builtin/cat-file.c

   1 /*
   2  * GIT - The information manager from hell
   3  *
   4  * Copyright (C) Linus Torvalds, 2005
   5  */
   6 #define USE_THE_INDEX_VARIABLE
   7 #include "cache.h"
   8 #include "alloc.h"
   9 #include "config.h"
  10 #include "convert.h"
  11 #include "builtin.h"
  12 #include "diff.h"
  13 #include "environment.h"
  14 #include "gettext.h"
  15 #include "hex.h"
  16 #include "ident.h"
  17 #include "parse-options.h"
  18 #include "userdiff.h"
  19 #include "streaming.h"
  20 #include "tree-walk.h"
  21 #include "oid-array.h"
  22 #include "packfile.h"
  23 #include "object-file.h"
  24 #include "object-name.h"
  25 #include "object-store.h"
  26 #include "replace-object.h"
  27 #include "promisor-remote.h"
  28 #include "mailmap.h"
  29 #include "write-or-die.h"
  30
  31 enum batch_mode {
  32         BATCH_MODE_CONTENTS,
  33         BATCH_MODE_INFO,
  34         BATCH_MODE_QUEUE_AND_DISPATCH,
  35 };
  36
  37 struct batch_options {
  38         int enabled;
  39         int follow_symlinks;
  40         enum batch_mode batch_mode;
  41         int buffer_output;
  42         int all_objects;
  43         int unordered;
  44         int transform_mode; /* may be 'w' or 'c' for --filters or --textconv */
  45         char input_delim;
  46         char output_delim;
  47         const char *format;
  48 };
  49
  50 static const char *force_path;
  51
  52 static struct string_list mailmap = STRING_LIST_INIT_NODUP;
  53 static int use_mailmap;
  54
  55 static char *replace_idents_using_mailmap(char *, size_t *);
  56
  57 static char *replace_idents_using_mailmap(char *object_buf, size_t *size)
  58 {
  59         struct strbuf sb = STRBUF_INIT;
  60         const char *headers[] = { "author ", "committer ", "tagger ", NULL };
  61
  62         strbuf_attach(&sb, object_buf, *size, *size + 1);
  63         apply_mailmap_to_header(&sb, headers, &mailmap);
  64         *size = sb.len;
  65         return strbuf_detach(&sb, NULL);
  66 }
  67
  68 static int filter_object(const char *path, unsigned mode,
  69                          const struct object_id *oid,
  70                          char **buf, unsigned long *size)
  71 {
  72         enum object_type type;
  73
  74         *buf = repo_read_object_file(the_repository, oid, &type, size);
  75         if (!*buf)
  76                 return error(_("cannot read object %s '%s'"),
  77                              oid_to_hex(oid), path);
  78         if ((type == OBJ_BLOB) && S_ISREG(mode)) {
  79                 struct strbuf strbuf = STRBUF_INIT;
  80                 struct checkout_metadata meta;
  81
  82                 init_checkout_metadata(&meta, NULL, NULL, oid);
  83                 if (convert_to_working_tree(&the_index, path, *buf, *size, &strbuf, &meta)) {
  84                         free(*buf);
  85                         *size = strbuf.len;
  86                         *buf = strbuf_detach(&strbuf, NULL);
  87                 }
  88         }
  89
  90         return 0;
  91 }
  92
  93 static int stream_blob(const struct object_id *oid)
  94 {
  95         if (stream_blob_to_fd(1, oid, NULL, 0))
  96                 die("unable to stream %s to stdout", oid_to_hex(oid));
  97         return 0;
  98 }
  99
 100 static int cat_one_file(int opt, const char *exp_type, const char *obj_name,
 101                         int unknown_type)
 102 {
 103         int ret;
 104         struct object_id oid;
 105         enum object_type type;
 106         char *buf;
 107         unsigned long size;
 108         struct object_context obj_context;
 109         struct object_info oi = OBJECT_INFO_INIT;
 110         struct strbuf sb = STRBUF_INIT;
 111         unsigned flags = OBJECT_INFO_LOOKUP_REPLACE;
 112         unsigned get_oid_flags = GET_OID_RECORD_PATH | GET_OID_ONLY_TO_DIE;
 113         const char *path = force_path;
 114         const int opt_cw = (opt == 'c' || opt == 'w');
 115         if (!path && opt_cw)
 116                 get_oid_flags |= GET_OID_REQUIRE_PATH;
 117
 118         if (unknown_type)
 119                 flags |= OBJECT_INFO_ALLOW_UNKNOWN_TYPE;
 120
 121         if (get_oid_with_context(the_repository, obj_name, get_oid_flags, &oid,
 122                                  &obj_context))
 123                 die("Not a valid object name %s", obj_name);
 124
 125         if (!path)
 126                 path = obj_context.path;
 127         if (obj_context.mode == S_IFINVALID)
 128                 obj_context.mode = 0100644;
 129
 130         buf = NULL;
 131         switch (opt) {
 132         case 't':
 133                 oi.type_name = &sb;
 134                 if (oid_object_info_extended(the_repository, &oid, &oi, flags) < 0)
 135                         die("git cat-file: could not get object info");
 136                 if (sb.len) {
 137                         printf("%s\n", sb.buf);
 138                         strbuf_release(&sb);
 139                         ret = 0;
 140                         goto cleanup;
 141                 }
 142                 break;
 143
 144         case 's':
 145                 oi.sizep = &size;
 146
 147                 if (use_mailmap) {
 148                         oi.typep = &type;
 149                         oi.contentp = (void**)&buf;
 150                 }
 151
 152                 if (oid_object_info_extended(the_repository, &oid, &oi, flags) < 0)
 153                         die("git cat-file: could not get object info");
 154
 155                 if (use_mailmap && (type == OBJ_COMMIT || type == OBJ_TAG)) {
 156                         size_t s = size;
 157                         buf = replace_idents_using_mailmap(buf, &s);
 158                         size = cast_size_t_to_ulong(s);
 159                 }
 160
 161                 printf("%"PRIuMAX"\n", (uintmax_t)size);
 162                 ret = 0;
 163                 goto cleanup;
 164
 165         case 'e':
 166                 return !repo_has_object_file(the_repository, &oid);
 167
 168         case 'w':
 169
 170                 if (filter_object(path, obj_context.mode,
 171                                   &oid, &buf, &size)) {
 172                         ret = -1;
 173                         goto cleanup;
 174                 }
 175                 break;
 176
 177         case 'c':
 178                 if (textconv_object(the_repository, path, obj_context.mode,
 179                                     &oid, 1, &buf, &size))
 180                         break;
 181                 /* else fallthrough */
 182
 183         case 'p':
 184                 type = oid_object_info(the_repository, &oid, NULL);
 185                 if (type < 0)
 186                         die("Not a valid object name %s", obj_name);
 187
 188                 /* custom pretty-print here */
 189                 if (type == OBJ_TREE) {
 190                         const char *ls_args[3] = { NULL };
 191                         ls_args[0] =  "ls-tree";
 192                         ls_args[1] =  obj_name;
 193                         ret = cmd_ls_tree(2, ls_args, NULL);
 194                         goto cleanup;
 195                 }
 196
 197                 if (type == OBJ_BLOB) {
 198                         ret = stream_blob(&oid);
 199                         goto cleanup;
 200                 }
 201                 buf = repo_read_object_file(the_repository, &oid, &type,
 202                                             &size);
 203                 if (!buf)
 204                         die("Cannot read object %s", obj_name);
 205
 206                 if (use_mailmap) {
 207                         size_t s = size;
 208                         buf = replace_idents_using_mailmap(buf, &s);
 209                         size = cast_size_t_to_ulong(s);
 210                 }
 211
 212                 /* otherwise just spit out the data */
 213                 break;
 214
 215         case 0:
 216         {
 217                 enum object_type exp_type_id = type_from_string(exp_type);
 218
 219                 if (exp_type_id == OBJ_BLOB) {
 220                         struct object_id blob_oid;
 221                         if (oid_object_info(the_repository, &oid, NULL) == OBJ_TAG) {
 222                                 char *buffer = repo_read_object_file(the_repository,
 223                                                                      &oid,
 224                                                                      &type,
 225                                                                      &size);
 226                                 const char *target;
 227                                 if (!skip_prefix(buffer, "object ", &target) ||
 228                                     get_oid_hex(target, &blob_oid))
 229                                         die("%s not a valid tag", oid_to_hex(&oid));
 230                                 free(buffer);
 231                         } else
 232                                 oidcpy(&blob_oid, &oid);
 233
 234                         if (oid_object_info(the_repository, &blob_oid, NULL) == OBJ_BLOB) {
 235                                 ret = stream_blob(&blob_oid);
 236                                 goto cleanup;
 237                         }
 238                         /*
 239                          * we attempted to dereference a tag to a blob
 240                          * and failed; there may be new dereference
 241                          * mechanisms this code is not aware of.
 242                          * fall-back to the usual case.
 243                          */
 244                 }
 245                 buf = read_object_with_reference(the_repository, &oid,
 246                                                  exp_type_id, &size, NULL);
 247
 248                 if (use_mailmap) {
 249                         size_t s = size;
 250                         buf = replace_idents_using_mailmap(buf, &s);
 251                         size = cast_size_t_to_ulong(s);
 252                 }
 253                 break;
 254         }
 255         default:
 256                 die("git cat-file: unknown option: %s", exp_type);
 257         }
 258
 259         if (!buf)
 260                 die("git cat-file %s: bad file", obj_name);
 261
 262         write_or_die(1, buf, size);
 263         ret = 0;
 264 cleanup:
 265         free(buf);
 266         free(obj_context.path);
 267         return ret;
 268 }
 269
 270 struct expand_data {
 271         struct object_id oid;
 272         enum object_type type;
 273         unsigned long size;
 274         off_t disk_size;
 275         const char *rest;
 276         struct object_id delta_base_oid;
 277
 278         /*
 279          * If mark_query is true, we do not expand anything, but rather
 280          * just mark the object_info with items we wish to query.
 281          */
 282         int mark_query;
 283
 284         /*
 285          * Whether to split the input on whitespace before feeding it to
 286          * get_sha1; this is decided during the mark_query phase based on
 287          * whether we have a %(rest) token in our format.
 288          */
 289         int split_on_whitespace;
 290
 291         /*
 292          * After a mark_query run, this object_info is set up to be
 293          * passed to oid_object_info_extended. It will point to the data
 294          * elements above, so you can retrieve the response from there.
 295          */
 296         struct object_info info;
 297
 298         /*
 299          * This flag will be true if the requested batch format and options
 300          * don't require us to call oid_object_info, which can then be
 301          * optimized out.
 302          */
 303         unsigned skip_object_info : 1;
 304 };
 305
 306 static int is_atom(const char *atom, const char *s, int slen)
 307 {
 308         int alen = strlen(atom);
 309         return alen == slen && !memcmp(atom, s, alen);
 310 }
 311
 312 static void expand_atom(struct strbuf *sb, const char *atom, int len,
 313                         void *vdata)
 314 {
 315         struct expand_data *data = vdata;
 316
 317         if (is_atom("objectname", atom, len)) {
 318                 if (!data->mark_query)
 319                         strbuf_addstr(sb, oid_to_hex(&data->oid));
 320         } else if (is_atom("objecttype", atom, len)) {
 321                 if (data->mark_query)
 322                         data->info.typep = &data->type;
 323                 else
 324                         strbuf_addstr(sb, type_name(data->type));
 325         } else if (is_atom("objectsize", atom, len)) {
 326                 if (data->mark_query)
 327                         data->info.sizep = &data->size;
 328                 else
 329                         strbuf_addf(sb, "%"PRIuMAX , (uintmax_t)data->size);
 330         } else if (is_atom("objectsize:disk", atom, len)) {
 331                 if (data->mark_query)
 332                         data->info.disk_sizep = &data->disk_size;
 333                 else
 334                         strbuf_addf(sb, "%"PRIuMAX, (uintmax_t)data->disk_size);
 335         } else if (is_atom("rest", atom, len)) {
 336                 if (data->mark_query)
 337                         data->split_on_whitespace = 1;
 338                 else if (data->rest)
 339                         strbuf_addstr(sb, data->rest);
 340         } else if (is_atom("deltabase", atom, len)) {
 341                 if (data->mark_query)
 342                         data->info.delta_base_oid = &data->delta_base_oid;
 343                 else
 344                         strbuf_addstr(sb,
 345                                       oid_to_hex(&data->delta_base_oid));
 346         } else
 347                 die("unknown format element: %.*s", len, atom);
 348 }
 349
 350 static size_t expand_format(struct strbuf *sb, const char *start, void *data)
 351 {
 352         const char *end;
 353
 354         if (*start != '(')
 355                 return 0;
 356         end = strchr(start + 1, ')');
 357         if (!end)
 358                 die("format element '%s' does not end in ')'", start);
 359
 360         expand_atom(sb, start + 1, end - start - 1, data);
 361
 362         return end - start + 1;
 363 }
 364
 365 static void batch_write(struct batch_options *opt, const void *data, int len)
 366 {
 367         if (opt->buffer_output) {
 368                 if (fwrite(data, 1, len, stdout) != len)
 369                         die_errno("unable to write to stdout");
 370         } else
 371                 write_or_die(1, data, len);
 372 }
 373
 374 static void print_object_or_die(struct batch_options *opt, struct expand_data *data)
 375 {
 376         const struct object_id *oid = &data->oid;
 377
 378         assert(data->info.typep);
 379
 380         if (data->type == OBJ_BLOB) {
 381                 if (opt->buffer_output)
 382                         fflush(stdout);
 383                 if (opt->transform_mode) {
 384                         char *contents;
 385                         unsigned long size;
 386
 387                         if (!data->rest)
 388                                 die("missing path for '%s'", oid_to_hex(oid));
 389
 390                         if (opt->transform_mode == 'w') {
 391                                 if (filter_object(data->rest, 0100644, oid,
 392                                                   &contents, &size))
 393                                         die("could not convert '%s' %s",
 394                                             oid_to_hex(oid), data->rest);
 395                         } else if (opt->transform_mode == 'c') {
 396                                 enum object_type type;
 397                                 if (!textconv_object(the_repository,
 398                                                      data->rest, 0100644, oid,
 399                                                      1, &contents, &size))
 400                                         contents = repo_read_object_file(the_repository,
 401                                                                          oid,
 402                                                                          &type,
 403                                                                          &size);
 404                                 if (!contents)
 405                                         die("could not convert '%s' %s",
 406                                             oid_to_hex(oid), data->rest);
 407                         } else
 408                                 BUG("invalid transform_mode: %c", opt->transform_mode);
 409                         batch_write(opt, contents, size);
 410                         free(contents);
 411                 } else {
 412                         stream_blob(oid);
 413                 }
 414         }
 415         else {
 416                 enum object_type type;
 417                 unsigned long size;
 418                 void *contents;
 419
 420                 contents = repo_read_object_file(the_repository, oid, &type,
 421                                                  &size);
 422
 423                 if (use_mailmap) {
 424                         size_t s = size;
 425                         contents = replace_idents_using_mailmap(contents, &s);
 426                         size = cast_size_t_to_ulong(s);
 427                 }
 428
 429                 if (!contents)
 430                         die("object %s disappeared", oid_to_hex(oid));
 431                 if (type != data->type)
 432                         die("object %s changed type!?", oid_to_hex(oid));
 433                 if (data->info.sizep && size != data->size && !use_mailmap)
 434                         die("object %s changed size!?", oid_to_hex(oid));
 435
 436                 batch_write(opt, contents, size);
 437                 free(contents);
 438         }
 439 }
 440
 441 static void print_default_format(struct strbuf *scratch, struct expand_data *data,
 442                                  struct batch_options *opt)
 443 {
 444         strbuf_addf(scratch, "%s %s %"PRIuMAX"%c", oid_to_hex(&data->oid),
 445                     type_name(data->type),
 446                     (uintmax_t)data->size, opt->output_delim);
 447 }
 448
 449 /*
 450  * If "pack" is non-NULL, then "offset" is the byte offset within the pack from
 451  * which the object may be accessed (though note that we may also rely on
 452  * data->oid, too). If "pack" is NULL, then offset is ignored.
 453  */
 454 static void batch_object_write(const char *obj_name,
 455                                struct strbuf *scratch,
 456                                struct batch_options *opt,
 457                                struct expand_data *data,
 458                                struct packed_git *pack,
 459                                off_t offset)
 460 {
 461         if (!data->skip_object_info) {
 462                 int ret;
 463
 464                 if (use_mailmap)
 465                         data->info.typep = &data->type;
 466
 467                 if (pack)
 468                         ret = packed_object_info(the_repository, pack, offset,
 469                                                  &data->info);
 470                 else
 471                         ret = oid_object_info_extended(the_repository,
 472                                                        &data->oid, &data->info,
 473                                                        OBJECT_INFO_LOOKUP_REPLACE);
 474                 if (ret < 0) {
 475                         printf("%s missing%c",
 476                                obj_name ? obj_name : oid_to_hex(&data->oid), opt->output_delim);
 477                         fflush(stdout);
 478                         return;
 479                 }
 480
 481                 if (use_mailmap && (data->type == OBJ_COMMIT || data->type == OBJ_TAG)) {
 482                         size_t s = data->size;
 483                         char *buf = NULL;
 484
 485                         buf = repo_read_object_file(the_repository, &data->oid, &data->type,
 486                                                     &data->size);
 487                         buf = replace_idents_using_mailmap(buf, &s);
 488                         data->size = cast_size_t_to_ulong(s);
 489
 490                         free(buf);
 491                 }
 492         }
 493
 494         strbuf_reset(scratch);
 495
 496         if (!opt->format) {
 497                 print_default_format(scratch, data, opt);
 498         } else {
 499                 strbuf_expand(scratch, opt->format, expand_format, data);
 500                 strbuf_addch(scratch, opt->output_delim);
 501         }
 502
 503         batch_write(opt, scratch->buf, scratch->len);
 504
 505         if (opt->batch_mode == BATCH_MODE_CONTENTS) {
 506                 print_object_or_die(opt, data);
 507                 batch_write(opt, &opt->output_delim, 1);
 508         }
 509 }
 510
 511 static void batch_one_object(const char *obj_name,
 512                              struct strbuf *scratch,
 513                              struct batch_options *opt,
 514                              struct expand_data *data)
 515 {
 516         struct object_context ctx;
 517         int flags = opt->follow_symlinks ? GET_OID_FOLLOW_SYMLINKS : 0;
 518         enum get_oid_result result;
 519
 520         result = get_oid_with_context(the_repository, obj_name,
 521                                       flags, &data->oid, &ctx);
 522         if (result != FOUND) {
 523                 switch (result) {
 524                 case MISSING_OBJECT:
 525                         printf("%s missing%c", obj_name, opt->output_delim);
 526                         break;
 527                 case SHORT_NAME_AMBIGUOUS:
 528                         printf("%s ambiguous%c", obj_name, opt->output_delim);
 529                         break;
 530                 case DANGLING_SYMLINK:
 531                         printf("dangling %"PRIuMAX"%c%s%c",
 532                                (uintmax_t)strlen(obj_name),
 533                                opt->output_delim, obj_name, opt->output_delim);
 534                         break;
 535                 case SYMLINK_LOOP:
 536                         printf("loop %"PRIuMAX"%c%s%c",
 537                                (uintmax_t)strlen(obj_name),
 538                                opt->output_delim, obj_name, opt->output_delim);
 539                         break;
 540                 case NOT_DIR:
 541                         printf("notdir %"PRIuMAX"%c%s%c",
 542                                (uintmax_t)strlen(obj_name),
 543                                opt->output_delim, obj_name, opt->output_delim);
 544                         break;
 545                 default:
 546                         BUG("unknown get_sha1_with_context result %d\n",
 547                                result);
 548                         break;
 549                 }
 550                 fflush(stdout);
 551                 return;
 552         }
 553
 554         if (ctx.mode == 0) {
 555                 printf("symlink %"PRIuMAX"%c%s%c",
 556                        (uintmax_t)ctx.symlink_path.len,
 557                        opt->output_delim, ctx.symlink_path.buf, opt->output_delim);
 558                 fflush(stdout);
 559                 return;
 560         }
 561
 562         batch_object_write(obj_name, scratch, opt, data, NULL, 0);
 563 }
 564
 565 struct object_cb_data {
 566         struct batch_options *opt;
 567         struct expand_data *expand;
 568         struct oidset *seen;
 569         struct strbuf *scratch;
 570 };
 571
 572 static int batch_object_cb(const struct object_id *oid, void *vdata)
 573 {
 574         struct object_cb_data *data = vdata;
 575         oidcpy(&data->expand->oid, oid);
 576         batch_object_write(NULL, data->scratch, data->opt, data->expand,
 577                            NULL, 0);
 578         return 0;
 579 }
 580
 581 static int collect_loose_object(const struct object_id *oid,
 582                                 const char *path UNUSED,
 583                                 void *data)
 584 {
 585         oid_array_append(data, oid);
 586         return 0;
 587 }
 588
 589 static int collect_packed_object(const struct object_id *oid,
 590                                  struct packed_git *pack UNUSED,
 591                                  uint32_t pos UNUSED,
 592                                  void *data)
 593 {
 594         oid_array_append(data, oid);
 595         return 0;
 596 }
 597
 598 static int batch_unordered_object(const struct object_id *oid,
 599                                   struct packed_git *pack, off_t offset,
 600                                   void *vdata)
 601 {
 602         struct object_cb_data *data = vdata;
 603
 604         if (oidset_insert(data->seen, oid))
 605                 return 0;
 606
 607         oidcpy(&data->expand->oid, oid);
 608         batch_object_write(NULL, data->scratch, data->opt, data->expand,
 609                            pack, offset);
 610         return 0;
 611 }
 612
 613 static int batch_unordered_loose(const struct object_id *oid,
 614                                  const char *path UNUSED,
 615                                  void *data)
 616 {
 617         return batch_unordered_object(oid, NULL, 0, data);
 618 }
 619
 620 static int batch_unordered_packed(const struct object_id *oid,
 621                                   struct packed_git *pack,
 622                                   uint32_t pos,
 623                                   void *data)
 624 {
 625         return batch_unordered_object(oid, pack,
 626                                       nth_packed_object_offset(pack, pos),
 627                                       data);
 628 }
 629
 630 typedef void (*parse_cmd_fn_t)(struct batch_options *, const char *,
 631                                struct strbuf *, struct expand_data *);
 632
 633 struct queued_cmd {
 634         parse_cmd_fn_t fn;
 635         char *line;
 636 };
 637
 638 static void parse_cmd_contents(struct batch_options *opt,
 639                              const char *line,
 640                              struct strbuf *output,
 641                              struct expand_data *data)
 642 {
 643         opt->batch_mode = BATCH_MODE_CONTENTS;
 644         batch_one_object(line, output, opt, data);
 645 }
 646
 647 static void parse_cmd_info(struct batch_options *opt,
 648                            const char *line,
 649                            struct strbuf *output,
 650                            struct expand_data *data)
 651 {
 652         opt->batch_mode = BATCH_MODE_INFO;
 653         batch_one_object(line, output, opt, data);
 654 }
 655
 656 static void dispatch_calls(struct batch_options *opt,
 657                 struct strbuf *output,
 658                 struct expand_data *data,
 659                 struct queued_cmd *cmd,
 660                 int nr)
 661 {
 662         int i;
 663
 664         if (!opt->buffer_output)
 665                 die(_("flush is only for --buffer mode"));
 666
 667         for (i = 0; i < nr; i++)
 668                 cmd[i].fn(opt, cmd[i].line, output, data);
 669
 670         fflush(stdout);
 671 }
 672
 673 static void free_cmds(struct queued_cmd *cmd, size_t *nr)
 674 {
 675         size_t i;
 676
 677         for (i = 0; i < *nr; i++)
 678                 FREE_AND_NULL(cmd[i].line);
 679
 680         *nr = 0;
 681 }
 682
 683
 684 static const struct parse_cmd {
 685         const char *name;
 686         parse_cmd_fn_t fn;
 687         unsigned takes_args;
 688 } commands[] = {
 689         { "contents", parse_cmd_contents, 1},
 690         { "info", parse_cmd_info, 1},
 691         { "flush", NULL, 0},
 692 };
 693
 694 static void batch_objects_command(struct batch_options *opt,
 695                                     struct strbuf *output,
 696                                     struct expand_data *data)
 697 {
 698         struct strbuf input = STRBUF_INIT;
 699         struct queued_cmd *queued_cmd = NULL;
 700         size_t alloc = 0, nr = 0;
 701
 702         while (strbuf_getdelim_strip_crlf(&input, stdin, opt->input_delim) != EOF) {
 703                 int i;
 704                 const struct parse_cmd *cmd = NULL;
 705                 const char *p = NULL, *cmd_end;
 706                 struct queued_cmd call = {0};
 707
 708                 if (!input.len)
 709                         die(_("empty command in input"));
 710                 if (isspace(*input.buf))
 711                         die(_("whitespace before command: '%s'"), input.buf);
 712
 713                 for (i = 0; i < ARRAY_SIZE(commands); i++) {
 714                         if (!skip_prefix(input.buf, commands[i].name, &cmd_end))
 715                                 continue;
 716
 717                         cmd = &commands[i];
 718                         if (cmd->takes_args) {
 719                                 if (*cmd_end != ' ')
 720                                         die(_("%s requires arguments"),
 721                                             commands[i].name);
 722
 723                                 p = cmd_end + 1;
 724                         } else if (*cmd_end) {
 725                                 die(_("%s takes no arguments"),
 726                                     commands[i].name);
 727                         }
 728
 729                         break;
 730                 }
 731
 732                 if (!cmd)
 733                         die(_("unknown command: '%s'"), input.buf);
 734
 735                 if (!strcmp(cmd->name, "flush")) {
 736                         dispatch_calls(opt, output, data, queued_cmd, nr);
 737                         free_cmds(queued_cmd, &nr);
 738                 } else if (!opt->buffer_output) {
 739                         cmd->fn(opt, p, output, data);
 740                 } else {
 741                         ALLOC_GROW(queued_cmd, nr + 1, alloc);
 742                         call.fn = cmd->fn;
 743                         call.line = xstrdup_or_null(p);
 744                         queued_cmd[nr++] = call;
 745                 }
 746         }
 747
 748         if (opt->buffer_output &&
 749             nr &&
 750             !git_env_bool("GIT_TEST_CAT_FILE_NO_FLUSH_ON_EXIT", 0)) {
 751                 dispatch_calls(opt, output, data, queued_cmd, nr);
 752                 free_cmds(queued_cmd, &nr);
 753         }
 754
 755         free_cmds(queued_cmd, &nr);
 756         free(queued_cmd);
 757         strbuf_release(&input);
 758 }
 759
 760 #define DEFAULT_FORMAT "%(objectname) %(objecttype) %(objectsize)"
 761
 762 static int batch_objects(struct batch_options *opt)
 763 {
 764         struct strbuf input = STRBUF_INIT;
 765         struct strbuf output = STRBUF_INIT;
 766         struct expand_data data;
 767         int save_warning;
 768         int retval = 0;
 769
 770         /*
 771          * Expand once with our special mark_query flag, which will prime the
 772          * object_info to be handed to oid_object_info_extended for each
 773          * object.
 774          */
 775         memset(&data, 0, sizeof(data));
 776         data.mark_query = 1;
 777         strbuf_expand(&output,
 778                       opt->format ? opt->format : DEFAULT_FORMAT,
 779                       expand_format,
 780                       &data);
 781         data.mark_query = 0;
 782         strbuf_release(&output);
 783         if (opt->transform_mode)
 784                 data.split_on_whitespace = 1;
 785
 786         if (opt->format && !strcmp(opt->format, DEFAULT_FORMAT))
 787                 opt->format = NULL;
 788         /*
 789          * If we are printing out the object, then always fill in the type,
 790          * since we will want to decide whether or not to stream.
 791          */
 792         if (opt->batch_mode == BATCH_MODE_CONTENTS)
 793                 data.info.typep = &data.type;
 794
 795         if (opt->all_objects) {
 796                 struct object_cb_data cb;
 797                 struct object_info empty = OBJECT_INFO_INIT;
 798
 799                 if (!memcmp(&data.info, &empty, sizeof(empty)))
 800                         data.skip_object_info = 1;
 801
 802                 if (repo_has_promisor_remote(the_repository))
 803                         warning("This repository uses promisor remotes. Some objects may not be loaded.");
 804
 805                 disable_replace_refs();
 806
 807                 cb.opt = opt;
 808                 cb.expand = &data;
 809                 cb.scratch = &output;
 810
 811                 if (opt->unordered) {
 812                         struct oidset seen = OIDSET_INIT;
 813
 814                         cb.seen = &seen;
 815
 816                         for_each_loose_object(batch_unordered_loose, &cb, 0);
 817                         for_each_packed_object(batch_unordered_packed, &cb,
 818                                                FOR_EACH_OBJECT_PACK_ORDER);
 819
 820                         oidset_clear(&seen);
 821                 } else {
 822                         struct oid_array sa = OID_ARRAY_INIT;
 823
 824                         for_each_loose_object(collect_loose_object, &sa, 0);
 825                         for_each_packed_object(collect_packed_object, &sa, 0);
 826
 827                         oid_array_for_each_unique(&sa, batch_object_cb, &cb);
 828
 829                         oid_array_clear(&sa);
 830                 }
 831
 832                 strbuf_release(&output);
 833                 return 0;
 834         }
 835
 836         /*
 837          * We are going to call get_sha1 on a potentially very large number of
 838          * objects. In most large cases, these will be actual object sha1s. The
 839          * cost to double-check that each one is not also a ref (just so we can
 840          * warn) ends up dwarfing the actual cost of the object lookups
 841          * themselves. We can work around it by just turning off the warning.
 842          */
 843         save_warning = warn_on_object_refname_ambiguity;
 844         warn_on_object_refname_ambiguity = 0;
 845
 846         if (opt->batch_mode == BATCH_MODE_QUEUE_AND_DISPATCH) {
 847                 batch_objects_command(opt, &output, &data);
 848                 goto cleanup;
 849         }
 850
 851         while (strbuf_getdelim_strip_crlf(&input, stdin, opt->input_delim) != EOF) {
 852                 if (data.split_on_whitespace) {
 853                         /*
 854                          * Split at first whitespace, tying off the beginning
 855                          * of the string and saving the remainder (or NULL) in
 856                          * data.rest.
 857                          */
 858                         char *p = strpbrk(input.buf, " \t");
 859                         if (p) {
 860                                 while (*p && strchr(" \t", *p))
 861                                         *p++ = '\0';
 862                         }
 863                         data.rest = p;
 864                 }
 865
 866                 batch_one_object(input.buf, &output, opt, &data);
 867         }
 868
 869  cleanup:
 870         strbuf_release(&input);
 871         strbuf_release(&output);
 872         warn_on_object_refname_ambiguity = save_warning;
 873         return retval;
 874 }
 875
 876 static int git_cat_file_config(const char *var, const char *value, void *cb)
 877 {
 878         if (userdiff_config(var, value) < 0)
 879                 return -1;
 880
 881         return git_default_config(var, value, cb);
 882 }
 883
 884 static int batch_option_callback(const struct option *opt,
 885                                  const char *arg,
 886                                  int unset)
 887 {
 888         struct batch_options *bo = opt->value;
 889
 890         BUG_ON_OPT_NEG(unset);
 891
 892         if (bo->enabled) {
 893                 return error(_("only one batch option may be specified"));
 894         }
 895
 896         bo->enabled = 1;
 897
 898         if (!strcmp(opt->long_name, "batch"))
 899                 bo->batch_mode = BATCH_MODE_CONTENTS;
 900         else if (!strcmp(opt->long_name, "batch-check"))
 901                 bo->batch_mode = BATCH_MODE_INFO;
 902         else if (!strcmp(opt->long_name, "batch-command"))
 903                 bo->batch_mode = BATCH_MODE_QUEUE_AND_DISPATCH;
 904         else
 905                 BUG("%s given to batch-option-callback", opt->long_name);
 906
 907         bo->format = arg;
 908
 909         return 0;
 910 }
 911
 912 int cmd_cat_file(int argc, const char **argv, const char *prefix)
 913 {
 914         int opt = 0;
 915         int opt_cw = 0;
 916         int opt_epts = 0;
 917         const char *exp_type = NULL, *obj_name = NULL;
 918         struct batch_options batch = {0};
 919         int unknown_type = 0;
 920         int input_nul_terminated = 0;
 921         int nul_terminated = 0;
 922
 923         const char * const usage[] = {
 924                 N_("git cat-file <type> <object>"),
 925                 N_("git cat-file (-e | -p) <object>"),
 926                 N_("git cat-file (-t | -s) [--allow-unknown-type] <object>"),
 927                 N_("git cat-file (--batch | --batch-check | --batch-command) [--batch-all-objects]\n"
 928                    "             [--buffer] [--follow-symlinks] [--unordered]\n"
 929                    "             [--textconv | --filters] [-Z]"),
 930                 N_("git cat-file (--textconv | --filters)\n"
 931                    "             [<rev>:<path|tree-ish> | --path=<path|tree-ish> <rev>]"),
 932                 NULL
 933         };
 934         const struct option options[] = {
 935                 /* Simple queries */
 936                 OPT_GROUP(N_("Check object existence or emit object contents")),
 937                 OPT_CMDMODE('e', NULL, &opt,
 938                             N_("check if <object> exists"), 'e'),
 939                 OPT_CMDMODE('p', NULL, &opt, N_("pretty-print <object> content"), 'p'),
 940
 941                 OPT_GROUP(N_("Emit [broken] object attributes")),
 942                 OPT_CMDMODE('t', NULL, &opt, N_("show object type (one of 'blob', 'tree', 'commit', 'tag', ...)"), 't'),
 943                 OPT_CMDMODE('s', NULL, &opt, N_("show object size"), 's'),
 944                 OPT_BOOL(0, "allow-unknown-type", &unknown_type,
 945                           N_("allow -s and -t to work with broken/corrupt objects")),
 946                 OPT_BOOL(0, "use-mailmap", &use_mailmap, N_("use mail map file")),
 947                 OPT_ALIAS(0, "mailmap", "use-mailmap"),
 948                 /* Batch mode */
 949                 OPT_GROUP(N_("Batch objects requested on stdin (or --batch-all-objects)")),
 950                 OPT_CALLBACK_F(0, "batch", &batch, N_("format"),
 951                         N_("show full <object> or <rev> contents"),
 952                         PARSE_OPT_OPTARG | PARSE_OPT_NONEG,
 953                         batch_option_callback),
 954                 OPT_CALLBACK_F(0, "batch-check", &batch, N_("format"),
 955                         N_("like --batch, but don't emit <contents>"),
 956                         PARSE_OPT_OPTARG | PARSE_OPT_NONEG,
 957                         batch_option_callback),
 958                 OPT_BOOL_F('z', NULL, &input_nul_terminated, N_("stdin is NUL-terminated"),
 959                         PARSE_OPT_HIDDEN),
 960                 OPT_BOOL('Z', NULL, &nul_terminated, N_("stdin and stdout is NUL-terminated")),
 961                 OPT_CALLBACK_F(0, "batch-command", &batch, N_("format"),
 962                         N_("read commands from stdin"),
 963                         PARSE_OPT_OPTARG | PARSE_OPT_NONEG,
 964                         batch_option_callback),
 965                 OPT_CMDMODE(0, "batch-all-objects", &opt,
 966                             N_("with --batch[-check]: ignores stdin, batches all known objects"), 'b'),
 967                 /* Batch-specific options */
 968                 OPT_GROUP(N_("Change or optimize batch output")),
 969                 OPT_BOOL(0, "buffer", &batch.buffer_output, N_("buffer --batch output")),
 970                 OPT_BOOL(0, "follow-symlinks", &batch.follow_symlinks,
 971                          N_("follow in-tree symlinks")),
 972                 OPT_BOOL(0, "unordered", &batch.unordered,
 973                          N_("do not order objects before emitting them")),
 974                 /* Textconv options, stand-ole*/
 975                 OPT_GROUP(N_("Emit object (blob or tree) with conversion or filter (stand-alone, or with batch)")),
 976                 OPT_CMDMODE(0, "textconv", &opt,
 977                             N_("run textconv on object's content"), 'c'),
 978                 OPT_CMDMODE(0, "filters", &opt,
 979                             N_("run filters on object's content"), 'w'),
 980                 OPT_STRING(0, "path", &force_path, N_("blob|tree"),
 981                            N_("use a <path> for (--textconv | --filters); Not with 'batch'")),
 982                 OPT_END()
 983         };
 984
 985         git_config(git_cat_file_config, NULL);
 986
 987         batch.buffer_output = -1;
 988
 989         argc = parse_options(argc, argv, prefix, options, usage, 0);
 990         opt_cw = (opt == 'c' || opt == 'w');
 991         opt_epts = (opt == 'e' || opt == 'p' || opt == 't' || opt == 's');
 992
 993         if (use_mailmap)
 994                 read_mailmap(&mailmap);
 995
 996         /* --batch-all-objects? */
 997         if (opt == 'b')
 998                 batch.all_objects = 1;
 999
1000         /* Option compatibility */
1001         if (force_path && !opt_cw)
1002                 usage_msg_optf(_("'%s=<%s>' needs '%s' or '%s'"),
1003                                usage, options,
1004                                "--path", _("path|tree-ish"), "--filters",
1005                                "--textconv");
1006
1007         /* Option compatibility with batch mode */
1008         if (batch.enabled)
1009                 ;
1010         else if (batch.follow_symlinks)
1011                 usage_msg_optf(_("'%s' requires a batch mode"), usage, options,
1012                                "--follow-symlinks");
1013         else if (batch.buffer_output >= 0)
1014                 usage_msg_optf(_("'%s' requires a batch mode"), usage, options,
1015                                "--buffer");
1016         else if (batch.all_objects)
1017                 usage_msg_optf(_("'%s' requires a batch mode"), usage, options,
1018                                "--batch-all-objects");
1019         else if (input_nul_terminated)
1020                 usage_msg_optf(_("'%s' requires a batch mode"), usage, options,
1021                                "-z");
1022         else if (nul_terminated)
1023                 usage_msg_optf(_("'%s' requires a batch mode"), usage, options,
1024                                "-Z");
1025
1026         batch.input_delim = batch.output_delim = '\n';
1027         if (input_nul_terminated)
1028                 batch.input_delim = '\0';
1029         if (nul_terminated)
1030                 batch.input_delim = batch.output_delim = '\0';
1031
1032         /* Batch defaults */
1033         if (batch.buffer_output < 0)
1034                 batch.buffer_output = batch.all_objects;
1035
1036         /* Return early if we're in batch mode? */
1037         if (batch.enabled) {
1038                 if (opt_cw)
1039                         batch.transform_mode = opt;
1040                 else if (opt && opt != 'b')
1041                         usage_msg_optf(_("'-%c' is incompatible with batch mode"),
1042                                        usage, options, opt);
1043                 else if (argc)
1044                         usage_msg_opt(_("batch modes take no arguments"), usage,
1045                                       options);
1046
1047                 return batch_objects(&batch);
1048         }
1049
1050         if (opt) {
1051                 if (!argc && opt == 'c')
1052                         usage_msg_optf(_("<rev> required with '%s'"),
1053                                        usage, options, "--textconv");
1054                 else if (!argc && opt == 'w')
1055                         usage_msg_optf(_("<rev> required with '%s'"),
1056                                        usage, options, "--filters");
1057                 else if (!argc && opt_epts)
1058                         usage_msg_optf(_("<object> required with '-%c'"),
1059                                        usage, options, opt);
1060                 else if (argc == 1)
1061                         obj_name = argv[0];
1062                 else
1063                         usage_msg_opt(_("too many arguments"), usage, options);
1064         } else if (!argc) {
1065                 usage_with_options(usage, options);
1066         } else if (argc != 2) {
1067                 usage_msg_optf(_("only two arguments allowed in <type> <object> mode, not %d"),
1068                               usage, options, argc);
1069         } else if (argc) {
1070                 exp_type = argv[0];
1071                 obj_name = argv[1];
1072         }
1073
1074         if (unknown_type && opt != 't' && opt != 's')
1075                 die("git cat-file --allow-unknown-type: use with -s or -t");
1076         return cat_one_file(opt, exp_type, obj_name, unknown_type);
1077 }