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 alternate_object_database {
  11         struct alternate_object_database *next;
  12
  13         /* see alt_scratch_buf() */
  14         struct strbuf scratch;
  15         size_t base_len;
  16
  17         /*
  18          * Used to store the results of readdir(3) calls when searching
  19          * for unique abbreviated hashes.  This cache is never
  20          * invalidated, thus it's racy and not necessarily accurate.
  21          * That's fine for its purpose; don't use it for tasks requiring
  22          * greater accuracy!
  23          */
  24         char loose_objects_subdir_seen[256];
  25         struct oid_array loose_objects_cache;
  26
  27         /*
  28          * Path to the alternative object store. If this is a relative path,
  29          * it is relative to the current working directory.
  30          */
  31         char path[FLEX_ARRAY];
  32 };
  33 void prepare_alt_odb(struct repository *r);
  34 char *compute_alternate_path(const char *path, struct strbuf *err);
  35 typedef int alt_odb_fn(struct alternate_object_database *, void *);
  36 int foreach_alt_odb(alt_odb_fn, void*);
  37
  38 /*
  39  * Allocate a "struct alternate_object_database" but do _not_ actually
  40  * add it to the list of alternates.
  41  */
  42 struct alternate_object_database *alloc_alt_odb(const char *dir);
  43
  44 /*
  45  * Add the directory to the on-disk alternates file; the new entry will also
  46  * take effect in the current process.
  47  */
  48 void add_to_alternates_file(const char *dir);
  49
  50 /*
  51  * Add the directory to the in-memory list of alternates (along with any
  52  * recursive alternates it points to), but do not modify the on-disk alternates
  53  * file.
  54  */
  55 void add_to_alternates_memory(const char *dir);
  56
  57 /*
  58  * Returns a scratch strbuf pre-filled with the alternate object directory,
  59  * including a trailing slash, which can be used to access paths in the
  60  * alternate. Always use this over direct access to alt->scratch, as it
  61  * cleans up any previous use of the scratch buffer.
  62  */
  63 struct strbuf *alt_scratch_buf(struct alternate_object_database *alt);
  64
  65 struct packed_git {
  66         struct packed_git *next;
  67         struct list_head mru;
  68         struct pack_window *windows;
  69         off_t pack_size;
  70         const void *index_data;
  71         size_t index_size;
  72         uint32_t num_objects;
  73         uint32_t num_bad_objects;
  74         unsigned char *bad_object_sha1;
  75         int index_version;
  76         time_t mtime;
  77         int pack_fd;
  78         int index;              /* for builtin/pack-objects.c */
  79         unsigned pack_local:1,
  80                  pack_keep:1,
  81                  pack_keep_in_core:1,
  82                  freshened:1,
  83                  do_not_close:1,
  84                  pack_promisor:1;
  85         unsigned char sha1[20];
  86         struct revindex_entry *revindex;
  87         /* something like ".git/objects/pack/xxxxx.pack" */
  88         char pack_name[FLEX_ARRAY]; /* more */
  89 };
  90
  91 struct multi_pack_index;
  92
  93 struct raw_object_store {
  94         /*
  95          * Path to the repository's object store.
  96          * Cannot be NULL after initialization.
  97          */
  98         char *objectdir;
  99
 100         /* Path to extra alternate object database if not NULL */
 101         char *alternate_db;
 102
 103         struct alternate_object_database *alt_odb_list;
 104         struct alternate_object_database **alt_odb_tail;
 105
 106         /*
 107          * Objects that should be substituted by other objects
 108          * (see git-replace(1)).
 109          */
 110         struct oidmap *replace_map;
 111
 112         struct commit_graph *commit_graph;
 113         unsigned commit_graph_attempted : 1; /* if loading has been attempted */
 114
 115         /*
 116          * private data
 117          *
 118          * should only be accessed directly by packfile.c and midx.c
 119          */
 120         struct multi_pack_index *multi_pack_index;
 121
 122         /*
 123          * private data
 124          *
 125          * should only be accessed directly by packfile.c
 126          */
 127
 128         struct packed_git *packed_git;
 129         /* A most-recently-used ordered version of the packed_git list. */
 130         struct list_head packed_git_mru;
 131
 132         /*
 133          * A linked list containing all packfiles, starting with those
 134          * contained in the multi_pack_index.
 135          */
 136         struct packed_git *all_packs;
 137
 138         /*
 139          * A fast, rough count of the number of objects in the repository.
 140          * These two fields are not meant for direct access. Use
 141          * approximate_object_count() instead.
 142          */
 143         unsigned long approximate_object_count;
 144         unsigned approximate_object_count_valid : 1;
 145
 146         /*
 147          * Whether packed_git has already been populated with this repository's
 148          * packs.
 149          */
 150         unsigned packed_git_initialized : 1;
 151 };
 152
 153 struct raw_object_store *raw_object_store_new(void);
 154 void raw_object_store_clear(struct raw_object_store *o);
 155
 156 /*
 157  * Put in `buf` the name of the file in the local object database that
 158  * would be used to store a loose object with the specified sha1.
 159  */
 160 void sha1_file_name(struct repository *r, struct strbuf *buf, const unsigned char *sha1);
 161
 162 void *map_sha1_file(struct repository *r, const unsigned char *sha1, unsigned long *size);
 163
 164 extern void *read_object_file_extended(struct repository *r,
 165                                        const struct object_id *oid,
 166                                        enum object_type *type,
 167                                        unsigned long *size, int lookup_replace);
 168 static inline void *repo_read_object_file(struct repository *r,
 169                                           const struct object_id *oid,
 170                                           enum object_type *type,
 171                                           unsigned long *size)
 172 {
 173         return read_object_file_extended(r, oid, type, size, 1);
 174 }
 175 #ifndef NO_THE_REPOSITORY_COMPATIBILITY_MACROS
 176 #define read_object_file(oid, type, size) repo_read_object_file(the_repository, oid, type, size)
 177 #endif
 178
 179 /* Read and unpack an object file into memory, write memory to an object file */
 180 int oid_object_info(struct repository *r, const struct object_id *, unsigned long *);
 181
 182 extern int hash_object_file(const void *buf, unsigned long len,
 183                             const char *type, struct object_id *oid);
 184
 185 extern int write_object_file(const void *buf, unsigned long len,
 186                              const char *type, struct object_id *oid);
 187
 188 extern int hash_object_file_literally(const void *buf, unsigned long len,
 189                                       const char *type, struct object_id *oid,
 190                                       unsigned flags);
 191
 192 extern int pretend_object_file(void *, unsigned long, enum object_type,
 193                                struct object_id *oid);
 194
 195 extern int force_object_loose(const struct object_id *oid, time_t mtime);
 196
 197 /*
 198  * Open the loose object at path, check its hash, and return the contents,
 199  * type, and size. If the object is a blob, then "contents" may return NULL,
 200  * to allow streaming of large blobs.
 201  *
 202  * Returns 0 on success, negative on error (details may be written to stderr).
 203  */
 204 int read_loose_object(const char *path,
 205                       const struct object_id *expected_oid,
 206                       enum object_type *type,
 207                       unsigned long *size,
 208                       void **contents);
 209
 210 /*
 211  * Convenience for sha1_object_info_extended() with a NULL struct
 212  * object_info. OBJECT_INFO_SKIP_CACHED is automatically set; pass
 213  * nonzero flags to also set other flags.
 214  */
 215 int repo_has_sha1_file_with_flags(struct repository *r,
 216                                   const unsigned char *sha1, int flags);
 217 static inline int repo_has_sha1_file(struct repository *r,
 218                                      const unsigned char *sha1)
 219 {
 220         return repo_has_sha1_file_with_flags(r, sha1, 0);
 221 }
 222
 223 #ifndef NO_THE_REPOSITORY_COMPATIBILITY_MACROS
 224 #define has_sha1_file_with_flags(sha1, flags) repo_has_sha1_file_with_flags(the_repository, sha1, flags)
 225 #define has_sha1_file(sha1) repo_has_sha1_file(the_repository, sha1)
 226 #endif
 227
 228 /* Same as the above, except for struct object_id. */
 229 int repo_has_object_file(struct repository *r, const struct object_id *oid);
 230 int repo_has_object_file_with_flags(struct repository *r,
 231                                     const struct object_id *oid, int flags);
 232 #ifndef NO_THE_REPOSITORY_COMPATIBILITY_MACROS
 233 #define has_object_file(oid) repo_has_object_file(the_repository, oid)
 234 #define has_object_file_with_flags(oid, flags) repo_has_object_file_with_flags(the_repository, oid, flags)
 235 #endif
 236
 237 /*
 238  * Return true iff an alternate object database has a loose object
 239  * with the specified name.  This function does not respect replace
 240  * references.
 241  */
 242 extern int has_loose_object_nonlocal(const struct object_id *);
 243
 244 extern void assert_oid_type(const struct object_id *oid, enum object_type expect);
 245
 246 struct object_info {
 247         /* Request */
 248         enum object_type *typep;
 249         unsigned long *sizep;
 250         off_t *disk_sizep;
 251         unsigned char *delta_base_sha1;
 252         struct strbuf *type_name;
 253         void **contentp;
 254
 255         /* Response */
 256         enum {
 257                 OI_CACHED,
 258                 OI_LOOSE,
 259                 OI_PACKED,
 260                 OI_DBCACHED
 261         } whence;
 262         union {
 263                 /*
 264                  * struct {
 265                  *      ... Nothing to expose in this case
 266                  * } cached;
 267                  * struct {
 268                  *      ... Nothing to expose in this case
 269                  * } loose;
 270                  */
 271                 struct {
 272                         struct packed_git *pack;
 273                         off_t offset;
 274                         unsigned int is_delta;
 275                 } packed;
 276         } u;
 277 };
 278
 279 /*
 280  * Initializer for a "struct object_info" that wants no items. You may
 281  * also memset() the memory to all-zeroes.
 282  */
 283 #define OBJECT_INFO_INIT {NULL}
 284
 285 /* Invoke lookup_replace_object() on the given hash */
 286 #define OBJECT_INFO_LOOKUP_REPLACE 1
 287 /* Allow reading from a loose object file of unknown/bogus type */
 288 #define OBJECT_INFO_ALLOW_UNKNOWN_TYPE 2
 289 /* Do not check cached storage */
 290 #define OBJECT_INFO_SKIP_CACHED 4
 291 /* Do not retry packed storage after checking packed and loose storage */
 292 #define OBJECT_INFO_QUICK 8
 293 /* Do not check loose object */
 294 #define OBJECT_INFO_IGNORE_LOOSE 16
 295
 296 int oid_object_info_extended(struct repository *r,
 297                              const struct object_id *,
 298                              struct object_info *, unsigned flags);
 299
 300 /*
 301  * Iterate over the files in the loose-object parts of the object
 302  * directory "path", triggering the following callbacks:
 303  *
 304  *  - loose_object is called for each loose object we find.
 305  *
 306  *  - loose_cruft is called for any files that do not appear to be
 307  *    loose objects. Note that we only look in the loose object
 308  *    directories "objects/[0-9a-f]{2}/", so we will not report
 309  *    "objects/foobar" as cruft.
 310  *
 311  *  - loose_subdir is called for each top-level hashed subdirectory
 312  *    of the object directory (e.g., "$OBJDIR/f0"). It is called
 313  *    after the objects in the directory are processed.
 314  *
 315  * Any callback that is NULL will be ignored. Callbacks returning non-zero
 316  * will end the iteration.
 317  *
 318  * In the "buf" variant, "path" is a strbuf which will also be used as a
 319  * scratch buffer, but restored to its original contents before
 320  * the function returns.
 321  */
 322 typedef int each_loose_object_fn(const struct object_id *oid,
 323                                  const char *path,
 324                                  void *data);
 325 typedef int each_loose_cruft_fn(const char *basename,
 326                                 const char *path,
 327                                 void *data);
 328 typedef int each_loose_subdir_fn(unsigned int nr,
 329                                  const char *path,
 330                                  void *data);
 331 int for_each_file_in_obj_subdir(unsigned int subdir_nr,
 332                                 struct strbuf *path,
 333                                 each_loose_object_fn obj_cb,
 334                                 each_loose_cruft_fn cruft_cb,
 335                                 each_loose_subdir_fn subdir_cb,
 336                                 void *data);
 337 int for_each_loose_file_in_objdir(const char *path,
 338                                   each_loose_object_fn obj_cb,
 339                                   each_loose_cruft_fn cruft_cb,
 340                                   each_loose_subdir_fn subdir_cb,
 341                                   void *data);
 342 int for_each_loose_file_in_objdir_buf(struct strbuf *path,
 343                                       each_loose_object_fn obj_cb,
 344                                       each_loose_cruft_fn cruft_cb,
 345                                       each_loose_subdir_fn subdir_cb,
 346                                       void *data);
 347
 348 /* Flags for for_each_*_object() below. */
 349 enum for_each_object_flags {
 350         /* Iterate only over local objects, not alternates. */
 351         FOR_EACH_OBJECT_LOCAL_ONLY = (1<<0),
 352
 353         /* Only iterate over packs obtained from the promisor remote. */
 354         FOR_EACH_OBJECT_PROMISOR_ONLY = (1<<1),
 355
 356         /*
 357          * Visit objects within a pack in packfile order rather than .idx order
 358          */
 359         FOR_EACH_OBJECT_PACK_ORDER = (1<<2),
 360 };
 361
 362 /*
 363  * Iterate over all accessible loose objects without respect to
 364  * reachability. By default, this includes both local and alternate objects.
 365  * The order in which objects are visited is unspecified.
 366  *
 367  * Any flags specific to packs are ignored.
 368  */
 369 int for_each_loose_object(each_loose_object_fn, void *,
 370                           enum for_each_object_flags flags);
 371
 372 /*
 373  * Iterate over all accessible packed objects without respect to reachability.
 374  * By default, this includes both local and alternate packs.
 375  *
 376  * Note that some objects may appear twice if they are found in multiple packs.
 377  * Each pack is visited in an unspecified order. By default, objects within a
 378  * pack are visited in pack-idx order (i.e., sorted by oid).
 379  */
 380 typedef int each_packed_object_fn(const struct object_id *oid,
 381                                   struct packed_git *pack,
 382                                   uint32_t pos,
 383                                   void *data);
 384 int for_each_object_in_pack(struct packed_git *p,
 385                             each_packed_object_fn, void *data,
 386                             enum for_each_object_flags flags);
 387 int for_each_packed_object(each_packed_object_fn, void *,
 388                            enum for_each_object_flags flags);
 389
 390 #endif /* OBJECT_STORE_H */