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