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;
253 #if DEBUG_DATABASE_SANITY
257 INTERNAL_ERROR ("Term %s:%s is not unique as expected.\n",
261 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
264 static Xapian::Document
265 find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id)
267 return notmuch->xapian_db->get_document (doc_id);
270 static notmuch_private_status_t
271 find_unique_document (notmuch_database_t *notmuch,
272 const char *prefix_name,
274 Xapian::Document *document,
275 unsigned int *doc_id)
277 notmuch_private_status_t status;
279 status = find_unique_doc_id (notmuch, prefix_name, value, doc_id);
282 *document = Xapian::Document ();
286 *document = find_document_for_doc_id (notmuch, *doc_id);
287 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
291 notmuch_database_find_message (notmuch_database_t *notmuch,
292 const char *message_id)
294 notmuch_private_status_t status;
297 status = find_unique_doc_id (notmuch, "id", message_id, &doc_id);
299 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
302 return _notmuch_message_create (notmuch, notmuch, doc_id, NULL);
305 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
306 * a (potentially nested) parenthesized sequence with '\' used to
307 * escape any character (including parentheses).
309 * If the sequence to be skipped continues to the end of the string,
310 * then 'str' will be left pointing at the final terminating '\0'
314 skip_space_and_comments (const char **str)
319 while (*s && (isspace (*s) || *s == '(')) {
320 while (*s && isspace (*s))
325 while (*s && nesting) {
328 } else if (*s == ')') {
330 } else if (*s == '\\') {
342 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
343 * comments, and the '<' and '>' delimeters.
345 * If not NULL, then *next will be made to point to the first character
346 * not parsed, (possibly pointing to the final '\0' terminator.
348 * Returns a newly talloc'ed string belonging to 'ctx'.
350 * Returns NULL if there is any error parsing the message-id. */
352 _parse_message_id (void *ctx, const char *message_id, const char **next)
357 if (message_id == NULL || *message_id == '\0')
362 skip_space_and_comments (&s);
364 /* Skip any unstructured text as well. */
365 while (*s && *s != '<')
376 skip_space_and_comments (&s);
379 while (*end && *end != '>')
388 if (end > s && *end == '>')
393 result = talloc_strndup (ctx, s, end - s + 1);
395 /* Finally, collapse any whitespace that is within the message-id
401 for (r = result, len = strlen (r); *r; r++, len--)
402 if (*r == ' ' || *r == '\t')
403 memmove (r, r+1, len);
409 /* Parse a References header value, putting a (talloc'ed under 'ctx')
410 * copy of each referenced message-id into 'hash'.
412 * We explicitly avoid including any reference identical to
413 * 'message_id' in the result (to avoid mass confusion when a single
414 * message references itself cyclically---and yes, mail messages are
415 * not infrequent in the wild that do this---don't ask me why).
418 parse_references (void *ctx,
419 const char *message_id,
425 if (refs == NULL || *refs == '\0')
429 ref = _parse_message_id (ctx, refs, &refs);
431 if (ref && strcmp (ref, message_id))
432 g_hash_table_insert (hash, ref, NULL);
437 notmuch_database_create (const char *path)
439 notmuch_database_t *notmuch = NULL;
440 char *notmuch_path = NULL;
445 fprintf (stderr, "Error: Cannot create a database for a NULL path.\n");
449 err = stat (path, &st);
451 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
452 path, strerror (errno));
456 if (! S_ISDIR (st.st_mode)) {
457 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
462 notmuch_path = talloc_asprintf (NULL, "%s/%s", path, ".notmuch");
464 err = mkdir (notmuch_path, 0755);
467 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
468 notmuch_path, strerror (errno));
472 notmuch = notmuch_database_open (path,
473 NOTMUCH_DATABASE_MODE_READ_WRITE);
477 talloc_free (notmuch_path);
483 notmuch_database_open (const char *path,
484 notmuch_database_mode_t mode)
486 notmuch_database_t *notmuch = NULL;
487 char *notmuch_path = NULL, *xapian_path = NULL;
492 if (asprintf (¬much_path, "%s/%s", path, ".notmuch") == -1) {
494 fprintf (stderr, "Out of memory\n");
498 err = stat (notmuch_path, &st);
500 fprintf (stderr, "Error opening database at %s: %s\n",
501 notmuch_path, strerror (errno));
505 if (asprintf (&xapian_path, "%s/%s", notmuch_path, "xapian") == -1) {
507 fprintf (stderr, "Out of memory\n");
511 notmuch = talloc (NULL, notmuch_database_t);
512 notmuch->exception_reported = FALSE;
513 notmuch->path = talloc_strdup (notmuch, path);
515 if (notmuch->path[strlen (notmuch->path) - 1] == '/')
516 notmuch->path[strlen (notmuch->path) - 1] = '\0';
518 notmuch->mode = mode;
520 if (mode == NOTMUCH_DATABASE_MODE_READ_WRITE) {
521 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
522 Xapian::DB_CREATE_OR_OPEN);
524 notmuch->xapian_db = new Xapian::Database (xapian_path);
526 notmuch->query_parser = new Xapian::QueryParser;
527 notmuch->term_gen = new Xapian::TermGenerator;
528 notmuch->term_gen->set_stemmer (Xapian::Stem ("english"));
529 notmuch->value_range_processor = new Xapian::NumberValueRangeProcessor (NOTMUCH_VALUE_TIMESTAMP);
531 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
532 notmuch->query_parser->set_database (*notmuch->xapian_db);
533 notmuch->query_parser->set_stemmer (Xapian::Stem ("english"));
534 notmuch->query_parser->set_stemming_strategy (Xapian::QueryParser::STEM_SOME);
535 notmuch->query_parser->add_valuerangeprocessor (notmuch->value_range_processor);
537 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
538 prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i];
539 notmuch->query_parser->add_boolean_prefix (prefix->name,
543 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
544 prefix_t *prefix = &PROBABILISTIC_PREFIX[i];
545 notmuch->query_parser->add_prefix (prefix->name, prefix->prefix);
547 } catch (const Xapian::Error &error) {
548 fprintf (stderr, "A Xapian exception occurred opening database: %s\n",
549 error.get_msg().c_str());
563 notmuch_database_close (notmuch_database_t *notmuch)
566 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_WRITE)
567 (static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db))->flush ();
568 } catch (const Xapian::Error &error) {
569 if (! notmuch->exception_reported) {
570 fprintf (stderr, "Error: A Xapian exception occurred flushing database: %s\n",
571 error.get_msg().c_str());
575 delete notmuch->term_gen;
576 delete notmuch->query_parser;
577 delete notmuch->xapian_db;
578 delete notmuch->value_range_processor;
579 talloc_free (notmuch);
583 notmuch_database_get_path (notmuch_database_t *notmuch)
585 return notmuch->path;
588 static notmuch_private_status_t
589 find_directory_document (notmuch_database_t *notmuch, const char *db_path,
590 Xapian::Document *doc, unsigned int *doc_id)
592 return find_unique_document (notmuch, "directory", db_path, doc, doc_id);
595 /* We allow the user to use arbitrarily long paths for directories. But
596 * we have a term-length limit. So if we exceed that, we'll use the
597 * SHA-1 of the path for the database term.
599 * Note: This function may return the original value of 'path'. If it
600 * does not, then the caller is responsible to free() the returned
604 directory_db_path (const char *path)
606 int term_len = strlen (_find_prefix ("directory")) + strlen (path);
608 if (term_len > NOTMUCH_TERM_MAX)
609 return notmuch_sha1_of_string (path);
614 /* Given a path, split it into two parts: the directory part is all
615 * components except for the last, and the basename is that last
616 * component. Getting the return-value for either part is optional
617 * (the caller can pass NULL).
619 * The original 'path' can represent either a regular file or a
620 * directory---the splitting will be carried out in the same way in
621 * either case. Trailing slashes on 'path' will be ignored, and any
622 * cases of multiple '/' characters appearing in series will be
623 * treated as a single '/'.
625 * Allocation (if any) will have 'ctx' as the talloc owner. But
626 * pointers will be returned within the original path string whenever
629 * Note: If 'path' is non-empty and contains no non-trailing slash,
630 * (that is, consists of a filename with no parent directory), then
631 * the directory returned will be an empty string. However, if 'path'
632 * is an empty string, then both directory and basename will be
636 _notmuch_database_split_path (void *ctx,
638 const char **directory,
639 const char **basename)
643 if (path == NULL || *path == '\0') {
648 return NOTMUCH_STATUS_SUCCESS;
651 /* Find the last slash (not counting a trailing slash), if any. */
653 slash = path + strlen (path) - 1;
655 /* First, skip trailing slashes. */
656 while (slash != path) {
663 /* Then, find a slash. */
664 while (slash != path) {
674 /* Finally, skip multiple slashes. */
675 while (slash != path) {
684 *directory = talloc_strdup (ctx, "");
689 *directory = talloc_strndup (ctx, path, slash - path + 1);
692 return NOTMUCH_STATUS_SUCCESS;
696 _notmuch_database_find_directory_id (notmuch_database_t *notmuch,
698 unsigned int *directory_id)
700 notmuch_private_status_t private_status;
701 notmuch_status_t status = NOTMUCH_STATUS_SUCCESS;
706 return NOTMUCH_STATUS_SUCCESS;
709 db_path = directory_db_path (path);
711 private_status = find_unique_doc_id (notmuch, "directory",
712 db_path, directory_id);
713 if (private_status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
714 status = notmuch_database_set_directory_mtime (notmuch,
719 private_status = find_unique_doc_id (notmuch, "directory",
720 db_path, directory_id);
721 status = COERCE_STATUS (private_status, "_find_directory_id");
726 free ((char *) db_path);
735 _notmuch_database_get_directory_path (void *ctx,
736 notmuch_database_t *notmuch,
739 Xapian::Document document;
741 document = find_document_for_doc_id (notmuch, doc_id);
743 return talloc_strdup (ctx, document.get_data ().c_str ());
746 /* Given a legal 'filename' for the database, (either relative to
747 * database path or absolute with initial components identical to
748 * database path), return a new string (with 'ctx' as the talloc
749 * owner) suitable for use as a direntry term value.
752 _notmuch_database_filename_to_direntry (void *ctx,
753 notmuch_database_t *notmuch,
754 const char *filename,
757 const char *relative, *directory, *basename;
758 Xapian::docid directory_id;
759 notmuch_status_t status;
761 relative = _notmuch_database_relative_path (notmuch, filename);
763 status = _notmuch_database_split_path (ctx, relative,
764 &directory, &basename);
768 status = _notmuch_database_find_directory_id (notmuch, directory,
773 *direntry = talloc_asprintf (ctx, "%u:%s", directory_id, basename);
775 return NOTMUCH_STATUS_SUCCESS;
778 /* Given a legal 'path' for the database, return the relative path.
780 * The return value will be a pointer to the originl path contents,
781 * and will be either the original string (if 'path' was relative) or
782 * a portion of the string (if path was absolute and begins with the
786 _notmuch_database_relative_path (notmuch_database_t *notmuch,
789 const char *db_path, *relative;
790 unsigned int db_path_len;
792 db_path = notmuch_database_get_path (notmuch);
793 db_path_len = strlen (db_path);
797 if (*relative == '/') {
798 while (*relative == '/' && *(relative+1) == '/')
801 if (strncmp (relative, db_path, db_path_len) == 0)
803 relative += db_path_len;
804 while (*relative == '/')
813 notmuch_database_set_directory_mtime (notmuch_database_t *notmuch,
817 Xapian::Document doc;
818 Xapian::WritableDatabase *db;
820 notmuch_private_status_t status;
821 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
822 const char *parent, *db_path = NULL;
823 unsigned int parent_id;
824 void *local = talloc_new (notmuch);
826 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY) {
827 fprintf (stderr, "Attempted to update a read-only database.\n");
828 return NOTMUCH_STATUS_READONLY_DATABASE;
831 path = _notmuch_database_relative_path (notmuch, path);
833 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
834 db_path = directory_db_path (path);
837 status = find_directory_document (notmuch, db_path, &doc, &doc_id);
839 doc.add_value (NOTMUCH_VALUE_TIMESTAMP,
840 Xapian::sortable_serialise (mtime));
842 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
843 char *term = talloc_asprintf (local, "%s%s",
844 _find_prefix ("directory"), db_path);
849 _notmuch_database_split_path (local, path, &parent, NULL);
851 _notmuch_database_find_directory_id (notmuch, parent, &parent_id);
853 term = talloc_asprintf (local, "%s%u",
854 _find_prefix ("parent"),
858 db->add_document (doc);
860 db->replace_document (doc_id, doc);
863 } catch (const Xapian::Error &error) {
864 fprintf (stderr, "A Xapian exception occurred setting directory mtime: %s.\n",
865 error.get_msg().c_str());
866 notmuch->exception_reported = TRUE;
867 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
871 free ((char *) db_path);
879 notmuch_database_get_directory_mtime (notmuch_database_t *notmuch,
882 Xapian::Document doc;
884 notmuch_private_status_t status;
885 const char *db_path = NULL;
888 db_path = directory_db_path (path);
891 status = find_directory_document (notmuch, db_path, &doc, &doc_id);
893 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
896 ret = Xapian::sortable_unserialise (doc.get_value (NOTMUCH_VALUE_TIMESTAMP));
897 } catch (Xapian::Error &error) {
904 free ((char *) db_path);
909 /* Find the thread ID to which the message with 'message_id' belongs.
911 * Returns NULL if no message with message ID 'message_id' is in the
914 * Otherwise, returns a newly talloced string belonging to 'ctx'.
917 _resolve_message_id_to_thread_id (notmuch_database_t *notmuch,
919 const char *message_id)
921 notmuch_message_t *message;
922 const char *ret = NULL;
924 message = notmuch_database_find_message (notmuch, message_id);
928 ret = talloc_steal (ctx, notmuch_message_get_thread_id (message));
932 notmuch_message_destroy (message);
937 static notmuch_status_t
938 _merge_threads (notmuch_database_t *notmuch,
939 const char *winner_thread_id,
940 const char *loser_thread_id)
942 Xapian::PostingIterator loser, loser_end;
943 notmuch_message_t *message = NULL;
944 notmuch_private_status_t private_status;
945 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
947 find_doc_ids (notmuch, "thread", loser_thread_id, &loser, &loser_end);
949 for ( ; loser != loser_end; loser++) {
950 message = _notmuch_message_create (notmuch, notmuch,
951 *loser, &private_status);
952 if (message == NULL) {
953 ret = COERCE_STATUS (private_status,
954 "Cannot find document for doc_id from query");
958 _notmuch_message_remove_term (message, "thread", loser_thread_id);
959 _notmuch_message_add_term (message, "thread", winner_thread_id);
960 _notmuch_message_sync (message);
962 notmuch_message_destroy (message);
968 notmuch_message_destroy (message);
974 _my_talloc_free_for_g_hash (void *ptr)
979 static notmuch_status_t
980 _notmuch_database_link_message_to_parents (notmuch_database_t *notmuch,
981 notmuch_message_t *message,
982 notmuch_message_file_t *message_file,
983 const char **thread_id)
985 GHashTable *parents = NULL;
986 const char *refs, *in_reply_to, *in_reply_to_message_id;
987 GList *l, *keys = NULL;
988 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
990 parents = g_hash_table_new_full (g_str_hash, g_str_equal,
991 _my_talloc_free_for_g_hash, NULL);
993 refs = notmuch_message_file_get_header (message_file, "references");
994 parse_references (message, notmuch_message_get_message_id (message),
997 in_reply_to = notmuch_message_file_get_header (message_file, "in-reply-to");
998 parse_references (message, notmuch_message_get_message_id (message),
999 parents, in_reply_to);
1001 /* Carefully avoid adding any self-referential in-reply-to term. */
1002 in_reply_to_message_id = _parse_message_id (message, in_reply_to, NULL);
1003 if (in_reply_to_message_id &&
1004 strcmp (in_reply_to_message_id,
1005 notmuch_message_get_message_id (message)))
1007 _notmuch_message_add_term (message, "replyto",
1008 _parse_message_id (message, in_reply_to, NULL));
1011 keys = g_hash_table_get_keys (parents);
1012 for (l = keys; l; l = l->next) {
1013 char *parent_message_id;
1014 const char *parent_thread_id;
1016 parent_message_id = (char *) l->data;
1017 parent_thread_id = _resolve_message_id_to_thread_id (notmuch,
1021 if (parent_thread_id == NULL) {
1022 _notmuch_message_add_term (message, "reference",
1025 if (*thread_id == NULL) {
1026 *thread_id = talloc_strdup (message, parent_thread_id);
1027 _notmuch_message_add_term (message, "thread", *thread_id);
1028 } else if (strcmp (*thread_id, parent_thread_id)) {
1029 ret = _merge_threads (notmuch, *thread_id, parent_thread_id);
1040 g_hash_table_unref (parents);
1045 static notmuch_status_t
1046 _notmuch_database_link_message_to_children (notmuch_database_t *notmuch,
1047 notmuch_message_t *message,
1048 const char **thread_id)
1050 const char *message_id = notmuch_message_get_message_id (message);
1051 Xapian::PostingIterator child, children_end;
1052 notmuch_message_t *child_message = NULL;
1053 const char *child_thread_id;
1054 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1055 notmuch_private_status_t private_status;
1057 find_doc_ids (notmuch, "reference", message_id, &child, &children_end);
1059 for ( ; child != children_end; child++) {
1061 child_message = _notmuch_message_create (message, notmuch,
1062 *child, &private_status);
1063 if (child_message == NULL) {
1064 ret = COERCE_STATUS (private_status,
1065 "Cannot find document for doc_id from query");
1069 child_thread_id = notmuch_message_get_thread_id (child_message);
1070 if (*thread_id == NULL) {
1071 *thread_id = talloc_strdup (message, child_thread_id);
1072 _notmuch_message_add_term (message, "thread", *thread_id);
1073 } else if (strcmp (*thread_id, child_thread_id)) {
1074 _notmuch_message_remove_term (child_message, "reference",
1076 _notmuch_message_sync (child_message);
1077 ret = _merge_threads (notmuch, *thread_id, child_thread_id);
1082 notmuch_message_destroy (child_message);
1083 child_message = NULL;
1088 notmuch_message_destroy (child_message);
1093 /* Given a (mostly empty) 'message' and its corresponding
1094 * 'message_file' link it to existing threads in the database.
1096 * We first look at 'message_file' and its link-relevant headers
1097 * (References and In-Reply-To) for message IDs. We also look in the
1098 * database for existing message that reference 'message'.
1100 * The end result is to call _notmuch_message_ensure_thread_id which
1101 * generates a new thread ID if the message doesn't connect to any
1104 static notmuch_status_t
1105 _notmuch_database_link_message (notmuch_database_t *notmuch,
1106 notmuch_message_t *message,
1107 notmuch_message_file_t *message_file)
1109 notmuch_status_t status;
1110 const char *thread_id = NULL;
1112 status = _notmuch_database_link_message_to_parents (notmuch, message,
1118 status = _notmuch_database_link_message_to_children (notmuch, message,
1123 if (thread_id == NULL)
1124 _notmuch_message_ensure_thread_id (message);
1126 return NOTMUCH_STATUS_SUCCESS;
1130 notmuch_database_add_message (notmuch_database_t *notmuch,
1131 const char *filename,
1132 notmuch_message_t **message_ret)
1134 notmuch_message_file_t *message_file;
1135 notmuch_message_t *message = NULL;
1136 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1137 notmuch_private_status_t private_status;
1139 const char *date, *header;
1140 const char *from, *to, *subject;
1141 char *message_id = NULL;
1144 *message_ret = NULL;
1146 message_file = notmuch_message_file_open (filename);
1147 if (message_file == NULL) {
1148 ret = NOTMUCH_STATUS_FILE_ERROR;
1152 notmuch_message_file_restrict_headers (message_file,
1163 /* Before we do any real work, (especially before doing a
1164 * potential SHA-1 computation on the entire file's contents),
1165 * let's make sure that what we're looking at looks like an
1166 * actual email message.
1168 from = notmuch_message_file_get_header (message_file, "from");
1169 subject = notmuch_message_file_get_header (message_file, "subject");
1170 to = notmuch_message_file_get_header (message_file, "to");
1172 if ((from == NULL || *from == '\0') &&
1173 (subject == NULL || *subject == '\0') &&
1174 (to == NULL || *to == '\0'))
1176 ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
1180 /* Now that we're sure it's mail, the first order of business
1181 * is to find a message ID (or else create one ourselves). */
1183 header = notmuch_message_file_get_header (message_file, "message-id");
1184 if (header && *header != '\0') {
1185 message_id = _parse_message_id (message_file, header, NULL);
1187 /* So the header value isn't RFC-compliant, but it's
1188 * better than no message-id at all. */
1189 if (message_id == NULL)
1190 message_id = talloc_strdup (message_file, header);
1192 /* Reject a Message ID that's too long. */
1193 if (message_id && strlen (message_id) + 1 > NOTMUCH_TERM_MAX) {
1194 talloc_free (message_id);
1199 if (message_id == NULL ) {
1200 /* No message-id at all, let's generate one by taking a
1201 * hash over the file's contents. */
1202 char *sha1 = notmuch_sha1_of_file (filename);
1204 /* If that failed too, something is really wrong. Give up. */
1206 ret = NOTMUCH_STATUS_FILE_ERROR;
1210 message_id = talloc_asprintf (message_file,
1211 "notmuch-sha1-%s", sha1);
1215 /* Now that we have a message ID, we get a message object,
1216 * (which may or may not reference an existing document in the
1219 message = _notmuch_message_create_for_message_id (notmuch,
1223 talloc_free (message_id);
1225 if (message == NULL) {
1226 ret = COERCE_STATUS (private_status,
1227 "Unexpected status value from _notmuch_message_create_for_message_id");
1231 _notmuch_message_add_filename (message, filename);
1233 /* Is this a newly created message object? */
1234 if (private_status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
1235 _notmuch_message_add_term (message, "type", "mail");
1237 ret = _notmuch_database_link_message (notmuch, message,
1242 date = notmuch_message_file_get_header (message_file, "date");
1243 _notmuch_message_set_date (message, date);
1245 _notmuch_message_index_file (message, filename);
1247 ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
1250 _notmuch_message_sync (message);
1251 } catch (const Xapian::Error &error) {
1252 fprintf (stderr, "A Xapian exception occurred adding message: %s.\n",
1253 error.get_description().c_str());
1254 notmuch->exception_reported = TRUE;
1255 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
1261 if (ret == NOTMUCH_STATUS_SUCCESS && message_ret)
1262 *message_ret = message;
1264 notmuch_message_destroy (message);
1268 notmuch_message_file_close (message_file);
1274 _notmuch_convert_tags (void *ctx, Xapian::TermIterator &i,
1275 Xapian::TermIterator &end)
1277 const char *prefix = _find_prefix ("tag");
1278 notmuch_tags_t *tags;
1281 /* Currently this iteration is written with the assumption that
1282 * "tag" has a single-character prefix. */
1283 assert (strlen (prefix) == 1);
1285 tags = _notmuch_tags_create (ctx);
1286 if (unlikely (tags == NULL))
1294 if (tag.empty () || tag[0] != *prefix)
1297 _notmuch_tags_add_tag (tags, tag.c_str () + 1);
1302 _notmuch_tags_prepare_iterator (tags);
1308 notmuch_database_get_all_tags (notmuch_database_t *db)
1310 Xapian::TermIterator i, end;
1311 i = db->xapian_db->allterms_begin();
1312 end = db->xapian_db->allterms_end();
1313 return _notmuch_convert_tags(db, i, end);