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