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