object-store.h

   1 #ifndef OBJECT_STORE_H
   2 #define OBJECT_STORE_H
   3
   4 #include "cache.h"
   5 #include "oidmap.h"
   6 #include "list.h"
   7 #include "sha1-array.h"
   8 #include "strbuf.h"
   9
  10 struct object_directory {
  11         struct object_directory *next;
  12
  13         /*
  14          * Used to store the results of readdir(3) calls when we are OK
  15          * sacrificing accuracy due to races for speed. That includes
  16          * object existence with OBJECT_INFO_QUICK, as well as
  17          * our search for unique abbreviated hashes. Don't use it for tasks
  18          * requiring greater accuracy!
  19          *
  20          * Be sure to call odb_load_loose_cache() before using.
  21          */
  22         char loose_objects_subdir_seen[256];
  23         struct oid_array loose_objects_cache[256];
  24
  25         /*
  26          * Path to the alternative object store. If this is a relative path,
  27          * it is relative to the current working directory.
  28          */
  29         char *path;
  30 };
  31
  32 void prepare_alt_odb(struct repository *r);
  33 char *compute_alternate_path(const char *path, struct strbuf *err);
  34 typedef int alt_odb_fn(struct object_directory *, void *);
  35 int foreach_alt_odb(alt_odb_fn, void*);
  36 typedef void alternate_ref_fn(const struct object_id *oid, void *);
  37 void for_each_alternate_ref(alternate_ref_fn, void *);
  38
  39 /*
  40  * Add the directory to the on-disk alternates file; the new entry will also
  41  * take effect in the current process.
  42  */
  43 void add_to_alternates_file(const char *dir);
  44
  45 /*
  46  * Add the directory to the in-memory list of alternates (along with any
  47  * recursive alternates it points to), but do not modify the on-disk alternates
  48  * file.
  49  */
  50 void add_to_alternates_memory(const char *dir);
  51
  52 /*
  53  * Populate and return the loose object cache array corresponding to the
  54  * given object ID.
  55  */
  56 struct oid_array *odb_loose_cache(struct object_directory *odb,
  57                                   const struct object_id *oid);
  58
  59 /* Empty the loose object cache for the specified object directory. */
  60 void odb_clear_loose_cache(struct object_directory *odb);
  61
  62 struct packed_git {
  63         struct hashmap_entry packmap_ent;
  64         struct packed_git *next;
  65         struct list_head mru;
  66         struct pack_window *windows;
  67         off_t pack_size;
  68         const void *index_data;
  69         size_t index_size;
  70         uint32_t num_objects;
  71         uint32_t num_bad_objects;
  72         unsigned char *bad_object_sha1;
  73         int index_version;
  74         time_t mtime;
  75         int pack_fd;
  76         int index;              /* for builtin/pack-objects.c */
  77         unsigned pack_local:1,
  78                  pack_keep:1,
  79                  pack_keep_in_core:1,
  80                  freshened:1,
  81                  do_not_close:1,
  82                  pack_promisor:1,
  83                  multi_pack_index:1;
  84         unsigned char hash[GIT_MAX_RAWSZ];
  85         struct revindex_entry *revindex;
  86         /* something like ".git/objects/pack/xxxxx.pack" */
  87         char pack_name[FLEX_ARRAY]; /* more */
  88 };
  89
  90 struct multi_pack_index;
  91
  92 static inline int pack_map_entry_cmp(const void *unused_cmp_data,
  93                                      const struct hashmap_entry *entry,
  94                                      const struct hashmap_entry *entry2,
  95                                      const void *keydata)
  96 {
  97         const char *key = keydata;
  98         const struct packed_git *pg1, *pg2;
  99
 100         pg1 = container_of(entry, const struct packed_git, packmap_ent);
 101         pg2 = container_of(entry2, const struct packed_git, packmap_ent);
 102
 103         return strcmp(pg1->pack_name, key ? key : pg2->pack_name);
 104 }
 105
 106 struct raw_object_store {
 107         /*
 108          * Set of all object directories; the main directory is first (and
 109          * cannot be NULL after initialization). Subsequent directories are
 110          * alternates.
 111          */
 112         struct object_directory *odb;
 113         struct object_directory **odb_tail;
 114         int loaded_alternates;
 115
 116         /*
 117          * A list of alternate object directories loaded from the environment;
 118          * this should not generally need to be accessed directly, but will
 119          * populate the "odb" list when prepare_alt_odb() is run.
 120          */
 121         char *alternate_db;
 122
 123         /*
 124          * Objects that should be substituted by other objects
 125          * (see git-replace(1)).
 126          */
 127         struct oidmap *replace_map;
 128
 129         struct commit_graph *commit_graph;
 130         unsigned commit_graph_attempted : 1; /* if loading has been attempted */
 131
 132         /*
 133          * private data
 134          *
 135          * should only be accessed directly by packfile.c and midx.c
 136          */
 137         struct multi_pack_index *multi_pack_index;
 138
 139         /*
 140          * private data
 141          *
 142          * should only be accessed directly by packfile.c
 143          */
 144
 145         struct packed_git *packed_git;
 146         /* A most-recently-used ordered version of the packed_git list. */
 147         struct list_head packed_git_mru;
 148
 149         /*
 150          * A map of packfiles to packed_git structs for tracking which
 151          * packs have been loaded already.
 152          */
 153         struct hashmap pack_map;
 154
 155         /*
 156          * A fast, rough count of the number of objects in the repository.
 157          * These two fields are not meant for direct access. Use
 158          * approximate_object_count() instead.
 159          */
 160         unsigned long approximate_object_count;
 161         unsigned approximate_object_count_valid : 1;
 162
 163         /*
 164          * Whether packed_git has already been populated with this repository's
 165          * packs.
 166          */
 167         unsigned packed_git_initialized : 1;
 168 };
 169
 170 struct raw_object_store *raw_object_store_new(void);
 171 void raw_object_store_clear(struct raw_object_store *o);
 172
 173 /*
 174  * Put in `buf` the name of the file in the local object database that
 175  * would be used to store a loose object with the specified oid.
 176  */
 177 const char *loose_object_path(struct repository *r, struct strbuf *buf,
 178                               const struct object_id *oid);
 179
 180 void *map_loose_object(struct repository *r, const struct object_id *oid,
 181                        unsigned long *size);
 182
 183 void *read_object_file_extended(struct repository *r,
 184                                 const struct object_id *oid,
 185                                 enum object_type *type,
 186                                 unsigned long *size, int lookup_replace);
 187 static inline void *repo_read_object_file(struct repository *r,
 188                                           const struct object_id *oid,
 189                                           enum object_type *type,
 190                                           unsigned long *size)
 191 {
 192         return read_object_file_extended(r, oid, type, size, 1);
 193 }
 194 #ifndef NO_THE_REPOSITORY_COMPATIBILITY_MACROS
 195 #define read_object_file(oid, type, size) repo_read_object_file(the_repository, oid, type, size)
 196 #endif
 197
 198 /* Read and unpack an object file into memory, write memory to an object file */
 199 int oid_object_info(struct repository *r, const struct object_id *, unsigned long *);
 200
 201 int hash_object_file(const void *buf, unsigned long len,
 202                      const char *type, struct object_id *oid);
 203
 204 int write_object_file(const void *buf, unsigned long len,
 205                       const char *type, struct object_id *oid);
 206
 207 int hash_object_file_literally(const void *buf, unsigned long len,
 208                                const char *type, struct object_id *oid,
 209                                unsigned flags);
 210
 211 int pretend_object_file(void *, unsigned long, enum object_type,
 212                         struct object_id *oid);
 213
 214 int force_object_loose(const struct object_id *oid, time_t mtime);
 215
 216 /*
 217  * Open the loose object at path, check its hash, and return the contents,
 218  * type, and size. If the object is a blob, then "contents" may return NULL,
 219  * to allow streaming of large blobs.
 220  *
 221  * Returns 0 on success, negative on error (details may be written to stderr).
 222  */
 223 int read_loose_object(const char *path,
 224                       const struct object_id *expected_oid,
 225                       enum object_type *type,
 226                       unsigned long *size,
 227                       void **contents);
 228
 229 #ifndef NO_THE_REPOSITORY_COMPATIBILITY_MACROS
 230 #define has_sha1_file_with_flags(sha1, flags) repo_has_sha1_file_with_flags(the_repository, sha1, flags)
 231 #define has_sha1_file(sha1) repo_has_sha1_file(the_repository, sha1)
 232 #endif
 233
 234 /* Same as the above, except for struct object_id. */
 235 int repo_has_object_file(struct repository *r, const struct object_id *oid);
 236 int repo_has_object_file_with_flags(struct repository *r,
 237                                     const struct object_id *oid, int flags);
 238 #ifndef NO_THE_REPOSITORY_COMPATIBILITY_MACROS
 239 #define has_object_file(oid) repo_has_object_file(the_repository, oid)
 240 #define has_object_file_with_flags(oid, flags) repo_has_object_file_with_flags(the_repository, oid, flags)
 241 #endif
 242
 243 /*
 244  * Return true iff an alternate object database has a loose object
 245  * with the specified name.  This function does not respect replace
 246  * references.
 247  */
 248 int has_loose_object_nonlocal(const struct object_id *);
 249
 250 void assert_oid_type(const struct object_id *oid, enum object_type expect);
 251
 252 struct object_info {
 253         /* Request */
 254         enum object_type *typep;
 255         unsigned long *sizep;
 256         off_t *disk_sizep;
 257         unsigned char *delta_base_sha1;
 258         struct strbuf *type_name;
 259         void **contentp;
 260
 261         /* Response */
 262         enum {
 263                 OI_CACHED,
 264                 OI_LOOSE,
 265                 OI_PACKED,
 266                 OI_DBCACHED
 267         } whence;
 268         union {
 269                 /*
 270                  * struct {
 271                  *      ... Nothing to expose in this case
 272                  * } cached;
 273                  * struct {
 274                  *      ... Nothing to expose in this case
 275                  * } loose;
 276                  */
 277                 struct {
 278                         struct packed_git *pack;
 279                         off_t offset;
 280                         unsigned int is_delta;
 281                 } packed;
 282         } u;
 283 };
 284
 285 /*
 286  * Initializer for a "struct object_info" that wants no items. You may
 287  * also memset() the memory to all-zeroes.
 288  */
 289 #define OBJECT_INFO_INIT {NULL}
 290
 291 /* Invoke lookup_replace_object() on the given hash */
 292 #define OBJECT_INFO_LOOKUP_REPLACE 1
 293 /* Allow reading from a loose object file of unknown/bogus type */
 294 #define OBJECT_INFO_ALLOW_UNKNOWN_TYPE 2
 295 /* Do not check cached storage */
 296 #define OBJECT_INFO_SKIP_CACHED 4
 297 /* Do not retry packed storage after checking packed and loose storage */
 298 #define OBJECT_INFO_QUICK 8
 299 /* Do not check loose object */
 300 #define OBJECT_INFO_IGNORE_LOOSE 16
 301 /*
 302  * Do not attempt to fetch the object if missing (even if fetch_is_missing is
 303  * nonzero).
 304  */
 305 #define OBJECT_INFO_SKIP_FETCH_OBJECT 32
 306 /*
 307  * This is meant for bulk prefetching of missing blobs in a partial
 308  * clone. Implies OBJECT_INFO_SKIP_FETCH_OBJECT and OBJECT_INFO_QUICK
 309  */
 310 #define OBJECT_INFO_FOR_PREFETCH (OBJECT_INFO_SKIP_FETCH_OBJECT | OBJECT_INFO_QUICK)
 311
 312 int oid_object_info_extended(struct repository *r,
 313                              const struct object_id *,
 314                              struct object_info *, unsigned flags);
 315
 316 /*
 317  * Iterate over the files in the loose-object parts of the object
 318  * directory "path", triggering the following callbacks:
 319  *
 320  *  - loose_object is called for each loose object we find.
 321  *
 322  *  - loose_cruft is called for any files that do not appear to be
 323  *    loose objects. Note that we only look in the loose object
 324  *    directories "objects/[0-9a-f]{2}/", so we will not report
 325  *    "objects/foobar" as cruft.
 326  *
 327  *  - loose_subdir is called for each top-level hashed subdirectory
 328  *    of the object directory (e.g., "$OBJDIR/f0"). It is called
 329  *    after the objects in the directory are processed.
 330  *
 331  * Any callback that is NULL will be ignored. Callbacks returning non-zero
 332  * will end the iteration.
 333  *
 334  * In the "buf" variant, "path" is a strbuf which will also be used as a
 335  * scratch buffer, but restored to its original contents before
 336  * the function returns.
 337  */
 338 typedef int each_loose_object_fn(const struct object_id *oid,
 339                                  const char *path,
 340                                  void *data);
 341 typedef int each_loose_cruft_fn(const char *basename,
 342                                 const char *path,
 343                                 void *data);
 344 typedef int each_loose_subdir_fn(unsigned int nr,
 345                                  const char *path,
 346                                  void *data);
 347 int for_each_file_in_obj_subdir(unsigned int subdir_nr,
 348                                 struct strbuf *path,
 349                                 each_loose_object_fn obj_cb,
 350                                 each_loose_cruft_fn cruft_cb,
 351                                 each_loose_subdir_fn subdir_cb,
 352                                 void *data);
 353 int for_each_loose_file_in_objdir(const char *path,
 354                                   each_loose_object_fn obj_cb,
 355                                   each_loose_cruft_fn cruft_cb,
 356                                   each_loose_subdir_fn subdir_cb,
 357                                   void *data);
 358 int for_each_loose_file_in_objdir_buf(struct strbuf *path,
 359                                       each_loose_object_fn obj_cb,
 360                                       each_loose_cruft_fn cruft_cb,
 361                                       each_loose_subdir_fn subdir_cb,
 362                                       void *data);
 363
 364 /* Flags for for_each_*_object() below. */
 365 enum for_each_object_flags {
 366         /* Iterate only over local objects, not alternates. */
 367         FOR_EACH_OBJECT_LOCAL_ONLY = (1<<0),
 368
 369         /* Only iterate over packs obtained from the promisor remote. */
 370         FOR_EACH_OBJECT_PROMISOR_ONLY = (1<<1),
 371
 372         /*
 373          * Visit objects within a pack in packfile order rather than .idx order
 374          */
 375         FOR_EACH_OBJECT_PACK_ORDER = (1<<2),
 376 };
 377
 378 /*
 379  * Iterate over all accessible loose objects without respect to
 380  * reachability. By default, this includes both local and alternate objects.
 381  * The order in which objects are visited is unspecified.
 382  *
 383  * Any flags specific to packs are ignored.
 384  */
 385 int for_each_loose_object(each_loose_object_fn, void *,
 386                           enum for_each_object_flags flags);
 387
 388 /*
 389  * Iterate over all accessible packed objects without respect to reachability.
 390  * By default, this includes both local and alternate packs.
 391  *
 392  * Note that some objects may appear twice if they are found in multiple packs.
 393  * Each pack is visited in an unspecified order. By default, objects within a
 394  * pack are visited in pack-idx order (i.e., sorted by oid).
 395  */
 396 typedef int each_packed_object_fn(const struct object_id *oid,
 397                                   struct packed_git *pack,
 398                                   uint32_t pos,
 399                                   void *data);
 400 int for_each_object_in_pack(struct packed_git *p,
 401                             each_packed_object_fn, void *data,
 402                             enum for_each_object_flags flags);
 403 int for_each_packed_object(each_packed_object_fn, void *,
 404                            enum for_each_object_flags flags);
 405
 406 #endif /* OBJECT_STORE_H */