notes.c

   1 #include "cache.h"
   2 #include "notes.h"
   3 #include "utf8.h"
   4 #include "strbuf.h"
   5 #include "tree-walk.h"
   6
   7 /*
   8  * Use a non-balancing simple 16-tree structure with struct int_node as
   9  * internal nodes, and struct leaf_node as leaf nodes. Each int_node has a
  10  * 16-array of pointers to its children.
  11  * The bottom 2 bits of each pointer is used to identify the pointer type
  12  * - ptr & 3 == 0 - NULL pointer, assert(ptr == NULL)
  13  * - ptr & 3 == 1 - pointer to next internal node - cast to struct int_node *
  14  * - ptr & 3 == 2 - pointer to note entry - cast to struct leaf_node *
  15  * - ptr & 3 == 3 - pointer to subtree entry - cast to struct leaf_node *
  16  *
  17  * The root node is a statically allocated struct int_node.
  18  */
  19 struct int_node {
  20         void *a[16];
  21 };
  22
  23 /*
  24  * Leaf nodes come in two variants, note entries and subtree entries,
  25  * distinguished by the LSb of the leaf node pointer (see above).
  26  * As a note entry, the key is the SHA1 of the referenced object, and the
  27  * value is the SHA1 of the note object.
  28  * As a subtree entry, the key is the prefix SHA1 (w/trailing NULs) of the
  29  * referenced object, using the last byte of the key to store the length of
  30  * the prefix. The value is the SHA1 of the tree object containing the notes
  31  * subtree.
  32  */
  33 struct leaf_node {
  34         unsigned char key_sha1[20];
  35         unsigned char val_sha1[20];
  36 };
  37
  38 #define PTR_TYPE_NULL     0
  39 #define PTR_TYPE_INTERNAL 1
  40 #define PTR_TYPE_NOTE     2
  41 #define PTR_TYPE_SUBTREE  3
  42
  43 #define GET_PTR_TYPE(ptr)       ((uintptr_t) (ptr) & 3)
  44 #define CLR_PTR_TYPE(ptr)       ((void *) ((uintptr_t) (ptr) & ~3))
  45 #define SET_PTR_TYPE(ptr, type) ((void *) ((uintptr_t) (ptr) | (type)))
  46
  47 #define GET_NIBBLE(n, sha1) (((sha1[n >> 1]) >> ((~n & 0x01) << 2)) & 0x0f)
  48
  49 #define SUBTREE_SHA1_PREFIXCMP(key_sha1, subtree_sha1) \
  50         (memcmp(key_sha1, subtree_sha1, subtree_sha1[19]))
  51
  52 static struct int_node root_node;
  53
  54 static int initialized;
  55
  56 static void load_subtree(struct leaf_node *subtree, struct int_node *node,
  57                 unsigned int n);
  58
  59 /*
  60  * Search the tree until the appropriate location for the given key is found:
  61  * 1. Start at the root node, with n = 0
  62  * 2. If a[0] at the current level is a matching subtree entry, unpack that
  63  *    subtree entry and remove it; restart search at the current level.
  64  * 3. Use the nth nibble of the key as an index into a:
  65  *    - If a[n] is an int_node, recurse from #2 into that node and increment n
  66  *    - If a matching subtree entry, unpack that subtree entry (and remove it);
  67  *      restart search at the current level.
  68  *    - Otherwise, we have found one of the following:
  69  *      - a subtree entry which does not match the key
  70  *      - a note entry which may or may not match the key
  71  *      - an unused leaf node (NULL)
  72  *      In any case, set *tree and *n, and return pointer to the tree location.
  73  */
  74 static void **note_tree_search(struct int_node **tree,
  75                 unsigned char *n, const unsigned char *key_sha1)
  76 {
  77         struct leaf_node *l;
  78         unsigned char i;
  79         void *p = (*tree)->a[0];
  80
  81         if (GET_PTR_TYPE(p) == PTR_TYPE_SUBTREE) {
  82                 l = (struct leaf_node *) CLR_PTR_TYPE(p);
  83                 if (!SUBTREE_SHA1_PREFIXCMP(key_sha1, l->key_sha1)) {
  84                         /* unpack tree and resume search */
  85                         (*tree)->a[0] = NULL;
  86                         load_subtree(l, *tree, *n);
  87                         free(l);
  88                         return note_tree_search(tree, n, key_sha1);
  89                 }
  90         }
  91
  92         i = GET_NIBBLE(*n, key_sha1);
  93         p = (*tree)->a[i];
  94         switch (GET_PTR_TYPE(p)) {
  95         case PTR_TYPE_INTERNAL:
  96                 *tree = CLR_PTR_TYPE(p);
  97                 (*n)++;
  98                 return note_tree_search(tree, n, key_sha1);
  99         case PTR_TYPE_SUBTREE:
 100                 l = (struct leaf_node *) CLR_PTR_TYPE(p);
 101                 if (!SUBTREE_SHA1_PREFIXCMP(key_sha1, l->key_sha1)) {
 102                         /* unpack tree and resume search */
 103                         (*tree)->a[i] = NULL;
 104                         load_subtree(l, *tree, *n);
 105                         free(l);
 106                         return note_tree_search(tree, n, key_sha1);
 107                 }
 108                 /* fall through */
 109         default:
 110                 return &((*tree)->a[i]);
 111         }
 112 }
 113
 114 /*
 115  * To find a leaf_node:
 116  * Search to the tree location appropriate for the given key:
 117  * If a note entry with matching key, return the note entry, else return NULL.
 118  */
 119 static struct leaf_node *note_tree_find(struct int_node *tree, unsigned char n,
 120                 const unsigned char *key_sha1)
 121 {
 122         void **p = note_tree_search(&tree, &n, key_sha1);
 123         if (GET_PTR_TYPE(*p) == PTR_TYPE_NOTE) {
 124                 struct leaf_node *l = (struct leaf_node *) CLR_PTR_TYPE(*p);
 125                 if (!hashcmp(key_sha1, l->key_sha1))
 126                         return l;
 127         }
 128         return NULL;
 129 }
 130
 131 /* Create a new blob object by concatenating the two given blob objects */
 132 static int concatenate_notes(unsigned char *cur_sha1,
 133                 const unsigned char *new_sha1)
 134 {
 135         char *cur_msg, *new_msg, *buf;
 136         unsigned long cur_len, new_len, buf_len;
 137         enum object_type cur_type, new_type;
 138         int ret;
 139
 140         /* read in both note blob objects */
 141         new_msg = read_sha1_file(new_sha1, &new_type, &new_len);
 142         if (!new_msg || !new_len || new_type != OBJ_BLOB) {
 143                 free(new_msg);
 144                 return 0;
 145         }
 146         cur_msg = read_sha1_file(cur_sha1, &cur_type, &cur_len);
 147         if (!cur_msg || !cur_len || cur_type != OBJ_BLOB) {
 148                 free(cur_msg);
 149                 free(new_msg);
 150                 hashcpy(cur_sha1, new_sha1);
 151                 return 0;
 152         }
 153
 154         /* we will separate the notes by a newline anyway */
 155         if (cur_msg[cur_len - 1] == '\n')
 156                 cur_len--;
 157
 158         /* concatenate cur_msg and new_msg into buf */
 159         buf_len = cur_len + 1 + new_len;
 160         buf = (char *) xmalloc(buf_len);
 161         memcpy(buf, cur_msg, cur_len);
 162         buf[cur_len] = '\n';
 163         memcpy(buf + cur_len + 1, new_msg, new_len);
 164
 165         free(cur_msg);
 166         free(new_msg);
 167
 168         /* create a new blob object from buf */
 169         ret = write_sha1_file(buf, buf_len, "blob", cur_sha1);
 170         free(buf);
 171         return ret;
 172 }
 173
 174 /*
 175  * To insert a leaf_node:
 176  * Search to the tree location appropriate for the given leaf_node's key:
 177  * - If location is unused (NULL), store the tweaked pointer directly there
 178  * - If location holds a note entry that matches the note-to-be-inserted, then
 179  *   concatenate the two notes.
 180  * - If location holds a note entry that matches the subtree-to-be-inserted,
 181  *   then unpack the subtree-to-be-inserted into the location.
 182  * - If location holds a matching subtree entry, unpack the subtree at that
 183  *   location, and restart the insert operation from that level.
 184  * - Else, create a new int_node, holding both the node-at-location and the
 185  *   node-to-be-inserted, and store the new int_node into the location.
 186  */
 187 static void note_tree_insert(struct int_node *tree, unsigned char n,
 188                 struct leaf_node *entry, unsigned char type)
 189 {
 190         struct int_node *new_node;
 191         struct leaf_node *l;
 192         void **p = note_tree_search(&tree, &n, entry->key_sha1);
 193
 194         assert(GET_PTR_TYPE(entry) == 0); /* no type bits set */
 195         l = (struct leaf_node *) CLR_PTR_TYPE(*p);
 196         switch (GET_PTR_TYPE(*p)) {
 197         case PTR_TYPE_NULL:
 198                 assert(!*p);
 199                 *p = SET_PTR_TYPE(entry, type);
 200                 return;
 201         case PTR_TYPE_NOTE:
 202                 switch (type) {
 203                 case PTR_TYPE_NOTE:
 204                         if (!hashcmp(l->key_sha1, entry->key_sha1)) {
 205                                 /* skip concatenation if l == entry */
 206                                 if (!hashcmp(l->val_sha1, entry->val_sha1))
 207                                         return;
 208
 209                                 if (concatenate_notes(l->val_sha1,
 210                                                 entry->val_sha1))
 211                                         die("failed to concatenate note %s "
 212                                             "into note %s for object %s",
 213                                             sha1_to_hex(entry->val_sha1),
 214                                             sha1_to_hex(l->val_sha1),
 215                                             sha1_to_hex(l->key_sha1));
 216                                 free(entry);
 217                                 return;
 218                         }
 219                         break;
 220                 case PTR_TYPE_SUBTREE:
 221                         if (!SUBTREE_SHA1_PREFIXCMP(l->key_sha1,
 222                                                     entry->key_sha1)) {
 223                                 /* unpack 'entry' */
 224                                 load_subtree(entry, tree, n);
 225                                 free(entry);
 226                                 return;
 227                         }
 228                         break;
 229                 }
 230                 break;
 231         case PTR_TYPE_SUBTREE:
 232                 if (!SUBTREE_SHA1_PREFIXCMP(entry->key_sha1, l->key_sha1)) {
 233                         /* unpack 'l' and restart insert */
 234                         *p = NULL;
 235                         load_subtree(l, tree, n);
 236                         free(l);
 237                         note_tree_insert(tree, n, entry, type);
 238                         return;
 239                 }
 240                 break;
 241         }
 242
 243         /* non-matching leaf_node */
 244         assert(GET_PTR_TYPE(*p) == PTR_TYPE_NOTE ||
 245                GET_PTR_TYPE(*p) == PTR_TYPE_SUBTREE);
 246         new_node = (struct int_node *) xcalloc(sizeof(struct int_node), 1);
 247         note_tree_insert(new_node, n + 1, l, GET_PTR_TYPE(*p));
 248         *p = SET_PTR_TYPE(new_node, PTR_TYPE_INTERNAL);
 249         note_tree_insert(new_node, n + 1, entry, type);
 250 }
 251
 252 /* Free the entire notes data contained in the given tree */
 253 static void note_tree_free(struct int_node *tree)
 254 {
 255         unsigned int i;
 256         for (i = 0; i < 16; i++) {
 257                 void *p = tree->a[i];
 258                 switch (GET_PTR_TYPE(p)) {
 259                 case PTR_TYPE_INTERNAL:
 260                         note_tree_free(CLR_PTR_TYPE(p));
 261                         /* fall through */
 262                 case PTR_TYPE_NOTE:
 263                 case PTR_TYPE_SUBTREE:
 264                         free(CLR_PTR_TYPE(p));
 265                 }
 266         }
 267 }
 268
 269 /*
 270  * Convert a partial SHA1 hex string to the corresponding partial SHA1 value.
 271  * - hex      - Partial SHA1 segment in ASCII hex format
 272  * - hex_len  - Length of above segment. Must be multiple of 2 between 0 and 40
 273  * - sha1     - Partial SHA1 value is written here
 274  * - sha1_len - Max #bytes to store in sha1, Must be >= hex_len / 2, and < 20
 275  * Returns -1 on error (invalid arguments or invalid SHA1 (not in hex format)).
 276  * Otherwise, returns number of bytes written to sha1 (i.e. hex_len / 2).
 277  * Pads sha1 with NULs up to sha1_len (not included in returned length).
 278  */
 279 static int get_sha1_hex_segment(const char *hex, unsigned int hex_len,
 280                 unsigned char *sha1, unsigned int sha1_len)
 281 {
 282         unsigned int i, len = hex_len >> 1;
 283         if (hex_len % 2 != 0 || len > sha1_len)
 284                 return -1;
 285         for (i = 0; i < len; i++) {
 286                 unsigned int val = (hexval(hex[0]) << 4) | hexval(hex[1]);
 287                 if (val & ~0xff)
 288                         return -1;
 289                 *sha1++ = val;
 290                 hex += 2;
 291         }
 292         for (; i < sha1_len; i++)
 293                 *sha1++ = 0;
 294         return len;
 295 }
 296
 297 static void load_subtree(struct leaf_node *subtree, struct int_node *node,
 298                 unsigned int n)
 299 {
 300         unsigned char object_sha1[20];
 301         unsigned int prefix_len;
 302         void *buf;
 303         struct tree_desc desc;
 304         struct name_entry entry;
 305
 306         buf = fill_tree_descriptor(&desc, subtree->val_sha1);
 307         if (!buf)
 308                 die("Could not read %s for notes-index",
 309                      sha1_to_hex(subtree->val_sha1));
 310
 311         prefix_len = subtree->key_sha1[19];
 312         assert(prefix_len * 2 >= n);
 313         memcpy(object_sha1, subtree->key_sha1, prefix_len);
 314         while (tree_entry(&desc, &entry)) {
 315                 int len = get_sha1_hex_segment(entry.path, strlen(entry.path),
 316                                 object_sha1 + prefix_len, 20 - prefix_len);
 317                 if (len < 0)
 318                         continue; /* entry.path is not a SHA1 sum. Skip */
 319                 len += prefix_len;
 320
 321                 /*
 322                  * If object SHA1 is complete (len == 20), assume note object
 323                  * If object SHA1 is incomplete (len < 20), assume note subtree
 324                  */
 325                 if (len <= 20) {
 326                         unsigned char type = PTR_TYPE_NOTE;
 327                         struct leaf_node *l = (struct leaf_node *)
 328                                 xcalloc(sizeof(struct leaf_node), 1);
 329                         hashcpy(l->key_sha1, object_sha1);
 330                         hashcpy(l->val_sha1, entry.sha1);
 331                         if (len < 20) {
 332                                 if (!S_ISDIR(entry.mode))
 333                                         continue; /* entry cannot be subtree */
 334                                 l->key_sha1[19] = (unsigned char) len;
 335                                 type = PTR_TYPE_SUBTREE;
 336                         }
 337                         note_tree_insert(node, n, l, type);
 338                 }
 339         }
 340         free(buf);
 341 }
 342
 343 static void initialize_notes(const char *notes_ref_name)
 344 {
 345         unsigned char sha1[20], object_sha1[20];
 346         unsigned mode;
 347         struct leaf_node root_tree;
 348
 349         if (!notes_ref_name || read_ref(notes_ref_name, object_sha1) ||
 350             get_tree_entry(object_sha1, "", sha1, &mode))
 351                 return;
 352
 353         hashclr(root_tree.key_sha1);
 354         hashcpy(root_tree.val_sha1, sha1);
 355         load_subtree(&root_tree, &root_node, 0);
 356 }
 357
 358 static unsigned char *lookup_notes(const unsigned char *object_sha1)
 359 {
 360         struct leaf_node *found = note_tree_find(&root_node, 0, object_sha1);
 361         if (found)
 362                 return found->val_sha1;
 363         return NULL;
 364 }
 365
 366 void free_notes(void)
 367 {
 368         note_tree_free(&root_node);
 369         memset(&root_node, 0, sizeof(struct int_node));
 370         initialized = 0;
 371 }
 372
 373 void format_note(const unsigned char *object_sha1, struct strbuf *sb,
 374                 const char *output_encoding, int flags)
 375 {
 376         static const char utf8[] = "utf-8";
 377         unsigned char *sha1;
 378         char *msg, *msg_p;
 379         unsigned long linelen, msglen;
 380         enum object_type type;
 381
 382         if (!initialized) {
 383                 const char *env = getenv(GIT_NOTES_REF_ENVIRONMENT);
 384                 if (env)
 385                         notes_ref_name = getenv(GIT_NOTES_REF_ENVIRONMENT);
 386                 else if (!notes_ref_name)
 387                         notes_ref_name = GIT_NOTES_DEFAULT_REF;
 388                 initialize_notes(notes_ref_name);
 389                 initialized = 1;
 390         }
 391
 392         sha1 = lookup_notes(object_sha1);
 393         if (!sha1)
 394                 return;
 395
 396         if (!(msg = read_sha1_file(sha1, &type, &msglen)) || !msglen ||
 397                         type != OBJ_BLOB) {
 398                 free(msg);
 399                 return;
 400         }
 401
 402         if (output_encoding && *output_encoding &&
 403                         strcmp(utf8, output_encoding)) {
 404                 char *reencoded = reencode_string(msg, output_encoding, utf8);
 405                 if (reencoded) {
 406                         free(msg);
 407                         msg = reencoded;
 408                         msglen = strlen(msg);
 409                 }
 410         }
 411
 412         /* we will end the annotation by a newline anyway */
 413         if (msglen && msg[msglen - 1] == '\n')
 414                 msglen--;
 415
 416         if (flags & NOTES_SHOW_HEADER)
 417                 strbuf_addstr(sb, "\nNotes:\n");
 418
 419         for (msg_p = msg; msg_p < msg + msglen; msg_p += linelen + 1) {
 420                 linelen = strchrnul(msg_p, '\n') - msg_p;
 421
 422                 if (flags & NOTES_INDENT)
 423                         strbuf_addstr(sb, "    ");
 424                 strbuf_add(sb, msg_p, linelen);
 425                 strbuf_addch(sb, '\n');
 426         }
 427
 428         free(msg);
 429 }