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_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 * TIMESTAMP: 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[] = {
107 { "ref", "XREFERENCE" },
108 { "replyto", "XREPLYTO" },
109 { "timestamp", "XTIMESTAMP" },
110 { "contact", "XCONTACT" }
113 prefix_t BOOLEAN_PREFIX_EXTERNAL[] = {
119 prefix_t PROBABILISTIC_PREFIX[]= {
122 { "attachment", "XATTACHMENT" },
123 { "subject", "XSUBJECT"}
127 _internal_error (const char *format, ...)
131 va_start (va_args, format);
133 fprintf (stderr, "Internal error: ");
134 vfprintf (stderr, format, va_args);
142 _find_prefix (const char *name)
146 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_INTERNAL); i++)
147 if (strcmp (name, BOOLEAN_PREFIX_INTERNAL[i].name) == 0)
148 return BOOLEAN_PREFIX_INTERNAL[i].prefix;
150 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++)
151 if (strcmp (name, BOOLEAN_PREFIX_EXTERNAL[i].name) == 0)
152 return BOOLEAN_PREFIX_EXTERNAL[i].prefix;
154 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++)
155 if (strcmp (name, PROBABILISTIC_PREFIX[i].name) == 0)
156 return PROBABILISTIC_PREFIX[i].prefix;
158 INTERNAL_ERROR ("No prefix exists for '%s'\n", name);
164 notmuch_status_to_string (notmuch_status_t status)
167 case NOTMUCH_STATUS_SUCCESS:
168 return "No error occurred";
169 case NOTMUCH_STATUS_OUT_OF_MEMORY:
170 return "Out of memory";
171 case NOTMUCH_STATUS_XAPIAN_EXCEPTION:
172 return "A Xapian exception occurred";
173 case NOTMUCH_STATUS_FILE_ERROR:
174 return "Something went wrong trying to read or write a file";
175 case NOTMUCH_STATUS_FILE_NOT_EMAIL:
176 return "File is not an email";
177 case NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID:
178 return "Message ID is identical to a message in database";
179 case NOTMUCH_STATUS_NULL_POINTER:
180 return "Erroneous NULL pointer";
181 case NOTMUCH_STATUS_TAG_TOO_LONG:
182 return "Tag value is too long (exceeds NOTMUCH_TAG_MAX)";
183 case NOTMUCH_STATUS_UNBALANCED_FREEZE_THAW:
184 return "Unblanced number of calls to notmuch_message_freeze/thaw";
186 case NOTMUCH_STATUS_LAST_STATUS:
187 return "Unknown error status value";
192 find_doc_ids (notmuch_database_t *notmuch,
193 const char *prefix_name,
195 Xapian::PostingIterator *begin,
196 Xapian::PostingIterator *end)
198 Xapian::PostingIterator i;
201 term = talloc_asprintf (notmuch, "%s%s",
202 _find_prefix (prefix_name), value);
204 *begin = notmuch->xapian_db->postlist_begin (term);
206 *end = notmuch->xapian_db->postlist_end (term);
211 static notmuch_private_status_t
212 find_unique_doc_id (notmuch_database_t *notmuch,
213 const char *prefix_name,
215 unsigned int *doc_id)
217 Xapian::PostingIterator i, end;
219 find_doc_ids (notmuch, prefix_name, value, &i, &end);
223 return NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND;
226 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
230 static Xapian::Document
231 find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id)
233 return notmuch->xapian_db->get_document (doc_id);
236 static notmuch_private_status_t
237 find_unique_document (notmuch_database_t *notmuch,
238 const char *prefix_name,
240 Xapian::Document *document,
241 unsigned int *doc_id)
243 notmuch_private_status_t status;
245 status = find_unique_doc_id (notmuch, prefix_name, value, doc_id);
248 *document = Xapian::Document ();
252 *document = find_document_for_doc_id (notmuch, *doc_id);
253 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
257 notmuch_database_find_message (notmuch_database_t *notmuch,
258 const char *message_id)
260 notmuch_private_status_t status;
263 status = find_unique_doc_id (notmuch, "id", message_id, &doc_id);
265 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
268 return _notmuch_message_create (notmuch, notmuch, doc_id, NULL);
271 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
272 * a (potentially nested) parenthesized sequence with '\' used to
273 * escape any character (including parentheses).
275 * If the sequence to be skipped continues to the end of the string,
276 * then 'str' will be left pointing at the final terminating '\0'
280 skip_space_and_comments (const char **str)
285 while (*s && (isspace (*s) || *s == '(')) {
286 while (*s && isspace (*s))
291 while (*s && nesting) {
307 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
308 * comments, and the '<' and '>' delimeters.
310 * If not NULL, then *next will be made to point to the first character
311 * not parsed, (possibly pointing to the final '\0' terminator.
313 * Returns a newly talloc'ed string belonging to 'ctx'.
315 * Returns NULL if there is any error parsing the message-id. */
317 parse_message_id (void *ctx, const char *message_id, const char **next)
322 if (message_id == NULL)
327 skip_space_and_comments (&s);
329 /* Skip any unstructured text as well. */
330 while (*s && *s != '<')
341 skip_space_and_comments (&s);
344 while (*end && *end != '>')
353 if (end > s && *end == '>')
358 result = talloc_strndup (ctx, s, end - s + 1);
360 /* Finally, collapse any whitespace that is within the message-id
366 for (r = result, len = strlen (r); *r; r++, len--)
367 if (*r == ' ' || *r == '\t')
368 memmove (r, r+1, len);
374 /* Parse a References header value, putting a (talloc'ed under 'ctx')
375 * copy of each referenced message-id into 'hash'. */
377 parse_references (void *ctx,
387 ref = parse_message_id (ctx, refs, &refs);
390 g_hash_table_insert (hash, ref, NULL);
395 notmuch_database_default_path (void)
399 if (getenv ("NOTMUCH_BASE"))
400 return strdup (getenv ("NOTMUCH_BASE"));
402 if (asprintf (&path, "%s/mail", getenv ("HOME")) == -1) {
403 fprintf (stderr, "Out of memory.\n");
411 notmuch_database_create (const char *path)
413 notmuch_database_t *notmuch = NULL;
414 char *notmuch_path = NULL;
417 char *local_path = NULL;
420 path = local_path = notmuch_database_default_path ();
422 err = stat (path, &st);
424 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
425 path, strerror (errno));
429 if (! S_ISDIR (st.st_mode)) {
430 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
435 notmuch_path = talloc_asprintf (NULL, "%s/%s", path, ".notmuch");
437 err = mkdir (notmuch_path, 0755);
440 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
441 notmuch_path, strerror (errno));
445 notmuch = notmuch_database_open (path);
449 talloc_free (notmuch_path);
457 notmuch_database_open (const char *path)
459 notmuch_database_t *notmuch = NULL;
460 char *notmuch_path = NULL, *xapian_path = NULL;
463 char *local_path = NULL;
467 path = local_path = notmuch_database_default_path ();
469 if (asprintf (¬much_path, "%s/%s", path, ".notmuch") == -1) {
471 fprintf (stderr, "Out of memory\n");
475 err = stat (notmuch_path, &st);
477 fprintf (stderr, "Error opening database at %s: %s\n",
478 notmuch_path, strerror (errno));
482 if (asprintf (&xapian_path, "%s/%s", notmuch_path, "xapian") == -1) {
484 fprintf (stderr, "Out of memory\n");
488 notmuch = talloc (NULL, notmuch_database_t);
489 notmuch->path = talloc_strdup (notmuch, path);
491 if (notmuch->path[strlen (notmuch->path) - 1] == '/')
492 notmuch->path[strlen (notmuch->path) - 1] = '\0';
495 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
496 Xapian::DB_CREATE_OR_OPEN);
497 notmuch->query_parser = new Xapian::QueryParser;
498 notmuch->term_gen = new Xapian::TermGenerator;
499 notmuch->term_gen->set_stemmer (Xapian::Stem ("english"));
501 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
502 notmuch->query_parser->set_database (*notmuch->xapian_db);
503 notmuch->query_parser->set_stemmer (Xapian::Stem ("english"));
504 notmuch->query_parser->set_stemming_strategy (Xapian::QueryParser::STEM_SOME);
506 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
507 prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i];
508 notmuch->query_parser->add_boolean_prefix (prefix->name,
512 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
513 prefix_t *prefix = &PROBABILISTIC_PREFIX[i];
514 notmuch->query_parser->add_prefix (prefix->name, prefix->prefix);
516 } catch (const Xapian::Error &error) {
517 fprintf (stderr, "A Xapian exception occurred: %s\n",
518 error.get_msg().c_str());
534 notmuch_database_close (notmuch_database_t *notmuch)
536 notmuch->xapian_db->flush ();
538 delete notmuch->term_gen;
539 delete notmuch->query_parser;
540 delete notmuch->xapian_db;
541 talloc_free (notmuch);
545 notmuch_database_get_path (notmuch_database_t *notmuch)
547 return notmuch->path;
550 static notmuch_private_status_t
551 find_timestamp_document (notmuch_database_t *notmuch, const char *db_key,
552 Xapian::Document *doc, unsigned int *doc_id)
554 return find_unique_document (notmuch, "timestamp", db_key, doc, doc_id);
557 /* We allow the user to use arbitrarily long keys for timestamps,
558 * (they're for filesystem paths after all, which have no limit we
559 * know about). But we have a term-length limit. So if we exceed that,
560 * we'll use the SHA-1 of the user's key as the actual key for
561 * constructing a database term.
563 * Caution: This function returns a newly allocated string which the
564 * caller should free() when finished.
567 timestamp_db_key (const char *key)
569 int term_len = strlen (_find_prefix ("timestamp")) + strlen (key);
571 if (term_len > NOTMUCH_TERM_MAX)
572 return notmuch_sha1_of_string (key);
578 notmuch_database_set_timestamp (notmuch_database_t *notmuch,
579 const char *key, time_t timestamp)
581 Xapian::Document doc;
583 notmuch_private_status_t status;
584 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
587 db_key = timestamp_db_key (key);
590 status = find_timestamp_document (notmuch, db_key, &doc, &doc_id);
592 doc.add_value (NOTMUCH_VALUE_TIMESTAMP,
593 Xapian::sortable_serialise (timestamp));
595 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
596 char *term = talloc_asprintf (NULL, "%s%s",
597 _find_prefix ("timestamp"), db_key);
601 notmuch->xapian_db->add_document (doc);
603 notmuch->xapian_db->replace_document (doc_id, doc);
606 } catch (Xapian::Error &error) {
607 fprintf (stderr, "A Xapian exception occurred: %s.\n",
608 error.get_msg().c_str());
609 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
619 notmuch_database_get_timestamp (notmuch_database_t *notmuch, const char *key)
621 Xapian::Document doc;
623 notmuch_private_status_t status;
627 db_key = timestamp_db_key (key);
630 status = find_timestamp_document (notmuch, db_key, &doc, &doc_id);
632 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
635 ret = Xapian::sortable_unserialise (doc.get_value (NOTMUCH_VALUE_TIMESTAMP));
636 } catch (Xapian::Error &error) {
647 /* Find the thread ID to which the message with 'message_id' belongs.
649 * Returns NULL if no message with message ID 'message_id' is in the
652 * Otherwise, returns a newly talloced string belonging to 'ctx'.
655 _resolve_message_id_to_thread_id (notmuch_database_t *notmuch,
657 const char *message_id)
659 notmuch_message_t *message;
660 const char *ret = NULL;
662 message = notmuch_database_find_message (notmuch, message_id);
666 ret = talloc_steal (ctx, notmuch_message_get_thread_id (message));
670 notmuch_message_destroy (message);
675 static notmuch_status_t
676 _merge_threads (notmuch_database_t *notmuch,
677 const char *winner_thread_id,
678 const char *loser_thread_id)
680 Xapian::PostingIterator loser, loser_end;
681 notmuch_message_t *message = NULL;
682 notmuch_private_status_t private_status;
683 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
685 find_doc_ids (notmuch, "thread", loser_thread_id, &loser, &loser_end);
687 for ( ; loser != loser_end; loser++) {
688 message = _notmuch_message_create (notmuch, notmuch,
689 *loser, &private_status);
690 if (message == NULL) {
691 ret = COERCE_STATUS (private_status,
692 "Cannot find document for doc_id from query");
696 _notmuch_message_remove_term (message, "thread", loser_thread_id);
697 _notmuch_message_add_term (message, "thread", winner_thread_id);
698 _notmuch_message_sync (message);
700 notmuch_message_destroy (message);
706 notmuch_message_destroy (message);
712 _my_talloc_free_for_g_hash (void *ptr)
717 static notmuch_status_t
718 _notmuch_database_link_message_to_parents (notmuch_database_t *notmuch,
719 notmuch_message_t *message,
720 notmuch_message_file_t *message_file,
721 const char **thread_id)
723 GHashTable *parents = NULL;
724 const char *refs, *in_reply_to;
725 GList *l, *keys = NULL;
726 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
728 parents = g_hash_table_new_full (g_str_hash, g_str_equal,
729 _my_talloc_free_for_g_hash, NULL);
731 refs = notmuch_message_file_get_header (message_file, "references");
732 parse_references (message, parents, refs);
734 in_reply_to = notmuch_message_file_get_header (message_file, "in-reply-to");
735 parse_references (message, parents, in_reply_to);
736 _notmuch_message_add_term (message, "replyto",
737 parse_message_id (message, in_reply_to, NULL));
739 keys = g_hash_table_get_keys (parents);
740 for (l = keys; l; l = l->next) {
741 char *parent_message_id;
742 const char *parent_thread_id;
744 parent_message_id = (char *) l->data;
745 parent_thread_id = _resolve_message_id_to_thread_id (notmuch,
749 if (parent_thread_id == NULL) {
750 _notmuch_message_add_term (message, "ref", parent_message_id);
752 if (*thread_id == NULL) {
753 *thread_id = talloc_strdup (message, parent_thread_id);
754 _notmuch_message_add_term (message, "thread", *thread_id);
755 } else if (strcmp (*thread_id, parent_thread_id)) {
756 ret = _merge_threads (notmuch, *thread_id, parent_thread_id);
767 g_hash_table_unref (parents);
772 static notmuch_status_t
773 _notmuch_database_link_message_to_children (notmuch_database_t *notmuch,
774 notmuch_message_t *message,
775 const char **thread_id)
777 const char *message_id = notmuch_message_get_message_id (message);
778 Xapian::PostingIterator child, children_end;
779 notmuch_message_t *child_message = NULL;
780 const char *child_thread_id;
781 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
782 notmuch_private_status_t private_status;
784 find_doc_ids (notmuch, "ref", message_id, &child, &children_end);
786 for ( ; child != children_end; child++) {
788 child_message = _notmuch_message_create (message, notmuch,
789 *child, &private_status);
790 if (child_message == NULL) {
791 ret = COERCE_STATUS (private_status,
792 "Cannot find document for doc_id from query");
796 child_thread_id = notmuch_message_get_thread_id (child_message);
797 if (*thread_id == NULL) {
798 *thread_id = talloc_strdup (message, child_thread_id);
799 _notmuch_message_add_term (message, "thread", *thread_id);
800 } else if (strcmp (*thread_id, child_thread_id)) {
801 _notmuch_message_remove_term (child_message, "ref",
803 _notmuch_message_sync (child_message);
804 ret = _merge_threads (notmuch, *thread_id, child_thread_id);
809 notmuch_message_destroy (child_message);
810 child_message = NULL;
815 notmuch_message_destroy (child_message);
820 /* Given a (mostly empty) 'message' and its corresponding
821 * 'message_file' link it to existing threads in the database.
823 * We first looke at 'message_file' and its link-relevant headers
824 * (References and In-Reply-To) for message IDs. We also look in the
825 * database for existing message that reference 'message'.p
827 * The end result is to call _notmuch_message_add_thread_id with one
828 * or more thread IDs to which this message belongs, (including
829 * generating a new thread ID if necessary if the message doesn't
830 * connect to any existing threads).
832 static notmuch_status_t
833 _notmuch_database_link_message (notmuch_database_t *notmuch,
834 notmuch_message_t *message,
835 notmuch_message_file_t *message_file)
837 notmuch_status_t status;
838 const char *thread_id = NULL;
840 status = _notmuch_database_link_message_to_parents (notmuch, message,
846 status = _notmuch_database_link_message_to_children (notmuch, message,
851 if (thread_id == NULL)
852 _notmuch_message_ensure_thread_id (message);
854 return NOTMUCH_STATUS_SUCCESS;
858 notmuch_database_add_message (notmuch_database_t *notmuch,
859 const char *filename,
860 notmuch_message_t **message_ret)
862 notmuch_message_file_t *message_file;
863 notmuch_message_t *message;
864 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
865 notmuch_private_status_t private_status;
867 const char *date, *header;
868 const char *from, *to, *subject;
874 message_file = notmuch_message_file_open (filename);
875 if (message_file == NULL) {
876 ret = NOTMUCH_STATUS_FILE_ERROR;
880 notmuch_message_file_restrict_headers (message_file,
891 /* Before we do any real work, (especially before doing a
892 * potential SHA-1 computation on the entire file's contents),
893 * let's make sure that what we're looking at looks like an
894 * actual email message.
896 from = notmuch_message_file_get_header (message_file, "from");
897 subject = notmuch_message_file_get_header (message_file, "subject");
898 to = notmuch_message_file_get_header (message_file, "to");
904 ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
908 /* Now that we're sure it's mail, the first order of business
909 * is to find a message ID (or else create one ourselves). */
911 header = notmuch_message_file_get_header (message_file, "message-id");
913 message_id = parse_message_id (message_file, header, NULL);
914 /* So the header value isn't RFC-compliant, but it's
915 * better than no message-id at all. */
916 if (message_id == NULL)
917 message_id = talloc_strdup (message_file, header);
919 /* No message-id at all, let's generate one by taking a
920 * hash over the file's contents. */
921 char *sha1 = notmuch_sha1_of_file (filename);
923 /* If that failed too, something is really wrong. Give up. */
925 ret = NOTMUCH_STATUS_FILE_ERROR;
929 message_id = talloc_asprintf (message_file,
930 "notmuch-sha1-%s", sha1);
934 /* Now that we have a message ID, we get a message object,
935 * (which may or may not reference an existing document in the
938 /* Use NULL for owner since we want to free this locally. */
939 message = _notmuch_message_create_for_message_id (NULL,
944 talloc_free (message_id);
949 /* Is this a newly created message object? */
950 if (private_status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
951 _notmuch_message_set_filename (message, filename);
952 _notmuch_message_add_term (message, "type", "mail");
954 ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
958 ret = _notmuch_database_link_message (notmuch, message, message_file);
962 date = notmuch_message_file_get_header (message_file, "date");
963 _notmuch_message_set_date (message, date);
965 _notmuch_message_index_file (message, filename);
967 _notmuch_message_sync (message);
968 } catch (const Xapian::Error &error) {
969 fprintf (stderr, "A Xapian exception occurred: %s.\n",
970 error.get_msg().c_str());
971 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
977 if (ret == NOTMUCH_STATUS_SUCCESS && message_ret)
978 *message_ret = message;
980 notmuch_message_destroy (message);
984 notmuch_message_file_close (message_file);