notes.c

   1 #include "cache.h"
   2 #include "commit.h"
   3 #include "notes.h"
   4 #include "utf8.h"
   5 #include "strbuf.h"
   6 #include "tree-walk.h"
   7
   8 /*
   9  * Use a non-balancing simple 16-tree structure with struct int_node as
  10  * internal nodes, and struct leaf_node as leaf nodes. Each int_node has a
  11  * 16-array of pointers to its children.
  12  * The bottom 2 bits of each pointer is used to identify the pointer type
  13  * - ptr & 3 == 0 - NULL pointer, assert(ptr == NULL)
  14  * - ptr & 3 == 1 - pointer to next internal node - cast to struct int_node *
  15  * - ptr & 3 == 2 - pointer to note entry - cast to struct leaf_node *
  16  * - ptr & 3 == 3 - pointer to subtree entry - cast to struct leaf_node *
  17  *
  18  * The root node is a statically allocated struct int_node.
  19  */
  20 struct int_node {
  21         void *a[16];
  22 };
  23
  24 /*
  25  * Leaf nodes come in two variants, note entries and subtree entries,
  26  * distinguished by the LSb of the leaf node pointer (see above).
  27  * As a note entry, the key is the SHA1 of the referenced commit, and the
  28  * value is the SHA1 of the note object.
  29  * As a subtree entry, the key is the prefix SHA1 (w/trailing NULs) of the
  30  * referenced commit, using the last byte of the key to store the length of
  31  * the prefix. The value is the SHA1 of the tree object containing the notes
  32  * subtree.
  33  */
  34 struct leaf_node {
  35         unsigned char key_sha1[20];
  36         unsigned char val_sha1[20];
  37 };
  38
  39 #define PTR_TYPE_NULL     0
  40 #define PTR_TYPE_INTERNAL 1
  41 #define PTR_TYPE_NOTE     2
  42 #define PTR_TYPE_SUBTREE  3
  43
  44 #define GET_PTR_TYPE(ptr)       ((uintptr_t) (ptr) & 3)
  45 #define CLR_PTR_TYPE(ptr)       ((void *) ((uintptr_t) (ptr) & ~3))
  46 #define SET_PTR_TYPE(ptr, type) ((void *) ((uintptr_t) (ptr) | (type)))
  47
  48 #define GET_NIBBLE(n, sha1) (((sha1[n >> 1]) >> ((~n & 0x01) << 2)) & 0x0f)
  49
  50 #define SUBTREE_SHA1_PREFIXCMP(key_sha1, subtree_sha1) \
  51         (memcmp(key_sha1, subtree_sha1, subtree_sha1[19]))
  52
  53 static struct int_node root_node;
  54
  55 static int initialized;
  56
  57 static void load_subtree(struct leaf_node *subtree, struct int_node *node,
  58                 unsigned int n);
  59
  60 /*
  61  * Search the tree until the appropriate location for the given key is found:
  62  * 1. Start at the root node, with n = 0
  63  * 2. If a[0] at the current level is a matching subtree entry, unpack that
  64  *    subtree entry and remove it; restart search at the current level.
  65  * 3. Use the nth nibble of the key as an index into a:
  66  *    - If a[n] is an int_node, recurse from #2 into that node and increment n
  67  *    - If a matching subtree entry, unpack that subtree entry (and remove it);
  68  *      restart search at the current level.
  69  *    - Otherwise, we have found one of the following:
  70  *      - a subtree entry which does not match the key
  71  *      - a note entry which may or may not match the key
  72  *      - an unused leaf node (NULL)
  73  *      In any case, set *tree and *n, and return pointer to the tree location.
  74  */
  75 static void **note_tree_search(struct int_node **tree,
  76                 unsigned char *n, const unsigned char *key_sha1)
  77 {
  78         struct leaf_node *l;
  79         unsigned char i;
  80         void *p = (*tree)->a[0];
  81
  82         if (GET_PTR_TYPE(p) == PTR_TYPE_SUBTREE) {
  83                 l = (struct leaf_node *) CLR_PTR_TYPE(p);
  84                 if (!SUBTREE_SHA1_PREFIXCMP(key_sha1, l->key_sha1)) {
  85                         /* unpack tree and resume search */
  86                         (*tree)->a[0] = NULL;
  87                         load_subtree(l, *tree, *n);
  88                         free(l);
  89                         return note_tree_search(tree, n, key_sha1);
  90                 }
  91         }
  92
  93         i = GET_NIBBLE(*n, key_sha1);
  94         p = (*tree)->a[i];
  95         switch (GET_PTR_TYPE(p)) {
  96         case PTR_TYPE_INTERNAL:
  97                 *tree = CLR_PTR_TYPE(p);
  98                 (*n)++;
  99                 return note_tree_search(tree, n, key_sha1);
 100         case PTR_TYPE_SUBTREE:
 101                 l = (struct leaf_node *) CLR_PTR_TYPE(p);
 102                 if (!SUBTREE_SHA1_PREFIXCMP(key_sha1, l->key_sha1)) {
 103                         /* unpack tree and resume search */
 104                         (*tree)->a[i] = NULL;
 105                         load_subtree(l, *tree, *n);
 106                         free(l);
 107                         return note_tree_search(tree, n, key_sha1);
 108                 }
 109                 /* fall through */
 110         default:
 111                 return &((*tree)->a[i]);
 112         }
 113 }
 114
 115 /*
 116  * To find a leaf_node:
 117  * Search to the tree location appropriate for the given key:
 118  * If a note entry with matching key, return the note entry, else return NULL.
 119  */
 120 static struct leaf_node *note_tree_find(struct int_node *tree, unsigned char n,
 121                 const unsigned char *key_sha1)
 122 {
 123         void **p = note_tree_search(&tree, &n, key_sha1);
 124         if (GET_PTR_TYPE(*p) == PTR_TYPE_NOTE) {
 125                 struct leaf_node *l = (struct leaf_node *) CLR_PTR_TYPE(*p);
 126                 if (!hashcmp(key_sha1, l->key_sha1))
 127                         return l;
 128         }
 129         return NULL;
 130 }
 131
 132 /* Create a new blob object by concatenating the two given blob objects */
 133 static int concatenate_notes(unsigned char *cur_sha1,
 134                 const unsigned char *new_sha1)
 135 {
 136         char *cur_msg, *new_msg, *buf;
 137         unsigned long cur_len, new_len, buf_len;
 138         enum object_type cur_type, new_type;
 139         int ret;
 140
 141         /* read in both note blob objects */
 142         new_msg = read_sha1_file(new_sha1, &new_type, &new_len);
 143         if (!new_msg || !new_len || new_type != OBJ_BLOB) {
 144                 free(new_msg);
 145                 return 0;
 146         }
 147         cur_msg = read_sha1_file(cur_sha1, &cur_type, &cur_len);
 148         if (!cur_msg || !cur_len || cur_type != OBJ_BLOB) {
 149                 free(cur_msg);
 150                 free(new_msg);
 151                 hashcpy(cur_sha1, new_sha1);
 152                 return 0;
 153         }
 154
 155         /* we will separate the notes by a newline anyway */
 156         if (cur_msg[cur_len - 1] == '\n')
 157                 cur_len--;
 158
 159         /* concatenate cur_msg and new_msg into buf */
 160         buf_len = cur_len + 1 + new_len;
 161         buf = (char *) xmalloc(buf_len);
 162         memcpy(buf, cur_msg, cur_len);
 163         buf[cur_len] = '\n';
 164         memcpy(buf + cur_len + 1, new_msg, new_len);
 165
 166         free(cur_msg);
 167         free(new_msg);
 168
 169         /* create a new blob object from buf */
 170         ret = write_sha1_file(buf, buf_len, "blob", cur_sha1);
 171         free(buf);
 172         return ret;
 173 }
 174
 175 /*
 176  * To insert a leaf_node:
 177  * Search to the tree location appropriate for the given leaf_node's key:
 178  * - If location is unused (NULL), store the tweaked pointer directly there
 179  * - If location holds a note entry that matches the note-to-be-inserted, then
 180  *   concatenate the two notes.
 181  * - If location holds a note entry that matches the subtree-to-be-inserted,
 182  *   then unpack the subtree-to-be-inserted into the location.
 183  * - If location holds a matching subtree entry, unpack the subtree at that
 184  *   location, and restart the insert operation from that level.
 185  * - Else, create a new int_node, holding both the node-at-location and the
 186  *   node-to-be-inserted, and store the new int_node into the location.
 187  */
 188 static void note_tree_insert(struct int_node *tree, unsigned char n,
 189                 struct leaf_node *entry, unsigned char type)
 190 {
 191         struct int_node *new_node;
 192         struct leaf_node *l;
 193         void **p = note_tree_search(&tree, &n, entry->key_sha1);
 194
 195         assert(GET_PTR_TYPE(entry) == 0); /* no type bits set */
 196         l = (struct leaf_node *) CLR_PTR_TYPE(*p);
 197         switch (GET_PTR_TYPE(*p)) {
 198         case PTR_TYPE_NULL:
 199                 assert(!*p);
 200                 *p = SET_PTR_TYPE(entry, type);
 201                 return;
 202         case PTR_TYPE_NOTE:
 203                 switch (type) {
 204                 case PTR_TYPE_NOTE:
 205                         if (!hashcmp(l->key_sha1, entry->key_sha1)) {
 206                                 /* skip concatenation if l == entry */
 207                                 if (!hashcmp(l->val_sha1, entry->val_sha1))
 208                                         return;
 209
 210                                 if (concatenate_notes(l->val_sha1,
 211                                                 entry->val_sha1))
 212                                         die("failed to concatenate note %s "
 213                                             "into note %s for commit %s",
 214                                             sha1_to_hex(entry->val_sha1),
 215                                             sha1_to_hex(l->val_sha1),
 216                                             sha1_to_hex(l->key_sha1));
 217                                 free(entry);
 218                                 return;
 219                         }
 220                         break;
 221                 case PTR_TYPE_SUBTREE:
 222                         if (!SUBTREE_SHA1_PREFIXCMP(l->key_sha1,
 223                                                     entry->key_sha1)) {
 224                                 /* unpack 'entry' */
 225                                 load_subtree(entry, tree, n);
 226                                 free(entry);
 227                                 return;
 228                         }
 229                         break;
 230                 }
 231                 break;
 232         case PTR_TYPE_SUBTREE:
 233                 if (!SUBTREE_SHA1_PREFIXCMP(entry->key_sha1, l->key_sha1)) {
 234                         /* unpack 'l' and restart insert */
 235                         *p = NULL;
 236                         load_subtree(l, tree, n);
 237                         free(l);
 238                         note_tree_insert(tree, n, entry, type);
 239                         return;
 240                 }
 241                 break;
 242         }
 243
 244         /* non-matching leaf_node */
 245         assert(GET_PTR_TYPE(*p) == PTR_TYPE_NOTE ||
 246                GET_PTR_TYPE(*p) == PTR_TYPE_SUBTREE);
 247         new_node = (struct int_node *) xcalloc(sizeof(struct int_node), 1);
 248         note_tree_insert(new_node, n + 1, l, GET_PTR_TYPE(*p));
 249         *p = SET_PTR_TYPE(new_node, PTR_TYPE_INTERNAL);
 250         note_tree_insert(new_node, n + 1, entry, type);
 251 }
 252
 253 /* Free the entire notes data contained in the given tree */
 254 static void note_tree_free(struct int_node *tree)
 255 {
 256         unsigned int i;
 257         for (i = 0; i < 16; i++) {
 258                 void *p = tree->a[i];
 259                 switch (GET_PTR_TYPE(p)) {
 260                 case PTR_TYPE_INTERNAL:
 261                         note_tree_free(CLR_PTR_TYPE(p));
 262                         /* fall through */
 263                 case PTR_TYPE_NOTE:
 264                 case PTR_TYPE_SUBTREE:
 265                         free(CLR_PTR_TYPE(p));
 266                 }
 267         }
 268 }
 269
 270 /*
 271  * Convert a partial SHA1 hex string to the corresponding partial SHA1 value.
 272  * - hex      - Partial SHA1 segment in ASCII hex format
 273  * - hex_len  - Length of above segment. Must be multiple of 2 between 0 and 40
 274  * - sha1     - Partial SHA1 value is written here
 275  * - sha1_len - Max #bytes to store in sha1, Must be >= hex_len / 2, and < 20
 276  * Returns -1 on error (invalid arguments or invalid SHA1 (not in hex format)).
 277  * Otherwise, returns number of bytes written to sha1 (i.e. hex_len / 2).
 278  * Pads sha1 with NULs up to sha1_len (not included in returned length).
 279  */
 280 static int get_sha1_hex_segment(const char *hex, unsigned int hex_len,
 281                 unsigned char *sha1, unsigned int sha1_len)
 282 {
 283         unsigned int i, len = hex_len >> 1;
 284         if (hex_len % 2 != 0 || len > sha1_len)
 285                 return -1;
 286         for (i = 0; i < len; i++) {
 287                 unsigned int val = (hexval(hex[0]) << 4) | hexval(hex[1]);
 288                 if (val & ~0xff)
 289                         return -1;
 290                 *sha1++ = val;
 291                 hex += 2;
 292         }
 293         for (; i < sha1_len; i++)
 294                 *sha1++ = 0;
 295         return len;
 296 }
 297
 298 static void load_subtree(struct leaf_node *subtree, struct int_node *node,
 299                 unsigned int n)
 300 {
 301         unsigned char commit_sha1[20];
 302         unsigned int prefix_len;
 303         void *buf;
 304         struct tree_desc desc;
 305         struct name_entry entry;
 306
 307         buf = fill_tree_descriptor(&desc, subtree->val_sha1);
 308         if (!buf)
 309                 die("Could not read %s for notes-index",
 310                      sha1_to_hex(subtree->val_sha1));
 311
 312         prefix_len = subtree->key_sha1[19];
 313         assert(prefix_len * 2 >= n);
 314         memcpy(commit_sha1, subtree->key_sha1, prefix_len);
 315         while (tree_entry(&desc, &entry)) {
 316                 int len = get_sha1_hex_segment(entry.path, strlen(entry.path),
 317                                 commit_sha1 + prefix_len, 20 - prefix_len);
 318                 if (len < 0)
 319                         continue; /* entry.path is not a SHA1 sum. Skip */
 320                 len += prefix_len;
 321
 322                 /*
 323                  * If commit SHA1 is complete (len == 20), assume note object
 324                  * If commit SHA1 is incomplete (len < 20), assume note subtree
 325                  */
 326                 if (len <= 20) {
 327                         unsigned char type = PTR_TYPE_NOTE;
 328                         struct leaf_node *l = (struct leaf_node *)
 329                                 xcalloc(sizeof(struct leaf_node), 1);
 330                         hashcpy(l->key_sha1, commit_sha1);
 331                         hashcpy(l->val_sha1, entry.sha1);
 332                         if (len < 20) {
 333                                 if (!S_ISDIR(entry.mode))
 334                                         continue; /* entry cannot be subtree */
 335                                 l->key_sha1[19] = (unsigned char) len;
 336                                 type = PTR_TYPE_SUBTREE;
 337                         }
 338                         note_tree_insert(node, n, l, type);
 339                 }
 340         }
 341         free(buf);
 342 }
 343
 344 static void initialize_notes(const char *notes_ref_name)
 345 {
 346         unsigned char sha1[20], commit_sha1[20];
 347         unsigned mode;
 348         struct leaf_node root_tree;
 349
 350         if (!notes_ref_name || read_ref(notes_ref_name, commit_sha1) ||
 351             get_tree_entry(commit_sha1, "", sha1, &mode))
 352                 return;
 353
 354         hashclr(root_tree.key_sha1);
 355         hashcpy(root_tree.val_sha1, sha1);
 356         load_subtree(&root_tree, &root_node, 0);
 357 }
 358
 359 static unsigned char *lookup_notes(const unsigned char *commit_sha1)
 360 {
 361         struct leaf_node *found = note_tree_find(&root_node, 0, commit_sha1);
 362         if (found)
 363                 return found->val_sha1;
 364         return NULL;
 365 }
 366
 367 void free_notes(void)
 368 {
 369         note_tree_free(&root_node);
 370         memset(&root_node, 0, sizeof(struct int_node));
 371         initialized = 0;
 372 }
 373
 374 void get_commit_notes(const struct commit *commit, struct strbuf *sb,
 375                 const char *output_encoding, int flags)
 376 {
 377         static const char utf8[] = "utf-8";
 378         unsigned char *sha1;
 379         char *msg, *msg_p;
 380         unsigned long linelen, msglen;
 381         enum object_type type;
 382
 383         if (!initialized) {
 384                 const char *env = getenv(GIT_NOTES_REF_ENVIRONMENT);
 385                 if (env)
 386                         notes_ref_name = getenv(GIT_NOTES_REF_ENVIRONMENT);
 387                 else if (!notes_ref_name)
 388                         notes_ref_name = GIT_NOTES_DEFAULT_REF;
 389                 initialize_notes(notes_ref_name);
 390                 initialized = 1;
 391         }
 392
 393         sha1 = lookup_notes(commit->object.sha1);
 394         if (!sha1)
 395                 return;
 396
 397         if (!(msg = read_sha1_file(sha1, &type, &msglen)) || !msglen ||
 398                         type != OBJ_BLOB) {
 399                 free(msg);
 400                 return;
 401         }
 402
 403         if (output_encoding && *output_encoding &&
 404                         strcmp(utf8, output_encoding)) {
 405                 char *reencoded = reencode_string(msg, output_encoding, utf8);
 406                 if (reencoded) {
 407                         free(msg);
 408                         msg = reencoded;
 409                         msglen = strlen(msg);
 410                 }
 411         }
 412
 413         /* we will end the annotation by a newline anyway */
 414         if (msglen && msg[msglen - 1] == '\n')
 415                 msglen--;
 416
 417         if (flags & NOTES_SHOW_HEADER)
 418                 strbuf_addstr(sb, "\nNotes:\n");
 419
 420         for (msg_p = msg; msg_p < msg + msglen; msg_p += linelen + 1) {
 421                 linelen = strchrnul(msg_p, '\n') - msg_p;
 422
 423                 if (flags & NOTES_INDENT)
 424                         strbuf_addstr(sb, "    ");
 425                 strbuf_add(sb, msg_p, linelen);
 426                 strbuf_addch(sb, '\n');
 427         }
 428
 429         free(msg);
 430 }