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"
28 #include <glib.h> /* g_free, GPtrArray, GHashTable */
32 #define ARRAY_SIZE(arr) (sizeof (arr) / sizeof (arr[0]))
39 #define NOTMUCH_DATABASE_VERSION 1
41 #define STRINGIFY(s) _SUB_STRINGIFY(s)
42 #define _SUB_STRINGIFY(s) #s
44 /* Here's the current schema for our database (for NOTMUCH_DATABASE_VERSION):
46 * We currently have two different types of documents (mail and
47 * directory) and also some metadata.
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. This is from the Message-ID header
60 * if present and not too long (see NOTMUCH_MESSAGE_ID_MAX).
61 * If it's present and too long, then we use
62 * "notmuch-sha1-<sha1_sum_of_message_id>".
63 * If this header is not present, we use
64 * "notmuch-sha1-<sha1_sum_of_entire_file>".
66 * thread: The ID of the thread to which the mail belongs
68 * replyto: The ID from the In-Reply-To header of the mail (if any).
70 * Multiple terms of given prefix:
72 * reference: All message IDs from In-Reply-To and Re ferences
73 * headers in the message.
75 * tag: Any tags associated with this message by the user.
77 * file-direntry: A colon-separated pair of values
78 * (INTEGER:STRING), where INTEGER is the
79 * document ID of a directory document, and
80 * STRING is the name of a file within that
81 * directory for this mail message.
83 * A mail document also has two values:
85 * TIMESTAMP: The time_t value corresponding to the message's
88 * MESSAGE_ID: The unique ID of the mail mess (see "id" above)
90 * In addition, terms from the content of the message are added with
91 * "from", "to", "attachment", and "subject" prefixes for use by the
92 * user in searching. But the database doesn't really care itself
95 * The data portion of a mail document is empty.
99 * A directory document is used by a client of the notmuch library to
100 * maintain data necessary to allow for efficient polling of mail
103 * All directory documents contain one term:
105 * directory: The directory path (relative to the database path)
106 * Or the SHA1 sum of the directory path (if the
107 * path itself is too long to fit in a Xapian
110 * And all directory documents for directories other than top-level
111 * directories also contain the following term:
113 * directory-direntry: A colon-separated pair of values
114 * (INTEGER:STRING), where INTEGER is the
115 * document ID of the parent directory
116 * document, and STRING is the name of this
117 * directory within that parent.
119 * All directory documents have a single value:
121 * TIMESTAMP: The mtime of the directory (at last scan)
123 * The data portion of a directory document contains the path of the
124 * directory (relative to the database path).
128 * Xapian allows us to store arbitrary name-value pairs as
129 * "metadata". We currently use the following metadata names with the
132 * version The database schema version, (which is distinct
133 * from both the notmuch package version (see
134 * notmuch --version) and the libnotmuch library
135 * version. The version is stored as an base-10
136 * ASCII integer. The initial database version
137 * was 1, (though a schema existed before that
138 * were no "version" database value existed at
139 * all). Succesive versions are allocated as
140 * changes are made to the database (such as by
141 * indexing new fields).
143 * last_thread_id The last thread ID generated. This is stored
144 * as a 16-byte hexadecimal ASCII representation
145 * of a 64-bit unsigned integer. The first ID
146 * generated is 1 and the value will be
147 * incremented for each thread ID.
149 * thread_id_* A pre-allocated thread ID for a particular
150 * message. This is actually an arbitarily large
151 * family of metadata name. Any particular name is
152 * formed by concatenating "thread_id_" with a message
153 * ID (or the SHA1 sum of a message ID if it is very
154 * long---see description of 'id' in the mail
155 * document). The value stored is a thread ID.
157 * These thread ID metadata values are stored
158 * whenever a message references a parent message
159 * that does not yet exist in the database. A
160 * thread ID will be allocated and stored, and if
161 * the message is later added, the stored thread
162 * ID will be used (and the metadata value will
165 * Even before a message is added, it's
166 * pre-allocated thread ID is useful so that all
167 * descendant messages that reference this common
168 * parent can be recognized as belonging to the
172 /* With these prefix values we follow the conventions published here:
174 * http://xapian.org/docs/omega/termprefixes.html
176 * as much as makes sense. Note that I took some liberty in matching
177 * the reserved prefix values to notmuch concepts, (for example, 'G'
178 * is documented as "newsGroup (or similar entity - e.g. a web forum
179 * name)", for which I think the thread is the closest analogue in
180 * notmuch. This in spite of the fact that we will eventually be
181 * storing mailing-list messages where 'G' for "mailing list name"
182 * might be even a closer analogue. I'm treating the single-character
183 * prefixes preferentially for core notmuch concepts (which will be
184 * nearly universal to all mail messages).
187 static prefix_t BOOLEAN_PREFIX_INTERNAL[] = {
189 { "reference", "XREFERENCE" },
190 { "replyto", "XREPLYTO" },
191 { "directory", "XDIRECTORY" },
192 { "file-direntry", "XFDIRENTRY" },
193 { "directory-direntry", "XDDIRENTRY" },
196 static prefix_t BOOLEAN_PREFIX_EXTERNAL[] = {
203 static prefix_t PROBABILISTIC_PREFIX[]= {
206 { "attachment", "XATTACHMENT" },
207 { "subject", "XSUBJECT"}
211 _internal_error (const char *format, ...)
215 va_start (va_args, format);
217 fprintf (stderr, "Internal error: ");
218 vfprintf (stderr, format, va_args);
226 _find_prefix (const char *name)
230 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_INTERNAL); i++) {
231 if (strcmp (name, BOOLEAN_PREFIX_INTERNAL[i].name) == 0)
232 return BOOLEAN_PREFIX_INTERNAL[i].prefix;
235 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
236 if (strcmp (name, BOOLEAN_PREFIX_EXTERNAL[i].name) == 0)
237 return BOOLEAN_PREFIX_EXTERNAL[i].prefix;
240 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
241 if (strcmp (name, PROBABILISTIC_PREFIX[i].name) == 0)
242 return PROBABILISTIC_PREFIX[i].prefix;
245 INTERNAL_ERROR ("No prefix exists for '%s'\n", name);
251 notmuch_status_to_string (notmuch_status_t status)
254 case NOTMUCH_STATUS_SUCCESS:
255 return "No error occurred";
256 case NOTMUCH_STATUS_OUT_OF_MEMORY:
257 return "Out of memory";
258 case NOTMUCH_STATUS_READ_ONLY_DATABASE:
259 return "Attempt to write to a read-only database";
260 case NOTMUCH_STATUS_XAPIAN_EXCEPTION:
261 return "A Xapian exception occurred";
262 case NOTMUCH_STATUS_FILE_ERROR:
263 return "Something went wrong trying to read or write a file";
264 case NOTMUCH_STATUS_FILE_NOT_EMAIL:
265 return "File is not an email";
266 case NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID:
267 return "Message ID is identical to a message in database";
268 case NOTMUCH_STATUS_NULL_POINTER:
269 return "Erroneous NULL pointer";
270 case NOTMUCH_STATUS_TAG_TOO_LONG:
271 return "Tag value is too long (exceeds NOTMUCH_TAG_MAX)";
272 case NOTMUCH_STATUS_UNBALANCED_FREEZE_THAW:
273 return "Unbalanced number of calls to notmuch_message_freeze/thaw";
275 case NOTMUCH_STATUS_LAST_STATUS:
276 return "Unknown error status value";
281 find_doc_ids_for_term (notmuch_database_t *notmuch,
283 Xapian::PostingIterator *begin,
284 Xapian::PostingIterator *end)
286 *begin = notmuch->xapian_db->postlist_begin (term);
288 *end = notmuch->xapian_db->postlist_end (term);
292 find_doc_ids (notmuch_database_t *notmuch,
293 const char *prefix_name,
295 Xapian::PostingIterator *begin,
296 Xapian::PostingIterator *end)
300 term = talloc_asprintf (notmuch, "%s%s",
301 _find_prefix (prefix_name), value);
303 find_doc_ids_for_term (notmuch, term, begin, end);
308 notmuch_private_status_t
309 _notmuch_database_find_unique_doc_id (notmuch_database_t *notmuch,
310 const char *prefix_name,
312 unsigned int *doc_id)
314 Xapian::PostingIterator i, end;
316 find_doc_ids (notmuch, prefix_name, value, &i, &end);
320 return NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND;
325 #if DEBUG_DATABASE_SANITY
329 INTERNAL_ERROR ("Term %s:%s is not unique as expected.\n",
333 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
336 static Xapian::Document
337 find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id)
339 return notmuch->xapian_db->get_document (doc_id);
342 /* Generate a compressed version of 'message_id' of the form:
344 * notmuch-sha1-<sha1_sum_of_message_id>
347 _message_id_compressed (void *ctx, const char *message_id)
349 char *sha1, *compressed;
351 sha1 = notmuch_sha1_of_string (message_id);
353 compressed = talloc_asprintf (ctx, "notmuch-sha1-%s", sha1);
360 notmuch_database_find_message (notmuch_database_t *notmuch,
361 const char *message_id)
363 notmuch_private_status_t status;
366 if (strlen (message_id) > NOTMUCH_MESSAGE_ID_MAX)
367 message_id = _message_id_compressed (notmuch, message_id);
370 status = _notmuch_database_find_unique_doc_id (notmuch, "id",
371 message_id, &doc_id);
373 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
376 return _notmuch_message_create (notmuch, notmuch, doc_id, NULL);
377 } catch (const Xapian::Error &error) {
378 fprintf (stderr, "A Xapian exception occurred finding message: %s.\n",
379 error.get_msg().c_str());
380 notmuch->exception_reported = TRUE;
385 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
386 * a (potentially nested) parenthesized sequence with '\' used to
387 * escape any character (including parentheses).
389 * If the sequence to be skipped continues to the end of the string,
390 * then 'str' will be left pointing at the final terminating '\0'
394 skip_space_and_comments (const char **str)
399 while (*s && (isspace (*s) || *s == '(')) {
400 while (*s && isspace (*s))
405 while (*s && nesting) {
408 } else if (*s == ')') {
410 } else if (*s == '\\') {
422 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
423 * comments, and the '<' and '>' delimeters.
425 * If not NULL, then *next will be made to point to the first character
426 * not parsed, (possibly pointing to the final '\0' terminator.
428 * Returns a newly talloc'ed string belonging to 'ctx'.
430 * Returns NULL if there is any error parsing the message-id. */
432 _parse_message_id (void *ctx, const char *message_id, const char **next)
437 if (message_id == NULL || *message_id == '\0')
442 skip_space_and_comments (&s);
444 /* Skip any unstructured text as well. */
445 while (*s && *s != '<')
456 skip_space_and_comments (&s);
459 while (*end && *end != '>')
468 if (end > s && *end == '>')
473 result = talloc_strndup (ctx, s, end - s + 1);
475 /* Finally, collapse any whitespace that is within the message-id
481 for (r = result, len = strlen (r); *r; r++, len--)
482 if (*r == ' ' || *r == '\t')
483 memmove (r, r+1, len);
489 /* Parse a References header value, putting a (talloc'ed under 'ctx')
490 * copy of each referenced message-id into 'hash'.
492 * We explicitly avoid including any reference identical to
493 * 'message_id' in the result (to avoid mass confusion when a single
494 * message references itself cyclically---and yes, mail messages are
495 * not infrequent in the wild that do this---don't ask me why).
498 parse_references (void *ctx,
499 const char *message_id,
505 if (refs == NULL || *refs == '\0')
509 ref = _parse_message_id (ctx, refs, &refs);
511 if (ref && strcmp (ref, message_id))
512 g_hash_table_insert (hash, ref, NULL);
517 notmuch_database_create (const char *path)
519 notmuch_database_t *notmuch = NULL;
520 char *notmuch_path = NULL;
525 fprintf (stderr, "Error: Cannot create a database for a NULL path.\n");
529 err = stat (path, &st);
531 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
532 path, strerror (errno));
536 if (! S_ISDIR (st.st_mode)) {
537 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
542 notmuch_path = talloc_asprintf (NULL, "%s/%s", path, ".notmuch");
544 err = mkdir (notmuch_path, 0755);
547 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
548 notmuch_path, strerror (errno));
552 notmuch = notmuch_database_open (path,
553 NOTMUCH_DATABASE_MODE_READ_WRITE);
554 notmuch_database_upgrade (notmuch, NULL, NULL);
558 talloc_free (notmuch_path);
564 _notmuch_database_ensure_writable (notmuch_database_t *notmuch)
566 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY) {
567 fprintf (stderr, "Cannot write to a read-only database.\n");
568 return NOTMUCH_STATUS_READ_ONLY_DATABASE;
571 return NOTMUCH_STATUS_SUCCESS;
575 notmuch_database_open (const char *path,
576 notmuch_database_mode_t mode)
578 notmuch_database_t *notmuch = NULL;
579 char *notmuch_path = NULL, *xapian_path = NULL;
582 unsigned int i, version;
584 if (asprintf (¬much_path, "%s/%s", path, ".notmuch") == -1) {
586 fprintf (stderr, "Out of memory\n");
590 err = stat (notmuch_path, &st);
592 fprintf (stderr, "Error opening database at %s: %s\n",
593 notmuch_path, strerror (errno));
597 if (asprintf (&xapian_path, "%s/%s", notmuch_path, "xapian") == -1) {
599 fprintf (stderr, "Out of memory\n");
603 notmuch = talloc (NULL, notmuch_database_t);
604 notmuch->exception_reported = FALSE;
605 notmuch->path = talloc_strdup (notmuch, path);
607 if (notmuch->path[strlen (notmuch->path) - 1] == '/')
608 notmuch->path[strlen (notmuch->path) - 1] = '\0';
610 notmuch->needs_upgrade = FALSE;
611 notmuch->mode = mode;
613 string last_thread_id;
615 if (mode == NOTMUCH_DATABASE_MODE_READ_WRITE) {
616 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
617 Xapian::DB_CREATE_OR_OPEN);
618 version = notmuch_database_get_version (notmuch);
620 if (version > NOTMUCH_DATABASE_VERSION) {
622 "Error: Notmuch database at %s\n"
623 " has a newer database format version (%u) than supported by this\n"
624 " version of notmuch (%u). Refusing to open this database in\n"
625 " read-write mode.\n",
626 notmuch_path, version, NOTMUCH_DATABASE_VERSION);
627 notmuch->mode = NOTMUCH_DATABASE_MODE_READ_ONLY;
628 notmuch_database_close (notmuch);
633 if (version < NOTMUCH_DATABASE_VERSION)
634 notmuch->needs_upgrade = TRUE;
636 notmuch->xapian_db = new Xapian::Database (xapian_path);
637 version = notmuch_database_get_version (notmuch);
638 if (version > NOTMUCH_DATABASE_VERSION)
641 "Warning: Notmuch database at %s\n"
642 " has a newer database format version (%u) than supported by this\n"
643 " version of notmuch (%u). Some operations may behave incorrectly,\n"
644 " (but the database will not be harmed since it is being opened\n"
645 " in read-only mode).\n",
646 notmuch_path, version, NOTMUCH_DATABASE_VERSION);
650 notmuch->last_doc_id = notmuch->xapian_db->get_lastdocid ();
651 last_thread_id = notmuch->xapian_db->get_metadata ("last_thread_id");
652 if (last_thread_id.empty ()) {
653 notmuch->last_thread_id = 0;
658 str = last_thread_id.c_str ();
659 notmuch->last_thread_id = strtoull (str, &end, 16);
661 INTERNAL_ERROR ("Malformed database last_thread_id: %s", str);
664 notmuch->query_parser = new Xapian::QueryParser;
665 notmuch->term_gen = new Xapian::TermGenerator;
666 notmuch->term_gen->set_stemmer (Xapian::Stem ("english"));
667 notmuch->value_range_processor = new Xapian::NumberValueRangeProcessor (NOTMUCH_VALUE_TIMESTAMP);
669 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
670 notmuch->query_parser->set_database (*notmuch->xapian_db);
671 notmuch->query_parser->set_stemmer (Xapian::Stem ("english"));
672 notmuch->query_parser->set_stemming_strategy (Xapian::QueryParser::STEM_SOME);
673 notmuch->query_parser->add_valuerangeprocessor (notmuch->value_range_processor);
675 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
676 prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i];
677 notmuch->query_parser->add_boolean_prefix (prefix->name,
681 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
682 prefix_t *prefix = &PROBABILISTIC_PREFIX[i];
683 notmuch->query_parser->add_prefix (prefix->name, prefix->prefix);
685 } catch (const Xapian::Error &error) {
686 fprintf (stderr, "A Xapian exception occurred opening database: %s\n",
687 error.get_msg().c_str());
691 notmuch_database_set_maildir_sync (notmuch, FALSE);
703 notmuch_database_close (notmuch_database_t *notmuch)
706 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_WRITE)
707 (static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db))->flush ();
708 } catch (const Xapian::Error &error) {
709 if (! notmuch->exception_reported) {
710 fprintf (stderr, "Error: A Xapian exception occurred flushing database: %s\n",
711 error.get_msg().c_str());
715 delete notmuch->term_gen;
716 delete notmuch->query_parser;
717 delete notmuch->xapian_db;
718 delete notmuch->value_range_processor;
719 talloc_free (notmuch);
723 notmuch_database_set_maildir_sync (notmuch_database_t *database,
724 notmuch_bool_t maildir_sync)
726 database->maildir_sync = maildir_sync;
730 notmuch_database_get_path (notmuch_database_t *notmuch)
732 return notmuch->path;
736 notmuch_database_get_version (notmuch_database_t *notmuch)
738 unsigned int version;
739 string version_string;
743 version_string = notmuch->xapian_db->get_metadata ("version");
744 if (version_string.empty ())
747 str = version_string.c_str ();
748 if (str == NULL || *str == '\0')
751 version = strtoul (str, &end, 10);
753 INTERNAL_ERROR ("Malformed database version: %s", str);
759 notmuch_database_needs_upgrade (notmuch_database_t *notmuch)
761 return notmuch->needs_upgrade;
764 static volatile sig_atomic_t do_progress_notify = 0;
767 handle_sigalrm (unused (int signal))
769 do_progress_notify = 1;
772 /* Upgrade the current database.
774 * After opening a database in read-write mode, the client should
775 * check if an upgrade is needed (notmuch_database_needs_upgrade) and
776 * if so, upgrade with this function before making any modifications.
778 * The optional progress_notify callback can be used by the caller to
779 * provide progress indication to the user. If non-NULL it will be
780 * called periodically with 'count' as the number of messages upgraded
781 * so far and 'total' the overall number of messages that will be
785 notmuch_database_upgrade (notmuch_database_t *notmuch,
786 void (*progress_notify) (void *closure,
790 Xapian::WritableDatabase *db;
791 struct sigaction action;
792 struct itimerval timerval;
793 notmuch_bool_t timer_is_active = FALSE;
794 unsigned int version;
795 notmuch_status_t status;
796 unsigned int count = 0, total = 0;
798 status = _notmuch_database_ensure_writable (notmuch);
802 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
804 version = notmuch_database_get_version (notmuch);
806 if (version >= NOTMUCH_DATABASE_VERSION)
807 return NOTMUCH_STATUS_SUCCESS;
809 if (progress_notify) {
810 /* Setup our handler for SIGALRM */
811 memset (&action, 0, sizeof (struct sigaction));
812 action.sa_handler = handle_sigalrm;
813 sigemptyset (&action.sa_mask);
814 action.sa_flags = SA_RESTART;
815 sigaction (SIGALRM, &action, NULL);
817 /* Then start a timer to send SIGALRM once per second. */
818 timerval.it_interval.tv_sec = 1;
819 timerval.it_interval.tv_usec = 0;
820 timerval.it_value.tv_sec = 1;
821 timerval.it_value.tv_usec = 0;
822 setitimer (ITIMER_REAL, &timerval, NULL);
824 timer_is_active = TRUE;
827 /* Before version 1, each message document had its filename in the
828 * data field. Copy that into the new format by calling
829 * notmuch_message_add_filename.
832 notmuch_query_t *query = notmuch_query_create (notmuch, "");
833 notmuch_messages_t *messages;
834 notmuch_message_t *message;
836 Xapian::TermIterator t, t_end;
838 total = notmuch_query_count_messages (query);
840 for (messages = notmuch_query_search_messages (query);
841 notmuch_messages_valid (messages);
842 notmuch_messages_move_to_next (messages))
844 if (do_progress_notify) {
845 progress_notify (closure, (double) count / total);
846 do_progress_notify = 0;
849 message = notmuch_messages_get (messages);
851 filename = _notmuch_message_talloc_copy_data (message);
852 if (filename && *filename != '\0') {
853 _notmuch_message_add_filename (message, filename);
854 _notmuch_message_sync (message);
856 talloc_free (filename);
858 notmuch_message_destroy (message);
863 notmuch_query_destroy (query);
865 /* Also, before version 1 we stored directory timestamps in
866 * XTIMESTAMP documents instead of the current XDIRECTORY
867 * documents. So copy those as well. */
869 t_end = notmuch->xapian_db->allterms_end ("XTIMESTAMP");
871 for (t = notmuch->xapian_db->allterms_begin ("XTIMESTAMP");
875 Xapian::PostingIterator p, p_end;
876 std::string term = *t;
878 p_end = notmuch->xapian_db->postlist_end (term);
880 for (p = notmuch->xapian_db->postlist_begin (term);
884 Xapian::Document document;
886 notmuch_directory_t *directory;
888 if (do_progress_notify) {
889 progress_notify (closure, (double) count / total);
890 do_progress_notify = 0;
893 document = find_document_for_doc_id (notmuch, *p);
894 mtime = Xapian::sortable_unserialise (
895 document.get_value (NOTMUCH_VALUE_TIMESTAMP));
897 directory = notmuch_database_get_directory (notmuch,
899 notmuch_directory_set_mtime (directory, mtime);
900 notmuch_directory_destroy (directory);
905 db->set_metadata ("version", STRINGIFY (NOTMUCH_DATABASE_VERSION));
908 /* Now that the upgrade is complete we can remove the old data
909 * and documents that are no longer needed. */
911 notmuch_query_t *query = notmuch_query_create (notmuch, "");
912 notmuch_messages_t *messages;
913 notmuch_message_t *message;
916 for (messages = notmuch_query_search_messages (query);
917 notmuch_messages_valid (messages);
918 notmuch_messages_move_to_next (messages))
920 if (do_progress_notify) {
921 progress_notify (closure, (double) count / total);
922 do_progress_notify = 0;
925 message = notmuch_messages_get (messages);
927 filename = _notmuch_message_talloc_copy_data (message);
928 if (filename && *filename != '\0') {
929 _notmuch_message_clear_data (message);
930 _notmuch_message_sync (message);
932 talloc_free (filename);
934 notmuch_message_destroy (message);
937 notmuch_query_destroy (query);
941 Xapian::TermIterator t, t_end;
943 t_end = notmuch->xapian_db->allterms_end ("XTIMESTAMP");
945 for (t = notmuch->xapian_db->allterms_begin ("XTIMESTAMP");
949 Xapian::PostingIterator p, p_end;
950 std::string term = *t;
952 p_end = notmuch->xapian_db->postlist_end (term);
954 for (p = notmuch->xapian_db->postlist_begin (term);
958 if (do_progress_notify) {
959 progress_notify (closure, (double) count / total);
960 do_progress_notify = 0;
963 db->delete_document (*p);
968 if (timer_is_active) {
969 /* Now stop the timer. */
970 timerval.it_interval.tv_sec = 0;
971 timerval.it_interval.tv_usec = 0;
972 timerval.it_value.tv_sec = 0;
973 timerval.it_value.tv_usec = 0;
974 setitimer (ITIMER_REAL, &timerval, NULL);
976 /* And disable the signal handler. */
977 action.sa_handler = SIG_IGN;
978 sigaction (SIGALRM, &action, NULL);
981 return NOTMUCH_STATUS_SUCCESS;
984 /* We allow the user to use arbitrarily long paths for directories. But
985 * we have a term-length limit. So if we exceed that, we'll use the
986 * SHA-1 of the path for the database term.
988 * Note: This function may return the original value of 'path'. If it
989 * does not, then the caller is responsible to free() the returned
993 _notmuch_database_get_directory_db_path (const char *path)
995 int term_len = strlen (_find_prefix ("directory")) + strlen (path);
997 if (term_len > NOTMUCH_TERM_MAX)
998 return notmuch_sha1_of_string (path);
1003 /* Given a path, split it into two parts: the directory part is all
1004 * components except for the last, and the basename is that last
1005 * component. Getting the return-value for either part is optional
1006 * (the caller can pass NULL).
1008 * The original 'path' can represent either a regular file or a
1009 * directory---the splitting will be carried out in the same way in
1010 * either case. Trailing slashes on 'path' will be ignored, and any
1011 * cases of multiple '/' characters appearing in series will be
1012 * treated as a single '/'.
1014 * Allocation (if any) will have 'ctx' as the talloc owner. But
1015 * pointers will be returned within the original path string whenever
1018 * Note: If 'path' is non-empty and contains no non-trailing slash,
1019 * (that is, consists of a filename with no parent directory), then
1020 * the directory returned will be an empty string. However, if 'path'
1021 * is an empty string, then both directory and basename will be
1025 _notmuch_database_split_path (void *ctx,
1027 const char **directory,
1028 const char **basename)
1032 if (path == NULL || *path == '\0') {
1037 return NOTMUCH_STATUS_SUCCESS;
1040 /* Find the last slash (not counting a trailing slash), if any. */
1042 slash = path + strlen (path) - 1;
1044 /* First, skip trailing slashes. */
1045 while (slash != path) {
1052 /* Then, find a slash. */
1053 while (slash != path) {
1063 /* Finally, skip multiple slashes. */
1064 while (slash != path) {
1071 if (slash == path) {
1073 *directory = talloc_strdup (ctx, "");
1078 *directory = talloc_strndup (ctx, path, slash - path + 1);
1081 return NOTMUCH_STATUS_SUCCESS;
1085 _notmuch_database_find_directory_id (notmuch_database_t *notmuch,
1087 unsigned int *directory_id)
1089 notmuch_directory_t *directory;
1090 notmuch_status_t status;
1094 return NOTMUCH_STATUS_SUCCESS;
1097 directory = _notmuch_directory_create (notmuch, path, &status);
1103 *directory_id = _notmuch_directory_get_document_id (directory);
1105 notmuch_directory_destroy (directory);
1107 return NOTMUCH_STATUS_SUCCESS;
1111 _notmuch_database_get_directory_path (void *ctx,
1112 notmuch_database_t *notmuch,
1113 unsigned int doc_id)
1115 Xapian::Document document;
1117 document = find_document_for_doc_id (notmuch, doc_id);
1119 return talloc_strdup (ctx, document.get_data ().c_str ());
1122 /* Given a legal 'filename' for the database, (either relative to
1123 * database path or absolute with initial components identical to
1124 * database path), return a new string (with 'ctx' as the talloc
1125 * owner) suitable for use as a direntry term value.
1127 * The necessary directory documents will be created in the database
1131 _notmuch_database_filename_to_direntry (void *ctx,
1132 notmuch_database_t *notmuch,
1133 const char *filename,
1136 const char *relative, *directory, *basename;
1137 Xapian::docid directory_id;
1138 notmuch_status_t status;
1140 relative = _notmuch_database_relative_path (notmuch, filename);
1142 status = _notmuch_database_split_path (ctx, relative,
1143 &directory, &basename);
1147 status = _notmuch_database_find_directory_id (notmuch, directory,
1152 *direntry = talloc_asprintf (ctx, "%u:%s", directory_id, basename);
1154 return NOTMUCH_STATUS_SUCCESS;
1157 /* Given a legal 'path' for the database, return the relative path.
1159 * The return value will be a pointer to the originl path contents,
1160 * and will be either the original string (if 'path' was relative) or
1161 * a portion of the string (if path was absolute and begins with the
1165 _notmuch_database_relative_path (notmuch_database_t *notmuch,
1168 const char *db_path, *relative;
1169 unsigned int db_path_len;
1171 db_path = notmuch_database_get_path (notmuch);
1172 db_path_len = strlen (db_path);
1176 if (*relative == '/') {
1177 while (*relative == '/' && *(relative+1) == '/')
1180 if (strncmp (relative, db_path, db_path_len) == 0)
1182 relative += db_path_len;
1183 while (*relative == '/')
1191 notmuch_directory_t *
1192 notmuch_database_get_directory (notmuch_database_t *notmuch,
1195 notmuch_status_t status;
1198 return _notmuch_directory_create (notmuch, path, &status);
1199 } catch (const Xapian::Error &error) {
1200 fprintf (stderr, "A Xapian exception occurred getting directory: %s.\n",
1201 error.get_msg().c_str());
1202 notmuch->exception_reported = TRUE;
1207 /* Allocate a document ID that satisfies the following criteria:
1209 * 1. The ID does not exist for any document in the Xapian database
1211 * 2. The ID was not previously returned from this function
1213 * 3. The ID is the smallest integer satisfying (1) and (2)
1215 * This function will trigger an internal error if these constraints
1216 * cannot all be satisfied, (that is, the pool of available document
1217 * IDs has been exhausted).
1220 _notmuch_database_generate_doc_id (notmuch_database_t *notmuch)
1222 assert (notmuch->last_doc_id >= notmuch->xapian_db->get_lastdocid ());
1224 notmuch->last_doc_id++;
1226 if (notmuch->last_doc_id == 0)
1227 INTERNAL_ERROR ("Xapian document IDs are exhausted.\n");
1229 return notmuch->last_doc_id;
1233 _notmuch_database_generate_thread_id (notmuch_database_t *notmuch)
1235 /* 16 bytes (+ terminator) for hexadecimal representation of
1236 * a 64-bit integer. */
1237 static char thread_id[17];
1238 Xapian::WritableDatabase *db;
1240 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1242 notmuch->last_thread_id++;
1244 sprintf (thread_id, "%016" PRIx64, notmuch->last_thread_id);
1246 db->set_metadata ("last_thread_id", thread_id);
1252 _get_metadata_thread_id_key (void *ctx, const char *message_id)
1254 if (strlen (message_id) > NOTMUCH_MESSAGE_ID_MAX)
1255 message_id = _message_id_compressed (ctx, message_id);
1257 return talloc_asprintf (ctx, NOTMUCH_METADATA_THREAD_ID_PREFIX "%s",
1261 /* Find the thread ID to which the message with 'message_id' belongs.
1263 * Always returns a newly talloced string belonging to 'ctx'.
1265 * Note: If there is no message in the database with the given
1266 * 'message_id' then a new thread_id will be allocated for this
1267 * message and stored in the database metadata, (where this same
1268 * thread ID can be looked up if the message is added to the database
1272 _resolve_message_id_to_thread_id (notmuch_database_t *notmuch,
1274 const char *message_id)
1276 notmuch_message_t *message;
1277 string thread_id_string;
1278 const char *thread_id;
1280 Xapian::WritableDatabase *db;
1282 message = notmuch_database_find_message (notmuch, message_id);
1285 thread_id = talloc_steal (ctx, notmuch_message_get_thread_id (message));
1287 notmuch_message_destroy (message);
1292 /* Message has not been seen yet.
1294 * We may have seen a reference to it already, in which case, we
1295 * can return the thread ID stored in the metadata. Otherwise, we
1296 * generate a new thread ID and store it there.
1298 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1299 metadata_key = _get_metadata_thread_id_key (ctx, message_id);
1300 thread_id_string = notmuch->xapian_db->get_metadata (metadata_key);
1302 if (thread_id_string.empty()) {
1303 thread_id = _notmuch_database_generate_thread_id (notmuch);
1304 db->set_metadata (metadata_key, thread_id);
1306 thread_id = thread_id_string.c_str();
1309 talloc_free (metadata_key);
1311 return talloc_strdup (ctx, thread_id);
1314 static notmuch_status_t
1315 _merge_threads (notmuch_database_t *notmuch,
1316 const char *winner_thread_id,
1317 const char *loser_thread_id)
1319 Xapian::PostingIterator loser, loser_end;
1320 notmuch_message_t *message = NULL;
1321 notmuch_private_status_t private_status;
1322 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1324 find_doc_ids (notmuch, "thread", loser_thread_id, &loser, &loser_end);
1326 for ( ; loser != loser_end; loser++) {
1327 message = _notmuch_message_create (notmuch, notmuch,
1328 *loser, &private_status);
1329 if (message == NULL) {
1330 ret = COERCE_STATUS (private_status,
1331 "Cannot find document for doc_id from query");
1335 _notmuch_message_remove_term (message, "thread", loser_thread_id);
1336 _notmuch_message_add_term (message, "thread", winner_thread_id);
1337 _notmuch_message_sync (message);
1339 notmuch_message_destroy (message);
1345 notmuch_message_destroy (message);
1351 _my_talloc_free_for_g_hash (void *ptr)
1356 static notmuch_status_t
1357 _notmuch_database_link_message_to_parents (notmuch_database_t *notmuch,
1358 notmuch_message_t *message,
1359 notmuch_message_file_t *message_file,
1360 const char **thread_id)
1362 GHashTable *parents = NULL;
1363 const char *refs, *in_reply_to, *in_reply_to_message_id;
1364 GList *l, *keys = NULL;
1365 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1367 parents = g_hash_table_new_full (g_str_hash, g_str_equal,
1368 _my_talloc_free_for_g_hash, NULL);
1370 refs = notmuch_message_file_get_header (message_file, "references");
1371 parse_references (message, notmuch_message_get_message_id (message),
1374 in_reply_to = notmuch_message_file_get_header (message_file, "in-reply-to");
1375 parse_references (message, notmuch_message_get_message_id (message),
1376 parents, in_reply_to);
1378 /* Carefully avoid adding any self-referential in-reply-to term. */
1379 in_reply_to_message_id = _parse_message_id (message, in_reply_to, NULL);
1380 if (in_reply_to_message_id &&
1381 strcmp (in_reply_to_message_id,
1382 notmuch_message_get_message_id (message)))
1384 _notmuch_message_add_term (message, "replyto",
1385 _parse_message_id (message, in_reply_to, NULL));
1388 keys = g_hash_table_get_keys (parents);
1389 for (l = keys; l; l = l->next) {
1390 char *parent_message_id;
1391 const char *parent_thread_id;
1393 parent_message_id = (char *) l->data;
1395 _notmuch_message_add_term (message, "reference",
1398 parent_thread_id = _resolve_message_id_to_thread_id (notmuch,
1402 if (*thread_id == NULL) {
1403 *thread_id = talloc_strdup (message, parent_thread_id);
1404 _notmuch_message_add_term (message, "thread", *thread_id);
1405 } else if (strcmp (*thread_id, parent_thread_id)) {
1406 ret = _merge_threads (notmuch, *thread_id, parent_thread_id);
1416 g_hash_table_unref (parents);
1421 static notmuch_status_t
1422 _notmuch_database_link_message_to_children (notmuch_database_t *notmuch,
1423 notmuch_message_t *message,
1424 const char **thread_id)
1426 const char *message_id = notmuch_message_get_message_id (message);
1427 Xapian::PostingIterator child, children_end;
1428 notmuch_message_t *child_message = NULL;
1429 const char *child_thread_id;
1430 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1431 notmuch_private_status_t private_status;
1433 find_doc_ids (notmuch, "reference", message_id, &child, &children_end);
1435 for ( ; child != children_end; child++) {
1437 child_message = _notmuch_message_create (message, notmuch,
1438 *child, &private_status);
1439 if (child_message == NULL) {
1440 ret = COERCE_STATUS (private_status,
1441 "Cannot find document for doc_id from query");
1445 child_thread_id = notmuch_message_get_thread_id (child_message);
1446 if (*thread_id == NULL) {
1447 *thread_id = talloc_strdup (message, child_thread_id);
1448 _notmuch_message_add_term (message, "thread", *thread_id);
1449 } else if (strcmp (*thread_id, child_thread_id)) {
1450 _notmuch_message_remove_term (child_message, "reference",
1452 _notmuch_message_sync (child_message);
1453 ret = _merge_threads (notmuch, *thread_id, child_thread_id);
1458 notmuch_message_destroy (child_message);
1459 child_message = NULL;
1464 notmuch_message_destroy (child_message);
1469 /* Given a (mostly empty) 'message' and its corresponding
1470 * 'message_file' link it to existing threads in the database.
1472 * The first check is in the metadata of the database to see if we
1473 * have pre-allocated a thread_id in advance for this message, (which
1474 * would have happened if a message was previously added that
1475 * referenced this one).
1477 * Second, we look at 'message_file' and its link-relevant headers
1478 * (References and In-Reply-To) for message IDs.
1480 * Finally, we look in the database for existing message that
1481 * reference 'message'.
1483 * In all cases, we assign to the current message the first thread_id
1484 * found (through either parent or child). We will also merge any
1485 * existing, distinct threads where this message belongs to both,
1486 * (which is not uncommon when mesages are processed out of order).
1488 * Finally, if no thread ID has been found through parent or child, we
1489 * call _notmuch_message_generate_thread_id to generate a new thread
1490 * ID. This should only happen for new, top-level messages, (no
1491 * References or In-Reply-To header in this message, and no previously
1492 * added message refers to this message).
1494 static notmuch_status_t
1495 _notmuch_database_link_message (notmuch_database_t *notmuch,
1496 notmuch_message_t *message,
1497 notmuch_message_file_t *message_file)
1499 notmuch_status_t status;
1500 const char *message_id, *thread_id = NULL;
1504 message_id = notmuch_message_get_message_id (message);
1505 metadata_key = _get_metadata_thread_id_key (message, message_id);
1507 /* Check if we have already seen related messages to this one.
1508 * If we have then use the thread_id that we stored at that time.
1510 stored_id = notmuch->xapian_db->get_metadata (metadata_key);
1511 if (! stored_id.empty()) {
1512 Xapian::WritableDatabase *db;
1514 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1516 /* Clear the metadata for this message ID. We don't need it
1518 db->set_metadata (metadata_key, "");
1519 thread_id = stored_id.c_str();
1521 _notmuch_message_add_term (message, "thread", thread_id);
1523 talloc_free (metadata_key);
1525 status = _notmuch_database_link_message_to_parents (notmuch, message,
1531 status = _notmuch_database_link_message_to_children (notmuch, message,
1536 /* If not part of any existing thread, generate a new thread ID. */
1537 if (thread_id == NULL) {
1538 thread_id = _notmuch_database_generate_thread_id (notmuch);
1540 _notmuch_message_add_term (message, "thread", thread_id);
1543 return NOTMUCH_STATUS_SUCCESS;
1547 notmuch_database_add_message (notmuch_database_t *notmuch,
1548 const char *filename,
1549 notmuch_message_t **message_ret)
1551 notmuch_message_file_t *message_file;
1552 notmuch_message_t *message = NULL;
1553 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1554 notmuch_private_status_t private_status;
1556 const char *date, *header;
1557 const char *from, *to, *subject;
1558 char *message_id = NULL;
1561 *message_ret = NULL;
1563 ret = _notmuch_database_ensure_writable (notmuch);
1567 message_file = notmuch_message_file_open (filename);
1568 if (message_file == NULL)
1569 return NOTMUCH_STATUS_FILE_ERROR;
1571 notmuch_message_file_restrict_headers (message_file,
1582 /* Before we do any real work, (especially before doing a
1583 * potential SHA-1 computation on the entire file's contents),
1584 * let's make sure that what we're looking at looks like an
1585 * actual email message.
1587 from = notmuch_message_file_get_header (message_file, "from");
1588 subject = notmuch_message_file_get_header (message_file, "subject");
1589 to = notmuch_message_file_get_header (message_file, "to");
1591 if ((from == NULL || *from == '\0') &&
1592 (subject == NULL || *subject == '\0') &&
1593 (to == NULL || *to == '\0'))
1595 ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
1599 /* Now that we're sure it's mail, the first order of business
1600 * is to find a message ID (or else create one ourselves). */
1602 header = notmuch_message_file_get_header (message_file, "message-id");
1603 if (header && *header != '\0') {
1604 message_id = _parse_message_id (message_file, header, NULL);
1606 /* So the header value isn't RFC-compliant, but it's
1607 * better than no message-id at all. */
1608 if (message_id == NULL)
1609 message_id = talloc_strdup (message_file, header);
1611 /* If a message ID is too long, substitute its sha1 instead. */
1612 if (message_id && strlen (message_id) > NOTMUCH_MESSAGE_ID_MAX) {
1613 char *compressed = _message_id_compressed (message_file,
1615 talloc_free (message_id);
1616 message_id = compressed;
1620 if (message_id == NULL ) {
1621 /* No message-id at all, let's generate one by taking a
1622 * hash over the file's contents. */
1623 char *sha1 = notmuch_sha1_of_file (filename);
1625 /* If that failed too, something is really wrong. Give up. */
1627 ret = NOTMUCH_STATUS_FILE_ERROR;
1631 message_id = talloc_asprintf (message_file,
1632 "notmuch-sha1-%s", sha1);
1636 /* Now that we have a message ID, we get a message object,
1637 * (which may or may not reference an existing document in the
1640 message = _notmuch_message_create_for_message_id (notmuch,
1644 talloc_free (message_id);
1646 if (message == NULL) {
1647 ret = COERCE_STATUS (private_status,
1648 "Unexpected status value from _notmuch_message_create_for_message_id");
1652 _notmuch_message_add_filename (message, filename);
1654 /* This is a new message or it has a new filename and as such,
1655 * its tags in database either do not exists or might be out
1656 * of date. We assign the tags later in notmuch new, but until
1657 * then we should not synchronize the tags back to the maildir
1658 * flags (if notmuch is configured to do so). */
1659 notmuch_message_set_flag(message, NOTMUCH_MESSAGE_FLAG_TAGS_INVALID, TRUE);
1661 /* Is this a newly created message object? */
1662 if (private_status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
1663 _notmuch_message_add_term (message, "type", "mail");
1665 ret = _notmuch_database_link_message (notmuch, message,
1670 date = notmuch_message_file_get_header (message_file, "date");
1671 _notmuch_message_set_date (message, date);
1673 _notmuch_message_index_file (message, filename);
1675 ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
1678 _notmuch_message_sync (message);
1679 } catch (const Xapian::Error &error) {
1680 fprintf (stderr, "A Xapian exception occurred adding message: %s.\n",
1681 error.get_msg().c_str());
1682 notmuch->exception_reported = TRUE;
1683 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
1689 if ((ret == NOTMUCH_STATUS_SUCCESS ||
1690 ret == NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID) && message_ret)
1691 *message_ret = message;
1693 notmuch_message_destroy (message);
1697 notmuch_message_file_close (message_file);
1703 notmuch_database_remove_message (notmuch_database_t *notmuch,
1704 const char *filename)
1706 Xapian::WritableDatabase *db;
1708 const char *prefix = _find_prefix ("file-direntry");
1709 char *direntry, *term;
1710 Xapian::PostingIterator i, end;
1711 Xapian::Document document;
1712 notmuch_status_t status;
1714 status = _notmuch_database_ensure_writable (notmuch);
1718 local = talloc_new (notmuch);
1720 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1724 status = _notmuch_database_filename_to_direntry (local, notmuch,
1725 filename, &direntry);
1729 term = talloc_asprintf (notmuch, "%s%s", prefix, direntry);
1731 find_doc_ids_for_term (notmuch, term, &i, &end);
1733 for ( ; i != end; i++) {
1734 Xapian::TermIterator j;
1736 document = find_document_for_doc_id (notmuch, *i);
1738 document.remove_term (term);
1740 j = document.termlist_begin ();
1743 /* Was this the last file-direntry in the message? */
1744 if (j == document.termlist_end () ||
1745 strncmp ((*j).c_str (), prefix, strlen (prefix)))
1747 db->delete_document (document.get_docid ());
1748 status = NOTMUCH_STATUS_SUCCESS;
1750 db->replace_document (document.get_docid (), document);
1751 status = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
1754 } catch (const Xapian::Error &error) {
1755 fprintf (stderr, "Error: A Xapian exception occurred removing message: %s\n",
1756 error.get_msg().c_str());
1757 notmuch->exception_reported = TRUE;
1758 status = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
1761 talloc_free (local);
1767 _notmuch_convert_tags (void *ctx, Xapian::TermIterator &i,
1768 Xapian::TermIterator &end)
1770 const char *prefix = _find_prefix ("tag");
1771 notmuch_tags_t *tags;
1774 /* Currently this iteration is written with the assumption that
1775 * "tag" has a single-character prefix. */
1776 assert (strlen (prefix) == 1);
1778 tags = _notmuch_tags_create (ctx);
1779 if (unlikely (tags == NULL))
1787 if (tag.empty () || tag[0] != *prefix)
1790 _notmuch_tags_add_tag (tags, tag.c_str () + 1);
1795 _notmuch_tags_prepare_iterator (tags);
1801 notmuch_database_get_all_tags (notmuch_database_t *db)
1803 Xapian::TermIterator i, end;
1806 i = db->xapian_db->allterms_begin();
1807 end = db->xapian_db->allterms_end();
1808 return _notmuch_convert_tags(db, i, end);
1809 } catch (const Xapian::Error &error) {
1810 fprintf (stderr, "A Xapian exception occurred getting tags: %s.\n",
1811 error.get_msg().c_str());
1812 db->exception_reported = TRUE;