len);
}

3061
static void parse_ls(const char *p, struct branch *b)
D
David Barr 已提交
3062 3063
{
	struct tree_entry *root = NULL;
S
Stephen Boyd 已提交
3064
	struct tree_entry leaf = {NULL};
D
David Barr 已提交
3065

3066
	/* ls SP (<tree-ish> SP)? <path> */
D
David Barr 已提交
3067 3068 3069 3070 3071 3072 3073 3074
	if (*p == '"') {
		if (!b)
			die("Not in a commit: %s", command_buf.buf);
		root = &b->branch_tree;
	} else {
		struct object_entry *e = parse_treeish_dataref(&p);
		root = new_tree_entry();
		hashcpy(root->versions[1].sha1, e->idx.sha1);
3075 3076
		if (!is_null_sha1(root->versions[1].sha1))
			root->versions[1].mode = S_IFDIR;
D
David Barr 已提交
3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088
		load_tree(root);
	}
	if (*p == '"') {
		static struct strbuf uq = STRBUF_INIT;
		const char *endp;
		strbuf_reset(&uq);
		if (unquote_c_style(&uq, p, &endp))
			die("Invalid path: %s", command_buf.buf);
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
		p = uq.buf;
	}
3089
	tree_content_get(root, p, &leaf, 1);
D
David Barr 已提交
3090 3091 3092 3093 3094 3095 3096 3097
	/*
	 * A directory in preparation would have a sha1 of zero
	 * until it is saved.  Save, for simplicity.
	 */
	if (S_ISDIR(leaf.versions[1].mode))
		store_tree(&leaf);

	print_ls(leaf.versions[1].mode, leaf.versions[1].sha1, p);
3098 3099
	if (leaf.tree)
		release_tree_content_recursive(leaf.tree);
D
David Barr 已提交
3100 3101 3102 3103
	if (!b || root != &b->branch_tree)
		release_tree_entry(root);
}

3104
static void checkpoint(void)
3105
{
3106
	checkpoint_requested = 0;
3107 3108 3109 3110 3111 3112
	if (object_count) {
		cycle_packfile();
		dump_branches();
		dump_tags();
		dump_marks();
	}
3113 3114 3115 3116 3117
}

static void parse_checkpoint(void)
{
	checkpoint_requested = 1;
3118
	skip_optional_lf();
3119 3120
}

3121
static void parse_progress(void)
3122
{
3123
	fwrite(command_buf.buf, 1, command_buf.len, stdout);
3124 3125 3126 3127 3128
	fputc('\n', stdout);
	fflush(stdout);
	skip_optional_lf();
}

3129
static char* make_fast_import_path(const char *path)
3130
{
3131
	struct strbuf abs_path = STRBUF_INIT;
3132

3133 3134 3135 3136 3137 3138
	if (!relative_marks_paths || is_absolute_path(path))
		return xstrdup(path);
	strbuf_addf(&abs_path, "%s/info/fast-import/%s", get_git_dir(), path);
	return strbuf_detach(&abs_path, NULL);
}

3139 3140
static void option_import_marks(const char *marks,
					int from_stream, int ignore_missing)
3141
{
3142 3143 3144 3145 3146 3147 3148
	if (import_marks_file) {
		if (from_stream)
			die("Only one import-marks command allowed per stream");

		/* read previous mark file */
		if(!import_marks_file_from_stream)
			read_marks();
3149
	}
3150

3151
	import_marks_file = make_fast_import_path(marks);
3152
	safe_create_leading_directories_const(import_marks_file);
3153
	import_marks_file_from_stream = from_stream;
3154
	import_marks_file_ignore_missing = ignore_missing;
3155 3156
}

3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168
static void option_date_format(const char *fmt)
{
	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);
}

3169 3170 3171 3172 3173 3174 3175 3176 3177
static unsigned long ulong_arg(const char *option, const char *arg)
{
	char *endptr;
	unsigned long rv = strtoul(arg, &endptr, 0);
	if (strchr(arg, '-') || endptr == arg || *endptr)
		die("%s: argument must be a non-negative integer", option);
	return rv;
}

3178 3179
static void option_depth(const char *depth)
{
3180
	max_depth = ulong_arg("--depth", depth);
3181 3182 3183 3184 3185 3186
	if (max_depth > MAX_DEPTH)
		die("--depth cannot exceed %u", MAX_DEPTH);
}

static void option_active_branches(const char *branches)
{
3187
	max_active_branches = ulong_arg("--active-branches", branches);
3188 3189 3190 3191
}

static void option_export_marks(const char *marks)
{
3192
	export_marks_file = make_fast_import_path(marks);
3193
	safe_create_leading_directories_const(export_marks_file);
3194 3195
}

3196 3197 3198 3199 3200 3201 3202 3203
static void option_cat_blob_fd(const char *fd)
{
	unsigned long n = ulong_arg("--cat-blob-fd", fd);
	if (n > (unsigned long) INT_MAX)
		die("--cat-blob-fd cannot exceed %d", INT_MAX);
	cat_blob_fd = (int) n;
}

3204 3205 3206 3207 3208 3209 3210 3211 3212
static void option_export_pack_edges(const char *edges)
{
	if (pack_edges)
		fclose(pack_edges);
	pack_edges = fopen(edges, "a");
	if (!pack_edges)
		die_errno("Cannot open '%s'", edges);
}

S
Sverre Rabbelier 已提交
3213
static int parse_one_option(const char *option)
3214
{
3215
	if (skip_prefix(option, "max-pack-size=", &option)) {
3216
		unsigned long v;
3217
		if (!git_parse_ulong(option, &v))
3218 3219 3220 3221 3222 3223 3224 3225 3226
			return 0;
		if (v < 8192) {
			warning("max-pack-size is now in bytes, assuming --max-pack-size=%lum", v);
			v *= 1024 * 1024;
		} else if (v < 1024 * 1024) {
			warning("minimum max-pack-size is 1 MiB");
			v = 1024 * 1024;
		}
		max_packsize = v;
3227
	} else if (skip_prefix(option, "big-file-threshold=", &option)) {
3228
		unsigned long v;
3229
		if (!git_parse_ulong(option, &v))
3230 3231
			return 0;
		big_file_threshold = v;
3232 3233 3234 3235 3236 3237
	} else if (skip_prefix(option, "depth=", &option)) {
		option_depth(option);
	} else if (skip_prefix(option, "active-branches=", &option)) {
		option_active_branches(option);
	} else if (skip_prefix(option, "export-pack-edges=", &option)) {
		option_export_pack_edges(option);
3238
	} else if (starts_with(option, "quiet")) {
3239
		show_stats = 0;
3240
	} else if (starts_with(option, "stats")) {
3241 3242
		show_stats = 1;
	} else {
S
Sverre Rabbelier 已提交
3243
		return 0;
3244
	}
S
Sverre Rabbelier 已提交
3245 3246

	return 1;
3247 3248
}

3249
static int parse_one_feature(const char *feature, int from_stream)
3250
{
3251 3252 3253 3254 3255 3256 3257 3258 3259 3260
	const char *arg;

	if (skip_prefix(feature, "date-format=", &arg)) {
		option_date_format(arg);
	} else if (skip_prefix(feature, "import-marks=", &arg)) {
		option_import_marks(arg, from_stream, 0);
	} else if (skip_prefix(feature, "import-marks-if-exists=", &arg)) {
		option_import_marks(arg, from_stream, 1);
	} else if (skip_prefix(feature, "export-marks=", &arg)) {
		option_export_marks(arg);
3261 3262
	} else if (!strcmp(feature, "cat-blob")) {
		; /* Don't die - this feature is supported */
3263
	} else if (!strcmp(feature, "relative-marks")) {
3264
		relative_marks_paths = 1;
3265
	} else if (!strcmp(feature, "no-relative-marks")) {
3266
		relative_marks_paths = 0;
3267 3268
	} else if (!strcmp(feature, "done")) {
		require_explicit_termination = 1;
3269
	} else if (!strcmp(feature, "force")) {
3270
		force_update = 1;
D
David Barr 已提交
3271
	} else if (!strcmp(feature, "notes") || !strcmp(feature, "ls")) {
3272
		; /* do nothing; we have the feature */
3273 3274 3275 3276 3277 3278 3279
	} else {
		return 0;
	}

	return 1;
}

3280
static void parse_feature(const char *feature)
3281 3282 3283 3284
{
	if (seen_data_command)
		die("Got feature command '%s' after data command", feature);

3285
	if (parse_one_feature(feature, 1))
3286 3287 3288 3289 3290
		return;

	die("This version of fast-import does not support feature %s.", feature);
}

3291
static void parse_option(const char *option)
S
Sverre Rabbelier 已提交
3292 3293 3294 3295 3296 3297 3298 3299
{
	if (seen_data_command)
		die("Got option command '%s' after data command", option);

	if (parse_one_option(option))
		return;

	die("This version of fast-import does not support option: %s", option);
3300 3301
}

3302
static void git_pack_config(void)
3303
{
3304 3305 3306 3307
	int indexversion_value;
	unsigned long packsizelimit_value;

	if (!git_config_get_ulong("pack.depth", &max_depth)) {
3308 3309 3310
		if (max_depth > MAX_DEPTH)
			max_depth = MAX_DEPTH;
	}
3311 3312 3313 3314 3315 3316 3317
	if (!git_config_get_int("pack.compression", &pack_compression_level)) {
		if (pack_compression_level == -1)
			pack_compression_level = Z_DEFAULT_COMPRESSION;
		else if (pack_compression_level < 0 ||
			 pack_compression_level > Z_BEST_COMPRESSION)
			git_die_config("pack.compression",
					"bad pack compression level %d", pack_compression_level);
3318 3319
		pack_compression_seen = 1;
	}
3320 3321
	if (!git_config_get_int("pack.indexversion", &indexversion_value)) {
		pack_idx_opts.version = indexversion_value;
3322
		if (pack_idx_opts.version > 2)
3323 3324
			git_die_config("pack.indexversion",
					"bad pack.indexversion=%"PRIu32, pack_idx_opts.version);
3325
	}
3326 3327 3328 3329
	if (!git_config_get_ulong("pack.packsizelimit", &packsizelimit_value))
		max_packsize = packsizelimit_value;

	git_config(git_default_config, NULL);
3330 3331
}

3332
static const char fast_import_usage[] =
3333
"git fast-import [--date-format=<f>] [--max-pack-size=<n>] [--big-file-threshold=<n>] [--depth=<n>] [--active-branches=<n>] [--export-marks=<marks.file>]";
3334

S
Sverre Rabbelier 已提交
3335 3336 3337 3338 3339 3340 3341 3342 3343 3344
static void parse_argv(void)
{
	unsigned int i;

	for (i = 1; i < global_argc; i++) {
		const char *a = global_argv[i];

		if (*a != '-' || !strcmp(a, "--"))
			break;

3345 3346 3347 3348
		if (!skip_prefix(a, "--", &a))
			die("unknown option %s", a);

		if (parse_one_option(a))
S
Sverre Rabbelier 已提交
3349 3350
			continue;

3351
		if (parse_one_feature(a, 0))
S
Sverre Rabbelier 已提交
3352 3353
			continue;

3354 3355
		if (skip_prefix(a, "cat-blob-fd=", &a)) {
			option_cat_blob_fd(a);
3356 3357 3358
			continue;
		}

3359
		die("unknown option --%s", a);
S
Sverre Rabbelier 已提交
3360 3361 3362 3363 3364 3365 3366 3367 3368
	}
	if (i != global_argc)
		usage(fast_import_usage);

	seen_data_command = 1;
	if (import_marks_file)
		read_marks();
}

3369
int main(int argc, char **argv)
3370
{
3371
	unsigned int i;
3372

3373 3374
	git_extract_argv0_path(argv[0]);

3375 3376
	git_setup_gettext();

3377 3378 3379
	if (argc == 2 && !strcmp(argv[1], "-h"))
		usage(fast_import_usage);

3380
	setup_git_directory();
3381
	reset_pack_idx_option(&pack_idx_opts);
3382
	git_pack_config();
3383 3384 3385
	if (!pack_compression_seen && core_compression_seen)
		pack_compression_level = core_compression_level;

3386
	alloc_objects(object_entry_alloc);
3387
	strbuf_init(&command_buf, 0);
3388 3389 3390 3391
	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));
3392

S
Sverre Rabbelier 已提交
3393 3394
	global_argc = argc;
	global_argv = argv;
3395

3396 3397 3398 3399 3400
	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;

3401
	prepare_packed_git();
3402
	start_packfile();
3403
	set_die_routine(die_nicely);
3404
	set_checkpoint_signal();
3405
	while (read_next_command() != EOF) {
3406
		const char *v;
3407
		if (!strcmp("blob", command_buf.buf))
3408
			parse_new_blob();
3409 3410 3411 3412 3413 3414 3415 3416
		else if (skip_prefix(command_buf.buf, "ls ", &v))
			parse_ls(v, NULL);
		else if (skip_prefix(command_buf.buf, "commit ", &v))
			parse_new_commit(v);
		else if (skip_prefix(command_buf.buf, "tag ", &v))
			parse_new_tag(v);
		else if (skip_prefix(command_buf.buf, "reset ", &v))
			parse_reset_branch(v);
3417
		else if (!strcmp("checkpoint", command_buf.buf))
3418
			parse_checkpoint();
3419 3420
		else if (!strcmp("done", command_buf.buf))
			break;
3421
		else if (starts_with(command_buf.buf, "progress "))
3422
			parse_progress();
3423 3424 3425 3426
		else if (skip_prefix(command_buf.buf, "feature ", &v))
			parse_feature(v);
		else if (skip_prefix(command_buf.buf, "option git ", &v))
			parse_option(v);
3427
		else if (starts_with(command_buf.buf, "option "))
S
Sverre Rabbelier 已提交
3428
			/* ignore non-git options*/;
3429 3430
		else
			die("Unsupported command: %s", command_buf.buf);
3431 3432 3433

		if (checkpoint_requested)
			checkpoint();
3434
	}
S
Sverre Rabbelier 已提交
3435 3436 3437 3438 3439

	/* argv hasn't been parsed yet, do so */
	if (!seen_data_command)
		parse_argv();

3440 3441 3442
	if (require_explicit_termination && feof(stdin))
		die("stream ends early");

3443
	end_packfile();
3444

3445
	dump_branches();
3446
	dump_tags();
3447
	unkeep_all_packs();
3448
	dump_marks();
3449

3450 3451 3452
	if (pack_edges)
		fclose(pack_edges);

3453 3454 3455 3456 3457 3458 3459 3460 3461
	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");
3462 3463
		fprintf(stderr, "Alloc'd objects: %10" PRIuMAX "\n", alloc_count);
		fprintf(stderr, "Total objects:   %10" PRIuMAX " (%10" PRIuMAX " duplicates                  )\n", total_count, duplicate_count);
3464 3465 3466 3467
		fprintf(stderr, "      blobs  :   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas of %10" PRIuMAX" attempts)\n", object_count_by_type[OBJ_BLOB], duplicate_count_by_type[OBJ_BLOB], delta_count_by_type[OBJ_BLOB], delta_count_attempts_by_type[OBJ_BLOB]);
		fprintf(stderr, "      trees  :   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas of %10" PRIuMAX" attempts)\n", object_count_by_type[OBJ_TREE], duplicate_count_by_type[OBJ_TREE], delta_count_by_type[OBJ_TREE], delta_count_attempts_by_type[OBJ_TREE]);
		fprintf(stderr, "      commits:   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas of %10" PRIuMAX" attempts)\n", object_count_by_type[OBJ_COMMIT], duplicate_count_by_type[OBJ_COMMIT], delta_count_by_type[OBJ_COMMIT], delta_count_attempts_by_type[OBJ_COMMIT]);
		fprintf(stderr, "      tags   :   %10" PRIuMAX " (%10" PRIuMAX " duplicates %10" PRIuMAX " deltas of %10" PRIuMAX" attempts)\n", object_count_by_type[OBJ_TAG], duplicate_count_by_type[OBJ_TAG], delta_count_by_type[OBJ_TAG], delta_count_attempts_by_type[OBJ_TAG]);
3468
		fprintf(stderr, "Total branches:  %10lu (%10lu loads     )\n", branch_count, branch_load_count);
3469
		fprintf(stderr, "      marks:     %10" PRIuMAX " (%10" PRIuMAX " unique    )\n", (((uintmax_t)1) << marks->shift) * 1024, marks_set_count);
3470
		fprintf(stderr, "      atoms:     %10u\n", atom_cnt);
3471
		fprintf(stderr, "Memory total:    %10" PRIuMAX " KiB\n", (total_allocd + alloc_count*sizeof(struct object_entry))/1024);
3472
		fprintf(stderr, "       pools:    %10lu KiB\n", (unsigned long)(total_allocd/1024));
3473
		fprintf(stderr, "     objects:    %10" PRIuMAX " KiB\n", (alloc_count*sizeof(struct object_entry))/1024);
3474 3475 3476 3477 3478
		fprintf(stderr, "---------------------------------------------------------------------\n");
		pack_report();
		fprintf(stderr, "---------------------------------------------------------------------\n");
		fprintf(stderr, "\n");
	}
3479

3480
	return failure ? 1 : 0;
3481
}