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"
29 #include <glib.h> /* g_free, GPtrArray, GHashTable */
33 #define ARRAY_SIZE(arr) (sizeof (arr) / sizeof (arr[0]))
40 #define NOTMUCH_DATABASE_VERSION 1
42 #define STRINGIFY(s) _SUB_STRINGIFY(s)
43 #define _SUB_STRINGIFY(s) #s
45 /* Here's the current schema for our database (for NOTMUCH_DATABASE_VERSION):
47 * We currently have two different types of documents: mail and directory.
51 * A mail document is associated with a particular email message file
52 * on disk. It is indexed with the following prefixed terms which the
53 * database uses to construct threads, etc.:
55 * Single terms of given prefix:
59 * id: Unique ID of mail, (from Message-ID header or generated
60 * as "notmuch-sha1-<sha1_sum_of_entire_file>.
62 * thread: The ID of the thread to which the mail belongs
64 * replyto: The ID from the In-Reply-To header of the mail (if any).
66 * Multiple terms of given prefix:
68 * reference: All message IDs from In-Reply-To and Re ferences
69 * headers in the message.
71 * tag: Any tags associated with this message by the user.
73 * file-direntry: A colon-separated pair of values
74 * (INTEGER:STRING), where INTEGER is the
75 * document ID of a directory document, and
76 * STRING is the name of a file within that
77 * directory for this mail message.
79 * A mail document also has two values:
81 * TIMESTAMP: The time_t value corresponding to the message's
84 * MESSAGE_ID: The unique ID of the mail mess (see "id" above)
86 * In addition, terms from the content of the message are added with
87 * "from", "to", "attachment", and "subject" prefixes for use by the
88 * user in searching. But the database doesn't really care itself
91 * The data portion of a mail document is empty.
95 * A directory document is used by a client of the notmuch library to
96 * maintain data necessary to allow for efficient polling of mail
99 * All directory documents contain one term:
101 * directory: The directory path (relative to the database path)
102 * Or the SHA1 sum of the directory path (if the
103 * path itself is too long to fit in a Xapian
106 * And all directory documents for directories other than top-level
107 * directories also contain the following term:
109 * directory-direntry: A colon-separated pair of values
110 * (INTEGER:STRING), where INTEGER is the
111 * document ID of the parent directory
112 * document, and STRING is the name of this
113 * directory within that parent.
115 * All directory documents have a single value:
117 * TIMESTAMP: The mtime of the directory (at last scan)
119 * The data portion of a directory document contains the path of the
120 * directory (relative to the database path).
123 /* With these prefix values we follow the conventions published here:
125 * http://xapian.org/docs/omega/termprefixes.html
127 * as much as makes sense. Note that I took some liberty in matching
128 * the reserved prefix values to notmuch concepts, (for example, 'G'
129 * is documented as "newsGroup (or similar entity - e.g. a web forum
130 * name)", for which I think the thread is the closest analogue in
131 * notmuch. This in spite of the fact that we will eventually be
132 * storing mailing-list messages where 'G' for "mailing list name"
133 * might be even a closer analogue. I'm treating the single-character
134 * prefixes preferentially for core notmuch concepts (which will be
135 * nearly universal to all mail messages).
138 prefix_t BOOLEAN_PREFIX_INTERNAL[] = {
140 { "reference", "XREFERENCE" },
141 { "replyto", "XREPLYTO" },
142 { "directory", "XDIRECTORY" },
143 { "file-direntry", "XFDIRENTRY" },
144 { "directory-direntry", "XDDIRENTRY" },
147 prefix_t BOOLEAN_PREFIX_EXTERNAL[] = {
153 prefix_t PROBABILISTIC_PREFIX[]= {
156 { "attachment", "XATTACHMENT" },
157 { "subject", "XSUBJECT"}
161 _internal_error (const char *format, ...)
165 va_start (va_args, format);
167 fprintf (stderr, "Internal error: ");
168 vfprintf (stderr, format, va_args);
176 _find_prefix (const char *name)
180 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_INTERNAL); i++) {
181 if (strcmp (name, BOOLEAN_PREFIX_INTERNAL[i].name) == 0)
182 return BOOLEAN_PREFIX_INTERNAL[i].prefix;
185 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
186 if (strcmp (name, BOOLEAN_PREFIX_EXTERNAL[i].name) == 0)
187 return BOOLEAN_PREFIX_EXTERNAL[i].prefix;
190 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
191 if (strcmp (name, PROBABILISTIC_PREFIX[i].name) == 0)
192 return PROBABILISTIC_PREFIX[i].prefix;
195 INTERNAL_ERROR ("No prefix exists for '%s'\n", name);
201 notmuch_status_to_string (notmuch_status_t status)
204 case NOTMUCH_STATUS_SUCCESS:
205 return "No error occurred";
206 case NOTMUCH_STATUS_OUT_OF_MEMORY:
207 return "Out of memory";
208 case NOTMUCH_STATUS_READ_ONLY_DATABASE:
209 return "Attempt to write to a read-only database";
210 case NOTMUCH_STATUS_XAPIAN_EXCEPTION:
211 return "A Xapian exception occurred";
212 case NOTMUCH_STATUS_FILE_ERROR:
213 return "Something went wrong trying to read or write a file";
214 case NOTMUCH_STATUS_FILE_NOT_EMAIL:
215 return "File is not an email";
216 case NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID:
217 return "Message ID is identical to a message in database";
218 case NOTMUCH_STATUS_NULL_POINTER:
219 return "Erroneous NULL pointer";
220 case NOTMUCH_STATUS_TAG_TOO_LONG:
221 return "Tag value is too long (exceeds NOTMUCH_TAG_MAX)";
222 case NOTMUCH_STATUS_UNBALANCED_FREEZE_THAW:
223 return "Unbalanced number of calls to notmuch_message_freeze/thaw";
225 case NOTMUCH_STATUS_LAST_STATUS:
226 return "Unknown error status value";
231 find_doc_ids_for_term (notmuch_database_t *notmuch,
233 Xapian::PostingIterator *begin,
234 Xapian::PostingIterator *end)
236 *begin = notmuch->xapian_db->postlist_begin (term);
238 *end = notmuch->xapian_db->postlist_end (term);
242 find_doc_ids (notmuch_database_t *notmuch,
243 const char *prefix_name,
245 Xapian::PostingIterator *begin,
246 Xapian::PostingIterator *end)
250 term = talloc_asprintf (notmuch, "%s%s",
251 _find_prefix (prefix_name), value);
253 find_doc_ids_for_term (notmuch, term, begin, end);
258 notmuch_private_status_t
259 _notmuch_database_find_unique_doc_id (notmuch_database_t *notmuch,
260 const char *prefix_name,
262 unsigned int *doc_id)
264 Xapian::PostingIterator i, end;
266 find_doc_ids (notmuch, prefix_name, value, &i, &end);
270 return NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND;
275 #if DEBUG_DATABASE_SANITY
279 INTERNAL_ERROR ("Term %s:%s is not unique as expected.\n",
283 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
286 static Xapian::Document
287 find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id)
289 return notmuch->xapian_db->get_document (doc_id);
293 notmuch_database_find_message (notmuch_database_t *notmuch,
294 const char *message_id)
296 notmuch_private_status_t status;
299 status = _notmuch_database_find_unique_doc_id (notmuch, "id",
300 message_id, &doc_id);
302 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
305 return _notmuch_message_create (notmuch, notmuch, doc_id, NULL);
308 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
309 * a (potentially nested) parenthesized sequence with '\' used to
310 * escape any character (including parentheses).
312 * If the sequence to be skipped continues to the end of the string,
313 * then 'str' will be left pointing at the final terminating '\0'
317 skip_space_and_comments (const char **str)
322 while (*s && (isspace (*s) || *s == '(')) {
323 while (*s && isspace (*s))
328 while (*s && nesting) {
331 } else if (*s == ')') {
333 } else if (*s == '\\') {
345 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
346 * comments, and the '<' and '>' delimeters.
348 * If not NULL, then *next will be made to point to the first character
349 * not parsed, (possibly pointing to the final '\0' terminator.
351 * Returns a newly talloc'ed string belonging to 'ctx'.
353 * Returns NULL if there is any error parsing the message-id. */
355 _parse_message_id (void *ctx, const char *message_id, const char **next)
360 if (message_id == NULL || *message_id == '\0')
365 skip_space_and_comments (&s);
367 /* Skip any unstructured text as well. */
368 while (*s && *s != '<')
379 skip_space_and_comments (&s);
382 while (*end && *end != '>')
391 if (end > s && *end == '>')
396 result = talloc_strndup (ctx, s, end - s + 1);
398 /* Finally, collapse any whitespace that is within the message-id
404 for (r = result, len = strlen (r); *r; r++, len--)
405 if (*r == ' ' || *r == '\t')
406 memmove (r, r+1, len);
412 /* Parse a References header value, putting a (talloc'ed under 'ctx')
413 * copy of each referenced message-id into 'hash'.
415 * We explicitly avoid including any reference identical to
416 * 'message_id' in the result (to avoid mass confusion when a single
417 * message references itself cyclically---and yes, mail messages are
418 * not infrequent in the wild that do this---don't ask me why).
421 parse_references (void *ctx,
422 const char *message_id,
428 if (refs == NULL || *refs == '\0')
432 ref = _parse_message_id (ctx, refs, &refs);
434 if (ref && strcmp (ref, message_id))
435 g_hash_table_insert (hash, ref, NULL);
440 notmuch_database_create (const char *path)
442 notmuch_database_t *notmuch = NULL;
443 char *notmuch_path = NULL;
448 fprintf (stderr, "Error: Cannot create a database for a NULL path.\n");
452 err = stat (path, &st);
454 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
455 path, strerror (errno));
459 if (! S_ISDIR (st.st_mode)) {
460 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
465 notmuch_path = talloc_asprintf (NULL, "%s/%s", path, ".notmuch");
467 err = mkdir (notmuch_path, 0755);
470 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
471 notmuch_path, strerror (errno));
475 notmuch = notmuch_database_open (path,
476 NOTMUCH_DATABASE_MODE_READ_WRITE);
477 notmuch_database_upgrade (notmuch, NULL, NULL);
481 talloc_free (notmuch_path);
487 _notmuch_database_ensure_writable (notmuch_database_t *notmuch)
489 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY) {
490 fprintf (stderr, "Cannot write to a read-only database.\n");
491 return NOTMUCH_STATUS_READ_ONLY_DATABASE;
494 return NOTMUCH_STATUS_SUCCESS;
498 notmuch_database_open (const char *path,
499 notmuch_database_mode_t mode)
501 notmuch_database_t *notmuch = NULL;
502 char *notmuch_path = NULL, *xapian_path = NULL;
505 unsigned int i, version;
507 if (asprintf (¬much_path, "%s/%s", path, ".notmuch") == -1) {
509 fprintf (stderr, "Out of memory\n");
513 err = stat (notmuch_path, &st);
515 fprintf (stderr, "Error opening database at %s: %s\n",
516 notmuch_path, strerror (errno));
520 if (asprintf (&xapian_path, "%s/%s", notmuch_path, "xapian") == -1) {
522 fprintf (stderr, "Out of memory\n");
526 notmuch = talloc (NULL, notmuch_database_t);
527 notmuch->exception_reported = FALSE;
528 notmuch->path = talloc_strdup (notmuch, path);
530 if (notmuch->path[strlen (notmuch->path) - 1] == '/')
531 notmuch->path[strlen (notmuch->path) - 1] = '\0';
533 notmuch->needs_upgrade = FALSE;
534 notmuch->mode = mode;
536 if (mode == NOTMUCH_DATABASE_MODE_READ_WRITE) {
537 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
538 Xapian::DB_CREATE_OR_OPEN);
539 version = notmuch_database_get_version (notmuch);
541 if (version > NOTMUCH_DATABASE_VERSION) {
543 "Error: Notmuch database at %s\n"
544 " has a newer database format version (%u) than supported by this\n"
545 " version of notmuch (%u). Refusing to open this database in\n"
546 " read-write mode.\n",
547 notmuch_path, version, NOTMUCH_DATABASE_VERSION);
548 notmuch->mode = NOTMUCH_DATABASE_MODE_READ_ONLY;
549 notmuch_database_close (notmuch);
554 if (version < NOTMUCH_DATABASE_VERSION)
555 notmuch->needs_upgrade = TRUE;
557 notmuch->xapian_db = new Xapian::Database (xapian_path);
558 version = notmuch_database_get_version (notmuch);
559 if (version > NOTMUCH_DATABASE_VERSION)
562 "Warning: Notmuch database at %s\n"
563 " has a newer database format version (%u) than supported by this\n"
564 " version of notmuch (%u). Some operations may behave incorrectly,\n"
565 " (but the database will not be harmed since it is being opened\n"
566 " in read-only mode).\n",
567 notmuch_path, version, NOTMUCH_DATABASE_VERSION);
570 notmuch->query_parser = new Xapian::QueryParser;
571 notmuch->term_gen = new Xapian::TermGenerator;
572 notmuch->term_gen->set_stemmer (Xapian::Stem ("english"));
573 notmuch->value_range_processor = new Xapian::NumberValueRangeProcessor (NOTMUCH_VALUE_TIMESTAMP);
575 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
576 notmuch->query_parser->set_database (*notmuch->xapian_db);
577 notmuch->query_parser->set_stemmer (Xapian::Stem ("english"));
578 notmuch->query_parser->set_stemming_strategy (Xapian::QueryParser::STEM_SOME);
579 notmuch->query_parser->add_valuerangeprocessor (notmuch->value_range_processor);
581 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
582 prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i];
583 notmuch->query_parser->add_boolean_prefix (prefix->name,
587 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
588 prefix_t *prefix = &PROBABILISTIC_PREFIX[i];
589 notmuch->query_parser->add_prefix (prefix->name, prefix->prefix);
591 } catch (const Xapian::Error &error) {
592 fprintf (stderr, "A Xapian exception occurred opening database: %s\n",
593 error.get_msg().c_str());
607 notmuch_database_close (notmuch_database_t *notmuch)
610 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_WRITE)
611 (static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db))->flush ();
612 } catch (const Xapian::Error &error) {
613 if (! notmuch->exception_reported) {
614 fprintf (stderr, "Error: A Xapian exception occurred flushing database: %s\n",
615 error.get_msg().c_str());
619 delete notmuch->term_gen;
620 delete notmuch->query_parser;
621 delete notmuch->xapian_db;
622 delete notmuch->value_range_processor;
623 talloc_free (notmuch);
627 notmuch_database_get_path (notmuch_database_t *notmuch)
629 return notmuch->path;
633 notmuch_database_get_version (notmuch_database_t *notmuch)
635 unsigned int version;
636 string version_string;
640 version_string = notmuch->xapian_db->get_metadata ("version");
641 if (version_string.empty ())
644 str = version_string.c_str ();
645 if (str == NULL || *str == '\0')
648 version = strtoul (str, &end, 10);
650 INTERNAL_ERROR ("Malformed database version: %s", str);
656 notmuch_database_needs_upgrade (notmuch_database_t *notmuch)
658 return notmuch->needs_upgrade;
661 static volatile sig_atomic_t do_progress_notify = 0;
664 handle_sigalrm (unused (int signal))
666 do_progress_notify = 1;
669 /* Upgrade the current database.
671 * After opening a database in read-write mode, the client should
672 * check if an upgrade is needed (notmuch_database_needs_upgrade) and
673 * if so, upgrade with this function before making any modifications.
675 * The optional progress_notify callback can be used by the caller to
676 * provide progress indication to the user. If non-NULL it will be
677 * called periodically with 'count' as the number of messages upgraded
678 * so far and 'total' the overall number of messages that will be
682 notmuch_database_upgrade (notmuch_database_t *notmuch,
683 void (*progress_notify) (void *closure,
688 Xapian::WritableDatabase *db;
689 struct sigaction action;
690 struct itimerval timerval;
691 notmuch_bool_t timer_is_active = FALSE;
692 unsigned int version;
693 notmuch_status_t status;
695 status = _notmuch_database_ensure_writable (notmuch);
699 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
701 version = notmuch_database_get_version (notmuch);
703 if (version >= NOTMUCH_DATABASE_VERSION)
704 return NOTMUCH_STATUS_SUCCESS;
706 if (progress_notify) {
707 /* Setup our handler for SIGALRM */
708 memset (&action, 0, sizeof (struct sigaction));
709 action.sa_handler = handle_sigalrm;
710 sigemptyset (&action.sa_mask);
711 action.sa_flags = SA_RESTART;
712 sigaction (SIGALRM, &action, NULL);
714 /* Then start a timer to send SIGALRM once per second. */
715 timerval.it_interval.tv_sec = 1;
716 timerval.it_interval.tv_usec = 0;
717 timerval.it_value.tv_sec = 1;
718 timerval.it_value.tv_usec = 0;
719 setitimer (ITIMER_REAL, &timerval, NULL);
721 timer_is_active = TRUE;
724 /* Before version 1, each message document had its filename in the
725 * data field. Copy that into the new format by calling
726 * notmuch_message_add_filename.
729 unsigned int count = 0, total;
730 notmuch_query_t *query = notmuch_query_create (notmuch, "");
731 notmuch_messages_t *messages;
732 notmuch_message_t *message;
735 total = notmuch_query_count_messages (query);
737 for (messages = notmuch_query_search_messages (query);
738 notmuch_messages_has_more (messages);
739 notmuch_messages_advance (messages))
741 if (do_progress_notify) {
742 progress_notify (closure, count, total);
743 do_progress_notify = 0;
746 message = notmuch_messages_get (messages);
748 filename = _notmuch_message_talloc_copy_data (message);
749 if (filename && *filename != '\0') {
750 _notmuch_message_add_filename (message, filename);
751 _notmuch_message_sync (message);
753 talloc_free (filename);
755 notmuch_message_destroy (message);
760 notmuch_query_destroy (query);
763 /* Also, before version 1 we stored directory timestamps in
764 * XTIMESTAMP documents instead of the current XDIRECTORY
765 * documents. So copy those as well. */
767 Xapian::TermIterator t, t_end;
769 t_end = notmuch->xapian_db->allterms_end ("XTIMESTAMP");
771 for (t = notmuch->xapian_db->allterms_begin ("XTIMESTAMP");
775 Xapian::PostingIterator p, p_end;
776 std::string term = *t;
778 p_end = notmuch->xapian_db->postlist_end (term);
780 for (p = notmuch->xapian_db->postlist_begin (term);
784 Xapian::Document document;
786 notmuch_directory_t *directory;
788 document = find_document_for_doc_id (notmuch, *p);
789 mtime = Xapian::sortable_unserialise (
790 document.get_value (NOTMUCH_VALUE_TIMESTAMP));
792 directory = notmuch_database_get_directory (notmuch,
794 notmuch_directory_set_mtime (directory, mtime);
795 notmuch_directory_destroy (directory);
800 db->set_metadata ("version", STRINGIFY (NOTMUCH_DATABASE_VERSION));
803 /* Now that the upgrade is complete we can remove the old data
804 * and documents that are no longer needed. */
806 unsigned int count = 0, total;
807 notmuch_query_t *query = notmuch_query_create (notmuch, "");
808 notmuch_messages_t *messages;
809 notmuch_message_t *message;
812 total = notmuch_query_count_messages (query);
814 for (messages = notmuch_query_search_messages (query);
815 notmuch_messages_has_more (messages);
816 notmuch_messages_advance (messages))
818 if (do_progress_notify) {
819 progress_notify (closure, count, total);
820 do_progress_notify = 0;
823 message = notmuch_messages_get (messages);
825 filename = _notmuch_message_talloc_copy_data (message);
826 if (filename && *filename != '\0') {
827 _notmuch_message_clear_data (message);
828 _notmuch_message_sync (message);
830 talloc_free (filename);
832 notmuch_message_destroy (message);
837 notmuch_query_destroy (query);
841 Xapian::TermIterator t, t_end;
843 t_end = notmuch->xapian_db->allterms_end ("XTIMESTAMP");
845 for (t = notmuch->xapian_db->allterms_begin ("XTIMESTAMP");
849 Xapian::PostingIterator p, p_end;
850 std::string term = *t;
852 p_end = notmuch->xapian_db->postlist_end (term);
854 for (p = notmuch->xapian_db->postlist_begin (term);
858 db->delete_document (*p);
863 if (timer_is_active) {
864 /* Now stop the timer. */
865 timerval.it_interval.tv_sec = 0;
866 timerval.it_interval.tv_usec = 0;
867 timerval.it_value.tv_sec = 0;
868 timerval.it_value.tv_usec = 0;
869 setitimer (ITIMER_REAL, &timerval, NULL);
871 /* And disable the signal handler. */
872 action.sa_handler = SIG_IGN;
873 sigaction (SIGALRM, &action, NULL);
876 return NOTMUCH_STATUS_SUCCESS;
879 /* We allow the user to use arbitrarily long paths for directories. But
880 * we have a term-length limit. So if we exceed that, we'll use the
881 * SHA-1 of the path for the database term.
883 * Note: This function may return the original value of 'path'. If it
884 * does not, then the caller is responsible to free() the returned
888 _notmuch_database_get_directory_db_path (const char *path)
890 int term_len = strlen (_find_prefix ("directory")) + strlen (path);
892 if (term_len > NOTMUCH_TERM_MAX)
893 return notmuch_sha1_of_string (path);
898 /* Given a path, split it into two parts: the directory part is all
899 * components except for the last, and the basename is that last
900 * component. Getting the return-value for either part is optional
901 * (the caller can pass NULL).
903 * The original 'path' can represent either a regular file or a
904 * directory---the splitting will be carried out in the same way in
905 * either case. Trailing slashes on 'path' will be ignored, and any
906 * cases of multiple '/' characters appearing in series will be
907 * treated as a single '/'.
909 * Allocation (if any) will have 'ctx' as the talloc owner. But
910 * pointers will be returned within the original path string whenever
913 * Note: If 'path' is non-empty and contains no non-trailing slash,
914 * (that is, consists of a filename with no parent directory), then
915 * the directory returned will be an empty string. However, if 'path'
916 * is an empty string, then both directory and basename will be
920 _notmuch_database_split_path (void *ctx,
922 const char **directory,
923 const char **basename)
927 if (path == NULL || *path == '\0') {
932 return NOTMUCH_STATUS_SUCCESS;
935 /* Find the last slash (not counting a trailing slash), if any. */
937 slash = path + strlen (path) - 1;
939 /* First, skip trailing slashes. */
940 while (slash != path) {
947 /* Then, find a slash. */
948 while (slash != path) {
958 /* Finally, skip multiple slashes. */
959 while (slash != path) {
968 *directory = talloc_strdup (ctx, "");
973 *directory = talloc_strndup (ctx, path, slash - path + 1);
976 return NOTMUCH_STATUS_SUCCESS;
980 _notmuch_database_find_directory_id (notmuch_database_t *notmuch,
982 unsigned int *directory_id)
984 notmuch_directory_t *directory;
985 notmuch_status_t status;
989 return NOTMUCH_STATUS_SUCCESS;
992 directory = _notmuch_directory_create (notmuch, path, &status);
998 *directory_id = _notmuch_directory_get_document_id (directory);
1000 notmuch_directory_destroy (directory);
1002 return NOTMUCH_STATUS_SUCCESS;
1006 _notmuch_database_get_directory_path (void *ctx,
1007 notmuch_database_t *notmuch,
1008 unsigned int doc_id)
1010 Xapian::Document document;
1012 document = find_document_for_doc_id (notmuch, doc_id);
1014 return talloc_strdup (ctx, document.get_data ().c_str ());
1017 /* Given a legal 'filename' for the database, (either relative to
1018 * database path or absolute with initial components identical to
1019 * database path), return a new string (with 'ctx' as the talloc
1020 * owner) suitable for use as a direntry term value.
1022 * The necessary directory documents will be created in the database
1026 _notmuch_database_filename_to_direntry (void *ctx,
1027 notmuch_database_t *notmuch,
1028 const char *filename,
1031 const char *relative, *directory, *basename;
1032 Xapian::docid directory_id;
1033 notmuch_status_t status;
1035 relative = _notmuch_database_relative_path (notmuch, filename);
1037 status = _notmuch_database_split_path (ctx, relative,
1038 &directory, &basename);
1042 status = _notmuch_database_find_directory_id (notmuch, directory,
1047 *direntry = talloc_asprintf (ctx, "%u:%s", directory_id, basename);
1049 return NOTMUCH_STATUS_SUCCESS;
1052 /* Given a legal 'path' for the database, return the relative path.
1054 * The return value will be a pointer to the originl path contents,
1055 * and will be either the original string (if 'path' was relative) or
1056 * a portion of the string (if path was absolute and begins with the
1060 _notmuch_database_relative_path (notmuch_database_t *notmuch,
1063 const char *db_path, *relative;
1064 unsigned int db_path_len;
1066 db_path = notmuch_database_get_path (notmuch);
1067 db_path_len = strlen (db_path);
1071 if (*relative == '/') {
1072 while (*relative == '/' && *(relative+1) == '/')
1075 if (strncmp (relative, db_path, db_path_len) == 0)
1077 relative += db_path_len;
1078 while (*relative == '/')
1086 notmuch_directory_t *
1087 notmuch_database_get_directory (notmuch_database_t *notmuch,
1090 notmuch_status_t status;
1092 return _notmuch_directory_create (notmuch, path, &status);
1095 /* Find the thread ID to which the message with 'message_id' belongs.
1097 * Returns NULL if no message with message ID 'message_id' is in the
1100 * Otherwise, returns a newly talloced string belonging to 'ctx'.
1103 _resolve_message_id_to_thread_id (notmuch_database_t *notmuch,
1105 const char *message_id)
1107 notmuch_message_t *message;
1108 const char *ret = NULL;
1110 message = notmuch_database_find_message (notmuch, message_id);
1111 if (message == NULL)
1114 ret = talloc_steal (ctx, notmuch_message_get_thread_id (message));
1118 notmuch_message_destroy (message);
1123 static notmuch_status_t
1124 _merge_threads (notmuch_database_t *notmuch,
1125 const char *winner_thread_id,
1126 const char *loser_thread_id)
1128 Xapian::PostingIterator loser, loser_end;
1129 notmuch_message_t *message = NULL;
1130 notmuch_private_status_t private_status;
1131 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1133 find_doc_ids (notmuch, "thread", loser_thread_id, &loser, &loser_end);
1135 for ( ; loser != loser_end; loser++) {
1136 message = _notmuch_message_create (notmuch, notmuch,
1137 *loser, &private_status);
1138 if (message == NULL) {
1139 ret = COERCE_STATUS (private_status,
1140 "Cannot find document for doc_id from query");
1144 _notmuch_message_remove_term (message, "thread", loser_thread_id);
1145 _notmuch_message_add_term (message, "thread", winner_thread_id);
1146 _notmuch_message_sync (message);
1148 notmuch_message_destroy (message);
1154 notmuch_message_destroy (message);
1160 _my_talloc_free_for_g_hash (void *ptr)
1165 static notmuch_status_t
1166 _notmuch_database_link_message_to_parents (notmuch_database_t *notmuch,
1167 notmuch_message_t *message,
1168 notmuch_message_file_t *message_file,
1169 const char **thread_id)
1171 GHashTable *parents = NULL;
1172 const char *refs, *in_reply_to, *in_reply_to_message_id;
1173 GList *l, *keys = NULL;
1174 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1176 parents = g_hash_table_new_full (g_str_hash, g_str_equal,
1177 _my_talloc_free_for_g_hash, NULL);
1179 refs = notmuch_message_file_get_header (message_file, "references");
1180 parse_references (message, notmuch_message_get_message_id (message),
1183 in_reply_to = notmuch_message_file_get_header (message_file, "in-reply-to");
1184 parse_references (message, notmuch_message_get_message_id (message),
1185 parents, in_reply_to);
1187 /* Carefully avoid adding any self-referential in-reply-to term. */
1188 in_reply_to_message_id = _parse_message_id (message, in_reply_to, NULL);
1189 if (in_reply_to_message_id &&
1190 strcmp (in_reply_to_message_id,
1191 notmuch_message_get_message_id (message)))
1193 _notmuch_message_add_term (message, "replyto",
1194 _parse_message_id (message, in_reply_to, NULL));
1197 keys = g_hash_table_get_keys (parents);
1198 for (l = keys; l; l = l->next) {
1199 char *parent_message_id;
1200 const char *parent_thread_id;
1202 parent_message_id = (char *) l->data;
1203 parent_thread_id = _resolve_message_id_to_thread_id (notmuch,
1207 if (parent_thread_id == NULL) {
1208 _notmuch_message_add_term (message, "reference",
1211 if (*thread_id == NULL) {
1212 *thread_id = talloc_strdup (message, parent_thread_id);
1213 _notmuch_message_add_term (message, "thread", *thread_id);
1214 } else if (strcmp (*thread_id, parent_thread_id)) {
1215 ret = _merge_threads (notmuch, *thread_id, parent_thread_id);
1226 g_hash_table_unref (parents);
1231 static notmuch_status_t
1232 _notmuch_database_link_message_to_children (notmuch_database_t *notmuch,
1233 notmuch_message_t *message,
1234 const char **thread_id)
1236 const char *message_id = notmuch_message_get_message_id (message);
1237 Xapian::PostingIterator child, children_end;
1238 notmuch_message_t *child_message = NULL;
1239 const char *child_thread_id;
1240 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1241 notmuch_private_status_t private_status;
1243 find_doc_ids (notmuch, "reference", message_id, &child, &children_end);
1245 for ( ; child != children_end; child++) {
1247 child_message = _notmuch_message_create (message, notmuch,
1248 *child, &private_status);
1249 if (child_message == NULL) {
1250 ret = COERCE_STATUS (private_status,
1251 "Cannot find document for doc_id from query");
1255 child_thread_id = notmuch_message_get_thread_id (child_message);
1256 if (*thread_id == NULL) {
1257 *thread_id = talloc_strdup (message, child_thread_id);
1258 _notmuch_message_add_term (message, "thread", *thread_id);
1259 } else if (strcmp (*thread_id, child_thread_id)) {
1260 _notmuch_message_remove_term (child_message, "reference",
1262 _notmuch_message_sync (child_message);
1263 ret = _merge_threads (notmuch, *thread_id, child_thread_id);
1268 notmuch_message_destroy (child_message);
1269 child_message = NULL;
1274 notmuch_message_destroy (child_message);
1279 /* Given a (mostly empty) 'message' and its corresponding
1280 * 'message_file' link it to existing threads in the database.
1282 * We first look at 'message_file' and its link-relevant headers
1283 * (References and In-Reply-To) for message IDs. We also look in the
1284 * database for existing message that reference 'message'.
1286 * The end result is to call _notmuch_message_ensure_thread_id which
1287 * generates a new thread ID if the message doesn't connect to any
1290 static notmuch_status_t
1291 _notmuch_database_link_message (notmuch_database_t *notmuch,
1292 notmuch_message_t *message,
1293 notmuch_message_file_t *message_file)
1295 notmuch_status_t status;
1296 const char *thread_id = NULL;
1298 status = _notmuch_database_link_message_to_parents (notmuch, message,
1304 status = _notmuch_database_link_message_to_children (notmuch, message,
1309 if (thread_id == NULL)
1310 _notmuch_message_ensure_thread_id (message);
1312 return NOTMUCH_STATUS_SUCCESS;
1316 notmuch_database_add_message (notmuch_database_t *notmuch,
1317 const char *filename,
1318 notmuch_message_t **message_ret)
1320 notmuch_message_file_t *message_file;
1321 notmuch_message_t *message = NULL;
1322 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1323 notmuch_private_status_t private_status;
1325 const char *date, *header;
1326 const char *from, *to, *subject;
1327 char *message_id = NULL;
1330 *message_ret = NULL;
1332 ret = _notmuch_database_ensure_writable (notmuch);
1336 message_file = notmuch_message_file_open (filename);
1337 if (message_file == NULL)
1338 return NOTMUCH_STATUS_FILE_ERROR;
1340 notmuch_message_file_restrict_headers (message_file,
1351 /* Before we do any real work, (especially before doing a
1352 * potential SHA-1 computation on the entire file's contents),
1353 * let's make sure that what we're looking at looks like an
1354 * actual email message.
1356 from = notmuch_message_file_get_header (message_file, "from");
1357 subject = notmuch_message_file_get_header (message_file, "subject");
1358 to = notmuch_message_file_get_header (message_file, "to");
1360 if ((from == NULL || *from == '\0') &&
1361 (subject == NULL || *subject == '\0') &&
1362 (to == NULL || *to == '\0'))
1364 ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
1368 /* Now that we're sure it's mail, the first order of business
1369 * is to find a message ID (or else create one ourselves). */
1371 header = notmuch_message_file_get_header (message_file, "message-id");
1372 if (header && *header != '\0') {
1373 message_id = _parse_message_id (message_file, header, NULL);
1375 /* So the header value isn't RFC-compliant, but it's
1376 * better than no message-id at all. */
1377 if (message_id == NULL)
1378 message_id = talloc_strdup (message_file, header);
1380 /* Reject a Message ID that's too long. */
1381 if (message_id && strlen (message_id) + 1 > NOTMUCH_TERM_MAX) {
1382 talloc_free (message_id);
1387 if (message_id == NULL ) {
1388 /* No message-id at all, let's generate one by taking a
1389 * hash over the file's contents. */
1390 char *sha1 = notmuch_sha1_of_file (filename);
1392 /* If that failed too, something is really wrong. Give up. */
1394 ret = NOTMUCH_STATUS_FILE_ERROR;
1398 message_id = talloc_asprintf (message_file,
1399 "notmuch-sha1-%s", sha1);
1403 /* Now that we have a message ID, we get a message object,
1404 * (which may or may not reference an existing document in the
1407 message = _notmuch_message_create_for_message_id (notmuch,
1411 talloc_free (message_id);
1413 if (message == NULL) {
1414 ret = COERCE_STATUS (private_status,
1415 "Unexpected status value from _notmuch_message_create_for_message_id");
1419 _notmuch_message_add_filename (message, filename);
1421 /* Is this a newly created message object? */
1422 if (private_status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
1423 _notmuch_message_add_term (message, "type", "mail");
1425 ret = _notmuch_database_link_message (notmuch, message,
1430 date = notmuch_message_file_get_header (message_file, "date");
1431 _notmuch_message_set_date (message, date);
1433 _notmuch_message_index_file (message, filename);
1435 ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
1438 _notmuch_message_sync (message);
1439 } catch (const Xapian::Error &error) {
1440 fprintf (stderr, "A Xapian exception occurred adding message: %s.\n",
1441 error.get_description().c_str());
1442 notmuch->exception_reported = TRUE;
1443 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
1449 if (ret == NOTMUCH_STATUS_SUCCESS && message_ret)
1450 *message_ret = message;
1452 notmuch_message_destroy (message);
1456 notmuch_message_file_close (message_file);
1462 notmuch_database_remove_message (notmuch_database_t *notmuch,
1463 const char *filename)
1465 Xapian::WritableDatabase *db;
1466 void *local = talloc_new (notmuch);
1467 const char *prefix = _find_prefix ("file-direntry");
1468 char *direntry, *term;
1469 Xapian::PostingIterator i, end;
1470 Xapian::Document document;
1471 notmuch_status_t status;
1473 status = _notmuch_database_ensure_writable (notmuch);
1477 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1479 status = _notmuch_database_filename_to_direntry (local, notmuch,
1480 filename, &direntry);
1484 term = talloc_asprintf (notmuch, "%s%s", prefix, direntry);
1486 find_doc_ids_for_term (notmuch, term, &i, &end);
1488 for ( ; i != end; i++) {
1489 Xapian::TermIterator j;
1491 document = find_document_for_doc_id (notmuch, *i);
1493 document.remove_term (term);
1495 j = document.termlist_begin ();
1498 /* Was this the last file-direntry in the message? */
1499 if (j == document.termlist_end () ||
1500 strncmp ((*j).c_str (), prefix, strlen (prefix)))
1502 db->delete_document (document.get_docid ());
1503 status = NOTMUCH_STATUS_SUCCESS;
1505 db->replace_document (document.get_docid (), document);
1506 status = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
1510 talloc_free (local);
1516 _notmuch_convert_tags (void *ctx, Xapian::TermIterator &i,
1517 Xapian::TermIterator &end)
1519 const char *prefix = _find_prefix ("tag");
1520 notmuch_tags_t *tags;
1523 /* Currently this iteration is written with the assumption that
1524 * "tag" has a single-character prefix. */
1525 assert (strlen (prefix) == 1);
1527 tags = _notmuch_tags_create (ctx);
1528 if (unlikely (tags == NULL))
1536 if (tag.empty () || tag[0] != *prefix)
1539 _notmuch_tags_add_tag (tags, tag.c_str () + 1);
1544 _notmuch_tags_prepare_iterator (tags);
1550 notmuch_database_get_all_tags (notmuch_database_t *db)
1552 Xapian::TermIterator i, end;
1553 i = db->xapian_db->allterms_begin();
1554 end = db->xapian_db->allterms_end();
1555 return _notmuch_convert_tags(db, i, end);