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 #define ARRAY_SIZE(arr) (sizeof (arr) / sizeof (arr[0]))
38 /* Here's the current schema for our database:
40 * We currently have two different types of documents: mail and timestamps.
44 * A mail document is associated with a particular email message file
45 * on disk. It is indexed with the following prefixed terms:
47 * Single terms of given prefix:
51 * id: Unique ID of mail, (from Message-ID header or generated
52 * as "notmuch-sha1-<sha1_sum_of_entire_file>.
54 * thread: The ID of the thread to which the mail belongs
56 * Multiple terms of given prefix:
58 * ref: All unresolved message IDs from In-Reply-To and
59 * References headers in the message. (Once a referenced
60 * message is added to the database and the thread IDs
61 * are linked the corresponding "ref" term is dropped
62 * from the message document.)
64 * tag: Any tags associated with this message by the user.
66 * A mail document also has two values:
68 * TIMESTAMP: The time_t value corresponding to the message's
71 * MESSAGE_ID: The unique ID of the mail mess (see "id" above)
75 * A timestamp document is used by a client of the notmuch library to
76 * maintain data necessary to allow for efficient polling of mail
77 * directories. The notmuch library does no interpretation of
78 * timestamps, but merely allows the user to store and retrieve
79 * timestamps as name/value pairs.
81 * The timestamp document is indexed with a single prefixed term:
83 * timestamp: The user's key value (likely a directory name)
85 * and has a single value:
87 * TIMETAMPS: The time_t value from the user.
90 /* With these prefix values we follow the conventions published here:
92 * http://xapian.org/docs/omega/termprefixes.html
94 * as much as makes sense. Note that I took some liberty in matching
95 * the reserved prefix values to notmuch concepts, (for example, 'G'
96 * is documented as "newsGroup (or similar entity - e.g. a web forum
97 * name)", for which I think the thread is the closest analogue in
98 * notmuch. This in spite of the fact that we will eventually be
99 * storing mailing-list messages where 'G' for "mailing list name"
100 * might be even a closer analogue. I'm treating the single-character
101 * prefixes preferentially for core notmuch concepts (which will be
102 * nearly universal to all mail messages).
105 prefix_t BOOLEAN_PREFIX_INTERNAL[] = {
108 { "ref", "XREFERENCE" },
109 { "timestamp", "XTIMESTAMP" },
112 prefix_t BOOLEAN_PREFIX_EXTERNAL[] = {
118 _internal_error (const char *format, ...)
122 va_start (va_args, format);
124 vfprintf (stderr, format, va_args);
132 _find_prefix (const char *name)
136 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_INTERNAL); i++)
137 if (strcmp (name, BOOLEAN_PREFIX_INTERNAL[i].name) == 0)
138 return BOOLEAN_PREFIX_INTERNAL[i].prefix;
140 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++)
141 if (strcmp (name, BOOLEAN_PREFIX_EXTERNAL[i].name) == 0)
142 return BOOLEAN_PREFIX_EXTERNAL[i].prefix;
144 INTERNAL_ERROR ("No prefix exists for '%s'\n", name);
150 notmuch_status_to_string (notmuch_status_t status)
153 case NOTMUCH_STATUS_SUCCESS:
154 return "No error occurred";
155 case NOTMUCH_STATUS_OUT_OF_MEMORY:
156 return "Out of memory";
157 case NOTMUCH_STATUS_XAPIAN_EXCEPTION:
158 return "A Xapian exception occurred";
159 case NOTMUCH_STATUS_FILE_ERROR:
160 return "Something went wrong trying to read or write a file";
161 case NOTMUCH_STATUS_FILE_NOT_EMAIL:
162 return "File is not an email";
163 case NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID:
164 return "Message ID is identical to a message in database";
165 case NOTMUCH_STATUS_NULL_POINTER:
166 return "Erroneous NULL pointer";
167 case NOTMUCH_STATUS_TAG_TOO_LONG:
168 return "Tag value is too long (exceeds NOTMUCH_TAG_MAX)";
170 case NOTMUCH_STATUS_LAST_STATUS:
171 return "Unknown error status value";
175 /* XXX: We should drop this function and convert all callers to call
176 * _notmuch_message_add_term instead. */
178 add_term (Xapian::Document doc,
179 const char *prefix_name,
188 prefix = _find_prefix (prefix_name);
190 term = g_strdup_printf ("%s%s", prefix, value);
192 if (strlen (term) <= NOTMUCH_TERM_MAX)
199 find_doc_ids (notmuch_database_t *notmuch,
200 const char *prefix_name,
202 Xapian::PostingIterator *begin,
203 Xapian::PostingIterator *end)
205 Xapian::PostingIterator i;
208 term = g_strdup_printf ("%s%s", _find_prefix (prefix_name), value);
210 *begin = notmuch->xapian_db->postlist_begin (term);
212 *end = notmuch->xapian_db->postlist_end (term);
217 static notmuch_private_status_t
218 find_unique_doc_id (notmuch_database_t *notmuch,
219 const char *prefix_name,
221 unsigned int *doc_id)
223 Xapian::PostingIterator i, end;
225 find_doc_ids (notmuch, prefix_name, value, &i, &end);
229 return NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND;
232 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
236 static Xapian::Document
237 find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id)
239 return notmuch->xapian_db->get_document (doc_id);
242 static notmuch_private_status_t
243 find_unique_document (notmuch_database_t *notmuch,
244 const char *prefix_name,
246 Xapian::Document *document,
247 unsigned int *doc_id)
249 notmuch_private_status_t status;
251 status = find_unique_doc_id (notmuch, prefix_name, value, doc_id);
254 *document = Xapian::Document ();
258 *document = find_document_for_doc_id (notmuch, *doc_id);
259 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
263 notmuch_database_find_message (notmuch_database_t *notmuch,
264 const char *message_id)
266 notmuch_private_status_t status;
269 status = find_unique_doc_id (notmuch, "id", message_id, &doc_id);
271 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
274 return _notmuch_message_create (notmuch, notmuch, doc_id, NULL);
277 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
278 * a (potentially nested) parenthesized sequence with '\' used to
279 * escape any character (including parentheses).
281 * If the sequence to be skipped continues to the end of the string,
282 * then 'str' will be left pointing at the final terminating '\0'
286 skip_space_and_comments (const char **str)
291 while (*s && (isspace (*s) || *s == '(')) {
292 while (*s && isspace (*s))
297 while (*s && nesting) {
313 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
314 * comments, and the '<' and '>' delimeters.
316 * If not NULL, then *next will be made to point to the first character
317 * not parsed, (possibly pointing to the final '\0' terminator.
319 * Returns a newly allocated string which the caller should free()
322 * Returns NULL if there is any error parsing the message-id. */
324 parse_message_id (const char *message_id, const char **next)
329 if (message_id == NULL)
334 skip_space_and_comments (&s);
336 /* Skip any unstructured text as well. */
337 while (*s && *s != '<')
348 skip_space_and_comments (&s);
351 while (*end && *end != '>')
360 if (end > s && *end == '>')
365 result = strndup (s, end - s + 1);
367 /* Finally, collapse any whitespace that is within the message-id
373 for (r = result, len = strlen (r); *r; r++, len--)
374 if (*r == ' ' || *r == '\t')
375 memmove (r, r+1, len);
381 /* Parse a References header value, putting a copy of each referenced
382 * message-id into 'hash'. */
384 parse_references (GHashTable *hash,
393 ref = parse_message_id (refs, &refs);
396 g_hash_table_insert (hash, ref, NULL);
401 notmuch_database_default_path (void)
403 if (getenv ("NOTMUCH_BASE"))
404 return strdup (getenv ("NOTMUCH_BASE"));
406 return g_strdup_printf ("%s/mail", getenv ("HOME"));
410 notmuch_database_create (const char *path)
412 notmuch_database_t *notmuch = NULL;
413 char *notmuch_path = NULL;
416 char *local_path = NULL;
419 path = local_path = notmuch_database_default_path ();
421 err = stat (path, &st);
423 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
424 path, strerror (errno));
428 if (! S_ISDIR (st.st_mode)) {
429 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
434 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
436 err = mkdir (notmuch_path, 0755);
439 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
440 notmuch_path, strerror (errno));
444 notmuch = notmuch_database_open (path);
456 notmuch_database_open (const char *path)
458 notmuch_database_t *notmuch = NULL;
459 char *notmuch_path = NULL, *xapian_path = NULL;
462 char *local_path = NULL;
466 path = local_path = notmuch_database_default_path ();
468 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
470 err = stat (notmuch_path, &st);
472 fprintf (stderr, "Error opening database at %s: %s\n",
473 notmuch_path, strerror (errno));
477 xapian_path = g_strdup_printf ("%s/%s", notmuch_path, "xapian");
479 notmuch = talloc (NULL, notmuch_database_t);
480 notmuch->path = talloc_strdup (notmuch, path);
483 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
484 Xapian::DB_CREATE_OR_OPEN);
485 notmuch->query_parser = new Xapian::QueryParser;
486 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
487 notmuch->query_parser->set_database (*notmuch->xapian_db);
489 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
490 prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i];
491 notmuch->query_parser->add_boolean_prefix (prefix->name,
494 } catch (const Xapian::Error &error) {
495 fprintf (stderr, "A Xapian exception occurred: %s\n",
496 error.get_msg().c_str());
511 notmuch_database_close (notmuch_database_t *notmuch)
513 delete notmuch->query_parser;
514 delete notmuch->xapian_db;
515 talloc_free (notmuch);
519 notmuch_database_get_path (notmuch_database_t *notmuch)
521 return notmuch->path;
524 static notmuch_private_status_t
525 find_timestamp_document (notmuch_database_t *notmuch, const char *db_key,
526 Xapian::Document *doc, unsigned int *doc_id)
528 return find_unique_document (notmuch, "timestamp", db_key, doc, doc_id);
531 /* We allow the user to use arbitrarily long keys for timestamps,
532 * (they're for filesystem paths after all, which have no limit we
533 * know about). But we have a term-length limit. So if we exceed that,
534 * we'll use the SHA-1 of the user's key as the actual key for
535 * constructing a database term.
537 * Caution: This function returns a newly allocated string which the
538 * caller should free() when finished.
541 timestamp_db_key (const char *key)
543 int term_len = strlen (_find_prefix ("timestamp")) + strlen (key);
545 if (term_len > NOTMUCH_TERM_MAX)
546 return notmuch_sha1_of_string (key);
552 notmuch_database_set_timestamp (notmuch_database_t *notmuch,
553 const char *key, time_t timestamp)
555 Xapian::Document doc;
557 notmuch_private_status_t status;
558 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
561 db_key = timestamp_db_key (key);
564 status = find_timestamp_document (notmuch, db_key, &doc, &doc_id);
566 doc.add_value (NOTMUCH_VALUE_TIMESTAMP,
567 Xapian::sortable_serialise (timestamp));
569 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
570 char *term = talloc_asprintf (NULL, "%s%s",
571 _find_prefix ("timestamp"), db_key);
575 notmuch->xapian_db->add_document (doc);
577 notmuch->xapian_db->replace_document (doc_id, doc);
580 } catch (Xapian::Error &error) {
581 fprintf (stderr, "A Xapian exception occurred: %s.\n",
582 error.get_msg().c_str());
583 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
593 notmuch_database_get_timestamp (notmuch_database_t *notmuch, const char *key)
595 Xapian::Document doc;
597 notmuch_private_status_t status;
601 db_key = timestamp_db_key (key);
604 status = find_timestamp_document (notmuch, db_key, &doc, &doc_id);
606 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
609 ret = Xapian::sortable_unserialise (doc.get_value (NOTMUCH_VALUE_TIMESTAMP));
610 } catch (Xapian::Error &error) {
621 /* Find the thread ID to which the message with 'message_id' belongs.
623 * Returns NULL if no message with message ID 'message_id' is in the
626 * Otherwise, returns a newly talloced string belonging to 'ctx'.
629 _resolve_message_id_to_thread_id (notmuch_database_t *notmuch,
631 const char *message_id)
633 notmuch_message_t *message;
634 const char *ret = NULL;
636 message = notmuch_database_find_message (notmuch, message_id);
640 ret = talloc_steal (ctx, notmuch_message_get_thread_id (message));
644 notmuch_message_destroy (message);
649 static notmuch_status_t
650 _merge_threads (notmuch_database_t *notmuch,
651 const char *winner_thread_id,
652 const char *loser_thread_id)
654 Xapian::PostingIterator loser, loser_end;
655 notmuch_message_t *message = NULL;
656 notmuch_private_status_t private_status;
657 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
659 find_doc_ids (notmuch, "thread", loser_thread_id, &loser, &loser_end);
661 for ( ; loser != loser_end; loser++) {
662 message = _notmuch_message_create (notmuch, notmuch,
663 *loser, &private_status);
664 if (message == NULL) {
665 ret = COERCE_STATUS (private_status,
666 "Cannot find document for doc_id from query");
670 _notmuch_message_remove_term (message, "thread", loser_thread_id);
671 _notmuch_message_add_term (message, "thread", winner_thread_id);
672 _notmuch_message_sync (message);
674 notmuch_message_destroy (message);
680 notmuch_message_destroy (message);
685 static notmuch_status_t
686 _notmuch_database_link_message_to_parents (notmuch_database_t *notmuch,
687 notmuch_message_t *message,
688 notmuch_message_file_t *message_file,
689 const char **thread_id)
691 GHashTable *parents = NULL;
692 const char *refs, *in_reply_to;
693 GList *l, *keys = NULL;
694 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
696 parents = g_hash_table_new_full (g_str_hash, g_str_equal,
699 refs = notmuch_message_file_get_header (message_file, "references");
700 parse_references (parents, refs);
702 in_reply_to = notmuch_message_file_get_header (message_file, "in-reply-to");
703 parse_references (parents, in_reply_to);
705 keys = g_hash_table_get_keys (parents);
706 for (l = keys; l; l = l->next) {
707 char *parent_message_id;
708 const char *parent_thread_id;
710 parent_message_id = (char *) l->data;
711 parent_thread_id = _resolve_message_id_to_thread_id (notmuch,
715 if (parent_thread_id == NULL) {
716 _notmuch_message_add_term (message, "ref", parent_message_id);
718 if (*thread_id == NULL) {
719 *thread_id = talloc_strdup (message, parent_thread_id);
720 _notmuch_message_add_term (message, "thread", *thread_id);
721 } else if (strcmp (*thread_id, parent_thread_id)) {
722 ret = _merge_threads (notmuch, *thread_id, parent_thread_id);
733 g_hash_table_unref (parents);
738 static notmuch_status_t
739 _notmuch_database_link_message_to_children (notmuch_database_t *notmuch,
740 notmuch_message_t *message,
741 const char **thread_id)
743 const char *message_id = notmuch_message_get_message_id (message);
744 Xapian::PostingIterator child, children_end;
745 notmuch_message_t *child_message = NULL;
746 const char *child_thread_id;
747 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
748 notmuch_private_status_t private_status;
750 find_doc_ids (notmuch, "ref", message_id, &child, &children_end);
752 for ( ; child != children_end; child++) {
754 child_message = _notmuch_message_create (message, notmuch,
755 *child, &private_status);
756 if (child_message == NULL) {
757 ret = COERCE_STATUS (private_status,
758 "Cannot find document for doc_id from query");
762 child_thread_id = notmuch_message_get_thread_id (child_message);
763 if (*thread_id == NULL) {
764 *thread_id = talloc_strdup (message, child_thread_id);
765 _notmuch_message_add_term (message, "thread", *thread_id);
766 } else if (strcmp (*thread_id, child_thread_id)) {
767 _notmuch_message_remove_term (child_message, "ref",
769 _notmuch_message_sync (child_message);
770 ret = _merge_threads (notmuch, *thread_id, child_thread_id);
775 notmuch_message_destroy (child_message);
776 child_message = NULL;
781 notmuch_message_destroy (child_message);
786 /* Given a (mostly empty) 'message' and its corresponding
787 * 'message_file' link it to existing threads in the database.
789 * We first looke at 'message_file' and its link-relevant headers
790 * (References and In-Reply-To) for message IDs. We also look in the
791 * database for existing message that reference 'message'.p
793 * The end result is to call _notmuch_message_add_thread_id with one
794 * or more thread IDs to which this message belongs, (including
795 * generating a new thread ID if necessary if the message doesn't
796 * connect to any existing threads).
798 static notmuch_status_t
799 _notmuch_database_link_message (notmuch_database_t *notmuch,
800 notmuch_message_t *message,
801 notmuch_message_file_t *message_file)
803 notmuch_status_t status;
804 const char *thread_id = NULL;
806 status = _notmuch_database_link_message_to_parents (notmuch, message,
812 status = _notmuch_database_link_message_to_children (notmuch, message,
817 if (thread_id == NULL)
818 _notmuch_message_ensure_thread_id (message);
820 return NOTMUCH_STATUS_SUCCESS;
824 notmuch_database_add_message (notmuch_database_t *notmuch,
825 const char *filename)
827 notmuch_message_file_t *message_file;
828 notmuch_message_t *message;
829 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
831 const char *date, *header;
832 const char *from, *to, *subject, *old_filename;
835 message_file = notmuch_message_file_open (filename);
836 if (message_file == NULL) {
837 ret = NOTMUCH_STATUS_FILE_ERROR;
841 notmuch_message_file_restrict_headers (message_file,
852 /* The first order of business is to find/create a message ID. */
854 header = notmuch_message_file_get_header (message_file, "message-id");
856 message_id = parse_message_id (header, NULL);
857 /* So the header value isn't RFC-compliant, but it's
858 * better than no message-id at all. */
859 if (message_id == NULL)
860 message_id = xstrdup (header);
862 /* No message-id at all, let's generate one by taking a
863 * hash over the file's contents. */
864 char *sha1 = notmuch_sha1_of_file (filename);
866 /* If that failed too, something is really wrong. Give up. */
868 ret = NOTMUCH_STATUS_FILE_ERROR;
872 message_id = g_strdup_printf ("notmuch-sha1-%s", sha1);
876 /* Now that we have a message ID, we get a message object,
877 * (which may or may not reference an existing document in the
880 /* Use NULL for owner since we want to free this locally. */
881 message = _notmuch_message_create_for_message_id (NULL,
890 /* Has a message previously been added with the same ID? */
891 old_filename = notmuch_message_get_filename (message);
892 if (old_filename && strlen (old_filename)) {
893 ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
896 _notmuch_message_set_filename (message, filename);
897 _notmuch_message_add_term (message, "type", "mail");
900 ret = _notmuch_database_link_message (notmuch, message, message_file);
904 date = notmuch_message_file_get_header (message_file, "date");
905 _notmuch_message_set_date (message, date);
907 from = notmuch_message_file_get_header (message_file, "from");
908 subject = notmuch_message_file_get_header (message_file, "subject");
909 to = notmuch_message_file_get_header (message_file, "to");
915 ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
918 _notmuch_message_sync (message);
920 } catch (const Xapian::Error &error) {
921 fprintf (stderr, "A Xapian exception occurred: %s.\n",
922 error.get_msg().c_str());
923 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
929 notmuch_message_destroy (message);
931 notmuch_message_file_close (message_file);