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