1 /* database.cc - The database interfaces of the notmuch mail library
3 * Copyright © 2009 Carl Worth
5 * This program is free software: you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation, either version 3 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License
16 * along with this program. If not, see http://www.gnu.org/licenses/ .
18 * Author: Carl Worth <cworth@cworth.org>
21 #include "database-private.h"
27 #include <glib.h> /* g_strdup_printf, g_free, GPtrArray, GHashTable */
31 /* "128 bits of thread-id ought to be enough for anybody" */
32 #define NOTMUCH_THREAD_ID_BITS 128
33 #define NOTMUCH_THREAD_ID_DIGITS (NOTMUCH_THREAD_ID_BITS / 4)
34 typedef struct _thread_id {
35 char str[NOTMUCH_THREAD_ID_DIGITS + 1];
39 thread_id_generate (thread_id_t *thread_id)
41 static int seeded = 0;
48 dev_random = fopen ("/dev/random", "r");
49 if (dev_random == NULL) {
52 fread ((void *) &value, sizeof (value), 1, dev_random);
60 for (i = 0; i < NOTMUCH_THREAD_ID_DIGITS; i += 8) {
62 sprintf (s, "%08x", value);
67 /* XXX: We should drop this function and convert all callers to call
68 * _notmuch_message_add_term instead. */
70 add_term (Xapian::Document doc,
71 const char *prefix_name,
80 prefix = _find_prefix (prefix_name);
82 term = g_strdup_printf ("%s%s", prefix, value);
84 if (strlen (term) <= NOTMUCH_TERM_MAX)
91 find_messages_by_term (Xapian::Database *db,
92 const char *prefix_name,
94 Xapian::PostingIterator *begin,
95 Xapian::PostingIterator *end)
97 Xapian::PostingIterator i;
100 term = g_strdup_printf ("%s%s", _find_prefix (prefix_name), value);
102 *begin = db->postlist_begin (term);
105 *end = db->postlist_end (term);
111 find_message_by_docid (Xapian::Database *db, Xapian::docid docid)
113 return db->get_document (docid);
117 notmuch_database_find_message (notmuch_database_t *notmuch,
118 const char *message_id)
120 Xapian::PostingIterator i, end;
122 find_messages_by_term (notmuch->xapian_db,
123 "msgid", message_id, &i, &end);
128 return _notmuch_message_create (notmuch, notmuch, *i);
131 /* Return one or more thread_ids, (as a GPtrArray of strings), for the
132 * given message based on looking into the database for any messages
133 * referenced in parents, and also for any messages in the database
134 * referencing message_id.
136 * Caller should free all strings in the array and the array itself,
137 * (g_ptr_array_free) when done. */
139 find_thread_ids (notmuch_database_t *notmuch,
141 const char *message_id)
143 Xapian::WritableDatabase *db = notmuch->xapian_db;
144 Xapian::PostingIterator child, children_end;
145 Xapian::Document doc;
146 GHashTable *thread_ids;
149 const char *parent_message_id;
152 thread_ids = g_hash_table_new_full (g_str_hash, g_str_equal,
155 find_messages_by_term (db, "ref", message_id, &child, &children_end);
156 for ( ; child != children_end; child++) {
157 const char *thread_id;
158 doc = find_message_by_docid (db, *child);
160 thread_id = doc.get_value (NOTMUCH_VALUE_THREAD).c_str ();
161 if (strlen (thread_id) == 0) {
162 fprintf (stderr, "Database error: Message with doc_id %u has empty thread-id value (value index %d)\n",
163 *child, NOTMUCH_VALUE_THREAD);
165 g_hash_table_insert (thread_ids, strdup (thread_id), NULL);
169 for (i = 0; i < parents->len; i++) {
170 notmuch_message_t *parent;
171 notmuch_thread_ids_t *ids;
173 parent_message_id = (char *) g_ptr_array_index (parents, i);
174 parent = notmuch_database_find_message (notmuch, parent_message_id);
178 for (ids = notmuch_message_get_thread_ids (parent);
179 notmuch_thread_ids_has_more (ids);
180 notmuch_thread_ids_advance (ids))
184 id = notmuch_thread_ids_get (ids);
185 g_hash_table_insert (thread_ids, strdup (id), NULL);
188 notmuch_message_destroy (parent);
191 result = g_ptr_array_new ();
193 keys = g_hash_table_get_keys (thread_ids);
194 for (l = keys; l; l = l->next) {
195 char *id = (char *) l->data;
196 g_ptr_array_add (result, id);
200 /* We're done with the hash table, but we've taken the pointers to
201 * the allocated strings and put them into our result array, so
202 * tell the hash not to free them on its way out. */
203 g_hash_table_steal_all (thread_ids);
204 g_hash_table_unref (thread_ids);
209 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
210 * a (potentially nested) parenthesized sequence with '\' used to
211 * escape any character (including parentheses).
213 * If the sequence to be skipped continues to the end of the string,
214 * then 'str' will be left pointing at the final terminating '\0'
218 skip_space_and_comments (const char **str)
223 while (*s && (isspace (*s) || *s == '(')) {
224 while (*s && isspace (*s))
229 while (*s && nesting) {
245 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
246 * comments, and the '<' and '>' delimeters.
248 * If not NULL, then *next will be made to point to the first character
249 * not parsed, (possibly pointing to the final '\0' terminator.
251 * Returns a newly allocated string which the caller should free()
254 * Returns NULL if there is any error parsing the message-id. */
256 parse_message_id (const char *message_id, const char **next)
261 if (message_id == NULL)
266 skip_space_and_comments (&s);
268 /* Skip any unstructured text as well. */
269 while (*s && *s != '<')
280 skip_space_and_comments (&s);
283 while (*end && *end != '>')
292 if (end > s && *end == '>')
297 result = strndup (s, end - s + 1);
299 /* Finally, collapse any whitespace that is within the message-id
305 for (r = result, len = strlen (r); *r; r++, len--)
306 if (*r == ' ' || *r == '\t')
307 memmove (r, r+1, len);
313 /* Parse a References header value, putting a copy of each referenced
314 * message-id into 'array'. */
316 parse_references (GPtrArray *array,
325 ref = parse_message_id (refs, &refs);
328 g_ptr_array_add (array, ref);
333 notmuch_database_default_path (void)
335 if (getenv ("NOTMUCH_BASE"))
336 return strdup (getenv ("NOTMUCH_BASE"));
338 return g_strdup_printf ("%s/mail", getenv ("HOME"));
342 notmuch_database_create (const char *path)
344 notmuch_database_t *notmuch = NULL;
345 char *notmuch_path = NULL;
348 char *local_path = NULL;
351 path = local_path = notmuch_database_default_path ();
353 err = stat (path, &st);
355 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
356 path, strerror (errno));
360 if (! S_ISDIR (st.st_mode)) {
361 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
366 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
368 err = mkdir (notmuch_path, 0755);
371 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
372 notmuch_path, strerror (errno));
376 notmuch = notmuch_database_open (path);
388 notmuch_database_open (const char *path)
390 notmuch_database_t *notmuch = NULL;
391 char *notmuch_path = NULL, *xapian_path = NULL;
394 char *local_path = NULL;
397 path = local_path = notmuch_database_default_path ();
399 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
401 err = stat (notmuch_path, &st);
403 fprintf (stderr, "Error opening database at %s: %s\n",
404 notmuch_path, strerror (errno));
408 xapian_path = g_strdup_printf ("%s/%s", notmuch_path, "xapian");
410 notmuch = talloc (NULL, notmuch_database_t);
411 notmuch->path = talloc_strdup (notmuch, path);
414 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
415 Xapian::DB_CREATE_OR_OPEN);
416 notmuch->query_parser = new Xapian::QueryParser;
417 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
418 notmuch->query_parser->set_database (*notmuch->xapian_db);
419 } catch (const Xapian::Error &error) {
420 fprintf (stderr, "A Xapian exception occurred: %s\n",
421 error.get_msg().c_str());
436 notmuch_database_close (notmuch_database_t *notmuch)
438 delete notmuch->query_parser;
439 delete notmuch->xapian_db;
440 talloc_free (notmuch);
444 notmuch_database_get_path (notmuch_database_t *notmuch)
446 return notmuch->path;
450 notmuch_database_add_message (notmuch_database_t *notmuch,
451 const char *filename)
453 Xapian::WritableDatabase *db = notmuch->xapian_db;
454 Xapian::Document doc;
455 notmuch_message_file_t *message;
457 GPtrArray *parents, *thread_ids;
459 const char *refs, *in_reply_to, *date, *header;
460 const char *from, *to, *subject;
466 message = notmuch_message_file_open (filename);
468 notmuch_message_file_restrict_headers (message,
478 doc.set_data (filename);
480 add_term (doc, "type", "mail");
482 parents = g_ptr_array_new ();
484 refs = notmuch_message_file_get_header (message, "references");
485 parse_references (parents, refs);
487 in_reply_to = notmuch_message_file_get_header (message, "in-reply-to");
488 parse_references (parents, in_reply_to);
490 for (i = 0; i < parents->len; i++)
491 add_term (doc, "ref", (char *) g_ptr_array_index (parents, i));
493 header = notmuch_message_file_get_header (message, "message-id");
495 message_id = parse_message_id (header, NULL);
496 /* So the header value isn't RFC-compliant, but it's
497 * better than no message-id at all. */
498 if (message_id == NULL)
499 message_id = xstrdup (header);
501 /* XXX: Should generate a message_id here, (such as a SHA1
502 * sum of the message itself) */
506 thread_ids = find_thread_ids (notmuch, parents, message_id);
508 for (i = 0; i < parents->len; i++)
509 g_free (g_ptr_array_index (parents, i));
510 g_ptr_array_free (parents, TRUE);
512 add_term (doc, "msgid", message_id);
513 doc.add_value (NOTMUCH_VALUE_MESSAGE_ID, message_id);
516 if (thread_ids->len) {
521 for (i = 0; i < thread_ids->len; i++) {
522 id = (char *) thread_ids->pdata[i];
523 add_term (doc, "thread", id);
525 thread_id = g_string_new (id);
527 g_string_append_printf (thread_id, ",%s", id);
531 doc.add_value (NOTMUCH_VALUE_THREAD, thread_id->str);
532 g_string_free (thread_id, TRUE);
533 } else if (message_id) {
534 /* If not part of any existing thread, generate a new thread_id. */
535 thread_id_t thread_id;
537 thread_id_generate (&thread_id);
538 add_term (doc, "thread", thread_id.str);
539 doc.add_value (NOTMUCH_VALUE_THREAD, thread_id.str);
542 g_ptr_array_free (thread_ids, TRUE);
546 date = notmuch_message_file_get_header (message, "date");
547 time_value = notmuch_parse_date (date, NULL);
549 doc.add_value (NOTMUCH_VALUE_DATE,
550 Xapian::sortable_serialise (time_value));
552 from = notmuch_message_file_get_header (message, "from");
553 subject = notmuch_message_file_get_header (message, "subject");
554 to = notmuch_message_file_get_header (message, "to");
560 notmuch_message_file_close (message);
561 return NOTMUCH_STATUS_FILE_NOT_EMAIL;
563 db->add_document (doc);
565 } catch (const Xapian::Error &error) {
566 fprintf (stderr, "A Xapian exception occurred: %s.\n",
567 error.get_msg().c_str());
568 return NOTMUCH_STATUS_XAPIAN_EXCEPTION;
571 notmuch_message_file_close (message);
573 return NOTMUCH_STATUS_SUCCESS;