fast-import.c 87.6 KB
Newer Older
1
/*
2
(See Documentation/git-fast-import.txt for maintained documentation.)
3 4 5 6 7
Format of STDIN stream:

  stream ::= cmd*;

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

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

20
  new_commit ::= 'commit' sp ref_str lf
21
    mark?
22 23
    ('author' (sp name)? sp '<' email '>' sp when lf)?
    'committer' (sp name)? sp '<' email '>' sp when lf
24
    commit_msg
25 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;
250 251
	unsigned active : 1;
	unsigned pack_id : PACK_ID_BITS;
252
	unsigned char sha1[20];
253 254
};

255
struct tag {
256 257
	struct tag *next_tag;
	const char *name;
258
	unsigned int pack_id;
259 260 261
	unsigned char sha1[20];
};

262
struct hash_list {
263 264 265
	struct hash_list *next;
	unsigned char sha1[20];
};
266

267 268 269
typedef enum {
	WHENSPEC_RAW = 1,
	WHENSPEC_RFC2822,
270
	WHENSPEC_NOW
271 272
} whenspec_type;

273
struct recent_command {
274 275 276 277 278
	struct recent_command *prev;
	struct recent_command *next;
	char *buf;
};

279
/* Configured limits on output */
280
static unsigned long max_depth = 10;
281
static off_t max_packsize;
282
static int force_update;
283 284
static int pack_compression_level = Z_DEFAULT_COMPRESSION;
static int pack_compression_seen;
285 286 287 288 289 290 291

/* 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];
292
static uintmax_t delta_count_attempts_by_type[1 << TYPE_BITS];
293
static unsigned long object_count;
294
static unsigned long branch_count;
295
static unsigned long branch_load_count;
296
static int failure;
297
static FILE *pack_edges;
298
static unsigned int show_stats = 1;
S
Sverre Rabbelier 已提交
299
static int global_argc;
300
static char **global_argv;
301

302 303 304 305 306
/* 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;

307
/* Atom management */
308 309 310 311 312
static unsigned int atom_table_sz = 4451;
static unsigned int atom_cnt;
static struct atom_str **atom_table;

/* The .pack file being generated */
313
static struct pack_idx_option pack_idx_opts;
314
static unsigned int pack_id;
315
static struct sha1file *pack_file;
316
static struct packed_git *pack_data;
317
static struct packed_git **all_packs;
318
static off_t pack_size;
319 320

/* Table of objects we've written. */
321
static unsigned int object_entry_alloc = 5000;
322 323
static struct object_entry_pool *blocks;
static struct object_entry *object_table[1 << 16];
324
static struct mark_set *marks;
325 326
static const char *export_marks_file;
static const char *import_marks_file;
327
static int import_marks_file_from_stream;
328
static int import_marks_file_ignore_missing;
329
static int relative_marks_paths;
330 331

/* Our last blob */
P
Pierre Habouzit 已提交
332
static struct last_object last_blob = { STRBUF_INIT, 0, 0, 0 };
333 334 335 336 337 338

/* 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;
339 340
static struct strbuf old_tree = STRBUF_INIT;
static struct strbuf new_tree = STRBUF_INIT;
341

342
/* Branch data */
343 344 345
static unsigned long max_active_branches = 5;
static unsigned long cur_active_branches;
static unsigned long branch_table_sz = 1039;
346 347 348
static struct branch **branch_table;
static struct branch *active_branches;

349 350 351 352
/* Tag data */
static struct tag *first_tag;
static struct tag *last_tag;

353
/* Input stream parsing */
354
static whenspec_type whenspec = WHENSPEC_RAW;
355
static struct strbuf command_buf = STRBUF_INIT;
356
static int unread_command_buf;
357 358 359 360
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;
361
static uintmax_t next_mark;
362
static struct strbuf new_data = STRBUF_INIT;
363
static int seen_data_command;
364
static int require_explicit_termination;
365

366 367 368
/* Signal handling */
static volatile sig_atomic_t checkpoint_requested;

369 370 371
/* Where to write output of cat-blob commands */
static int cat_blob_fd = STDOUT_FILENO;

S
Sverre Rabbelier 已提交
372
static void parse_argv(void);
373
static void parse_cat_blob(void);
D
David Barr 已提交
374
static void parse_ls(struct branch *b);
375

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
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);
}

402 403
static void dump_marks_helper(FILE *, uintmax_t, struct mark_set *);

404
static void write_crash_report(const char *err)
405
{
406
	char *loc = git_path("fast_import_crash_%"PRIuMAX, (uintmax_t) getpid());
407 408 409
	FILE *rpt = fopen(loc, "w");
	struct branch *b;
	unsigned long lu;
410
	struct recent_command *rc;
411 412 413 414 415 416 417 418 419

	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");
420 421
	fprintf(rpt, "    fast-import process: %"PRIuMAX"\n", (uintmax_t) getpid());
	fprintf(rpt, "    parent process     : %"PRIuMAX"\n", (uintmax_t) getppid());
422 423 424 425
	fprintf(rpt, "    at %s\n", show_date(time(NULL), 0, DATE_LOCAL));
	fputc('\n', rpt);

	fputs("fatal: ", rpt);
426
	fputs(err, rpt);
427 428
	fputc('\n', rpt);

429 430 431 432 433 434 435 436 437 438 439 440
	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);
	}

441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461
	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);
	}

462 463 464 465 466 467 468 469 470 471 472 473 474
	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);
		}
	}

475 476 477
	fputc('\n', rpt);
	fputs("Marks\n", rpt);
	fputs("-----\n", rpt);
478 479
	if (export_marks_file)
		fprintf(rpt, "  exported to %s\n", export_marks_file);
480 481 482
	else
		dump_marks_helper(rpt, 0, marks);

483 484 485 486 487 488
	fputc('\n', rpt);
	fputs("-------------------\n", rpt);
	fputs("END OF CRASH REPORT\n", rpt);
	fclose(rpt);
}

489 490 491 492
static void end_packfile(void);
static void unkeep_all_packs(void);
static void dump_marks(void);

493 494 495
static NORETURN void die_nicely(const char *err, va_list params)
{
	static int zombie;
496
	char message[2 * PATH_MAX];
497

498
	vsnprintf(message, sizeof(message), err, params);
499
	fputs("fatal: ", stderr);
500
	fputs(message, stderr);
501 502 503 504
	fputc('\n', stderr);

	if (!zombie) {
		zombie = 1;
505
		write_crash_report(message);
506 507 508
		end_packfile();
		unkeep_all_packs();
		dump_marks();
509 510 511
	}
	exit(128);
}
512

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
#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

539
static void alloc_objects(unsigned int cnt)
540
{
541
	struct object_entry_pool *b;
542

543
	b = xmalloc(sizeof(struct object_entry_pool)
544
		+ cnt * sizeof(struct object_entry));
545
	b->next_pool = blocks;
546 547 548 549 550
	b->next_free = b->entries;
	b->end = b->entries + cnt;
	blocks = b;
	alloc_count += cnt;
}
551

552
static struct object_entry *new_object(unsigned char *sha1)
553
{
554
	struct object_entry *e;
555

556
	if (blocks->next_free == blocks->end)
557
		alloc_objects(object_entry_alloc);
558

559
	e = blocks->next_free++;
560
	hashcpy(e->idx.sha1, sha1);
561
	return e;
562 563
}

564
static struct object_entry *find_object(unsigned char *sha1)
565 566 567 568
{
	unsigned int h = sha1[0] << 8 | sha1[1];
	struct object_entry *e;
	for (e = object_table[h]; e; e = e->next)
569
		if (!hashcmp(sha1, e->idx.sha1))
570 571 572 573
			return e;
	return NULL;
}

574
static struct object_entry *insert_object(unsigned char *sha1)
575 576
{
	unsigned int h = sha1[0] << 8 | sha1[1];
577
	struct object_entry *e = object_table[h];
578 579

	while (e) {
580
		if (!hashcmp(sha1, e->idx.sha1))
581 582 583 584 585
			return e;
		e = e->next;
	}

	e = new_object(sha1);
586
	e->next = object_table[h];
587
	e->idx.offset = 0;
588
	object_table[h] = e;
589 590
	return e;
}
591

592 593 594 595 596 597 598 599
static unsigned int hc_str(const char *s, size_t len)
{
	unsigned int r = 0;
	while (len-- > 0)
		r = r * 31 + *s++;
	return r;
}

600
static void *pool_alloc(size_t len)
601 602 603 604
{
	struct mem_pool *p;
	void *r;

605 606 607 608
	/* round up to a 'uintmax_t' alignment */
	if (len & (sizeof(uintmax_t) - 1))
		len += sizeof(uintmax_t) - (len & (sizeof(uintmax_t) - 1));

609 610 611 612 613 614 615 616 617 618 619 620
	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;
621
		p->next_free = (char *) p->space;
622 623 624 625 626 627 628 629 630
		p->end = p->next_free + mem_pool_alloc;
		mem_pool = p;
	}

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

631
static void *pool_calloc(size_t count, size_t size)
632 633 634 635 636 637 638
{
	size_t len = count * size;
	void *r = pool_alloc(len);
	memset(r, 0, len);
	return r;
}

639
static char *pool_strdup(const char *s)
640 641 642 643 644 645
{
	char *r = pool_alloc(strlen(s) + 1);
	strcpy(r, s);
	return r;
}

646
static void insert_mark(uintmax_t idnum, struct object_entry *oe)
647 648 649 650 651 652 653 654 655
{
	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) {
656
		uintmax_t i = idnum >> s->shift;
657 658 659 660 661 662 663 664 665 666 667 668
		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;
}

669
static struct object_entry *find_mark(uintmax_t idnum)
670
{
671
	uintmax_t orig_idnum = idnum;
672 673 674 675
	struct mark_set *s = marks;
	struct object_entry *oe = NULL;
	if ((idnum >> s->shift) < 1024) {
		while (s && s->shift) {
676
			uintmax_t i = idnum >> s->shift;
677 678 679 680 681 682 683
			idnum -= i << s->shift;
			s = s->data.sets[i];
		}
		if (s)
			oe = s->data.marked[idnum];
	}
	if (!oe)
684
		die("mark :%" PRIuMAX " not declared", orig_idnum);
685 686 687
	return oe;
}

688
static struct atom_str *to_atom(const char *s, unsigned short len)
689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706
{
	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;
}

707
static struct branch *lookup_branch(const char *name)
708 709 710 711 712 713 714 715 716 717
{
	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;
}

718
static struct branch *new_branch(const char *name)
719 720
{
	unsigned int hc = hc_str(name, strlen(name)) % branch_table_sz;
721
	struct branch *b = lookup_branch(name);
722 723 724

	if (b)
		die("Invalid attempt to create duplicate branch: %s", name);
725
	if (check_refname_format(name, REFNAME_ALLOW_ONELEVEL))
726
		die("Branch name doesn't conform to GIT standards: %s", name);
727 728 729 730

	b = pool_calloc(1, sizeof(struct branch));
	b->name = pool_strdup(name);
	b->table_next_branch = branch_table[hc];
731 732
	b->branch_tree.versions[0].mode = S_IFDIR;
	b->branch_tree.versions[1].mode = S_IFDIR;
733
	b->num_notes = 0;
734
	b->active = 0;
735
	b->pack_id = MAX_PACK_ID;
736 737 738 739 740 741 742 743 744 745 746
	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;
}

747
static struct tree_content *new_tree_content(unsigned int cnt)
748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769
{
	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;
770
	t->delta_depth = 0;
771 772 773 774 775 776 777 778
	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);
779 780 781 782 783 784
	f->next_avail = avail_tree_table[hc];
	avail_tree_table[hc] = f;
}

static void release_tree_content_recursive(struct tree_content *t)
{
785 786 787
	unsigned int i;
	for (i = 0; i < t->entry_count; i++)
		release_tree_entry(t->entries[i]);
788
	release_tree_content(t);
789 790
}

791
static struct tree_content *grow_tree_content(
792 793 794 795 796
	struct tree_content *t,
	int amt)
{
	struct tree_content *r = new_tree_content(t->entry_count + amt);
	r->entry_count = t->entry_count;
797
	r->delta_depth = t->delta_depth;
798 799 800 801 802
	memcpy(r->entries,t->entries,t->entry_count*sizeof(t->entries[0]));
	release_tree_content(t);
	return r;
}

803
static struct tree_entry *new_tree_entry(void)
804 805 806 807 808
{
	struct tree_entry *e;

	if (!avail_tree_entry) {
		unsigned int n = tree_entry_alloc;
809
		total_allocd += n * sizeof(struct tree_entry);
810
		avail_tree_entry = e = xmalloc(n * sizeof(struct tree_entry));
811
		while (n-- > 1) {
812 813 814
			*((void**)e) = e + 1;
			e++;
		}
815
		*((void**)e) = NULL;
816 817 818 819 820 821 822 823 824 825
	}

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

static void release_tree_entry(struct tree_entry *e)
{
	if (e->tree)
826
		release_tree_content_recursive(e->tree);
827 828 829 830
	*((void**)e) = avail_tree_entry;
	avail_tree_entry = e;
}

831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855
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;
}

856
static void start_packfile(void)
857
{
858
	static char tmp_file[PATH_MAX];
859
	struct packed_git *p;
860
	struct pack_header hdr;
861
	int pack_fd;
862

863
	pack_fd = odb_mkstemp(tmp_file, sizeof(tmp_file),
864
			      "pack/tmp_pack_XXXXXX");
865 866
	p = xcalloc(1, sizeof(*p) + strlen(tmp_file) + 2);
	strcpy(p->pack_name, tmp_file);
867
	p->pack_fd = pack_fd;
868
	p->do_not_close = 1;
869
	pack_file = sha1fd(pack_fd, p->pack_name);
870 871 872 873

	hdr.hdr_signature = htonl(PACK_SIGNATURE);
	hdr.hdr_version = htonl(2);
	hdr.hdr_entries = 0;
874
	sha1write(pack_file, &hdr, sizeof(hdr));
875 876

	pack_data = p;
877 878
	pack_size = sizeof(hdr);
	object_count = 0;
879 880 881

	all_packs = xrealloc(all_packs, sizeof(*all_packs) * (pack_id + 1));
	all_packs[pack_id] = p;
882 883
}

884
static const char *create_index(void)
885
{
886 887 888
	const char *tmpfile;
	struct pack_idx_entry **idx, **c, **last;
	struct object_entry *e;
889 890
	struct object_entry_pool *o;

891 892
	/* Build the table of object IDs. */
	idx = xmalloc(object_count * sizeof(*idx));
893 894
	c = idx;
	for (o = blocks; o; o = o->next_pool)
895 896
		for (e = o->next_free; e-- != o->entries;)
			if (pack_id == e->pack_id)
897
				*c++ = &e->idx;
898
	last = idx + object_count;
899 900
	if (c != last)
		die("internal consistency error creating the index");
901

902
	tmpfile = write_idx_file(NULL, idx, object_count, &pack_idx_opts, pack_data->sha1);
903
	free(idx);
904 905 906
	return tmpfile;
}

907
static char *keep_pack(const char *curr_index_name)
908 909
{
	static char name[PATH_MAX];
910
	static const char *keep_msg = "fast-import";
911 912
	int keep_fd;

913
	keep_fd = odb_pack_keep(name, sizeof(name), pack_data->sha1);
914
	if (keep_fd < 0)
915
		die_errno("cannot create keep file");
916 917
	write_or_die(keep_fd, keep_msg, strlen(keep_msg));
	if (close(keep_fd))
918
		die_errno("failed to write keep file");
919 920 921 922 923 924 925 926 927 928

	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");
929
	free((void *)curr_index_name);
930 931 932
	return name;
}

933
static void unkeep_all_packs(void)
934 935 936 937 938 939 940 941
{
	static char name[PATH_MAX];
	int k;

	for (k = 0; k < pack_id; k++) {
		struct packed_git *p = all_packs[k];
		snprintf(name, sizeof(name), "%s/pack/pack-%s.keep",
			 get_object_directory(), sha1_to_hex(p->sha1));
942
		unlink_or_warn(name);
943
	}
944 945
}

946
static void end_packfile(void)
947
{
948 949
	struct packed_git *old_p = pack_data, *new_p;

950
	clear_delta_base_cache();
951
	if (object_count) {
952
		unsigned char cur_pack_sha1[20];
953
		char *idx_name;
954 955 956
		int i;
		struct branch *b;
		struct tag *t;
957

958
		close_pack_windows(pack_data);
959
		sha1close(pack_file, cur_pack_sha1, 0);
960
		fixup_pack_header_footer(pack_data->pack_fd, pack_data->sha1,
961
				    pack_data->pack_name, object_count,
962
				    cur_pack_sha1, pack_size);
963
		close(pack_data->pack_fd);
964
		idx_name = keep_pack(create_index());
965

M
Mike Ralphson 已提交
966
		/* Register the packfile with core git's machinery. */
967 968 969
		new_p = add_packed_git(idx_name, strlen(idx_name), 1);
		if (!new_p)
			die("core git rejected index %s", idx_name);
970
		all_packs[pack_id] = new_p;
971
		install_packed_git(new_p);
972 973

		/* Print the boundary */
974 975 976 977 978 979 980
		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));
				}
981
			}
982 983 984 985 986 987
			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);
988 989 990
		}

		pack_id++;
991
	}
992 993
	else {
		close(old_p->pack_fd);
994
		unlink_or_warn(old_p->pack_name);
995
	}
996 997 998
	free(old_p);

	/* We can't carry a delta across packfiles. */
P
Pierre Habouzit 已提交
999
	strbuf_release(&last_blob.data);
1000 1001
	last_blob.offset = 0;
	last_blob.depth = 0;
1002 1003
}

1004
static void cycle_packfile(void)
1005 1006 1007 1008 1009
{
	end_packfile();
	start_packfile();
}

1010 1011
static int store_object(
	enum object_type type,
1012
	struct strbuf *dat,
1013
	struct last_object *last,
1014
	unsigned char *sha1out,
1015
	uintmax_t mark)
1016 1017
{
	void *out, *delta;
1018 1019 1020
	struct object_entry *e;
	unsigned char hdr[96];
	unsigned char sha1[20];
1021
	unsigned long hdrlen, deltalen;
1022
	git_SHA_CTX c;
1023
	git_zstream s;
1024

1025
	hdrlen = sprintf((char *)hdr,"%s %lu", typename(type),
1026
		(unsigned long)dat->len) + 1;
1027 1028 1029 1030
	git_SHA1_Init(&c);
	git_SHA1_Update(&c, hdr, hdrlen);
	git_SHA1_Update(&c, dat->buf, dat->len);
	git_SHA1_Final(sha1, &c);
1031
	if (sha1out)
1032
		hashcpy(sha1out, sha1);
1033 1034

	e = insert_object(sha1);
1035 1036
	if (mark)
		insert_mark(mark, e);
1037
	if (e->idx.offset) {
1038
		duplicate_count_by_type[type]++;
1039
		return 1;
1040 1041 1042
	} else if (find_sha1_pack(sha1, packed_git)) {
		e->type = type;
		e->pack_id = MAX_PACK_ID;
1043
		e->idx.offset = 1; /* just not zero! */
1044 1045
		duplicate_count_by_type[type]++;
		return 1;
1046
	}
1047

1048
	if (last && last->data.buf && last->depth < max_depth && dat->len > 20) {
1049
		delta_count_attempts_by_type[type]++;
P
Pierre Habouzit 已提交
1050
		delta = diff_delta(last->data.buf, last->data.len,
1051
			dat->buf, dat->len,
1052
			&deltalen, dat->len - 20);
1053 1054
	} else
		delta = NULL;
1055 1056

	memset(&s, 0, sizeof(s));
1057
	git_deflate_init(&s, pack_compression_level);
1058 1059 1060 1061
	if (delta) {
		s.next_in = delta;
		s.avail_in = deltalen;
	} else {
1062 1063
		s.next_in = (void *)dat->buf;
		s.avail_in = dat->len;
1064
	}
J
Junio C Hamano 已提交
1065
	s.avail_out = git_deflate_bound(&s, s.avail_in);
1066
	s.next_out = out = xmalloc(s.avail_out);
1067 1068 1069
	while (git_deflate(&s, Z_FINISH) == Z_OK)
		; /* nothing */
	git_deflate_end(&s);
1070 1071

	/* Determine if we should auto-checkpoint. */
1072
	if ((max_packsize && (pack_size + 60 + s.total_out) > max_packsize)
1073 1074 1075 1076
		|| (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;
1077
		cycle_packfile();
1078 1079 1080 1081 1082

		/* We cannot carry a delta into the new pack. */
		if (delta) {
			free(delta);
			delta = NULL;
1083 1084

			memset(&s, 0, sizeof(s));
1085
			git_deflate_init(&s, pack_compression_level);
1086 1087
			s.next_in = (void *)dat->buf;
			s.avail_in = dat->len;
J
Junio C Hamano 已提交
1088
			s.avail_out = git_deflate_bound(&s, s.avail_in);
1089
			s.next_out = out = xrealloc(out, s.avail_out);
1090 1091 1092
			while (git_deflate(&s, Z_FINISH) == Z_OK)
				; /* nothing */
			git_deflate_end(&s);
1093 1094 1095 1096 1097
		}
	}

	e->type = type;
	e->pack_id = pack_id;
1098
	e->idx.offset = pack_size;
1099 1100
	object_count++;
	object_count_by_type[type]++;
1101

1102 1103
	crc32_begin(pack_file);

1104
	if (delta) {
1105
		off_t ofs = e->idx.offset - last->offset;
1106 1107
		unsigned pos = sizeof(hdr) - 1;

1108
		delta_count_by_type[type]++;
1109
		e->depth = last->depth + 1;
1110

1111
		hdrlen = encode_in_pack_object_header(OBJ_OFS_DELTA, deltalen, hdr);
1112
		sha1write(pack_file, hdr, hdrlen);
1113 1114 1115 1116 1117
		pack_size += hdrlen;

		hdr[pos] = ofs & 127;
		while (ofs >>= 7)
			hdr[--pos] = 128 | (--ofs & 127);
1118
		sha1write(pack_file, hdr + pos, sizeof(hdr) - pos);
1119
		pack_size += sizeof(hdr) - pos;
1120
	} else {
1121
		e->depth = 0;
1122
		hdrlen = encode_in_pack_object_header(type, dat->len, hdr);
1123
		sha1write(pack_file, hdr, hdrlen);
1124
		pack_size += hdrlen;
1125 1126
	}

1127
	sha1write(pack_file, out, s.total_out);
1128
	pack_size += s.total_out;
1129

1130 1131
	e->idx.crc32 = crc32_end(pack_file);

1132
	free(out);
1133
	free(delta);
1134
	if (last) {
P
Pierre Habouzit 已提交
1135 1136 1137
		if (last->no_swap) {
			last->data = *dat;
		} else {
1138
			strbuf_swap(&last->data, dat);
P
Pierre Habouzit 已提交
1139
		}
1140
		last->offset = e->idx.offset;
1141
		last->depth = e->depth;
1142 1143 1144 1145
	}
	return 0;
}

1146
static void truncate_pack(struct sha1file_checkpoint *checkpoint)
1147
{
1148
	if (sha1file_truncate(pack_file, checkpoint))
1149
		die_errno("cannot truncate pack to skip duplicate");
1150
	pack_size = checkpoint->offset;
1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162
}

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;
1163
	git_zstream s;
1164
	struct sha1file_checkpoint checkpoint;
1165 1166 1167
	int status = Z_OK;

	/* Determine if we should auto-checkpoint. */
1168
	if ((max_packsize && (pack_size + 60 + len) > max_packsize)
1169 1170 1171
		|| (pack_size + 60 + len) < pack_size)
		cycle_packfile();

1172 1173
	sha1file_checkpoint(pack_file, &checkpoint);
	offset = checkpoint.offset;
1174

1175 1176 1177 1178 1179 1180 1181
	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);

1182 1183
	crc32_begin(pack_file);

1184
	memset(&s, 0, sizeof(s));
1185
	git_deflate_init(&s, pack_compression_level);
1186

1187
	hdrlen = encode_in_pack_object_header(OBJ_BLOB, len, out_buf);
1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206
	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;
		}

1207
		status = git_deflate(&s, len ? 0 : Z_FINISH);
1208 1209 1210

		if (!s.avail_out || status == Z_STREAM_END) {
			size_t n = s.next_out - out_buf;
1211
			sha1write(pack_file, out_buf, n);
1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225
			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);
		}
	}
1226
	git_deflate_end(&s);
1227 1228 1229 1230 1231 1232 1233 1234 1235 1236
	git_SHA1_Final(sha1, &c);

	if (sha1out)
		hashcpy(sha1out, sha1);

	e = insert_object(sha1);

	if (mark)
		insert_mark(mark, e);

1237
	if (e->idx.offset) {
1238
		duplicate_count_by_type[OBJ_BLOB]++;
1239
		truncate_pack(&checkpoint);
1240 1241 1242 1243

	} else if (find_sha1_pack(sha1, packed_git)) {
		e->type = OBJ_BLOB;
		e->pack_id = MAX_PACK_ID;
1244
		e->idx.offset = 1; /* just not zero! */
1245
		duplicate_count_by_type[OBJ_BLOB]++;
1246
		truncate_pack(&checkpoint);
1247 1248 1249 1250 1251

	} else {
		e->depth = 0;
		e->type = OBJ_BLOB;
		e->pack_id = pack_id;
1252
		e->idx.offset = offset;
1253
		e->idx.crc32 = crc32_end(pack_file);
1254 1255 1256 1257 1258 1259 1260 1261
		object_count++;
		object_count_by_type[OBJ_BLOB]++;
	}

	free(in_buf);
	free(out_buf);
}

1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279
/* 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.
 */
1280 1281 1282
static void *gfi_unpack_entry(
	struct object_entry *oe,
	unsigned long *sizep)
1283
{
1284
	enum object_type type;
1285
	struct packed_git *p = all_packs[oe->pack_id];
1286
	if (p == pack_data && p->pack_size < (pack_size + 20)) {
1287 1288 1289 1290 1291 1292 1293 1294
		/* 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.
		 */
1295
		close_pack_windows(p);
1296
		sha1flush(pack_file);
1297 1298 1299 1300 1301 1302 1303

		/* 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.
		 */
1304
		p->pack_size = pack_size + 20;
1305
	}
1306
	return unpack_entry(p, oe->idx.offset, &type, sizep);
1307 1308
}

1309
static const char *get_mode(const char *str, uint16_t *modep)
1310 1311
{
	unsigned char c;
1312
	uint16_t mode = 0;
1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324

	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)
{
1325
	unsigned char *sha1 = root->versions[1].sha1;
1326 1327 1328 1329 1330 1331 1332
	struct object_entry *myoe;
	struct tree_content *t;
	unsigned long size;
	char *buf;
	const char *c;

	root->tree = t = new_tree_content(8);
1333
	if (is_null_sha1(sha1))
1334 1335
		return;

1336
	myoe = find_object(sha1);
1337
	if (myoe && myoe->pack_id != MAX_PACK_ID) {
1338
		if (myoe->type != OBJ_TREE)
1339
			die("Not a tree: %s", sha1_to_hex(sha1));
1340
		t->delta_depth = myoe->depth;
1341
		buf = gfi_unpack_entry(myoe, &size);
1342 1343
		if (!buf)
			die("Can't load tree %s", sha1_to_hex(sha1));
1344
	} else {
1345 1346 1347
		enum object_type type;
		buf = read_sha1_file(sha1, &type, &size);
		if (!buf || type != OBJ_TREE)
1348
			die("Can't load tree %s", sha1_to_hex(sha1));
1349 1350 1351 1352 1353 1354 1355
	}

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

		if (t->entry_count == t->entry_capacity)
1356
			root->tree = t = grow_tree_content(t, t->entry_count);
1357 1358 1359
		t->entries[t->entry_count++] = e;

		e->tree = NULL;
1360
		c = get_mode(c, &e->versions[1].mode);
1361
		if (!c)
1362 1363
			die("Corrupt mode in %s", sha1_to_hex(sha1));
		e->versions[0].mode = e->versions[1].mode;
1364
		e->name = to_atom(c, strlen(c));
1365
		c += e->name->str_len + 1;
1366 1367
		hashcpy(e->versions[0].sha1, (unsigned char *)c);
		hashcpy(e->versions[1].sha1, (unsigned char *)c);
1368 1369 1370 1371 1372
		c += 20;
	}
	free(buf);
}

1373
static int tecmp0 (const void *_a, const void *_b)
1374 1375 1376 1377
{
	struct tree_entry *a = *((struct tree_entry**)_a);
	struct tree_entry *b = *((struct tree_entry**)_b);
	return base_name_compare(
1378 1379
		a->name->str_dat, a->name->str_len, a->versions[0].mode,
		b->name->str_dat, b->name->str_len, b->versions[0].mode);
1380 1381
}

1382
static int tecmp1 (const void *_a, const void *_b)
1383
{
1384 1385 1386 1387 1388 1389 1390
	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);
}

1391
static void mktree(struct tree_content *t, int v, struct strbuf *b)
1392 1393
{
	size_t maxlen = 0;
1394 1395
	unsigned int i;

1396 1397 1398 1399
	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);
1400 1401

	for (i = 0; i < t->entry_count; i++) {
1402 1403
		if (t->entries[i]->versions[v].mode)
			maxlen += t->entries[i]->name->str_len + 34;
1404 1405
	}

1406 1407
	strbuf_reset(b);
	strbuf_grow(b, maxlen);
1408 1409
	for (i = 0; i < t->entry_count; i++) {
		struct tree_entry *e = t->entries[i];
1410 1411
		if (!e->versions[v].mode)
			continue;
1412 1413 1414
		strbuf_addf(b, "%o %s%c",
			(unsigned int)(e->versions[v].mode & ~NO_DELTA),
			e->name->str_dat, '\0');
1415
		strbuf_add(b, e->versions[v].sha1, 20);
1416
	}
1417 1418 1419 1420 1421 1422
}

static void store_tree(struct tree_entry *root)
{
	struct tree_content *t = root->tree;
	unsigned int i, j, del;
P
Pierre Habouzit 已提交
1423
	struct last_object lo = { STRBUF_INIT, 0, 0, /* no_swap */ 1 };
1424
	struct object_entry *le = NULL;
1425 1426 1427 1428 1429 1430 1431 1432 1433

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

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

1434 1435
	if (!(root->versions[0].mode & NO_DELTA))
		le = find_object(root->versions[0].sha1);
P
Pierre Habouzit 已提交
1436
	if (S_ISDIR(root->versions[0].mode) && le && le->pack_id == pack_id) {
1437
		mktree(t, 0, &old_tree);
P
Pierre Habouzit 已提交
1438
		lo.data = old_tree;
1439
		lo.offset = le->idx.offset;
1440 1441 1442
		lo.depth = t->delta_depth;
	}

1443 1444
	mktree(t, 1, &new_tree);
	store_object(OBJ_TREE, &new_tree, &lo, root->versions[1].sha1, 0);
1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458

	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;
1459 1460
}

1461 1462 1463 1464 1465 1466 1467 1468
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");
1469
	hashclr(root->versions[0].sha1);
1470 1471 1472 1473 1474 1475
	hashcpy(root->versions[1].sha1, sha1);
	if (root->tree)
		release_tree_content_recursive(root->tree);
	root->tree = newtree;
}

1476 1477 1478 1479
static int tree_content_set(
	struct tree_entry *root,
	const char *p,
	const unsigned char *sha1,
1480 1481
	const uint16_t mode,
	struct tree_content *subtree)
1482
{
1483
	struct tree_content *t;
1484 1485 1486 1487
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

1488 1489
	slash1 = strchrnul(p, '/');
	n = slash1 - p;
1490 1491
	if (!n)
		die("Empty path component found in input");
1492
	if (!*slash1 && !S_ISDIR(mode) && subtree)
1493
		die("Non-directories cannot have subtrees");
1494

1495 1496 1497
	if (!root->tree)
		load_tree(root);
	t = root->tree;
1498 1499
	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
1500
		if (e->name->str_len == n && !strncmp_icase(p, e->name->str_dat, n)) {
1501
			if (!*slash1) {
1502 1503
				if (!S_ISDIR(mode)
						&& e->versions[1].mode == mode
1504
						&& !hashcmp(e->versions[1].sha1, sha1))
1505
					return 0;
1506 1507
				e->versions[1].mode = mode;
				hashcpy(e->versions[1].sha1, sha1);
1508
				if (e->tree)
1509
					release_tree_content_recursive(e->tree);
1510
				e->tree = subtree;
1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527

				/*
				 * 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;

1528
				hashclr(root->versions[1].sha1);
1529 1530
				return 1;
			}
1531
			if (!S_ISDIR(e->versions[1].mode)) {
1532
				e->tree = new_tree_content(8);
1533
				e->versions[1].mode = S_IFDIR;
1534 1535 1536
			}
			if (!e->tree)
				load_tree(e);
1537
			if (tree_content_set(e, slash1 + 1, sha1, mode, subtree)) {
1538
				hashclr(root->versions[1].sha1);
1539 1540 1541 1542 1543 1544 1545
				return 1;
			}
			return 0;
		}
	}

	if (t->entry_count == t->entry_capacity)
1546
		root->tree = t = grow_tree_content(t, t->entry_count);
1547
	e = new_tree_entry();
1548
	e->name = to_atom(p, n);
1549 1550
	e->versions[0].mode = 0;
	hashclr(e->versions[0].sha1);
1551
	t->entries[t->entry_count++] = e;
1552
	if (*slash1) {
1553
		e->tree = new_tree_content(8);
1554
		e->versions[1].mode = S_IFDIR;
1555
		tree_content_set(e, slash1 + 1, sha1, mode, subtree);
1556
	} else {
1557
		e->tree = subtree;
1558 1559
		e->versions[1].mode = mode;
		hashcpy(e->versions[1].sha1, sha1);
1560
	}
1561
	hashclr(root->versions[1].sha1);
1562 1563 1564
	return 1;
}

1565 1566 1567
static int tree_content_remove(
	struct tree_entry *root,
	const char *p,
1568 1569
	struct tree_entry *backup_leaf,
	int allow_root)
1570
{
1571
	struct tree_content *t;
1572 1573 1574 1575
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

1576 1577
	slash1 = strchrnul(p, '/');
	n = slash1 - p;
1578

1579 1580
	if (!root->tree)
		load_tree(root);
1581 1582 1583 1584 1585 1586

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

1587
	t = root->tree;
1588 1589
	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
1590
		if (e->name->str_len == n && !strncmp_icase(p, e->name->str_dat, n)) {
1591
			if (*slash1 && !S_ISDIR(e->versions[1].mode))
1592 1593 1594 1595 1596 1597 1598
				/*
				 * 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;
1599
			if (!*slash1 || !S_ISDIR(e->versions[1].mode))
1600 1601 1602
				goto del_entry;
			if (!e->tree)
				load_tree(e);
1603
			if (tree_content_remove(e, slash1 + 1, backup_leaf, 0)) {
1604 1605 1606 1607 1608 1609
				for (n = 0; n < e->tree->entry_count; n++) {
					if (e->tree->entries[n]->versions[1].mode) {
						hashclr(root->versions[1].sha1);
						return 1;
					}
				}
1610
				backup_leaf = NULL;
1611
				goto del_entry;
1612 1613 1614 1615 1616 1617 1618
			}
			return 0;
		}
	}
	return 0;

del_entry:
1619 1620 1621
	if (backup_leaf)
		memcpy(backup_leaf, e, sizeof(*backup_leaf));
	else if (e->tree)
1622
		release_tree_content_recursive(e->tree);
1623
	e->tree = NULL;
1624 1625 1626
	e->versions[1].mode = 0;
	hashclr(e->versions[1].sha1);
	hashclr(root->versions[1].sha1);
1627
	return 1;
1628 1629
}

1630 1631 1632
static int tree_content_get(
	struct tree_entry *root,
	const char *p,
1633 1634
	struct tree_entry *leaf,
	int allow_root)
1635
{
1636
	struct tree_content *t;
1637 1638 1639 1640
	const char *slash1;
	unsigned int i, n;
	struct tree_entry *e;

1641 1642
	slash1 = strchrnul(p, '/');
	n = slash1 - p;
1643
	if (!n && !allow_root)
1644
		die("Empty path component found in input");
1645

1646 1647
	if (!root->tree)
		load_tree(root);
1648 1649 1650 1651 1652 1653

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

1654
	t = root->tree;
1655 1656
	for (i = 0; i < t->entry_count; i++) {
		e = t->entries[i];
1657
		if (e->name->str_len == n && !strncmp_icase(p, e->name->str_dat, n)) {
1658
			if (!*slash1)
1659
				goto found_entry;
1660 1661 1662 1663
			if (!S_ISDIR(e->versions[1].mode))
				return 0;
			if (!e->tree)
				load_tree(e);
1664
			return tree_content_get(e, slash1 + 1, leaf, 0);
1665 1666 1667
		}
	}
	return 0;
1668 1669 1670 1671 1672 1673 1674 1675

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;
1676 1677
}

1678
static int update_branch(struct branch *b)
1679 1680
{
	static const char *msg = "fast-import";
1681 1682 1683
	struct ref_lock *lock;
	unsigned char old_sha1[20];

1684 1685
	if (is_null_sha1(b->sha1))
		return 0;
1686 1687
	if (read_ref(b->name, old_sha1))
		hashclr(old_sha1);
1688
	lock = lock_any_ref_for_update(b->name, old_sha1, 0, NULL);
1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700
	if (!lock)
		return error("Unable to lock %s", b->name);
	if (!force_update && !is_null_sha1(old_sha1)) {
		struct commit *old_cmit, *new_cmit;

		old_cmit = lookup_commit_reference_gently(old_sha1, 0);
		new_cmit = lookup_commit_reference_gently(b->sha1, 0);
		if (!old_cmit || !new_cmit) {
			unlock_ref(lock);
			return error("Branch %s is missing commits.", b->name);
		}

1701
		if (!in_merge_bases(old_cmit, new_cmit)) {
1702
			unlock_ref(lock);
1703
			warning("Not updating %s"
1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715
				" (new tip %s does not contain %s)",
				b->name, sha1_to_hex(b->sha1), sha1_to_hex(old_sha1));
			return -1;
		}
	}
	if (write_ref_sha1(lock, b->sha1, msg) < 0)
		return error("Unable to update %s", b->name);
	return 0;
}

static void dump_branches(void)
{
1716 1717 1718 1719
	unsigned int i;
	struct branch *b;

	for (i = 0; i < branch_table_sz; i++) {
1720 1721
		for (b = branch_table[i]; b; b = b->table_next_branch)
			failure |= update_branch(b);
1722 1723 1724
	}
}

1725
static void dump_tags(void)
1726 1727 1728 1729
{
	static const char *msg = "fast-import";
	struct tag *t;
	struct ref_lock *lock;
1730
	char ref_name[PATH_MAX];
1731 1732

	for (t = first_tag; t; t = t->next_tag) {
1733 1734
		sprintf(ref_name, "tags/%s", t->name);
		lock = lock_ref_sha1(ref_name, NULL);
1735
		if (!lock || write_ref_sha1(lock, t->sha1, msg) < 0)
1736
			failure |= error("Unable to update %s", ref_name);
1737 1738 1739
	}
}

1740
static void dump_marks_helper(FILE *f,
1741
	uintmax_t base,
1742 1743
	struct mark_set *m)
{
1744
	uintmax_t k;
1745 1746 1747
	if (m->shift) {
		for (k = 0; k < 1024; k++) {
			if (m->data.sets[k])
1748
				dump_marks_helper(f, base + (k << m->shift),
1749 1750 1751 1752 1753
					m->data.sets[k]);
		}
	} else {
		for (k = 0; k < 1024; k++) {
			if (m->data.marked[k])
1754
				fprintf(f, ":%" PRIuMAX " %s\n", base + k,
1755
					sha1_to_hex(m->data.marked[k]->idx.sha1));
1756 1757 1758 1759
		}
	}
}

1760
static void dump_marks(void)
1761
{
1762 1763 1764 1765
	static struct lock_file mark_lock;
	int mark_fd;
	FILE *f;

1766
	if (!export_marks_file)
1767 1768
		return;

1769
	mark_fd = hold_lock_file_for_update(&mark_lock, export_marks_file, 0);
1770 1771
	if (mark_fd < 0) {
		failure |= error("Unable to write marks file %s: %s",
1772
			export_marks_file, strerror(errno));
1773
		return;
1774
	}
1775 1776 1777

	f = fdopen(mark_fd, "w");
	if (!f) {
1778
		int saved_errno = errno;
1779 1780
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1781
			export_marks_file, strerror(saved_errno));
1782
		return;
1783
	}
1784

B
Brandon Casey 已提交
1785
	/*
1786 1787
	 * 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
1788 1789 1790
	 * won't try to close() it.
	 */
	mark_lock.fd = -1;
1791 1792 1793

	dump_marks_helper(f, 0, marks);
	if (ferror(f) || fclose(f)) {
1794
		int saved_errno = errno;
1795 1796
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to write marks file %s: %s",
1797
			export_marks_file, strerror(saved_errno));
1798 1799 1800 1801
		return;
	}

	if (commit_lock_file(&mark_lock)) {
1802
		int saved_errno = errno;
1803 1804
		rollback_lock_file(&mark_lock);
		failure |= error("Unable to commit marks file %s: %s",
1805
			export_marks_file, strerror(saved_errno));
1806 1807
		return;
	}
1808 1809
}

1810 1811 1812 1813
static void read_marks(void)
{
	char line[512];
	FILE *f = fopen(import_marks_file, "r");
1814 1815 1816 1817 1818
	if (f)
		;
	else if (import_marks_file_ignore_missing && errno == ENOENT)
		return; /* Marks file does not exist */
	else
1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831
		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
1832
			|| *end != ' ' || get_sha1_hex(end + 1, sha1))
1833 1834 1835 1836 1837 1838 1839 1840 1841
			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;
1842
			e->idx.offset = 1; /* just not zero! */
1843 1844 1845 1846 1847 1848 1849
		}
		insert_mark(mark, e);
	}
	fclose(f);
}


1850
static int read_next_command(void)
1851
{
1852 1853 1854 1855 1856 1857 1858
	static int stdin_eof = 0;

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

1859
	for (;;) {
1860
		if (unread_command_buf) {
1861
			unread_command_buf = 0;
1862 1863 1864
		} else {
			struct recent_command *rc;

1865
			strbuf_detach(&command_buf, NULL);
1866 1867 1868
			stdin_eof = strbuf_getline(&command_buf, stdin, '\n');
			if (stdin_eof)
				return EOF;
1869

1870
			if (!seen_data_command
1871 1872
				&& !starts_with(command_buf.buf, "feature ")
				&& !starts_with(command_buf.buf, "option ")) {
S
Sverre Rabbelier 已提交
1873
				parse_argv();
1874 1875
			}

1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891
			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;
		}
1892
		if (starts_with(command_buf.buf, "cat-blob ")) {
1893 1894 1895 1896 1897 1898 1899
			parse_cat_blob();
			continue;
		}
		if (command_buf.buf[0] == '#')
			continue;
		return 0;
	}
1900 1901
}

J
Junio C Hamano 已提交
1902
static void skip_optional_lf(void)
1903 1904 1905 1906 1907 1908
{
	int term_char = fgetc(stdin);
	if (term_char != '\n' && term_char != EOF)
		ungetc(term_char, stdin);
}

1909
static void parse_mark(void)
1910
{
1911
	if (starts_with(command_buf.buf, "mark :")) {
1912
		next_mark = strtoumax(command_buf.buf + 6, NULL, 10);
1913 1914 1915
		read_next_command();
	}
	else
1916
		next_mark = 0;
1917 1918
}

1919
static int parse_data(struct strbuf *sb, uintmax_t limit, uintmax_t *len_res)
1920
{
1921
	strbuf_reset(sb);
1922

1923
	if (!starts_with(command_buf.buf, "data "))
1924 1925
		die("Expected 'data n' command, found: %s", command_buf.buf);

1926
	if (starts_with(command_buf.buf + 5, "<<")) {
1927
		char *term = xstrdup(command_buf.buf + 5 + 2);
1928 1929
		size_t term_len = command_buf.len - 5 - 2;

1930
		strbuf_detach(&command_buf, NULL);
1931
		for (;;) {
1932
			if (strbuf_getline(&command_buf, stdin, '\n') == EOF)
1933 1934 1935 1936
				die("EOF in data (terminator '%s' not found)", term);
			if (term_len == command_buf.len
				&& !strcmp(term, command_buf.buf))
				break;
1937 1938
			strbuf_addbuf(sb, &command_buf);
			strbuf_addch(sb, '\n');
1939 1940 1941 1942
		}
		free(term);
	}
	else {
1943 1944
		uintmax_t len = strtoumax(command_buf.buf + 5, NULL, 10);
		size_t n = 0, length = (size_t)len;
1945

1946 1947 1948 1949 1950 1951
		if (limit && limit < len) {
			*len_res = len;
			return 0;
		}
		if (length < len)
			die("data is too large to use in this context");
1952

1953
		while (n < length) {
1954
			size_t s = strbuf_fread(sb, length - n, stdin);
1955
			if (!s && feof(stdin))
1956 1957
				die("EOF in data (%lu bytes remaining)",
					(unsigned long)(length - n));
1958 1959
			n += s;
		}
1960 1961
	}

1962
	skip_optional_lf();
1963
	return 1;
1964 1965
}

1966 1967 1968
static int validate_raw_date(const char *src, char *result, int maxlen)
{
	const char *orig_src = src;
1969
	char *endp;
1970
	unsigned long num;
1971

1972 1973
	errno = 0;

1974 1975
	num = strtoul(src, &endp, 10);
	/* NEEDSWORK: perhaps check for reasonable values? */
1976
	if (errno || endp == src || *endp != ' ')
1977 1978 1979 1980 1981 1982
		return -1;

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

1983 1984 1985
	num = strtoul(src + 1, &endp, 10);
	if (errno || endp == src + 1 || *endp || (endp - orig_src) >= maxlen ||
	    1400 < num)
1986 1987 1988 1989 1990 1991 1992 1993
		return -1;

	strcpy(result, orig_src);
	return 0;
}

static char *parse_ident(const char *buf)
{
1994
	const char *ltgt;
1995 1996 1997
	size_t name_len;
	char *ident;

1998 1999 2000 2001
	/* ensure there is a space delimiter even if there is no name */
	if (*buf == '<')
		--buf;

2002 2003 2004 2005 2006 2007 2008
	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 != '>')
2009
		die("Missing > in ident string: %s", buf);
2010 2011
	ltgt++;
	if (*ltgt != ' ')
2012
		die("Missing space after > in ident string: %s", buf);
2013 2014
	ltgt++;
	name_len = ltgt - buf;
2015 2016 2017 2018 2019
	ident = xmalloc(name_len + 24);
	strncpy(ident, buf, name_len);

	switch (whenspec) {
	case WHENSPEC_RAW:
2020 2021
		if (validate_raw_date(ltgt, ident + name_len, 24) < 0)
			die("Invalid raw date \"%s\" in ident: %s", ltgt, buf);
2022 2023
		break;
	case WHENSPEC_RFC2822:
2024 2025
		if (parse_date(ltgt, ident + name_len, 24) < 0)
			die("Invalid rfc2822 date \"%s\" in ident: %s", ltgt, buf);
2026 2027
		break;
	case WHENSPEC_NOW:
2028
		if (strcmp("now", ltgt))
2029 2030 2031 2032 2033 2034 2035 2036
			die("Date in ident must be 'now': %s", buf);
		datestamp(ident + name_len, 24);
		break;
	}

	return ident;
}

2037 2038 2039 2040
static void parse_and_store_blob(
	struct last_object *last,
	unsigned char *sha1out,
	uintmax_t mark)
2041
{
P
Pierre Habouzit 已提交
2042
	static struct strbuf buf = STRBUF_INIT;
2043
	uintmax_t len;
2044

2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059
	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)
{
2060
	read_next_command();
2061
	parse_mark();
2062
	parse_and_store_blob(&last_blob, NULL, next_mark);
2063 2064
}

2065
static void unload_one_branch(void)
2066
{
2067 2068
	while (cur_active_branches
		&& cur_active_branches >= max_active_branches) {
2069
		uintmax_t min_commit = ULONG_MAX;
2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086
		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;
		}
2087
		e->active = 0;
2088 2089
		e->active_next_branch = NULL;
		if (e->branch_tree.tree) {
2090
			release_tree_content_recursive(e->branch_tree.tree);
2091 2092 2093
			e->branch_tree.tree = NULL;
		}
		cur_active_branches--;
2094 2095 2096
	}
}

2097
static void load_branch(struct branch *b)
2098
{
2099
	load_tree(&b->branch_tree);
2100 2101 2102 2103 2104 2105 2106
	if (!b->active) {
		b->active = 1;
		b->active_next_branch = active_branches;
		active_branches = b;
		cur_active_branches++;
		branch_load_count++;
	}
2107 2108
}

2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 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
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 */
2176 2177 2178 2179 2180
			if (fanout == 0xff) {
				/* Counting mode, no rename */
				num_notes++;
				continue;
			}
2181 2182 2183 2184 2185 2186 2187 2188
			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 */
2189
			if (!tree_content_remove(orig_root, fullpath, &leaf, 0))
2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216
				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);
}

2217 2218 2219 2220 2221 2222 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 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269
/*
 * 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);
	if (*end != ' ')
		die("Missing space after mark: %s", command_buf.buf);
	*p = end;
	return mark;
}

2270
static void file_change_m(struct branch *b)
2271
{
2272
	const char *p = command_buf.buf + 2;
2273
	static struct strbuf uq = STRBUF_INIT;
2274
	const char *endp;
2275
	struct object_entry *oe;
2276
	unsigned char sha1[20];
2277
	uint16_t mode, inline_data = 0;
2278

2279 2280 2281 2282
	p = get_mode(p, &mode);
	if (!p)
		die("Corrupt mode: %s", command_buf.buf);
	switch (mode) {
2283 2284 2285
	case 0644:
	case 0755:
		mode |= S_IFREG;
2286 2287
	case S_IFREG | 0644:
	case S_IFREG | 0755:
2288
	case S_IFLNK:
2289
	case S_IFDIR:
2290
	case S_IFGITLINK:
2291 2292 2293 2294 2295 2296
		/* ok */
		break;
	default:
		die("Corrupt mode: %s", command_buf.buf);
	}

2297
	if (*p == ':') {
2298
		oe = find_mark(parse_mark_ref_space(&p));
2299
		hashcpy(sha1, oe->idx.sha1);
2300
	} else if (starts_with(p, "inline ")) {
2301
		inline_data = 1;
2302
		oe = NULL; /* not used with inline_data, but makes gcc happy */
2303
		p += strlen("inline");  /* advance to space */
2304 2305
	} else {
		if (get_sha1_hex(p, sha1))
2306
			die("Invalid dataref: %s", command_buf.buf);
2307 2308
		oe = find_object(sha1);
		p += 40;
2309 2310
		if (*p != ' ')
			die("Missing space after SHA1: %s", command_buf.buf);
2311
	}
2312 2313
	assert(*p == ' ');
	p++;  /* skip space */
2314

2315 2316
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
2317 2318
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
2319
		p = uq.buf;
2320
	}
2321

2322 2323
	/* Git does not track empty, non-toplevel directories. */
	if (S_ISDIR(mode) && !memcmp(sha1, EMPTY_TREE_SHA1_BIN, 20) && *p) {
2324
		tree_content_remove(&b->branch_tree, p, NULL, 0);
2325 2326 2327
		return;
	}

2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341
	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) {
2342 2343 2344
		if (S_ISDIR(mode))
			die("Directories cannot be specified 'inline': %s",
				command_buf.buf);
2345 2346 2347 2348
		if (p != uq.buf) {
			strbuf_addstr(&uq, p);
			p = uq.buf;
		}
2349
		read_next_command();
2350
		parse_and_store_blob(&last_blob, sha1, 0);
2351
	} else {
2352 2353 2354 2355
		enum object_type expected = S_ISDIR(mode) ?
						OBJ_TREE: OBJ_BLOB;
		enum object_type type = oe ? oe->type :
					sha1_object_info(sha1, NULL);
2356
		if (type < 0)
2357 2358 2359 2360 2361 2362 2363
			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);
2364
	}
2365

2366 2367 2368 2369
	if (!*p) {
		tree_content_replace(&b->branch_tree, sha1, mode, NULL);
		return;
	}
2370
	tree_content_set(&b->branch_tree, p, sha1, mode, NULL);
2371
}
2372

2373 2374
static void file_change_d(struct branch *b)
{
2375
	const char *p = command_buf.buf + 2;
2376
	static struct strbuf uq = STRBUF_INIT;
2377 2378
	const char *endp;

2379 2380
	strbuf_reset(&uq);
	if (!unquote_c_style(&uq, p, &endp)) {
2381 2382
		if (*endp)
			die("Garbage after path in: %s", command_buf.buf);
2383
		p = uq.buf;
2384
	}
2385
	tree_content_remove(&b->branch_tree, p, NULL, 1);
2386 2387
}

2388
static void file_change_cr(struct branch *b, int rename)
2389 2390
{
	const char *s, *d;
2391 2392
	static struct strbuf s_uq = STRBUF_INIT;
	static struct strbuf d_uq = STRBUF_INIT;
2393 2394 2395 2396
	const char *endp;
	struct tree_entry leaf;

	s = command_buf.buf + 2;
2397 2398
	strbuf_reset(&s_uq);
	if (!unquote_c_style(&s_uq, s, &endp)) {
2399 2400
		if (*endp != ' ')
			die("Missing space after source: %s", command_buf.buf);
2401
	} else {
2402 2403 2404
		endp = strchr(s, ' ');
		if (!endp)
			die("Missing space after source: %s", command_buf.buf);
2405
		strbuf_add(&s_uq, s, endp - s);
2406
	}
2407
	s = s_uq.buf;
2408 2409 2410 2411 2412 2413

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

	d = endp;
2414 2415
	strbuf_reset(&d_uq);
	if (!unquote_c_style(&d_uq, d, &endp)) {
2416 2417
		if (*endp)
			die("Garbage after dest in: %s", command_buf.buf);
2418
		d = d_uq.buf;
2419 2420 2421
	}

	memset(&leaf, 0, sizeof(leaf));
2422
	if (rename)
2423
		tree_content_remove(&b->branch_tree, s, &leaf, 1);
2424
	else
2425
		tree_content_get(&b->branch_tree, s, &leaf, 1);
2426 2427
	if (!leaf.versions[1].mode)
		die("Path %s not in branch", s);
2428 2429 2430 2431 2432 2433 2434
	if (!*d) {	/* C "path/to/subdir" "" */
		tree_content_replace(&b->branch_tree,
			leaf.versions[1].sha1,
			leaf.versions[1].mode,
			leaf.tree);
		return;
	}
2435 2436 2437 2438 2439 2440
	tree_content_set(&b->branch_tree, d,
		leaf.versions[1].sha1,
		leaf.versions[1].mode,
		leaf.tree);
}

2441
static void note_change_n(struct branch *b, unsigned char *old_fanout)
2442 2443 2444
{
	const char *p = command_buf.buf + 2;
	static struct strbuf uq = STRBUF_INIT;
2445
	struct object_entry *oe;
2446 2447
	struct branch *s;
	unsigned char sha1[20], commit_sha1[20];
2448
	char path[60];
2449
	uint16_t inline_data = 0;
2450
	unsigned char new_fanout;
2451

2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468
	/*
	 * 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. */
2469 2470
	/* <dataref> or 'inline' */
	if (*p == ':') {
2471
		oe = find_mark(parse_mark_ref_space(&p));
2472
		hashcpy(sha1, oe->idx.sha1);
2473
	} else if (starts_with(p, "inline ")) {
2474
		inline_data = 1;
2475
		oe = NULL; /* not used with inline_data, but makes gcc happy */
2476
		p += strlen("inline");  /* advance to space */
2477 2478
	} else {
		if (get_sha1_hex(p, sha1))
2479
			die("Invalid dataref: %s", command_buf.buf);
2480 2481
		oe = find_object(sha1);
		p += 40;
2482 2483
		if (*p != ' ')
			die("Missing space after SHA1: %s", command_buf.buf);
2484
	}
2485 2486
	assert(*p == ' ');
	p++;  /* skip space */
2487

2488
	/* <commit-ish> */
2489 2490
	s = lookup_branch(p);
	if (s) {
2491 2492
		if (is_null_sha1(s->sha1))
			die("Can't add a note on empty branch.");
2493 2494
		hashcpy(commit_sha1, s->sha1);
	} else if (*p == ':') {
2495
		uintmax_t commit_mark = parse_mark_ref_eol(p);
2496 2497 2498
		struct object_entry *commit_oe = find_mark(commit_mark);
		if (commit_oe->type != OBJ_COMMIT)
			die("Mark :%" PRIuMAX " not a commit", commit_mark);
2499
		hashcpy(commit_sha1, commit_oe->idx.sha1);
2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515
	} 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();
2516
		parse_and_store_blob(&last_blob, sha1, 0);
2517 2518 2519 2520
	} else if (oe) {
		if (oe->type != OBJ_BLOB)
			die("Not a blob (actually a %s): %s",
				typename(oe->type), command_buf.buf);
2521
	} else if (!is_null_sha1(sha1)) {
2522 2523 2524 2525 2526 2527 2528 2529
		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);
	}

2530
	construct_path_with_fanout(sha1_to_hex(commit_sha1), *old_fanout, path);
2531
	if (tree_content_remove(&b->branch_tree, path, NULL, 0))
2532 2533 2534 2535 2536 2537 2538 2539 2540
		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);
2541 2542
}

2543 2544 2545 2546 2547 2548
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);
2549
	b->num_notes = 0;
2550 2551
}

2552
static void parse_from_commit(struct branch *b, char *buf, unsigned long size)
2553 2554 2555 2556 2557 2558 2559 2560 2561 2562
{
	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);
}

2563
static void parse_from_existing(struct branch *b)
2564 2565 2566 2567 2568 2569 2570 2571 2572 2573
{
	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);
2574
		parse_from_commit(b, buf, size);
2575 2576 2577 2578
		free(buf);
	}
}

2579
static int parse_from(struct branch *b)
2580
{
2581
	const char *from;
2582 2583
	struct branch *s;

2584
	if (!starts_with(command_buf.buf, "from "))
2585
		return 0;
2586

2587 2588 2589 2590
	if (b->branch_tree.tree) {
		release_tree_content_recursive(b->branch_tree.tree);
		b->branch_tree.tree = NULL;
	}
2591 2592 2593 2594 2595 2596

	from = strchr(command_buf.buf, ' ') + 1;
	s = lookup_branch(from);
	if (b == s)
		die("Can't create a branch from itself: %s", b->name);
	else if (s) {
2597
		unsigned char *t = s->branch_tree.versions[1].sha1;
2598
		hashcpy(b->sha1, s->sha1);
2599 2600
		hashcpy(b->branch_tree.versions[0].sha1, t);
		hashcpy(b->branch_tree.versions[1].sha1, t);
2601
	} else if (*from == ':') {
2602
		uintmax_t idnum = parse_mark_ref_eol(from);
2603 2604
		struct object_entry *oe = find_mark(idnum);
		if (oe->type != OBJ_COMMIT)
2605
			die("Mark :%" PRIuMAX " not a commit", idnum);
2606
		hashcpy(b->sha1, oe->idx.sha1);
2607
		if (oe->pack_id != MAX_PACK_ID) {
2608
			unsigned long size;
2609
			char *buf = gfi_unpack_entry(oe, &size);
2610
			parse_from_commit(b, buf, size);
2611
			free(buf);
2612
		} else
2613
			parse_from_existing(b);
2614
	} else if (!get_sha1(from, b->sha1))
2615
		parse_from_existing(b);
2616
	else
2617 2618 2619
		die("Invalid ref name or SHA1 expression: %s", from);

	read_next_command();
2620
	return 1;
2621 2622
}

2623
static struct hash_list *parse_merge(unsigned int *count)
2624
{
2625
	struct hash_list *list = NULL, **tail = &list, *n;
2626
	const char *from;
2627 2628 2629
	struct branch *s;

	*count = 0;
2630
	while (starts_with(command_buf.buf, "merge ")) {
2631 2632 2633 2634 2635 2636
		from = strchr(command_buf.buf, ' ') + 1;
		n = xmalloc(sizeof(*n));
		s = lookup_branch(from);
		if (s)
			hashcpy(n->sha1, s->sha1);
		else if (*from == ':') {
2637
			uintmax_t idnum = parse_mark_ref_eol(from);
2638 2639
			struct object_entry *oe = find_mark(idnum);
			if (oe->type != OBJ_COMMIT)
2640
				die("Mark :%" PRIuMAX " not a commit", idnum);
2641
			hashcpy(n->sha1, oe->idx.sha1);
2642 2643 2644
		} 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
2645
				commit_type, &size, n->sha1);
2646 2647 2648 2649
			if (!buf || size < 46)
				die("Not a valid commit: %s", from);
			free(buf);
		} else
2650 2651 2652
			die("Invalid ref name or SHA1 expression: %s", from);

		n->next = NULL;
2653 2654 2655
		*tail = n;
		tail = &n->next;

2656
		(*count)++;
2657 2658 2659 2660 2661
		read_next_command();
	}
	return list;
}

2662
static void parse_new_commit(void)
2663
{
2664
	static struct strbuf msg = STRBUF_INIT;
2665 2666 2667 2668
	struct branch *b;
	char *sp;
	char *author = NULL;
	char *committer = NULL;
2669 2670
	struct hash_list *merge_list = NULL;
	unsigned int merge_count;
2671
	unsigned char prev_fanout, new_fanout;
2672 2673 2674 2675

	/* Obtain the branch name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	b = lookup_branch(sp);
2676
	if (!b)
2677
		b = new_branch(sp);
2678 2679

	read_next_command();
2680
	parse_mark();
2681
	if (starts_with(command_buf.buf, "author ")) {
2682
		author = parse_ident(command_buf.buf + 7);
2683 2684
		read_next_command();
	}
2685
	if (starts_with(command_buf.buf, "committer ")) {
2686
		committer = parse_ident(command_buf.buf + 10);
2687 2688 2689 2690
		read_next_command();
	}
	if (!committer)
		die("Expected committer but didn't get one");
2691
	parse_data(&msg, 0, NULL);
2692
	read_next_command();
2693 2694
	parse_from(b);
	merge_list = parse_merge(&merge_count);
2695 2696

	/* ensure the branch is active/loaded */
2697
	if (!b->branch_tree.tree || !max_active_branches) {
2698 2699 2700
		unload_one_branch();
		load_branch(b);
	}
2701

2702 2703
	prev_fanout = convert_num_notes_to_fanout(b->num_notes);

2704
	/* file_change* */
2705
	while (command_buf.len > 0) {
2706
		if (starts_with(command_buf.buf, "M "))
2707
			file_change_m(b);
2708
		else if (starts_with(command_buf.buf, "D "))
2709
			file_change_d(b);
2710
		else if (starts_with(command_buf.buf, "R "))
2711
			file_change_cr(b, 1);
2712
		else if (starts_with(command_buf.buf, "C "))
2713
			file_change_cr(b, 0);
2714
		else if (starts_with(command_buf.buf, "N "))
2715
			note_change_n(b, &prev_fanout);
2716 2717
		else if (!strcmp("deleteall", command_buf.buf))
			file_change_deleteall(b);
2718
		else if (starts_with(command_buf.buf, "ls "))
D
David Barr 已提交
2719
			parse_ls(b);
2720 2721 2722 2723
		else {
			unread_command_buf = 1;
			break;
		}
2724 2725
		if (read_next_command() == EOF)
			break;
2726 2727
	}

2728 2729 2730 2731
	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);

2732
	/* build the tree and the commit */
2733
	store_tree(&b->branch_tree);
2734 2735
	hashcpy(b->branch_tree.versions[0].sha1,
		b->branch_tree.versions[1].sha1);
2736 2737 2738

	strbuf_reset(&new_data);
	strbuf_addf(&new_data, "tree %s\n",
2739
		sha1_to_hex(b->branch_tree.versions[1].sha1));
2740
	if (!is_null_sha1(b->sha1))
2741
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(b->sha1));
2742 2743
	while (merge_list) {
		struct hash_list *next = merge_list->next;
2744
		strbuf_addf(&new_data, "parent %s\n", sha1_to_hex(merge_list->sha1));
2745 2746 2747
		free(merge_list);
		merge_list = next;
	}
2748 2749 2750 2751 2752 2753
	strbuf_addf(&new_data,
		"author %s\n"
		"committer %s\n"
		"\n",
		author ? author : committer, committer);
	strbuf_addbuf(&new_data, &msg);
2754
	free(author);
2755 2756
	free(committer);

2757
	if (!store_object(OBJ_COMMIT, &new_data, NULL, b->sha1, next_mark))
2758
		b->pack_id = pack_id;
2759
	b->last_commit = object_count_by_type[OBJ_COMMIT];
2760 2761
}

2762
static void parse_new_tag(void)
2763
{
2764
	static struct strbuf msg = STRBUF_INIT;
2765 2766 2767 2768 2769
	char *sp;
	const char *from;
	char *tagger;
	struct branch *s;
	struct tag *t;
2770
	uintmax_t from_mark = 0;
2771
	unsigned char sha1[20];
2772
	enum object_type type;
2773 2774 2775 2776

	/* Obtain the new tag name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	t = pool_alloc(sizeof(struct tag));
2777
	memset(t, 0, sizeof(struct tag));
2778 2779 2780 2781 2782 2783 2784 2785 2786
	t->name = pool_strdup(sp);
	if (last_tag)
		last_tag->next_tag = t;
	else
		first_tag = t;
	last_tag = t;
	read_next_command();

	/* from ... */
2787
	if (!starts_with(command_buf.buf, "from "))
2788 2789 2790 2791
		die("Expected from command, got %s", command_buf.buf);
	from = strchr(command_buf.buf, ' ') + 1;
	s = lookup_branch(from);
	if (s) {
2792 2793
		if (is_null_sha1(s->sha1))
			die("Can't tag an empty branch.");
2794
		hashcpy(sha1, s->sha1);
2795
		type = OBJ_COMMIT;
2796
	} else if (*from == ':') {
2797
		struct object_entry *oe;
2798
		from_mark = parse_mark_ref_eol(from);
2799
		oe = find_mark(from_mark);
2800
		type = oe->type;
2801
		hashcpy(sha1, oe->idx.sha1);
2802
	} else if (!get_sha1(from, sha1)) {
2803 2804 2805 2806 2807 2808 2809
		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;
2810 2811 2812 2813 2814
	} else
		die("Invalid ref name or SHA1 expression: %s", from);
	read_next_command();

	/* tagger ... */
2815
	if (starts_with(command_buf.buf, "tagger ")) {
2816 2817 2818 2819
		tagger = parse_ident(command_buf.buf + 7);
		read_next_command();
	} else
		tagger = NULL;
2820 2821

	/* tag payload/message */
2822
	parse_data(&msg, 0, NULL);
2823 2824

	/* build the tag object */
2825
	strbuf_reset(&new_data);
2826

2827
	strbuf_addf(&new_data,
2828 2829 2830
		    "object %s\n"
		    "type %s\n"
		    "tag %s\n",
2831
		    sha1_to_hex(sha1), typename(type), t->name);
2832 2833 2834 2835
	if (tagger)
		strbuf_addf(&new_data,
			    "tagger %s\n", tagger);
	strbuf_addch(&new_data, '\n');
2836
	strbuf_addbuf(&new_data, &msg);
2837 2838
	free(tagger);

2839
	if (store_object(OBJ_TAG, &new_data, NULL, t->sha1, 0))
2840 2841 2842
		t->pack_id = MAX_PACK_ID;
	else
		t->pack_id = pack_id;
2843 2844
}

2845
static void parse_reset_branch(void)
2846 2847 2848 2849 2850 2851 2852 2853
{
	struct branch *b;
	char *sp;

	/* Obtain the branch name from the rest of our command */
	sp = strchr(command_buf.buf, ' ') + 1;
	b = lookup_branch(sp);
	if (b) {
2854 2855 2856
		hashclr(b->sha1);
		hashclr(b->branch_tree.versions[0].sha1);
		hashclr(b->branch_tree.versions[1].sha1);
2857 2858 2859 2860 2861
		if (b->branch_tree.tree) {
			release_tree_content_recursive(b->branch_tree.tree);
			b->branch_tree.tree = NULL;
		}
	}
2862 2863 2864
	else
		b = new_branch(sp);
	read_next_command();
2865
	parse_from(b);
2866
	if (command_buf.len > 0)
2867
		unread_command_buf = 1;
2868 2869
}

2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 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
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);
2913 2914 2915 2916 2917 2918
	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);
2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929
}

static void parse_cat_blob(void)
{
	const char *p;
	struct object_entry *oe = oe;
	unsigned char sha1[20];

	/* cat-blob SP <object> LF */
	p = command_buf.buf + strlen("cat-blob ");
	if (*p == ':') {
2930
		oe = find_mark(parse_mark_ref_eol(p));
2931 2932 2933 2934 2935
		if (!oe)
			die("Unknown mark: %s", command_buf.buf);
		hashcpy(sha1, oe->idx.sha1);
	} else {
		if (get_sha1_hex(p, sha1))
2936
			die("Invalid dataref: %s", command_buf.buf);
2937 2938 2939 2940 2941 2942 2943 2944
		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
2945 2946 2947 2948
static struct object_entry *dereference(struct object_entry *oe,
					unsigned char sha1[20])
{
	unsigned long size;
2949
	char *buf = NULL;
D
David Barr 已提交
2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966
	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:
2967
		die("Not a tree-ish: %s", command_buf.buf);
D
David Barr 已提交
2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001
	}

	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> */
3002
		e = find_mark(parse_mark_ref_space(p));
D
David Barr 已提交
3003 3004 3005 3006 3007
		if (!e)
			die("Unknown mark: %s", command_buf.buf);
		hashcpy(sha1, e->idx.sha1);
	} else {	/* <sha1> */
		if (get_sha1_hex(*p, sha1))
3008
			die("Invalid dataref: %s", command_buf.buf);
D
David Barr 已提交
3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037
		e = find_object(sha1);
		*p += 40;
	}

	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",
3038
				mode & ~NO_DELTA, type, sha1_to_hex(sha1));
D
David Barr 已提交
3039 3040 3041 3042 3043 3044 3045 3046 3047 3048
		quote_c_style(path, &line, NULL, 0);
		strbuf_addch(&line, '\n');
	}
	cat_blob_write(line.buf, line.len);
}

static void parse_ls(struct branch *b)
{
	const char *p;
	struct tree_entry *root = NULL;
S
Stephen Boyd 已提交
3049
	struct tree_entry leaf = {NULL};
D
David Barr 已提交
3050

3051
	/* ls SP (<tree-ish> SP)? <path> */
D
David Barr 已提交
3052 3053 3054 3055 3056 3057 3058 3059 3060
	p = command_buf.buf + strlen("ls ");
	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);
3061 3062
		if (!is_null_sha1(root->versions[1].sha1))
			root->versions[1].mode = S_IFDIR;
D
David Barr 已提交
3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076
		load_tree(root);
		if (*p++ != ' ')
			die("Missing space after tree-ish: %s", command_buf.buf);
	}
	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;
	}
3077
	tree_content_get(root, p, &leaf, 1);
D
David Barr 已提交
3078 3079 3080 3081 3082 3083 3084 3085
	/*
	 * 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);
3086 3087
	if (leaf.tree)
		release_tree_content_recursive(leaf.tree);
D
David Barr 已提交
3088 3089 3090 3091
	if (!b || root != &b->branch_tree)
		release_tree_entry(root);
}

3092
static void checkpoint(void)
3093
{
3094
	checkpoint_requested = 0;
3095 3096 3097 3098 3099 3100
	if (object_count) {
		cycle_packfile();
		dump_branches();
		dump_tags();
		dump_marks();
	}
3101 3102 3103 3104 3105
}

static void parse_checkpoint(void)
{
	checkpoint_requested = 1;
3106
	skip_optional_lf();
3107 3108
}

3109
static void parse_progress(void)
3110
{
3111
	fwrite(command_buf.buf, 1, command_buf.len, stdout);
3112 3113 3114 3115 3116
	fputc('\n', stdout);
	fflush(stdout);
	skip_optional_lf();
}

3117
static char* make_fast_import_path(const char *path)
3118
{
3119
	struct strbuf abs_path = STRBUF_INIT;
3120

3121 3122 3123 3124 3125 3126
	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);
}

3127 3128
static void option_import_marks(const char *marks,
					int from_stream, int ignore_missing)
3129
{
3130 3131 3132 3133 3134 3135 3136
	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();
3137
	}
3138

3139
	import_marks_file = make_fast_import_path(marks);
3140
	safe_create_leading_directories_const(import_marks_file);
3141
	import_marks_file_from_stream = from_stream;
3142
	import_marks_file_ignore_missing = ignore_missing;
3143 3144
}

3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156
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);
}

3157 3158 3159 3160 3161 3162 3163 3164 3165
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;
}

3166 3167
static void option_depth(const char *depth)
{
3168
	max_depth = ulong_arg("--depth", depth);
3169 3170 3171 3172 3173 3174
	if (max_depth > MAX_DEPTH)
		die("--depth cannot exceed %u", MAX_DEPTH);
}

static void option_active_branches(const char *branches)
{
3175
	max_active_branches = ulong_arg("--active-branches", branches);
3176 3177 3178 3179
}

static void option_export_marks(const char *marks)
{
3180
	export_marks_file = make_fast_import_path(marks);
3181
	safe_create_leading_directories_const(export_marks_file);
3182 3183
}

3184 3185 3186 3187 3188 3189 3190 3191
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;
}

3192 3193 3194 3195 3196 3197 3198 3199 3200
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
3201
static int parse_one_option(const char *option)
3202
{
3203
	if (starts_with(option, "max-pack-size=")) {
3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214
		unsigned long v;
		if (!git_parse_ulong(option + 14, &v))
			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;
3215
	} else if (starts_with(option, "big-file-threshold=")) {
3216 3217 3218 3219
		unsigned long v;
		if (!git_parse_ulong(option + 19, &v))
			return 0;
		big_file_threshold = v;
3220
	} else if (starts_with(option, "depth=")) {
3221
		option_depth(option + 6);
3222
	} else if (starts_with(option, "active-branches=")) {
3223
		option_active_branches(option + 16);
3224
	} else if (starts_with(option, "export-pack-edges=")) {
3225
		option_export_pack_edges(option + 18);
3226
	} else if (starts_with(option, "quiet")) {
3227
		show_stats = 0;
3228
	} else if (starts_with(option, "stats")) {
3229 3230
		show_stats = 1;
	} else {
S
Sverre Rabbelier 已提交
3231
		return 0;
3232
	}
S
Sverre Rabbelier 已提交
3233 3234

	return 1;
3235 3236
}

3237
static int parse_one_feature(const char *feature, int from_stream)
3238
{
3239
	if (starts_with(feature, "date-format=")) {
3240
		option_date_format(feature + 12);
3241
	} else if (starts_with(feature, "import-marks=")) {
3242
		option_import_marks(feature + 13, from_stream, 0);
3243
	} else if (starts_with(feature, "import-marks-if-exists=")) {
3244 3245
		option_import_marks(feature + strlen("import-marks-if-exists="),
					from_stream, 1);
3246
	} else if (starts_with(feature, "export-marks=")) {
3247
		option_export_marks(feature + 13);
3248 3249
	} else if (!strcmp(feature, "cat-blob")) {
		; /* Don't die - this feature is supported */
3250
	} else if (!strcmp(feature, "relative-marks")) {
3251
		relative_marks_paths = 1;
3252
	} else if (!strcmp(feature, "no-relative-marks")) {
3253
		relative_marks_paths = 0;
3254 3255
	} else if (!strcmp(feature, "done")) {
		require_explicit_termination = 1;
3256
	} else if (!strcmp(feature, "force")) {
3257
		force_update = 1;
D
David Barr 已提交
3258
	} else if (!strcmp(feature, "notes") || !strcmp(feature, "ls")) {
3259
		; /* do nothing; we have the feature */
3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273
	} else {
		return 0;
	}

	return 1;
}

static void parse_feature(void)
{
	char *feature = command_buf.buf + 8;

	if (seen_data_command)
		die("Got feature command '%s' after data command", feature);

3274
	if (parse_one_feature(feature, 1))
3275 3276 3277 3278 3279
		return;

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

S
Sverre Rabbelier 已提交
3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290
static void parse_option(void)
{
	char *option = command_buf.buf + 11;

	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);
3291 3292
}

3293
static int git_pack_config(const char *k, const char *v, void *cb)
3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310
{
	if (!strcmp(k, "pack.depth")) {
		max_depth = git_config_int(k, v);
		if (max_depth > MAX_DEPTH)
			max_depth = MAX_DEPTH;
		return 0;
	}
	if (!strcmp(k, "pack.compression")) {
		int level = git_config_int(k, v);
		if (level == -1)
			level = Z_DEFAULT_COMPRESSION;
		else if (level < 0 || level > Z_BEST_COMPRESSION)
			die("bad pack compression level %d", level);
		pack_compression_level = level;
		pack_compression_seen = 1;
		return 0;
	}
3311
	if (!strcmp(k, "pack.indexversion")) {
3312 3313
		pack_idx_opts.version = git_config_int(k, v);
		if (pack_idx_opts.version > 2)
3314
			die("bad pack.indexversion=%"PRIu32,
3315
			    pack_idx_opts.version);
3316 3317 3318 3319 3320 3321
		return 0;
	}
	if (!strcmp(k, "pack.packsizelimit")) {
		max_packsize = git_config_ulong(k, v);
		return 0;
	}
3322
	return git_default_config(k, v, cb);
3323 3324
}

3325
static const char fast_import_usage[] =
3326
"git fast-import [--date-format=<f>] [--max-pack-size=<n>] [--big-file-threshold=<n>] [--depth=<n>] [--active-branches=<n>] [--export-marks=<marks.file>]";
3327

S
Sverre Rabbelier 已提交
3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340
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;

		if (parse_one_option(a + 2))
			continue;

3341
		if (parse_one_feature(a + 2, 0))
S
Sverre Rabbelier 已提交
3342 3343
			continue;

3344
		if (starts_with(a + 2, "cat-blob-fd=")) {
3345 3346 3347 3348
			option_cat_blob_fd(a + 2 + strlen("cat-blob-fd="));
			continue;
		}

S
Sverre Rabbelier 已提交
3349 3350 3351 3352 3353 3354 3355 3356 3357 3358
		die("unknown option %s", a);
	}
	if (i != global_argc)
		usage(fast_import_usage);

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

3359
int main(int argc, char **argv)
3360
{
3361
	unsigned int i;
3362

3363 3364
	git_extract_argv0_path(argv[0]);

3365 3366
	git_setup_gettext();

3367 3368 3369
	if (argc == 2 && !strcmp(argv[1], "-h"))
		usage(fast_import_usage);

3370
	setup_git_directory();
3371
	reset_pack_idx_option(&pack_idx_opts);
3372
	git_config(git_pack_config, NULL);
3373 3374 3375
	if (!pack_compression_seen && core_compression_seen)
		pack_compression_level = core_compression_level;

3376
	alloc_objects(object_entry_alloc);
3377
	strbuf_init(&command_buf, 0);
3378 3379 3380 3381
	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));
3382

S
Sverre Rabbelier 已提交
3383 3384
	global_argc = argc;
	global_argv = argv;
3385

3386 3387 3388 3389 3390
	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;

3391
	prepare_packed_git();
3392
	start_packfile();
3393
	set_die_routine(die_nicely);
3394
	set_checkpoint_signal();
3395 3396
	while (read_next_command() != EOF) {
		if (!strcmp("blob", command_buf.buf))
3397
			parse_new_blob();
3398
		else if (starts_with(command_buf.buf, "ls "))
D
David Barr 已提交
3399
			parse_ls(NULL);
3400
		else if (starts_with(command_buf.buf, "commit "))
3401
			parse_new_commit();
3402
		else if (starts_with(command_buf.buf, "tag "))
3403
			parse_new_tag();
3404
		else if (starts_with(command_buf.buf, "reset "))
3405
			parse_reset_branch();
3406
		else if (!strcmp("checkpoint", command_buf.buf))
3407
			parse_checkpoint();
3408 3409
		else if (!strcmp("done", command_buf.buf))
			break;
3410
		else if (starts_with(command_buf.buf, "progress "))
3411
			parse_progress();
3412
		else if (starts_with(command_buf.buf, "feature "))
3413
			parse_feature();
3414
		else if (starts_with(command_buf.buf, "option git "))
S
Sverre Rabbelier 已提交
3415
			parse_option();
3416
		else if (starts_with(command_buf.buf, "option "))
S
Sverre Rabbelier 已提交
3417
			/* ignore non-git options*/;
3418 3419
		else
			die("Unsupported command: %s", command_buf.buf);
3420 3421 3422

		if (checkpoint_requested)
			checkpoint();
3423
	}
S
Sverre Rabbelier 已提交
3424 3425 3426 3427 3428

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

3429 3430 3431
	if (require_explicit_termination && feof(stdin))
		die("stream ends early");

3432
	end_packfile();
3433

3434
	dump_branches();
3435
	dump_tags();
3436
	unkeep_all_packs();
3437
	dump_marks();
3438

3439 3440 3441
	if (pack_edges)
		fclose(pack_edges);

3442 3443 3444 3445 3446 3447 3448 3449 3450
	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");
3451 3452
		fprintf(stderr, "Alloc'd objects: %10" PRIuMAX "\n", alloc_count);
		fprintf(stderr, "Total objects:   %10" PRIuMAX " (%10" PRIuMAX " duplicates                  )\n", total_count, duplicate_count);
3453 3454 3455 3456
		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]);
3457
		fprintf(stderr, "Total branches:  %10lu (%10lu loads     )\n", branch_count, branch_load_count);
3458
		fprintf(stderr, "      marks:     %10" PRIuMAX " (%10" PRIuMAX " unique    )\n", (((uintmax_t)1) << marks->shift) * 1024, marks_set_count);
3459
		fprintf(stderr, "      atoms:     %10u\n", atom_cnt);
3460
		fprintf(stderr, "Memory total:    %10" PRIuMAX " KiB\n", (total_allocd + alloc_count*sizeof(struct object_entry))/1024);
3461
		fprintf(stderr, "       pools:    %10lu KiB\n", (unsigned long)(total_allocd/1024));
3462
		fprintf(stderr, "     objects:    %10" PRIuMAX " KiB\n", (alloc_count*sizeof(struct object_entry))/1024);
3463 3464 3465 3466 3467
		fprintf(stderr, "---------------------------------------------------------------------\n");
		pack_report();
		fprintf(stderr, "---------------------------------------------------------------------\n");
		fprintf(stderr, "\n");
	}
3468

3469
	return failure ? 1 : 0;
3470
}