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