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 directory.
44 * A mail document is associated with a particular email message file
45 * on disk. It is indexed with the following prefixed terms which the
46 * database uses to construct threads, etc.:
48 * Single terms of given prefix:
52 * id: Unique ID of mail, (from Message-ID header or generated
53 * as "notmuch-sha1-<sha1_sum_of_entire_file>.
55 * thread: The ID of the thread to which the mail belongs
57 * replyto: The ID from the In-Reply-To header of the mail (if any).
59 * Multiple terms of given prefix:
61 * reference: All message IDs from In-Reply-To and Re ferences
62 * headers in the message.
64 * tag: Any tags associated with this message by the user.
66 * direntry: A colon-separated pair of values (INTEGER:STRING),
67 * where INTEGER is the document ID of a directory
68 * document, and STRING is the name of a file within
69 * that directory for this mail message.
71 * A mail document also has two values:
73 * TIMESTAMP: The time_t value corresponding to the message's
76 * MESSAGE_ID: The unique ID of the mail mess (see "id" above)
78 * In addition, terms from the content of the message are added with
79 * "from", "to", "attachment", and "subject" prefixes for use by the
80 * user in searching. But the database doesn't really care itself
83 * The data portion of a mail document is empty.
87 * A directory document is used by a client of the notmuch library to
88 * maintain data necessary to allow for efficient polling of mail
91 * The directory document contains the following terms:
93 * directory: The directory path (relative to the database path)
94 * Or the SHA1 sum of the directory path (if the
95 * path itself is too long to fit in a Xapian
98 * parent: The document ID of the parent directory document.
99 * Top-level directories will have a parent value of 0.
101 * and has a single value:
103 * TIMESTAMP: The mtime of the directory (at last scan)
105 * The data portion of a directory document contains the path of the
106 * directory (relative to the datbase path).
109 /* With these prefix values we follow the conventions published here:
111 * http://xapian.org/docs/omega/termprefixes.html
113 * as much as makes sense. Note that I took some liberty in matching
114 * the reserved prefix values to notmuch concepts, (for example, 'G'
115 * is documented as "newsGroup (or similar entity - e.g. a web forum
116 * name)", for which I think the thread is the closest analogue in
117 * notmuch. This in spite of the fact that we will eventually be
118 * storing mailing-list messages where 'G' for "mailing list name"
119 * might be even a closer analogue. I'm treating the single-character
120 * prefixes preferentially for core notmuch concepts (which will be
121 * nearly universal to all mail messages).
124 prefix_t BOOLEAN_PREFIX_INTERNAL[] = {
126 { "reference", "XREFERENCE" },
127 { "replyto", "XREPLYTO" },
128 { "directory", "XDIRECTORY" },
129 { "direntry", "XDIRENTRY" },
130 { "parent", "XPARENT" },
133 prefix_t BOOLEAN_PREFIX_EXTERNAL[] = {
139 prefix_t PROBABILISTIC_PREFIX[]= {
142 { "attachment", "XATTACHMENT" },
143 { "subject", "XSUBJECT"}
147 _internal_error (const char *format, ...)
151 va_start (va_args, format);
153 fprintf (stderr, "Internal error: ");
154 vfprintf (stderr, format, va_args);
162 _find_prefix (const char *name)
166 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_INTERNAL); i++) {
167 if (strcmp (name, BOOLEAN_PREFIX_INTERNAL[i].name) == 0)
168 return BOOLEAN_PREFIX_INTERNAL[i].prefix;
171 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
172 if (strcmp (name, BOOLEAN_PREFIX_EXTERNAL[i].name) == 0)
173 return BOOLEAN_PREFIX_EXTERNAL[i].prefix;
176 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
177 if (strcmp (name, PROBABILISTIC_PREFIX[i].name) == 0)
178 return PROBABILISTIC_PREFIX[i].prefix;
181 INTERNAL_ERROR ("No prefix exists for '%s'\n", name);
187 notmuch_status_to_string (notmuch_status_t status)
190 case NOTMUCH_STATUS_SUCCESS:
191 return "No error occurred";
192 case NOTMUCH_STATUS_OUT_OF_MEMORY:
193 return "Out of memory";
194 case NOTMUCH_STATUS_READONLY_DATABASE:
195 return "The database is read-only";
196 case NOTMUCH_STATUS_XAPIAN_EXCEPTION:
197 return "A Xapian exception occurred";
198 case NOTMUCH_STATUS_FILE_ERROR:
199 return "Something went wrong trying to read or write a file";
200 case NOTMUCH_STATUS_FILE_NOT_EMAIL:
201 return "File is not an email";
202 case NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID:
203 return "Message ID is identical to a message in database";
204 case NOTMUCH_STATUS_NULL_POINTER:
205 return "Erroneous NULL pointer";
206 case NOTMUCH_STATUS_TAG_TOO_LONG:
207 return "Tag value is too long (exceeds NOTMUCH_TAG_MAX)";
208 case NOTMUCH_STATUS_UNBALANCED_FREEZE_THAW:
209 return "Unbalanced number of calls to notmuch_message_freeze/thaw";
211 case NOTMUCH_STATUS_LAST_STATUS:
212 return "Unknown error status value";
217 find_doc_ids (notmuch_database_t *notmuch,
218 const char *prefix_name,
220 Xapian::PostingIterator *begin,
221 Xapian::PostingIterator *end)
223 Xapian::PostingIterator i;
226 term = talloc_asprintf (notmuch, "%s%s",
227 _find_prefix (prefix_name), value);
229 *begin = notmuch->xapian_db->postlist_begin (term);
231 *end = notmuch->xapian_db->postlist_end (term);
236 static notmuch_private_status_t
237 find_unique_doc_id (notmuch_database_t *notmuch,
238 const char *prefix_name,
240 unsigned int *doc_id)
242 Xapian::PostingIterator i, end;
244 find_doc_ids (notmuch, prefix_name, value, &i, &end);
248 return NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND;
251 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
255 static Xapian::Document
256 find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id)
258 return notmuch->xapian_db->get_document (doc_id);
261 static notmuch_private_status_t
262 find_unique_document (notmuch_database_t *notmuch,
263 const char *prefix_name,
265 Xapian::Document *document,
266 unsigned int *doc_id)
268 notmuch_private_status_t status;
270 status = find_unique_doc_id (notmuch, prefix_name, value, doc_id);
273 *document = Xapian::Document ();
277 *document = find_document_for_doc_id (notmuch, *doc_id);
278 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
282 notmuch_database_find_message (notmuch_database_t *notmuch,
283 const char *message_id)
285 notmuch_private_status_t status;
288 status = find_unique_doc_id (notmuch, "id", message_id, &doc_id);
290 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
293 return _notmuch_message_create (notmuch, notmuch, doc_id, NULL);
296 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
297 * a (potentially nested) parenthesized sequence with '\' used to
298 * escape any character (including parentheses).
300 * If the sequence to be skipped continues to the end of the string,
301 * then 'str' will be left pointing at the final terminating '\0'
305 skip_space_and_comments (const char **str)
310 while (*s && (isspace (*s) || *s == '(')) {
311 while (*s && isspace (*s))
316 while (*s && nesting) {
319 } else if (*s == ')') {
321 } else if (*s == '\\') {
333 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
334 * comments, and the '<' and '>' delimeters.
336 * If not NULL, then *next will be made to point to the first character
337 * not parsed, (possibly pointing to the final '\0' terminator.
339 * Returns a newly talloc'ed string belonging to 'ctx'.
341 * Returns NULL if there is any error parsing the message-id. */
343 _parse_message_id (void *ctx, const char *message_id, const char **next)
348 if (message_id == NULL || *message_id == '\0')
353 skip_space_and_comments (&s);
355 /* Skip any unstructured text as well. */
356 while (*s && *s != '<')
367 skip_space_and_comments (&s);
370 while (*end && *end != '>')
379 if (end > s && *end == '>')
384 result = talloc_strndup (ctx, s, end - s + 1);
386 /* Finally, collapse any whitespace that is within the message-id
392 for (r = result, len = strlen (r); *r; r++, len--)
393 if (*r == ' ' || *r == '\t')
394 memmove (r, r+1, len);
400 /* Parse a References header value, putting a (talloc'ed under 'ctx')
401 * copy of each referenced message-id into 'hash'.
403 * We explicitly avoid including any reference identical to
404 * 'message_id' in the result (to avoid mass confusion when a single
405 * message references itself cyclically---and yes, mail messages are
406 * not infrequent in the wild that do this---don't ask me why).
409 parse_references (void *ctx,
410 const char *message_id,
416 if (refs == NULL || *refs == '\0')
420 ref = _parse_message_id (ctx, refs, &refs);
422 if (ref && strcmp (ref, message_id))
423 g_hash_table_insert (hash, ref, NULL);
428 notmuch_database_create (const char *path)
430 notmuch_database_t *notmuch = NULL;
431 char *notmuch_path = NULL;
436 fprintf (stderr, "Error: Cannot create a database for a NULL path.\n");
440 err = stat (path, &st);
442 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
443 path, strerror (errno));
447 if (! S_ISDIR (st.st_mode)) {
448 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
453 notmuch_path = talloc_asprintf (NULL, "%s/%s", path, ".notmuch");
455 err = mkdir (notmuch_path, 0755);
458 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
459 notmuch_path, strerror (errno));
463 notmuch = notmuch_database_open (path,
464 NOTMUCH_DATABASE_MODE_READ_WRITE);
468 talloc_free (notmuch_path);
474 notmuch_database_open (const char *path,
475 notmuch_database_mode_t mode)
477 notmuch_database_t *notmuch = NULL;
478 char *notmuch_path = NULL, *xapian_path = NULL;
483 if (asprintf (¬much_path, "%s/%s", path, ".notmuch") == -1) {
485 fprintf (stderr, "Out of memory\n");
489 err = stat (notmuch_path, &st);
491 fprintf (stderr, "Error opening database at %s: %s\n",
492 notmuch_path, strerror (errno));
496 if (asprintf (&xapian_path, "%s/%s", notmuch_path, "xapian") == -1) {
498 fprintf (stderr, "Out of memory\n");
502 notmuch = talloc (NULL, notmuch_database_t);
503 notmuch->exception_reported = FALSE;
504 notmuch->path = talloc_strdup (notmuch, path);
506 if (notmuch->path[strlen (notmuch->path) - 1] == '/')
507 notmuch->path[strlen (notmuch->path) - 1] = '\0';
509 notmuch->mode = mode;
511 if (mode == NOTMUCH_DATABASE_MODE_READ_WRITE) {
512 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
513 Xapian::DB_CREATE_OR_OPEN);
515 notmuch->xapian_db = new Xapian::Database (xapian_path);
517 notmuch->query_parser = new Xapian::QueryParser;
518 notmuch->term_gen = new Xapian::TermGenerator;
519 notmuch->term_gen->set_stemmer (Xapian::Stem ("english"));
520 notmuch->value_range_processor = new Xapian::NumberValueRangeProcessor (NOTMUCH_VALUE_TIMESTAMP);
522 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
523 notmuch->query_parser->set_database (*notmuch->xapian_db);
524 notmuch->query_parser->set_stemmer (Xapian::Stem ("english"));
525 notmuch->query_parser->set_stemming_strategy (Xapian::QueryParser::STEM_SOME);
526 notmuch->query_parser->add_valuerangeprocessor (notmuch->value_range_processor);
528 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
529 prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i];
530 notmuch->query_parser->add_boolean_prefix (prefix->name,
534 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
535 prefix_t *prefix = &PROBABILISTIC_PREFIX[i];
536 notmuch->query_parser->add_prefix (prefix->name, prefix->prefix);
538 } catch (const Xapian::Error &error) {
539 fprintf (stderr, "A Xapian exception occurred opening database: %s\n",
540 error.get_msg().c_str());
554 notmuch_database_close (notmuch_database_t *notmuch)
557 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_WRITE)
558 (static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db))->flush ();
559 } catch (const Xapian::Error &error) {
560 if (! notmuch->exception_reported) {
561 fprintf (stderr, "Error: A Xapian exception occurred flushing database: %s\n",
562 error.get_msg().c_str());
566 delete notmuch->term_gen;
567 delete notmuch->query_parser;
568 delete notmuch->xapian_db;
569 delete notmuch->value_range_processor;
570 talloc_free (notmuch);
574 notmuch_database_get_path (notmuch_database_t *notmuch)
576 return notmuch->path;
579 static notmuch_private_status_t
580 find_directory_document (notmuch_database_t *notmuch, const char *db_path,
581 Xapian::Document *doc, unsigned int *doc_id)
583 return find_unique_document (notmuch, "directory", db_path, doc, doc_id);
586 /* We allow the user to use arbitrarily long paths for directories. But
587 * we have a term-length limit. So if we exceed that, we'll use the
588 * SHA-1 of the path for the database term.
590 * Note: This function may return the original value of 'path'. If it
591 * does not, then the caller is responsible to free() the returned
595 directory_db_path (const char *path)
597 int term_len = strlen (_find_prefix ("directory")) + strlen (path);
599 if (term_len > NOTMUCH_TERM_MAX)
600 return notmuch_sha1_of_string (path);
605 /* Given a path, split it into two parts: the directory part is all
606 * components except for the last, and the basename is that last
607 * component. Getting the return-value for either part is optional
608 * (the caller can pass NULL).
610 * The original 'path' can represent either a regular file or a
611 * directory---the splitting will be carried out in the same way in
612 * either case. Trailing slashes on 'path' will be ignored, and any
613 * cases of multiple '/' characters appearing in series will be
614 * treated as a single '/'.
616 * Allocation (if any) will have 'ctx' as the talloc owner. But
617 * pointers will be returned within the original path string whenever
620 * Note: If 'path' is non-empty and contains no non-trailing slash,
621 * (that is, consists of a filename with no parent directory), then
622 * the directory returned will be an empty string. However, if 'path'
623 * is an empty string, then both directory and basename will be
627 _notmuch_database_split_path (void *ctx,
629 const char **directory,
630 const char **basename)
634 if (path == NULL || *path == '\0') {
639 return NOTMUCH_STATUS_SUCCESS;
642 /* Find the last slash (not counting a trailing slash), if any. */
644 slash = path + strlen (path) - 1;
646 /* First, skip trailing slashes. */
647 while (slash != path) {
654 /* Then, find a slash. */
655 while (slash != path) {
665 /* Finally, skip multiple slashes. */
666 while (slash != path) {
675 *directory = talloc_strdup (ctx, "");
680 *directory = talloc_strndup (ctx, path, slash - path + 1);
683 return NOTMUCH_STATUS_SUCCESS;
687 _notmuch_database_find_directory_id (notmuch_database_t *notmuch,
689 unsigned int *directory_id)
691 notmuch_private_status_t private_status;
692 notmuch_status_t status = NOTMUCH_STATUS_SUCCESS;
697 return NOTMUCH_STATUS_SUCCESS;
700 db_path = directory_db_path (path);
702 private_status = find_unique_doc_id (notmuch, "directory",
703 db_path, directory_id);
704 if (private_status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
705 status = notmuch_database_set_directory_mtime (notmuch,
710 private_status = find_unique_doc_id (notmuch, "directory",
711 db_path, directory_id);
712 status = COERCE_STATUS (private_status, "_find_directory_id");
717 free ((char *) db_path);
726 _notmuch_database_get_directory_path (void *ctx,
727 notmuch_database_t *notmuch,
730 Xapian::Document document;
732 document = find_document_for_doc_id (notmuch, doc_id);
734 return talloc_strdup (ctx, document.get_data ().c_str ());
737 /* Given a legal 'filename' for the database, (either relative to
738 * database path or absolute with initial components identical to
739 * database path), return a new string (with 'ctx' as the talloc
740 * owner) suitable for use as a direntry term value.
743 _notmuch_database_filename_to_direntry (void *ctx,
744 notmuch_database_t *notmuch,
745 const char *filename,
748 const char *relative, *directory, *basename;
749 Xapian::docid directory_id;
750 notmuch_status_t status;
752 relative = _notmuch_database_relative_path (notmuch, filename);
754 status = _notmuch_database_split_path (ctx, relative,
755 &directory, &basename);
759 status = _notmuch_database_find_directory_id (notmuch, directory,
764 *direntry = talloc_asprintf (ctx, "%u:%s", directory_id, basename);
766 return NOTMUCH_STATUS_SUCCESS;
769 /* Given a legal 'path' for the database, return the relative path.
771 * The return value will be a pointer to the originl path contents,
772 * and will be either the original string (if 'path' was relative) or
773 * a portion of the string (if path was absolute and begins with the
777 _notmuch_database_relative_path (notmuch_database_t *notmuch,
780 const char *db_path, *relative;
781 unsigned int db_path_len;
783 db_path = notmuch_database_get_path (notmuch);
784 db_path_len = strlen (db_path);
788 if (*relative == '/') {
789 while (*relative == '/' && *(relative+1) == '/')
792 if (strncmp (relative, db_path, db_path_len) == 0)
794 relative += db_path_len;
795 while (*relative == '/')
804 notmuch_database_set_directory_mtime (notmuch_database_t *notmuch,
808 Xapian::Document doc;
809 Xapian::WritableDatabase *db;
811 notmuch_private_status_t status;
812 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
813 const char *parent, *db_path = NULL;
814 unsigned int parent_id;
815 void *local = talloc_new (notmuch);
817 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY) {
818 fprintf (stderr, "Attempted to update a read-only database.\n");
819 return NOTMUCH_STATUS_READONLY_DATABASE;
822 path = _notmuch_database_relative_path (notmuch, path);
824 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
825 db_path = directory_db_path (path);
828 status = find_directory_document (notmuch, db_path, &doc, &doc_id);
830 doc.add_value (NOTMUCH_VALUE_TIMESTAMP,
831 Xapian::sortable_serialise (mtime));
833 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
834 char *term = talloc_asprintf (local, "%s%s",
835 _find_prefix ("directory"), db_path);
840 _notmuch_database_split_path (local, path, &parent, NULL);
842 _notmuch_database_find_directory_id (notmuch, parent, &parent_id);
844 term = talloc_asprintf (local, "%s%u",
845 _find_prefix ("parent"),
849 db->add_document (doc);
851 db->replace_document (doc_id, doc);
854 } catch (const Xapian::Error &error) {
855 fprintf (stderr, "A Xapian exception occurred setting directory mtime: %s.\n",
856 error.get_msg().c_str());
857 notmuch->exception_reported = TRUE;
858 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
862 free ((char *) db_path);
870 notmuch_database_get_directory_mtime (notmuch_database_t *notmuch,
873 Xapian::Document doc;
875 notmuch_private_status_t status;
876 const char *db_path = NULL;
879 db_path = directory_db_path (path);
882 status = find_directory_document (notmuch, db_path, &doc, &doc_id);
884 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
887 ret = Xapian::sortable_unserialise (doc.get_value (NOTMUCH_VALUE_TIMESTAMP));
888 } catch (Xapian::Error &error) {
895 free ((char *) db_path);
900 /* Find the thread ID to which the message with 'message_id' belongs.
902 * Returns NULL if no message with message ID 'message_id' is in the
905 * Otherwise, returns a newly talloced string belonging to 'ctx'.
908 _resolve_message_id_to_thread_id (notmuch_database_t *notmuch,
910 const char *message_id)
912 notmuch_message_t *message;
913 const char *ret = NULL;
915 message = notmuch_database_find_message (notmuch, message_id);
919 ret = talloc_steal (ctx, notmuch_message_get_thread_id (message));
923 notmuch_message_destroy (message);
928 static notmuch_status_t
929 _merge_threads (notmuch_database_t *notmuch,
930 const char *winner_thread_id,
931 const char *loser_thread_id)
933 Xapian::PostingIterator loser, loser_end;
934 notmuch_message_t *message = NULL;
935 notmuch_private_status_t private_status;
936 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
938 find_doc_ids (notmuch, "thread", loser_thread_id, &loser, &loser_end);
940 for ( ; loser != loser_end; loser++) {
941 message = _notmuch_message_create (notmuch, notmuch,
942 *loser, &private_status);
943 if (message == NULL) {
944 ret = COERCE_STATUS (private_status,
945 "Cannot find document for doc_id from query");
949 _notmuch_message_remove_term (message, "thread", loser_thread_id);
950 _notmuch_message_add_term (message, "thread", winner_thread_id);
951 _notmuch_message_sync (message);
953 notmuch_message_destroy (message);
959 notmuch_message_destroy (message);
965 _my_talloc_free_for_g_hash (void *ptr)
970 static notmuch_status_t
971 _notmuch_database_link_message_to_parents (notmuch_database_t *notmuch,
972 notmuch_message_t *message,
973 notmuch_message_file_t *message_file,
974 const char **thread_id)
976 GHashTable *parents = NULL;
977 const char *refs, *in_reply_to, *in_reply_to_message_id;
978 GList *l, *keys = NULL;
979 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
981 parents = g_hash_table_new_full (g_str_hash, g_str_equal,
982 _my_talloc_free_for_g_hash, NULL);
984 refs = notmuch_message_file_get_header (message_file, "references");
985 parse_references (message, notmuch_message_get_message_id (message),
988 in_reply_to = notmuch_message_file_get_header (message_file, "in-reply-to");
989 parse_references (message, notmuch_message_get_message_id (message),
990 parents, in_reply_to);
992 /* Carefully avoid adding any self-referential in-reply-to term. */
993 in_reply_to_message_id = _parse_message_id (message, in_reply_to, NULL);
994 if (in_reply_to_message_id &&
995 strcmp (in_reply_to_message_id,
996 notmuch_message_get_message_id (message)))
998 _notmuch_message_add_term (message, "replyto",
999 _parse_message_id (message, in_reply_to, NULL));
1002 keys = g_hash_table_get_keys (parents);
1003 for (l = keys; l; l = l->next) {
1004 char *parent_message_id;
1005 const char *parent_thread_id;
1007 parent_message_id = (char *) l->data;
1008 parent_thread_id = _resolve_message_id_to_thread_id (notmuch,
1012 if (parent_thread_id == NULL) {
1013 _notmuch_message_add_term (message, "reference",
1016 if (*thread_id == NULL) {
1017 *thread_id = talloc_strdup (message, parent_thread_id);
1018 _notmuch_message_add_term (message, "thread", *thread_id);
1019 } else if (strcmp (*thread_id, parent_thread_id)) {
1020 ret = _merge_threads (notmuch, *thread_id, parent_thread_id);
1031 g_hash_table_unref (parents);
1036 static notmuch_status_t
1037 _notmuch_database_link_message_to_children (notmuch_database_t *notmuch,
1038 notmuch_message_t *message,
1039 const char **thread_id)
1041 const char *message_id = notmuch_message_get_message_id (message);
1042 Xapian::PostingIterator child, children_end;
1043 notmuch_message_t *child_message = NULL;
1044 const char *child_thread_id;
1045 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1046 notmuch_private_status_t private_status;
1048 find_doc_ids (notmuch, "reference", message_id, &child, &children_end);
1050 for ( ; child != children_end; child++) {
1052 child_message = _notmuch_message_create (message, notmuch,
1053 *child, &private_status);
1054 if (child_message == NULL) {
1055 ret = COERCE_STATUS (private_status,
1056 "Cannot find document for doc_id from query");
1060 child_thread_id = notmuch_message_get_thread_id (child_message);
1061 if (*thread_id == NULL) {
1062 *thread_id = talloc_strdup (message, child_thread_id);
1063 _notmuch_message_add_term (message, "thread", *thread_id);
1064 } else if (strcmp (*thread_id, child_thread_id)) {
1065 _notmuch_message_remove_term (child_message, "reference",
1067 _notmuch_message_sync (child_message);
1068 ret = _merge_threads (notmuch, *thread_id, child_thread_id);
1073 notmuch_message_destroy (child_message);
1074 child_message = NULL;
1079 notmuch_message_destroy (child_message);
1084 /* Given a (mostly empty) 'message' and its corresponding
1085 * 'message_file' link it to existing threads in the database.
1087 * We first look at 'message_file' and its link-relevant headers
1088 * (References and In-Reply-To) for message IDs. We also look in the
1089 * database for existing message that reference 'message'.
1091 * The end result is to call _notmuch_message_ensure_thread_id which
1092 * generates a new thread ID if the message doesn't connect to any
1095 static notmuch_status_t
1096 _notmuch_database_link_message (notmuch_database_t *notmuch,
1097 notmuch_message_t *message,
1098 notmuch_message_file_t *message_file)
1100 notmuch_status_t status;
1101 const char *thread_id = NULL;
1103 status = _notmuch_database_link_message_to_parents (notmuch, message,
1109 status = _notmuch_database_link_message_to_children (notmuch, message,
1114 if (thread_id == NULL)
1115 _notmuch_message_ensure_thread_id (message);
1117 return NOTMUCH_STATUS_SUCCESS;
1121 notmuch_database_add_message (notmuch_database_t *notmuch,
1122 const char *filename,
1123 notmuch_message_t **message_ret)
1125 notmuch_message_file_t *message_file;
1126 notmuch_message_t *message = NULL;
1127 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1128 notmuch_private_status_t private_status;
1130 const char *date, *header;
1131 const char *from, *to, *subject;
1132 char *message_id = NULL;
1135 *message_ret = NULL;
1137 message_file = notmuch_message_file_open (filename);
1138 if (message_file == NULL) {
1139 ret = NOTMUCH_STATUS_FILE_ERROR;
1143 notmuch_message_file_restrict_headers (message_file,
1154 /* Before we do any real work, (especially before doing a
1155 * potential SHA-1 computation on the entire file's contents),
1156 * let's make sure that what we're looking at looks like an
1157 * actual email message.
1159 from = notmuch_message_file_get_header (message_file, "from");
1160 subject = notmuch_message_file_get_header (message_file, "subject");
1161 to = notmuch_message_file_get_header (message_file, "to");
1163 if ((from == NULL || *from == '\0') &&
1164 (subject == NULL || *subject == '\0') &&
1165 (to == NULL || *to == '\0'))
1167 ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
1171 /* Now that we're sure it's mail, the first order of business
1172 * is to find a message ID (or else create one ourselves). */
1174 header = notmuch_message_file_get_header (message_file, "message-id");
1175 if (header && *header != '\0') {
1176 message_id = _parse_message_id (message_file, header, NULL);
1178 /* So the header value isn't RFC-compliant, but it's
1179 * better than no message-id at all. */
1180 if (message_id == NULL)
1181 message_id = talloc_strdup (message_file, header);
1183 /* Reject a Message ID that's too long. */
1184 if (message_id && strlen (message_id) + 1 > NOTMUCH_TERM_MAX) {
1185 talloc_free (message_id);
1190 if (message_id == NULL ) {
1191 /* No message-id at all, let's generate one by taking a
1192 * hash over the file's contents. */
1193 char *sha1 = notmuch_sha1_of_file (filename);
1195 /* If that failed too, something is really wrong. Give up. */
1197 ret = NOTMUCH_STATUS_FILE_ERROR;
1201 message_id = talloc_asprintf (message_file,
1202 "notmuch-sha1-%s", sha1);
1206 /* Now that we have a message ID, we get a message object,
1207 * (which may or may not reference an existing document in the
1210 message = _notmuch_message_create_for_message_id (notmuch,
1214 talloc_free (message_id);
1216 if (message == NULL) {
1217 ret = COERCE_STATUS (private_status,
1218 "Unexpected status value from _notmuch_message_create_for_message_id");
1222 _notmuch_message_add_filename (message, filename);
1224 /* Is this a newly created message object? */
1225 if (private_status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
1226 _notmuch_message_add_term (message, "type", "mail");
1228 ret = _notmuch_database_link_message (notmuch, message,
1233 date = notmuch_message_file_get_header (message_file, "date");
1234 _notmuch_message_set_date (message, date);
1236 _notmuch_message_index_file (message, filename);
1238 ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
1241 _notmuch_message_sync (message);
1242 } catch (const Xapian::Error &error) {
1243 fprintf (stderr, "A Xapian exception occurred adding message: %s.\n",
1244 error.get_description().c_str());
1245 notmuch->exception_reported = TRUE;
1246 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
1252 if (ret == NOTMUCH_STATUS_SUCCESS && message_ret)
1253 *message_ret = message;
1255 notmuch_message_destroy (message);
1259 notmuch_message_file_close (message_file);
1265 _notmuch_convert_tags (void *ctx, Xapian::TermIterator &i,
1266 Xapian::TermIterator &end)
1268 const char *prefix = _find_prefix ("tag");
1269 notmuch_tags_t *tags;
1272 /* Currently this iteration is written with the assumption that
1273 * "tag" has a single-character prefix. */
1274 assert (strlen (prefix) == 1);
1276 tags = _notmuch_tags_create (ctx);
1277 if (unlikely (tags == NULL))
1285 if (tag.empty () || tag[0] != *prefix)
1288 _notmuch_tags_add_tag (tags, tag.c_str () + 1);
1293 _notmuch_tags_prepare_iterator (tags);
1299 notmuch_database_get_all_tags (notmuch_database_t *db)
1301 Xapian::TermIterator i, end;
1302 i = db->xapian_db->allterms_begin();
1303 end = db->xapian_db->allterms_end();
1304 return _notmuch_convert_tags(db, i, end);