notes.c

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