builtin/cat-file.c

   1 /*
   2  * GIT - The information manager from hell
   3  *
   4  * Copyright (C) Linus Torvalds, 2005
   5  */
   6 #include "cache.h"
   7 #include "exec_cmd.h"
   8 #include "tag.h"
   9 #include "tree.h"
  10 #include "builtin.h"
  11 #include "parse-options.h"
  12 #include "diff.h"
  13 #include "userdiff.h"
  14 #include "streaming.h"
  15
  16 static int cat_one_file(int opt, const char *exp_type, const char *obj_name)
  17 {
  18         unsigned char sha1[20];
  19         enum object_type type;
  20         char *buf;
  21         unsigned long size;
  22         struct object_context obj_context;
  23
  24         if (get_sha1_with_context(obj_name, 0, sha1, &obj_context))
  25                 die("Not a valid object name %s", obj_name);
  26
  27         buf = NULL;
  28         switch (opt) {
  29         case 't':
  30                 type = sha1_object_info(sha1, NULL);
  31                 if (type > 0) {
  32                         printf("%s\n", typename(type));
  33                         return 0;
  34                 }
  35                 break;
  36
  37         case 's':
  38                 type = sha1_object_info(sha1, &size);
  39                 if (type > 0) {
  40                         printf("%lu\n", size);
  41                         return 0;
  42                 }
  43                 break;
  44
  45         case 'e':
  46                 return !has_sha1_file(sha1);
  47
  48         case 'c':
  49                 if (!obj_context.path[0])
  50                         die("git cat-file --textconv %s: <object> must be <sha1:path>",
  51                             obj_name);
  52
  53                 if (textconv_object(obj_context.path, obj_context.mode, sha1, 1, &buf, &size))
  54                         break;
  55
  56         case 'p':
  57                 type = sha1_object_info(sha1, NULL);
  58                 if (type < 0)
  59                         die("Not a valid object name %s", obj_name);
  60
  61                 /* custom pretty-print here */
  62                 if (type == OBJ_TREE) {
  63                         const char *ls_args[3] = { NULL };
  64                         ls_args[0] =  "ls-tree";
  65                         ls_args[1] =  obj_name;
  66                         return cmd_ls_tree(2, ls_args, NULL);
  67                 }
  68
  69                 if (type == OBJ_BLOB)
  70                         return stream_blob_to_fd(1, sha1, NULL, 0);
  71                 buf = read_sha1_file(sha1, &type, &size);
  72                 if (!buf)
  73                         die("Cannot read object %s", obj_name);
  74
  75                 /* otherwise just spit out the data */
  76                 break;
  77
  78         case 0:
  79                 if (type_from_string(exp_type) == OBJ_BLOB) {
  80                         unsigned char blob_sha1[20];
  81                         if (sha1_object_info(sha1, NULL) == OBJ_TAG) {
  82                                 enum object_type type;
  83                                 unsigned long size;
  84                                 char *buffer = read_sha1_file(sha1, &type, &size);
  85                                 const char *target;
  86                                 if (!skip_prefix(buffer, "object ", &target) ||
  87                                     get_sha1_hex(target, blob_sha1))
  88                                         die("%s not a valid tag", sha1_to_hex(sha1));
  89                                 free(buffer);
  90                         } else
  91                                 hashcpy(blob_sha1, sha1);
  92
  93                         if (sha1_object_info(blob_sha1, NULL) == OBJ_BLOB)
  94                                 return stream_blob_to_fd(1, blob_sha1, NULL, 0);
  95                         /*
  96                          * we attempted to dereference a tag to a blob
  97                          * and failed; there may be new dereference
  98                          * mechanisms this code is not aware of.
  99                          * fall-back to the usual case.
 100                          */
 101                 }
 102                 buf = read_object_with_reference(sha1, exp_type, &size, NULL);
 103                 break;
 104
 105         default:
 106                 die("git cat-file: unknown option: %s", exp_type);
 107         }
 108
 109         if (!buf)
 110                 die("git cat-file %s: bad file", obj_name);
 111
 112         write_or_die(1, buf, size);
 113         return 0;
 114 }
 115
 116 struct expand_data {
 117         unsigned char sha1[20];
 118         enum object_type type;
 119         unsigned long size;
 120         unsigned long disk_size;
 121         const char *rest;
 122         unsigned char delta_base_sha1[20];
 123
 124         /*
 125          * If mark_query is true, we do not expand anything, but rather
 126          * just mark the object_info with items we wish to query.
 127          */
 128         int mark_query;
 129
 130         /*
 131          * Whether to split the input on whitespace before feeding it to
 132          * get_sha1; this is decided during the mark_query phase based on
 133          * whether we have a %(rest) token in our format.
 134          */
 135         int split_on_whitespace;
 136
 137         /*
 138          * After a mark_query run, this object_info is set up to be
 139          * passed to sha1_object_info_extended. It will point to the data
 140          * elements above, so you can retrieve the response from there.
 141          */
 142         struct object_info info;
 143 };
 144
 145 static int is_atom(const char *atom, const char *s, int slen)
 146 {
 147         int alen = strlen(atom);
 148         return alen == slen && !memcmp(atom, s, alen);
 149 }
 150
 151 static void expand_atom(struct strbuf *sb, const char *atom, int len,
 152                         void *vdata)
 153 {
 154         struct expand_data *data = vdata;
 155
 156         if (is_atom("objectname", atom, len)) {
 157                 if (!data->mark_query)
 158                         strbuf_addstr(sb, sha1_to_hex(data->sha1));
 159         } else if (is_atom("objecttype", atom, len)) {
 160                 if (data->mark_query)
 161                         data->info.typep = &data->type;
 162                 else
 163                         strbuf_addstr(sb, typename(data->type));
 164         } else if (is_atom("objectsize", atom, len)) {
 165                 if (data->mark_query)
 166                         data->info.sizep = &data->size;
 167                 else
 168                         strbuf_addf(sb, "%lu", data->size);
 169         } else if (is_atom("objectsize:disk", atom, len)) {
 170                 if (data->mark_query)
 171                         data->info.disk_sizep = &data->disk_size;
 172                 else
 173                         strbuf_addf(sb, "%lu", data->disk_size);
 174         } else if (is_atom("rest", atom, len)) {
 175                 if (data->mark_query)
 176                         data->split_on_whitespace = 1;
 177                 else if (data->rest)
 178                         strbuf_addstr(sb, data->rest);
 179         } else if (is_atom("deltabase", atom, len)) {
 180                 if (data->mark_query)
 181                         data->info.delta_base_sha1 = data->delta_base_sha1;
 182                 else
 183                         strbuf_addstr(sb, sha1_to_hex(data->delta_base_sha1));
 184         } else
 185                 die("unknown format element: %.*s", len, atom);
 186 }
 187
 188 static size_t expand_format(struct strbuf *sb, const char *start, void *data)
 189 {
 190         const char *end;
 191
 192         if (*start != '(')
 193                 return 0;
 194         end = strchr(start + 1, ')');
 195         if (!end)
 196                 die("format element '%s' does not end in ')'", start);
 197
 198         expand_atom(sb, start + 1, end - start - 1, data);
 199
 200         return end - start + 1;
 201 }
 202
 203 static void print_object_or_die(int fd, struct expand_data *data)
 204 {
 205         const unsigned char *sha1 = data->sha1;
 206
 207         assert(data->info.typep);
 208
 209         if (data->type == OBJ_BLOB) {
 210                 if (stream_blob_to_fd(fd, sha1, NULL, 0) < 0)
 211                         die("unable to stream %s to stdout", sha1_to_hex(sha1));
 212         }
 213         else {
 214                 enum object_type type;
 215                 unsigned long size;
 216                 void *contents;
 217
 218                 contents = read_sha1_file(sha1, &type, &size);
 219                 if (!contents)
 220                         die("object %s disappeared", sha1_to_hex(sha1));
 221                 if (type != data->type)
 222                         die("object %s changed type!?", sha1_to_hex(sha1));
 223                 if (data->info.sizep && size != data->size)
 224                         die("object %s changed size!?", sha1_to_hex(sha1));
 225
 226                 write_or_die(fd, contents, size);
 227                 free(contents);
 228         }
 229 }
 230
 231 struct batch_options {
 232         int enabled;
 233         int print_contents;
 234         const char *format;
 235 };
 236
 237 static int batch_one_object(const char *obj_name, struct batch_options *opt,
 238                             struct expand_data *data)
 239 {
 240         struct strbuf buf = STRBUF_INIT;
 241
 242         if (!obj_name)
 243            return 1;
 244
 245         if (get_sha1(obj_name, data->sha1)) {
 246                 printf("%s missing\n", obj_name);
 247                 fflush(stdout);
 248                 return 0;
 249         }
 250
 251         if (sha1_object_info_extended(data->sha1, &data->info, LOOKUP_REPLACE_OBJECT) < 0) {
 252                 printf("%s missing\n", obj_name);
 253                 fflush(stdout);
 254                 return 0;
 255         }
 256
 257         strbuf_expand(&buf, opt->format, expand_format, data);
 258         strbuf_addch(&buf, '\n');
 259         write_or_die(1, buf.buf, buf.len);
 260         strbuf_release(&buf);
 261
 262         if (opt->print_contents) {
 263                 print_object_or_die(1, data);
 264                 write_or_die(1, "\n", 1);
 265         }
 266         return 0;
 267 }
 268
 269 static int batch_objects(struct batch_options *opt)
 270 {
 271         struct strbuf buf = STRBUF_INIT;
 272         struct expand_data data;
 273         int save_warning;
 274         int retval = 0;
 275
 276         if (!opt->format)
 277                 opt->format = "%(objectname) %(objecttype) %(objectsize)";
 278
 279         /*
 280          * Expand once with our special mark_query flag, which will prime the
 281          * object_info to be handed to sha1_object_info_extended for each
 282          * object.
 283          */
 284         memset(&data, 0, sizeof(data));
 285         data.mark_query = 1;
 286         strbuf_expand(&buf, opt->format, expand_format, &data);
 287         data.mark_query = 0;
 288
 289         /*
 290          * If we are printing out the object, then always fill in the type,
 291          * since we will want to decide whether or not to stream.
 292          */
 293         if (opt->print_contents)
 294                 data.info.typep = &data.type;
 295
 296         /*
 297          * We are going to call get_sha1 on a potentially very large number of
 298          * objects. In most large cases, these will be actual object sha1s. The
 299          * cost to double-check that each one is not also a ref (just so we can
 300          * warn) ends up dwarfing the actual cost of the object lookups
 301          * themselves. We can work around it by just turning off the warning.
 302          */
 303         save_warning = warn_on_object_refname_ambiguity;
 304         warn_on_object_refname_ambiguity = 0;
 305
 306         while (strbuf_getline(&buf, stdin, '\n') != EOF) {
 307                 if (data.split_on_whitespace) {
 308                         /*
 309                          * Split at first whitespace, tying off the beginning
 310                          * of the string and saving the remainder (or NULL) in
 311                          * data.rest.
 312                          */
 313                         char *p = strpbrk(buf.buf, " \t");
 314                         if (p) {
 315                                 while (*p && strchr(" \t", *p))
 316                                         *p++ = '\0';
 317                         }
 318                         data.rest = p;
 319                 }
 320
 321                 retval = batch_one_object(buf.buf, opt, &data);
 322                 if (retval)
 323                         break;
 324         }
 325
 326         strbuf_release(&buf);
 327         warn_on_object_refname_ambiguity = save_warning;
 328         return retval;
 329 }
 330
 331 static const char * const cat_file_usage[] = {
 332         N_("git cat-file (-t | -s | -e | -p | <type> | --textconv) <object>"),
 333         N_("git cat-file (--batch | --batch-check) < <list-of-objects>"),
 334         NULL
 335 };
 336
 337 static int git_cat_file_config(const char *var, const char *value, void *cb)
 338 {
 339         if (userdiff_config(var, value) < 0)
 340                 return -1;
 341
 342         return git_default_config(var, value, cb);
 343 }
 344
 345 static int batch_option_callback(const struct option *opt,
 346                                  const char *arg,
 347                                  int unset)
 348 {
 349         struct batch_options *bo = opt->value;
 350
 351         if (unset) {
 352                 memset(bo, 0, sizeof(*bo));
 353                 return 0;
 354         }
 355
 356         bo->enabled = 1;
 357         bo->print_contents = !strcmp(opt->long_name, "batch");
 358         bo->format = arg;
 359
 360         return 0;
 361 }
 362
 363 int cmd_cat_file(int argc, const char **argv, const char *prefix)
 364 {
 365         int opt = 0;
 366         const char *exp_type = NULL, *obj_name = NULL;
 367         struct batch_options batch = {0};
 368
 369         const struct option options[] = {
 370                 OPT_GROUP(N_("<type> can be one of: blob, tree, commit, tag")),
 371                 OPT_SET_INT('t', NULL, &opt, N_("show object type"), 't'),
 372                 OPT_SET_INT('s', NULL, &opt, N_("show object size"), 's'),
 373                 OPT_SET_INT('e', NULL, &opt,
 374                             N_("exit with zero when there's no error"), 'e'),
 375                 OPT_SET_INT('p', NULL, &opt, N_("pretty-print object's content"), 'p'),
 376                 OPT_SET_INT(0, "textconv", &opt,
 377                             N_("for blob objects, run textconv on object's content"), 'c'),
 378                 { OPTION_CALLBACK, 0, "batch", &batch, "format",
 379                         N_("show info and content of objects fed from the standard input"),
 380                         PARSE_OPT_OPTARG, batch_option_callback },
 381                 { OPTION_CALLBACK, 0, "batch-check", &batch, "format",
 382                         N_("show info about objects fed from the standard input"),
 383                         PARSE_OPT_OPTARG, batch_option_callback },
 384                 OPT_END()
 385         };
 386
 387         git_config(git_cat_file_config, NULL);
 388
 389         if (argc != 3 && argc != 2)
 390                 usage_with_options(cat_file_usage, options);
 391
 392         argc = parse_options(argc, argv, prefix, options, cat_file_usage, 0);
 393
 394         if (opt) {
 395                 if (argc == 1)
 396                         obj_name = argv[0];
 397                 else
 398                         usage_with_options(cat_file_usage, options);
 399         }
 400         if (!opt && !batch.enabled) {
 401                 if (argc == 2) {
 402                         exp_type = argv[0];
 403                         obj_name = argv[1];
 404                 } else
 405                         usage_with_options(cat_file_usage, options);
 406         }
 407         if (batch.enabled && (opt || argc)) {
 408                 usage_with_options(cat_file_usage, options);
 409         }
 410
 411         if (batch.enabled)
 412                 return batch_objects(&batch);
 413
 414         return cat_one_file(opt, exp_type, obj_name);
 415 }