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"
22 #include "parse-time-vrp.h"
23 #include "string-util.h"
32 #include <glib.h> /* g_free, GPtrArray, GHashTable */
33 #include <glib-object.h> /* g_type_init */
35 #include <gmime/gmime.h> /* g_mime_init */
39 #define ARRAY_SIZE(arr) (sizeof (arr) / sizeof (arr[0]))
46 #define NOTMUCH_DATABASE_VERSION 3
48 #define STRINGIFY(s) _SUB_STRINGIFY(s)
49 #define _SUB_STRINGIFY(s) #s
51 /* Here's the current schema for our database (for NOTMUCH_DATABASE_VERSION):
53 * We currently have two different types of documents (mail and
54 * directory) and also some metadata.
58 * A mail document is associated with a particular email message. It
59 * is stored in one or more files on disk (though only one has its
60 * content indexed) and is uniquely identified by its "id" field
61 * (which is generally the message ID). It is indexed with the
62 * following prefixed terms which the database uses to construct
65 * Single terms of given prefix:
69 * id: Unique ID of mail. This is from the Message-ID header
70 * if present and not too long (see NOTMUCH_MESSAGE_ID_MAX).
71 * If it's present and too long, then we use
72 * "notmuch-sha1-<sha1_sum_of_message_id>".
73 * If this header is not present, we use
74 * "notmuch-sha1-<sha1_sum_of_entire_file>".
76 * thread: The ID of the thread to which the mail belongs
78 * replyto: The ID from the In-Reply-To header of the mail (if any).
80 * Multiple terms of given prefix:
82 * reference: All message IDs from In-Reply-To and References
83 * headers in the message.
85 * tag: Any tags associated with this message by the user.
87 * file-direntry: A colon-separated pair of values
88 * (INTEGER:STRING), where INTEGER is the
89 * document ID of a directory document, and
90 * STRING is the name of a file within that
91 * directory for this mail message.
93 * A mail document also has four values:
95 * TIMESTAMP: The time_t value corresponding to the message's
98 * MESSAGE_ID: The unique ID of the mail mess (see "id" above)
100 * FROM: The value of the "From" header
102 * SUBJECT: The value of the "Subject" header
104 * In addition, terms from the content of the message are added with
105 * "from", "to", "attachment", and "subject" prefixes for use by the
106 * user in searching. Similarly, terms from the path of the mail
107 * message are added with "folder" and "path" prefixes. But the
108 * database doesn't really care itself about any of these.
110 * The data portion of a mail document is empty.
114 * A directory document is used by a client of the notmuch library to
115 * maintain data necessary to allow for efficient polling of mail
118 * All directory documents contain one term:
120 * directory: The directory path (relative to the database path)
121 * Or the SHA1 sum of the directory path (if the
122 * path itself is too long to fit in a Xapian
125 * And all directory documents for directories other than top-level
126 * directories also contain the following term:
128 * directory-direntry: A colon-separated pair of values
129 * (INTEGER:STRING), where INTEGER is the
130 * document ID of the parent directory
131 * document, and STRING is the name of this
132 * directory within that parent.
134 * All directory documents have a single value:
136 * TIMESTAMP: The mtime of the directory (at last scan)
138 * The data portion of a directory document contains the path of the
139 * directory (relative to the database path).
143 * Xapian allows us to store arbitrary name-value pairs as
144 * "metadata". We currently use the following metadata names with the
147 * version The database schema version, (which is distinct
148 * from both the notmuch package version (see
149 * notmuch --version) and the libnotmuch library
150 * version. The version is stored as an base-10
151 * ASCII integer. The initial database version
152 * was 1, (though a schema existed before that
153 * were no "version" database value existed at
154 * all). Successive versions are allocated as
155 * changes are made to the database (such as by
156 * indexing new fields).
158 * features The set of features supported by this
159 * database. This consists of a set of
160 * '\n'-separated lines, where each is a feature
161 * name, a '\t', and compatibility flags. If the
162 * compatibility flags contain 'w', then the
163 * opener must support this feature to safely
164 * write this database. If the compatibility
165 * flags contain 'r', then the opener must
166 * support this feature to read this database.
167 * Introduced in database version 3.
169 * last_thread_id The last thread ID generated. This is stored
170 * as a 16-byte hexadecimal ASCII representation
171 * of a 64-bit unsigned integer. The first ID
172 * generated is 1 and the value will be
173 * incremented for each thread ID.
175 * thread_id_* A pre-allocated thread ID for a particular
176 * message. This is actually an arbitrarily large
177 * family of metadata name. Any particular name is
178 * formed by concatenating "thread_id_" with a message
179 * ID (or the SHA1 sum of a message ID if it is very
180 * long---see description of 'id' in the mail
181 * document). The value stored is a thread ID.
183 * These thread ID metadata values are stored
184 * whenever a message references a parent message
185 * that does not yet exist in the database. A
186 * thread ID will be allocated and stored, and if
187 * the message is later added, the stored thread
188 * ID will be used (and the metadata value will
191 * Even before a message is added, it's
192 * pre-allocated thread ID is useful so that all
193 * descendant messages that reference this common
194 * parent can be recognized as belonging to the
198 /* With these prefix values we follow the conventions published here:
200 * http://xapian.org/docs/omega/termprefixes.html
202 * as much as makes sense. Note that I took some liberty in matching
203 * the reserved prefix values to notmuch concepts, (for example, 'G'
204 * is documented as "newsGroup (or similar entity - e.g. a web forum
205 * name)", for which I think the thread is the closest analogue in
206 * notmuch. This in spite of the fact that we will eventually be
207 * storing mailing-list messages where 'G' for "mailing list name"
208 * might be even a closer analogue. I'm treating the single-character
209 * prefixes preferentially for core notmuch concepts (which will be
210 * nearly universal to all mail messages).
213 static prefix_t BOOLEAN_PREFIX_INTERNAL[] = {
215 { "reference", "XREFERENCE" },
216 { "replyto", "XREPLYTO" },
217 { "directory", "XDIRECTORY" },
218 { "file-direntry", "XFDIRENTRY" },
219 { "directory-direntry", "XDDIRENTRY" },
222 static prefix_t BOOLEAN_PREFIX_EXTERNAL[] = {
229 * Without the ":", since this is a multi-letter prefix, Xapian
230 * will add a colon itself if the first letter of the path is
231 * upper-case ASCII. Including the ":" forces there to always be a
232 * colon, which keeps our own logic simpler.
234 { "folder", "XFOLDER:" },
237 static prefix_t PROBABILISTIC_PREFIX[]= {
240 { "attachment", "XATTACHMENT" },
241 { "subject", "XSUBJECT"},
245 _find_prefix (const char *name)
249 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_INTERNAL); i++) {
250 if (strcmp (name, BOOLEAN_PREFIX_INTERNAL[i].name) == 0)
251 return BOOLEAN_PREFIX_INTERNAL[i].prefix;
254 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
255 if (strcmp (name, BOOLEAN_PREFIX_EXTERNAL[i].name) == 0)
256 return BOOLEAN_PREFIX_EXTERNAL[i].prefix;
259 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
260 if (strcmp (name, PROBABILISTIC_PREFIX[i].name) == 0)
261 return PROBABILISTIC_PREFIX[i].prefix;
264 INTERNAL_ERROR ("No prefix exists for '%s'\n", name);
269 static const struct {
270 /* NOTMUCH_FEATURE_* value. */
271 _notmuch_features value;
272 /* Feature name as it appears in the database. This name should
273 * be appropriate for displaying to the user if an older version
274 * of notmuch doesn't support this feature. */
276 /* Compatibility flags when this feature is declared. */
278 } feature_names[] = {
279 { NOTMUCH_FEATURE_FILE_TERMS,
280 "multiple paths per message", "rw" },
281 { NOTMUCH_FEATURE_DIRECTORY_DOCS,
282 "relative directory paths", "rw" },
283 /* Header values are not required for reading a database because a
284 * reader can just refer to the message file. */
285 { NOTMUCH_FEATURE_FROM_SUBJECT_ID_VALUES,
286 "from/subject/message-ID in database", "w" },
287 { NOTMUCH_FEATURE_BOOL_FOLDER,
288 "exact folder:/path: search", "rw" },
289 { NOTMUCH_FEATURE_GHOSTS,
290 "mail documents for missing messages", "w"},
294 notmuch_status_to_string (notmuch_status_t status)
297 case NOTMUCH_STATUS_SUCCESS:
298 return "No error occurred";
299 case NOTMUCH_STATUS_OUT_OF_MEMORY:
300 return "Out of memory";
301 case NOTMUCH_STATUS_READ_ONLY_DATABASE:
302 return "Attempt to write to a read-only database";
303 case NOTMUCH_STATUS_XAPIAN_EXCEPTION:
304 return "A Xapian exception occurred";
305 case NOTMUCH_STATUS_FILE_ERROR:
306 return "Something went wrong trying to read or write a file";
307 case NOTMUCH_STATUS_FILE_NOT_EMAIL:
308 return "File is not an email";
309 case NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID:
310 return "Message ID is identical to a message in database";
311 case NOTMUCH_STATUS_NULL_POINTER:
312 return "Erroneous NULL pointer";
313 case NOTMUCH_STATUS_TAG_TOO_LONG:
314 return "Tag value is too long (exceeds NOTMUCH_TAG_MAX)";
315 case NOTMUCH_STATUS_UNBALANCED_FREEZE_THAW:
316 return "Unbalanced number of calls to notmuch_message_freeze/thaw";
317 case NOTMUCH_STATUS_UNBALANCED_ATOMIC:
318 return "Unbalanced number of calls to notmuch_database_begin_atomic/end_atomic";
319 case NOTMUCH_STATUS_UNSUPPORTED_OPERATION:
320 return "Unsupported operation";
321 case NOTMUCH_STATUS_UPGRADE_REQUIRED:
322 return "Operation requires a database upgrade";
324 case NOTMUCH_STATUS_LAST_STATUS:
325 return "Unknown error status value";
330 find_doc_ids_for_term (notmuch_database_t *notmuch,
332 Xapian::PostingIterator *begin,
333 Xapian::PostingIterator *end)
335 *begin = notmuch->xapian_db->postlist_begin (term);
337 *end = notmuch->xapian_db->postlist_end (term);
341 find_doc_ids (notmuch_database_t *notmuch,
342 const char *prefix_name,
344 Xapian::PostingIterator *begin,
345 Xapian::PostingIterator *end)
349 term = talloc_asprintf (notmuch, "%s%s",
350 _find_prefix (prefix_name), value);
352 find_doc_ids_for_term (notmuch, term, begin, end);
357 notmuch_private_status_t
358 _notmuch_database_find_unique_doc_id (notmuch_database_t *notmuch,
359 const char *prefix_name,
361 unsigned int *doc_id)
363 Xapian::PostingIterator i, end;
365 find_doc_ids (notmuch, prefix_name, value, &i, &end);
369 return NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND;
374 #if DEBUG_DATABASE_SANITY
378 INTERNAL_ERROR ("Term %s:%s is not unique as expected.\n",
382 return NOTMUCH_PRIVATE_STATUS_SUCCESS;
385 static Xapian::Document
386 find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id)
388 return notmuch->xapian_db->get_document (doc_id);
391 /* Generate a compressed version of 'message_id' of the form:
393 * notmuch-sha1-<sha1_sum_of_message_id>
396 _notmuch_message_id_compressed (void *ctx, const char *message_id)
398 char *sha1, *compressed;
400 sha1 = _notmuch_sha1_of_string (message_id);
402 compressed = talloc_asprintf (ctx, "notmuch-sha1-%s", sha1);
409 notmuch_database_find_message (notmuch_database_t *notmuch,
410 const char *message_id,
411 notmuch_message_t **message_ret)
413 notmuch_private_status_t status;
416 if (message_ret == NULL)
417 return NOTMUCH_STATUS_NULL_POINTER;
419 if (strlen (message_id) > NOTMUCH_MESSAGE_ID_MAX)
420 message_id = _notmuch_message_id_compressed (notmuch, message_id);
423 status = _notmuch_database_find_unique_doc_id (notmuch, "id",
424 message_id, &doc_id);
426 if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
429 *message_ret = _notmuch_message_create (notmuch, notmuch, doc_id,
431 if (*message_ret == NULL)
432 return NOTMUCH_STATUS_OUT_OF_MEMORY;
435 return NOTMUCH_STATUS_SUCCESS;
436 } catch (const Xapian::Error &error) {
437 fprintf (stderr, "A Xapian exception occurred finding message: %s.\n",
438 error.get_msg().c_str());
439 notmuch->exception_reported = TRUE;
441 return NOTMUCH_STATUS_XAPIAN_EXCEPTION;
445 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
446 * a (potentially nested) parenthesized sequence with '\' used to
447 * escape any character (including parentheses).
449 * If the sequence to be skipped continues to the end of the string,
450 * then 'str' will be left pointing at the final terminating '\0'
454 skip_space_and_comments (const char **str)
459 while (*s && (isspace (*s) || *s == '(')) {
460 while (*s && isspace (*s))
465 while (*s && nesting) {
468 } else if (*s == ')') {
470 } else if (*s == '\\') {
482 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
483 * comments, and the '<' and '>' delimiters.
485 * If not NULL, then *next will be made to point to the first character
486 * not parsed, (possibly pointing to the final '\0' terminator.
488 * Returns a newly talloc'ed string belonging to 'ctx'.
490 * Returns NULL if there is any error parsing the message-id. */
492 _parse_message_id (void *ctx, const char *message_id, const char **next)
497 if (message_id == NULL || *message_id == '\0')
502 skip_space_and_comments (&s);
504 /* Skip any unstructured text as well. */
505 while (*s && *s != '<')
516 skip_space_and_comments (&s);
519 while (*end && *end != '>')
528 if (end > s && *end == '>')
533 result = talloc_strndup (ctx, s, end - s + 1);
535 /* Finally, collapse any whitespace that is within the message-id
541 for (r = result, len = strlen (r); *r; r++, len--)
542 if (*r == ' ' || *r == '\t')
543 memmove (r, r+1, len);
549 /* Parse a References header value, putting a (talloc'ed under 'ctx')
550 * copy of each referenced message-id into 'hash'.
552 * We explicitly avoid including any reference identical to
553 * 'message_id' in the result (to avoid mass confusion when a single
554 * message references itself cyclically---and yes, mail messages are
555 * not infrequent in the wild that do this---don't ask me why).
557 * Return the last reference parsed, if it is not equal to message_id.
560 parse_references (void *ctx,
561 const char *message_id,
565 char *ref, *last_ref = NULL;
567 if (refs == NULL || *refs == '\0')
571 ref = _parse_message_id (ctx, refs, &refs);
573 if (ref && strcmp (ref, message_id)) {
574 g_hash_table_insert (hash, ref, NULL);
579 /* The return value of this function is used to add a parent
580 * reference to the database. We should avoid making a message
581 * its own parent, thus the above check.
587 notmuch_database_create (const char *path, notmuch_database_t **database)
589 notmuch_status_t status = NOTMUCH_STATUS_SUCCESS;
590 notmuch_database_t *notmuch = NULL;
591 char *notmuch_path = NULL;
596 fprintf (stderr, "Error: Cannot create a database for a NULL path.\n");
597 status = NOTMUCH_STATUS_NULL_POINTER;
601 err = stat (path, &st);
603 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
604 path, strerror (errno));
605 status = NOTMUCH_STATUS_FILE_ERROR;
609 if (! S_ISDIR (st.st_mode)) {
610 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
612 status = NOTMUCH_STATUS_FILE_ERROR;
616 notmuch_path = talloc_asprintf (NULL, "%s/%s", path, ".notmuch");
618 err = mkdir (notmuch_path, 0755);
621 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
622 notmuch_path, strerror (errno));
623 status = NOTMUCH_STATUS_FILE_ERROR;
627 status = notmuch_database_open (path,
628 NOTMUCH_DATABASE_MODE_READ_WRITE,
633 /* Upgrade doesn't add this feature to existing databases, but new
634 * databases have it. */
635 notmuch->features |= NOTMUCH_FEATURE_FROM_SUBJECT_ID_VALUES;
637 status = notmuch_database_upgrade (notmuch, NULL, NULL);
639 notmuch_database_close(notmuch);
645 talloc_free (notmuch_path);
650 talloc_free (notmuch);
655 _notmuch_database_ensure_writable (notmuch_database_t *notmuch)
657 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY) {
658 fprintf (stderr, "Cannot write to a read-only database.\n");
659 return NOTMUCH_STATUS_READ_ONLY_DATABASE;
662 return NOTMUCH_STATUS_SUCCESS;
665 /* Parse a database features string from the given database version.
666 * Returns the feature bit set.
668 * For version < 3, this ignores the features string and returns a
669 * hard-coded set of features.
671 * If there are unrecognized features that are required to open the
672 * database in mode (which should be 'r' or 'w'), return a
673 * comma-separated list of unrecognized but required features in
674 * *incompat_out suitable for presenting to the user. *incompat_out
675 * will be allocated from ctx.
677 static _notmuch_features
678 _parse_features (const void *ctx, const char *features, unsigned int version,
679 char mode, char **incompat_out)
681 _notmuch_features res = static_cast<_notmuch_features>(0);
682 unsigned int namelen, i;
686 /* Prior to database version 3, features were implied by the
689 return NOTMUCH_FEATURES_V0;
690 else if (version == 1)
691 return NOTMUCH_FEATURES_V1;
692 else if (version == 2)
693 return NOTMUCH_FEATURES_V2;
695 /* Parse the features string */
696 while ((features = strtok_len_c (features + llen, "\n", &llen)) != NULL) {
697 flags = strchr (features, '\t');
698 if (! flags || flags > features + llen)
700 namelen = flags - features;
702 for (i = 0; i < ARRAY_SIZE (feature_names); ++i) {
703 if (strlen (feature_names[i].name) == namelen &&
704 strncmp (feature_names[i].name, features, namelen) == 0) {
705 res |= feature_names[i].value;
710 if (i == ARRAY_SIZE (feature_names) && incompat_out) {
711 /* Unrecognized feature */
712 const char *have = strchr (flags, mode);
713 if (have && have < features + llen) {
714 /* This feature is required to access this database in
715 * 'mode', but we don't understand it. */
717 *incompat_out = talloc_strdup (ctx, "");
718 *incompat_out = talloc_asprintf_append_buffer (
719 *incompat_out, "%s%.*s", **incompat_out ? ", " : "",
729 _print_features (const void *ctx, unsigned int features)
732 char *res = talloc_strdup (ctx, "");
734 for (i = 0; i < ARRAY_SIZE (feature_names); ++i)
735 if (features & feature_names[i].value)
736 res = talloc_asprintf_append_buffer (
737 res, "%s\t%s\n", feature_names[i].name, feature_names[i].flags);
743 notmuch_database_open (const char *path,
744 notmuch_database_mode_t mode,
745 notmuch_database_t **database)
747 notmuch_status_t status = NOTMUCH_STATUS_SUCCESS;
748 void *local = talloc_new (NULL);
749 notmuch_database_t *notmuch = NULL;
750 char *notmuch_path, *xapian_path, *incompat_features;
753 unsigned int i, version;
754 static int initialized = 0;
757 fprintf (stderr, "Error: Cannot open a database for a NULL path.\n");
758 status = NOTMUCH_STATUS_NULL_POINTER;
762 if (! (notmuch_path = talloc_asprintf (local, "%s/%s", path, ".notmuch"))) {
763 fprintf (stderr, "Out of memory\n");
764 status = NOTMUCH_STATUS_OUT_OF_MEMORY;
768 err = stat (notmuch_path, &st);
770 fprintf (stderr, "Error opening database at %s: %s\n",
771 notmuch_path, strerror (errno));
772 status = NOTMUCH_STATUS_FILE_ERROR;
776 if (! (xapian_path = talloc_asprintf (local, "%s/%s", notmuch_path, "xapian"))) {
777 fprintf (stderr, "Out of memory\n");
778 status = NOTMUCH_STATUS_OUT_OF_MEMORY;
782 /* Initialize the GLib type system and threads */
783 #if !GLIB_CHECK_VERSION(2, 35, 1)
787 /* Initialize gmime */
789 g_mime_init (GMIME_ENABLE_RFC2047_WORKAROUNDS);
793 notmuch = talloc_zero (NULL, notmuch_database_t);
794 notmuch->exception_reported = FALSE;
795 notmuch->path = talloc_strdup (notmuch, path);
797 if (notmuch->path[strlen (notmuch->path) - 1] == '/')
798 notmuch->path[strlen (notmuch->path) - 1] = '\0';
800 notmuch->mode = mode;
801 notmuch->atomic_nesting = 0;
803 string last_thread_id;
805 if (mode == NOTMUCH_DATABASE_MODE_READ_WRITE) {
806 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
807 Xapian::DB_CREATE_OR_OPEN);
809 notmuch->xapian_db = new Xapian::Database (xapian_path);
812 /* Check version. As of database version 3, we represent
813 * changes in terms of features, so assume a version bump
814 * means a dramatically incompatible change. */
815 version = notmuch_database_get_version (notmuch);
816 if (version > NOTMUCH_DATABASE_VERSION) {
818 "Error: Notmuch database at %s\n"
819 " has a newer database format version (%u) than supported by this\n"
820 " version of notmuch (%u).\n",
821 notmuch_path, version, NOTMUCH_DATABASE_VERSION);
822 notmuch->mode = NOTMUCH_DATABASE_MODE_READ_ONLY;
823 notmuch_database_destroy (notmuch);
825 status = NOTMUCH_STATUS_FILE_ERROR;
829 /* Check features. */
830 incompat_features = NULL;
831 notmuch->features = _parse_features (
832 local, notmuch->xapian_db->get_metadata ("features").c_str (),
833 version, mode == NOTMUCH_DATABASE_MODE_READ_WRITE ? 'w' : 'r',
835 if (incompat_features) {
837 "Error: Notmuch database at %s\n"
838 " requires features (%s)\n"
839 " not supported by this version of notmuch.\n",
840 notmuch_path, incompat_features);
841 notmuch->mode = NOTMUCH_DATABASE_MODE_READ_ONLY;
842 notmuch_database_destroy (notmuch);
844 status = NOTMUCH_STATUS_FILE_ERROR;
848 notmuch->last_doc_id = notmuch->xapian_db->get_lastdocid ();
849 last_thread_id = notmuch->xapian_db->get_metadata ("last_thread_id");
850 if (last_thread_id.empty ()) {
851 notmuch->last_thread_id = 0;
856 str = last_thread_id.c_str ();
857 notmuch->last_thread_id = strtoull (str, &end, 16);
859 INTERNAL_ERROR ("Malformed database last_thread_id: %s", str);
862 notmuch->query_parser = new Xapian::QueryParser;
863 notmuch->term_gen = new Xapian::TermGenerator;
864 notmuch->term_gen->set_stemmer (Xapian::Stem ("english"));
865 notmuch->value_range_processor = new Xapian::NumberValueRangeProcessor (NOTMUCH_VALUE_TIMESTAMP);
866 notmuch->date_range_processor = new ParseTimeValueRangeProcessor (NOTMUCH_VALUE_TIMESTAMP);
868 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
869 notmuch->query_parser->set_database (*notmuch->xapian_db);
870 notmuch->query_parser->set_stemmer (Xapian::Stem ("english"));
871 notmuch->query_parser->set_stemming_strategy (Xapian::QueryParser::STEM_SOME);
872 notmuch->query_parser->add_valuerangeprocessor (notmuch->value_range_processor);
873 notmuch->query_parser->add_valuerangeprocessor (notmuch->date_range_processor);
875 for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
876 prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i];
877 notmuch->query_parser->add_boolean_prefix (prefix->name,
881 for (i = 0; i < ARRAY_SIZE (PROBABILISTIC_PREFIX); i++) {
882 prefix_t *prefix = &PROBABILISTIC_PREFIX[i];
883 notmuch->query_parser->add_prefix (prefix->name, prefix->prefix);
885 } catch (const Xapian::Error &error) {
886 fprintf (stderr, "A Xapian exception occurred opening database: %s\n",
887 error.get_msg().c_str());
888 notmuch_database_destroy (notmuch);
890 status = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
899 talloc_free (notmuch);
904 notmuch_database_close (notmuch_database_t *notmuch)
906 notmuch_status_t status = NOTMUCH_STATUS_SUCCESS;
908 /* Many Xapian objects (and thus notmuch objects) hold references to
909 * the database, so merely deleting the database may not suffice to
910 * close it. Thus, we explicitly close it here. */
911 if (notmuch->xapian_db != NULL) {
913 /* If there's an outstanding transaction, it's unclear if
914 * closing the Xapian database commits everything up to
915 * that transaction, or may discard committed (but
916 * unflushed) transactions. To be certain, explicitly
917 * cancel any outstanding transaction before closing. */
918 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_WRITE &&
919 notmuch->atomic_nesting)
920 (static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db))
921 ->cancel_transaction ();
923 /* Close the database. This implicitly flushes
924 * outstanding changes. */
925 notmuch->xapian_db->close();
926 } catch (const Xapian::Error &error) {
927 status = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
928 if (! notmuch->exception_reported) {
929 fprintf (stderr, "Error: A Xapian exception occurred closing database: %s\n",
930 error.get_msg().c_str());
935 delete notmuch->term_gen;
936 notmuch->term_gen = NULL;
937 delete notmuch->query_parser;
938 notmuch->query_parser = NULL;
939 delete notmuch->xapian_db;
940 notmuch->xapian_db = NULL;
941 delete notmuch->value_range_processor;
942 notmuch->value_range_processor = NULL;
943 delete notmuch->date_range_processor;
944 notmuch->date_range_processor = NULL;
949 #if HAVE_XAPIAN_COMPACT
951 unlink_cb (const char *path,
952 unused (const struct stat *sb),
954 unused (struct FTW *ftw))
956 return remove (path);
960 rmtree (const char *path)
962 return nftw (path, unlink_cb, 64, FTW_DEPTH | FTW_PHYS);
965 class NotmuchCompactor : public Xapian::Compactor
967 notmuch_compact_status_cb_t status_cb;
968 void *status_closure;
971 NotmuchCompactor(notmuch_compact_status_cb_t cb, void *closure) :
972 status_cb (cb), status_closure (closure) { }
975 set_status (const std::string &table, const std::string &status)
979 if (status_cb == NULL)
982 if (status.length () == 0)
983 msg = talloc_asprintf (NULL, "compacting table %s", table.c_str());
985 msg = talloc_asprintf (NULL, " %s", status.c_str());
991 status_cb (msg, status_closure);
996 /* Compacts the given database, optionally saving the original database
997 * in backup_path. Additionally, a callback function can be provided to
998 * give the user feedback on the progress of the (likely long-lived)
999 * compaction process.
1001 * The backup path must point to a directory on the same volume as the
1002 * original database. Passing a NULL backup_path will result in the
1003 * uncompacted database being deleted after compaction has finished.
1004 * Note that the database write lock will be held during the
1005 * compaction process to protect data integrity.
1008 notmuch_database_compact (const char *path,
1009 const char *backup_path,
1010 notmuch_compact_status_cb_t status_cb,
1014 char *notmuch_path, *xapian_path, *compact_xapian_path;
1015 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1016 notmuch_database_t *notmuch = NULL;
1017 struct stat statbuf;
1018 notmuch_bool_t keep_backup;
1020 local = talloc_new (NULL);
1022 return NOTMUCH_STATUS_OUT_OF_MEMORY;
1024 ret = notmuch_database_open (path, NOTMUCH_DATABASE_MODE_READ_WRITE, ¬much);
1029 if (! (notmuch_path = talloc_asprintf (local, "%s/%s", path, ".notmuch"))) {
1030 ret = NOTMUCH_STATUS_OUT_OF_MEMORY;
1034 if (! (xapian_path = talloc_asprintf (local, "%s/%s", notmuch_path, "xapian"))) {
1035 ret = NOTMUCH_STATUS_OUT_OF_MEMORY;
1039 if (! (compact_xapian_path = talloc_asprintf (local, "%s.compact", xapian_path))) {
1040 ret = NOTMUCH_STATUS_OUT_OF_MEMORY;
1044 if (backup_path == NULL) {
1045 if (! (backup_path = talloc_asprintf (local, "%s.old", xapian_path))) {
1046 ret = NOTMUCH_STATUS_OUT_OF_MEMORY;
1049 keep_backup = FALSE;
1055 if (stat (backup_path, &statbuf) != -1) {
1056 fprintf (stderr, "Path already exists: %s\n", backup_path);
1057 ret = NOTMUCH_STATUS_FILE_ERROR;
1060 if (errno != ENOENT) {
1061 fprintf (stderr, "Unknown error while stat()ing path: %s\n",
1063 ret = NOTMUCH_STATUS_FILE_ERROR;
1067 /* Unconditionally attempt to remove old work-in-progress database (if
1068 * any). This is "protected" by database lock. If this fails due to write
1069 * errors (etc), the following code will fail and provide error message.
1071 (void) rmtree (compact_xapian_path);
1074 NotmuchCompactor compactor (status_cb, closure);
1076 compactor.set_renumber (false);
1077 compactor.add_source (xapian_path);
1078 compactor.set_destdir (compact_xapian_path);
1079 compactor.compact ();
1080 } catch (const Xapian::Error &error) {
1081 fprintf (stderr, "Error while compacting: %s\n", error.get_msg().c_str());
1082 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
1086 if (rename (xapian_path, backup_path)) {
1087 fprintf (stderr, "Error moving %s to %s: %s\n",
1088 xapian_path, backup_path, strerror (errno));
1089 ret = NOTMUCH_STATUS_FILE_ERROR;
1093 if (rename (compact_xapian_path, xapian_path)) {
1094 fprintf (stderr, "Error moving %s to %s: %s\n",
1095 compact_xapian_path, xapian_path, strerror (errno));
1096 ret = NOTMUCH_STATUS_FILE_ERROR;
1100 if (! keep_backup) {
1101 if (rmtree (backup_path)) {
1102 fprintf (stderr, "Error removing old database %s: %s\n",
1103 backup_path, strerror (errno));
1104 ret = NOTMUCH_STATUS_FILE_ERROR;
1111 notmuch_status_t ret2;
1113 ret2 = notmuch_database_destroy (notmuch);
1115 /* don't clobber previous error status */
1116 if (ret == NOTMUCH_STATUS_SUCCESS && ret2 != NOTMUCH_STATUS_SUCCESS)
1120 talloc_free (local);
1126 notmuch_database_compact (unused (const char *path),
1127 unused (const char *backup_path),
1128 unused (notmuch_compact_status_cb_t status_cb),
1129 unused (void *closure))
1131 fprintf (stderr, "notmuch was compiled against a xapian version lacking compaction support.\n");
1132 return NOTMUCH_STATUS_UNSUPPORTED_OPERATION;
1137 notmuch_database_destroy (notmuch_database_t *notmuch)
1139 notmuch_status_t status;
1141 status = notmuch_database_close (notmuch);
1142 talloc_free (notmuch);
1148 notmuch_database_get_path (notmuch_database_t *notmuch)
1150 return notmuch->path;
1154 notmuch_database_get_version (notmuch_database_t *notmuch)
1156 unsigned int version;
1157 string version_string;
1161 version_string = notmuch->xapian_db->get_metadata ("version");
1162 if (version_string.empty ())
1165 str = version_string.c_str ();
1166 if (str == NULL || *str == '\0')
1169 version = strtoul (str, &end, 10);
1171 INTERNAL_ERROR ("Malformed database version: %s", str);
1177 notmuch_database_needs_upgrade (notmuch_database_t *notmuch)
1179 return notmuch->mode == NOTMUCH_DATABASE_MODE_READ_WRITE &&
1180 ((NOTMUCH_FEATURES_CURRENT & ~notmuch->features) ||
1181 (notmuch_database_get_version (notmuch) < NOTMUCH_DATABASE_VERSION));
1184 static volatile sig_atomic_t do_progress_notify = 0;
1187 handle_sigalrm (unused (int signal))
1189 do_progress_notify = 1;
1192 /* Upgrade the current database.
1194 * After opening a database in read-write mode, the client should
1195 * check if an upgrade is needed (notmuch_database_needs_upgrade) and
1196 * if so, upgrade with this function before making any modifications.
1198 * The optional progress_notify callback can be used by the caller to
1199 * provide progress indication to the user. If non-NULL it will be
1200 * called periodically with 'count' as the number of messages upgraded
1201 * so far and 'total' the overall number of messages that will be
1205 notmuch_database_upgrade (notmuch_database_t *notmuch,
1206 void (*progress_notify) (void *closure,
1210 void *local = talloc_new (NULL);
1211 Xapian::TermIterator t, t_end;
1212 Xapian::WritableDatabase *db;
1213 struct sigaction action;
1214 struct itimerval timerval;
1215 notmuch_bool_t timer_is_active = FALSE;
1216 enum _notmuch_features target_features, new_features;
1217 notmuch_status_t status;
1218 unsigned int count = 0, total = 0;
1220 status = _notmuch_database_ensure_writable (notmuch);
1224 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1226 target_features = notmuch->features | NOTMUCH_FEATURES_CURRENT;
1227 new_features = NOTMUCH_FEATURES_CURRENT & ~notmuch->features;
1229 if (! notmuch_database_needs_upgrade (notmuch))
1230 return NOTMUCH_STATUS_SUCCESS;
1232 if (progress_notify) {
1233 /* Setup our handler for SIGALRM */
1234 memset (&action, 0, sizeof (struct sigaction));
1235 action.sa_handler = handle_sigalrm;
1236 sigemptyset (&action.sa_mask);
1237 action.sa_flags = SA_RESTART;
1238 sigaction (SIGALRM, &action, NULL);
1240 /* Then start a timer to send SIGALRM once per second. */
1241 timerval.it_interval.tv_sec = 1;
1242 timerval.it_interval.tv_usec = 0;
1243 timerval.it_value.tv_sec = 1;
1244 timerval.it_value.tv_usec = 0;
1245 setitimer (ITIMER_REAL, &timerval, NULL);
1247 timer_is_active = TRUE;
1250 /* Figure out how much total work we need to do. */
1252 (NOTMUCH_FEATURE_FILE_TERMS | NOTMUCH_FEATURE_BOOL_FOLDER)) {
1253 notmuch_query_t *query = notmuch_query_create (notmuch, "");
1254 total += notmuch_query_count_messages (query);
1255 notmuch_query_destroy (query);
1257 if (new_features & NOTMUCH_FEATURE_DIRECTORY_DOCS) {
1258 t_end = db->allterms_end ("XTIMESTAMP");
1259 for (t = db->allterms_begin ("XTIMESTAMP"); t != t_end; t++)
1263 /* Perform the upgrade in a transaction. */
1264 db->begin_transaction (true);
1266 /* Set the target features so we write out changes in the desired
1268 notmuch->features = target_features;
1270 /* Perform per-message upgrades. */
1272 (NOTMUCH_FEATURE_FILE_TERMS | NOTMUCH_FEATURE_BOOL_FOLDER)) {
1273 notmuch_query_t *query = notmuch_query_create (notmuch, "");
1274 notmuch_messages_t *messages;
1275 notmuch_message_t *message;
1278 for (messages = notmuch_query_search_messages (query);
1279 notmuch_messages_valid (messages);
1280 notmuch_messages_move_to_next (messages))
1282 if (do_progress_notify) {
1283 progress_notify (closure, (double) count / total);
1284 do_progress_notify = 0;
1287 message = notmuch_messages_get (messages);
1289 /* Before version 1, each message document had its
1290 * filename in the data field. Copy that into the new
1291 * format by calling notmuch_message_add_filename.
1293 if (new_features & NOTMUCH_FEATURE_FILE_TERMS) {
1294 filename = _notmuch_message_talloc_copy_data (message);
1295 if (filename && *filename != '\0') {
1296 _notmuch_message_add_filename (message, filename);
1297 _notmuch_message_clear_data (message);
1299 talloc_free (filename);
1302 /* Prior to version 2, the "folder:" prefix was
1303 * probabilistic and stemmed. Change it to the current
1304 * boolean prefix. Add "path:" prefixes while at it.
1306 if (new_features & NOTMUCH_FEATURE_BOOL_FOLDER)
1307 _notmuch_message_upgrade_folder (message);
1309 _notmuch_message_sync (message);
1311 notmuch_message_destroy (message);
1316 notmuch_query_destroy (query);
1319 /* Perform per-directory upgrades. */
1321 /* Before version 1 we stored directory timestamps in
1322 * XTIMESTAMP documents instead of the current XDIRECTORY
1323 * documents. So copy those as well. */
1324 if (new_features & NOTMUCH_FEATURE_DIRECTORY_DOCS) {
1325 t_end = notmuch->xapian_db->allterms_end ("XTIMESTAMP");
1327 for (t = notmuch->xapian_db->allterms_begin ("XTIMESTAMP");
1331 Xapian::PostingIterator p, p_end;
1332 std::string term = *t;
1334 p_end = notmuch->xapian_db->postlist_end (term);
1336 for (p = notmuch->xapian_db->postlist_begin (term);
1340 Xapian::Document document;
1342 notmuch_directory_t *directory;
1344 if (do_progress_notify) {
1345 progress_notify (closure, (double) count / total);
1346 do_progress_notify = 0;
1349 document = find_document_for_doc_id (notmuch, *p);
1350 mtime = Xapian::sortable_unserialise (
1351 document.get_value (NOTMUCH_VALUE_TIMESTAMP));
1353 directory = _notmuch_directory_create (notmuch, term.c_str() + 10,
1354 NOTMUCH_FIND_CREATE, &status);
1355 notmuch_directory_set_mtime (directory, mtime);
1356 notmuch_directory_destroy (directory);
1358 db->delete_document (*p);
1365 db->set_metadata ("features", _print_features (local, notmuch->features));
1366 db->set_metadata ("version", STRINGIFY (NOTMUCH_DATABASE_VERSION));
1368 db->commit_transaction ();
1370 if (timer_is_active) {
1371 /* Now stop the timer. */
1372 timerval.it_interval.tv_sec = 0;
1373 timerval.it_interval.tv_usec = 0;
1374 timerval.it_value.tv_sec = 0;
1375 timerval.it_value.tv_usec = 0;
1376 setitimer (ITIMER_REAL, &timerval, NULL);
1378 /* And disable the signal handler. */
1379 action.sa_handler = SIG_IGN;
1380 sigaction (SIGALRM, &action, NULL);
1383 talloc_free (local);
1384 return NOTMUCH_STATUS_SUCCESS;
1388 notmuch_database_begin_atomic (notmuch_database_t *notmuch)
1390 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY ||
1391 notmuch->atomic_nesting > 0)
1395 (static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db))->begin_transaction (false);
1396 } catch (const Xapian::Error &error) {
1397 fprintf (stderr, "A Xapian exception occurred beginning transaction: %s.\n",
1398 error.get_msg().c_str());
1399 notmuch->exception_reported = TRUE;
1400 return NOTMUCH_STATUS_XAPIAN_EXCEPTION;
1404 notmuch->atomic_nesting++;
1405 return NOTMUCH_STATUS_SUCCESS;
1409 notmuch_database_end_atomic (notmuch_database_t *notmuch)
1411 Xapian::WritableDatabase *db;
1413 if (notmuch->atomic_nesting == 0)
1414 return NOTMUCH_STATUS_UNBALANCED_ATOMIC;
1416 if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY ||
1417 notmuch->atomic_nesting > 1)
1420 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1422 db->commit_transaction ();
1424 /* This is a hack for testing. Xapian never flushes on a
1425 * non-flushed commit, even if the flush threshold is 1.
1426 * However, we rely on flushing to test atomicity. */
1427 const char *thresh = getenv ("XAPIAN_FLUSH_THRESHOLD");
1428 if (thresh && atoi (thresh) == 1)
1430 } catch (const Xapian::Error &error) {
1431 fprintf (stderr, "A Xapian exception occurred committing transaction: %s.\n",
1432 error.get_msg().c_str());
1433 notmuch->exception_reported = TRUE;
1434 return NOTMUCH_STATUS_XAPIAN_EXCEPTION;
1438 notmuch->atomic_nesting--;
1439 return NOTMUCH_STATUS_SUCCESS;
1442 /* We allow the user to use arbitrarily long paths for directories. But
1443 * we have a term-length limit. So if we exceed that, we'll use the
1444 * SHA-1 of the path for the database term.
1446 * Note: This function may return the original value of 'path'. If it
1447 * does not, then the caller is responsible to free() the returned
1451 _notmuch_database_get_directory_db_path (const char *path)
1453 int term_len = strlen (_find_prefix ("directory")) + strlen (path);
1455 if (term_len > NOTMUCH_TERM_MAX)
1456 return _notmuch_sha1_of_string (path);
1461 /* Given a path, split it into two parts: the directory part is all
1462 * components except for the last, and the basename is that last
1463 * component. Getting the return-value for either part is optional
1464 * (the caller can pass NULL).
1466 * The original 'path' can represent either a regular file or a
1467 * directory---the splitting will be carried out in the same way in
1468 * either case. Trailing slashes on 'path' will be ignored, and any
1469 * cases of multiple '/' characters appearing in series will be
1470 * treated as a single '/'.
1472 * Allocation (if any) will have 'ctx' as the talloc owner. But
1473 * pointers will be returned within the original path string whenever
1476 * Note: If 'path' is non-empty and contains no non-trailing slash,
1477 * (that is, consists of a filename with no parent directory), then
1478 * the directory returned will be an empty string. However, if 'path'
1479 * is an empty string, then both directory and basename will be
1483 _notmuch_database_split_path (void *ctx,
1485 const char **directory,
1486 const char **basename)
1490 if (path == NULL || *path == '\0') {
1495 return NOTMUCH_STATUS_SUCCESS;
1498 /* Find the last slash (not counting a trailing slash), if any. */
1500 slash = path + strlen (path) - 1;
1502 /* First, skip trailing slashes. */
1503 while (slash != path) {
1510 /* Then, find a slash. */
1511 while (slash != path) {
1521 /* Finally, skip multiple slashes. */
1522 while (slash != path) {
1529 if (slash == path) {
1531 *directory = talloc_strdup (ctx, "");
1536 *directory = talloc_strndup (ctx, path, slash - path + 1);
1539 return NOTMUCH_STATUS_SUCCESS;
1542 /* Find the document ID of the specified directory.
1544 * If (flags & NOTMUCH_FIND_CREATE), a new directory document will be
1545 * created if one does not exist for 'path'. Otherwise, if the
1546 * directory document does not exist, this sets *directory_id to
1547 * ((unsigned int)-1) and returns NOTMUCH_STATUS_SUCCESS.
1550 _notmuch_database_find_directory_id (notmuch_database_t *notmuch,
1552 notmuch_find_flags_t flags,
1553 unsigned int *directory_id)
1555 notmuch_directory_t *directory;
1556 notmuch_status_t status;
1560 return NOTMUCH_STATUS_SUCCESS;
1563 directory = _notmuch_directory_create (notmuch, path, flags, &status);
1564 if (status || !directory) {
1569 *directory_id = _notmuch_directory_get_document_id (directory);
1571 notmuch_directory_destroy (directory);
1573 return NOTMUCH_STATUS_SUCCESS;
1577 _notmuch_database_get_directory_path (void *ctx,
1578 notmuch_database_t *notmuch,
1579 unsigned int doc_id)
1581 Xapian::Document document;
1583 document = find_document_for_doc_id (notmuch, doc_id);
1585 return talloc_strdup (ctx, document.get_data ().c_str ());
1588 /* Given a legal 'filename' for the database, (either relative to
1589 * database path or absolute with initial components identical to
1590 * database path), return a new string (with 'ctx' as the talloc
1591 * owner) suitable for use as a direntry term value.
1593 * If (flags & NOTMUCH_FIND_CREATE), the necessary directory documents
1594 * will be created in the database as needed. Otherwise, if the
1595 * necessary directory documents do not exist, this sets
1596 * *direntry to NULL and returns NOTMUCH_STATUS_SUCCESS.
1599 _notmuch_database_filename_to_direntry (void *ctx,
1600 notmuch_database_t *notmuch,
1601 const char *filename,
1602 notmuch_find_flags_t flags,
1605 const char *relative, *directory, *basename;
1606 Xapian::docid directory_id;
1607 notmuch_status_t status;
1609 relative = _notmuch_database_relative_path (notmuch, filename);
1611 status = _notmuch_database_split_path (ctx, relative,
1612 &directory, &basename);
1616 status = _notmuch_database_find_directory_id (notmuch, directory, flags,
1618 if (status || directory_id == (unsigned int)-1) {
1623 *direntry = talloc_asprintf (ctx, "%u:%s", directory_id, basename);
1625 return NOTMUCH_STATUS_SUCCESS;
1628 /* Given a legal 'path' for the database, return the relative path.
1630 * The return value will be a pointer to the original path contents,
1631 * and will be either the original string (if 'path' was relative) or
1632 * a portion of the string (if path was absolute and begins with the
1636 _notmuch_database_relative_path (notmuch_database_t *notmuch,
1639 const char *db_path, *relative;
1640 unsigned int db_path_len;
1642 db_path = notmuch_database_get_path (notmuch);
1643 db_path_len = strlen (db_path);
1647 if (*relative == '/') {
1648 while (*relative == '/' && *(relative+1) == '/')
1651 if (strncmp (relative, db_path, db_path_len) == 0)
1653 relative += db_path_len;
1654 while (*relative == '/')
1663 notmuch_database_get_directory (notmuch_database_t *notmuch,
1665 notmuch_directory_t **directory)
1667 notmuch_status_t status;
1669 if (directory == NULL)
1670 return NOTMUCH_STATUS_NULL_POINTER;
1674 *directory = _notmuch_directory_create (notmuch, path,
1675 NOTMUCH_FIND_LOOKUP, &status);
1676 } catch (const Xapian::Error &error) {
1677 fprintf (stderr, "A Xapian exception occurred getting directory: %s.\n",
1678 error.get_msg().c_str());
1679 notmuch->exception_reported = TRUE;
1680 status = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
1685 /* Allocate a document ID that satisfies the following criteria:
1687 * 1. The ID does not exist for any document in the Xapian database
1689 * 2. The ID was not previously returned from this function
1691 * 3. The ID is the smallest integer satisfying (1) and (2)
1693 * This function will trigger an internal error if these constraints
1694 * cannot all be satisfied, (that is, the pool of available document
1695 * IDs has been exhausted).
1698 _notmuch_database_generate_doc_id (notmuch_database_t *notmuch)
1700 assert (notmuch->last_doc_id >= notmuch->xapian_db->get_lastdocid ());
1702 notmuch->last_doc_id++;
1704 if (notmuch->last_doc_id == 0)
1705 INTERNAL_ERROR ("Xapian document IDs are exhausted.\n");
1707 return notmuch->last_doc_id;
1711 _notmuch_database_generate_thread_id (notmuch_database_t *notmuch)
1713 /* 16 bytes (+ terminator) for hexadecimal representation of
1714 * a 64-bit integer. */
1715 static char thread_id[17];
1716 Xapian::WritableDatabase *db;
1718 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1720 notmuch->last_thread_id++;
1722 sprintf (thread_id, "%016" PRIx64, notmuch->last_thread_id);
1724 db->set_metadata ("last_thread_id", thread_id);
1730 _get_metadata_thread_id_key (void *ctx, const char *message_id)
1732 if (strlen (message_id) > NOTMUCH_MESSAGE_ID_MAX)
1733 message_id = _notmuch_message_id_compressed (ctx, message_id);
1735 return talloc_asprintf (ctx, NOTMUCH_METADATA_THREAD_ID_PREFIX "%s",
1739 /* Find the thread ID to which the message with 'message_id' belongs.
1741 * Note: 'thread_id_ret' must not be NULL!
1742 * On success '*thread_id_ret' is set to a newly talloced string belonging to
1745 * Note: If there is no message in the database with the given
1746 * 'message_id' then a new thread_id will be allocated for this
1747 * message and stored in the database metadata, (where this same
1748 * thread ID can be looked up if the message is added to the database
1751 static notmuch_status_t
1752 _resolve_message_id_to_thread_id (notmuch_database_t *notmuch,
1754 const char *message_id,
1755 const char **thread_id_ret)
1757 notmuch_status_t status;
1758 notmuch_message_t *message;
1759 string thread_id_string;
1761 Xapian::WritableDatabase *db;
1763 status = notmuch_database_find_message (notmuch, message_id, &message);
1769 *thread_id_ret = talloc_steal (ctx,
1770 notmuch_message_get_thread_id (message));
1772 notmuch_message_destroy (message);
1774 return NOTMUCH_STATUS_SUCCESS;
1777 /* Message has not been seen yet.
1779 * We may have seen a reference to it already, in which case, we
1780 * can return the thread ID stored in the metadata. Otherwise, we
1781 * generate a new thread ID and store it there.
1783 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1784 metadata_key = _get_metadata_thread_id_key (ctx, message_id);
1785 thread_id_string = notmuch->xapian_db->get_metadata (metadata_key);
1787 if (thread_id_string.empty()) {
1788 *thread_id_ret = talloc_strdup (ctx,
1789 _notmuch_database_generate_thread_id (notmuch));
1790 db->set_metadata (metadata_key, *thread_id_ret);
1792 *thread_id_ret = talloc_strdup (ctx, thread_id_string.c_str());
1795 talloc_free (metadata_key);
1797 return NOTMUCH_STATUS_SUCCESS;
1800 static notmuch_status_t
1801 _merge_threads (notmuch_database_t *notmuch,
1802 const char *winner_thread_id,
1803 const char *loser_thread_id)
1805 Xapian::PostingIterator loser, loser_end;
1806 notmuch_message_t *message = NULL;
1807 notmuch_private_status_t private_status;
1808 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1810 find_doc_ids (notmuch, "thread", loser_thread_id, &loser, &loser_end);
1812 for ( ; loser != loser_end; loser++) {
1813 message = _notmuch_message_create (notmuch, notmuch,
1814 *loser, &private_status);
1815 if (message == NULL) {
1816 ret = COERCE_STATUS (private_status,
1817 "Cannot find document for doc_id from query");
1821 _notmuch_message_remove_term (message, "thread", loser_thread_id);
1822 _notmuch_message_add_term (message, "thread", winner_thread_id);
1823 _notmuch_message_sync (message);
1825 notmuch_message_destroy (message);
1831 notmuch_message_destroy (message);
1837 _my_talloc_free_for_g_hash (void *ptr)
1842 static notmuch_status_t
1843 _notmuch_database_link_message_to_parents (notmuch_database_t *notmuch,
1844 notmuch_message_t *message,
1845 notmuch_message_file_t *message_file,
1846 const char **thread_id)
1848 GHashTable *parents = NULL;
1849 const char *refs, *in_reply_to, *in_reply_to_message_id;
1850 const char *last_ref_message_id, *this_message_id;
1851 GList *l, *keys = NULL;
1852 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1854 parents = g_hash_table_new_full (g_str_hash, g_str_equal,
1855 _my_talloc_free_for_g_hash, NULL);
1856 this_message_id = notmuch_message_get_message_id (message);
1858 refs = _notmuch_message_file_get_header (message_file, "references");
1859 last_ref_message_id = parse_references (message,
1863 in_reply_to = _notmuch_message_file_get_header (message_file, "in-reply-to");
1864 in_reply_to_message_id = parse_references (message,
1866 parents, in_reply_to);
1868 /* For the parent of this message, use the last message ID of the
1869 * References header, if available. If not, fall back to the
1870 * first message ID in the In-Reply-To header. */
1871 if (last_ref_message_id) {
1872 _notmuch_message_add_term (message, "replyto",
1873 last_ref_message_id);
1874 } else if (in_reply_to_message_id) {
1875 _notmuch_message_add_term (message, "replyto",
1876 in_reply_to_message_id);
1879 keys = g_hash_table_get_keys (parents);
1880 for (l = keys; l; l = l->next) {
1881 char *parent_message_id;
1882 const char *parent_thread_id = NULL;
1884 parent_message_id = (char *) l->data;
1886 _notmuch_message_add_term (message, "reference",
1889 ret = _resolve_message_id_to_thread_id (notmuch,
1896 if (*thread_id == NULL) {
1897 *thread_id = talloc_strdup (message, parent_thread_id);
1898 _notmuch_message_add_term (message, "thread", *thread_id);
1899 } else if (strcmp (*thread_id, parent_thread_id)) {
1900 ret = _merge_threads (notmuch, *thread_id, parent_thread_id);
1910 g_hash_table_unref (parents);
1915 static notmuch_status_t
1916 _notmuch_database_link_message_to_children (notmuch_database_t *notmuch,
1917 notmuch_message_t *message,
1918 const char **thread_id)
1920 const char *message_id = notmuch_message_get_message_id (message);
1921 Xapian::PostingIterator child, children_end;
1922 notmuch_message_t *child_message = NULL;
1923 const char *child_thread_id;
1924 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
1925 notmuch_private_status_t private_status;
1927 find_doc_ids (notmuch, "reference", message_id, &child, &children_end);
1929 for ( ; child != children_end; child++) {
1931 child_message = _notmuch_message_create (message, notmuch,
1932 *child, &private_status);
1933 if (child_message == NULL) {
1934 ret = COERCE_STATUS (private_status,
1935 "Cannot find document for doc_id from query");
1939 child_thread_id = notmuch_message_get_thread_id (child_message);
1940 if (*thread_id == NULL) {
1941 *thread_id = talloc_strdup (message, child_thread_id);
1942 _notmuch_message_add_term (message, "thread", *thread_id);
1943 } else if (strcmp (*thread_id, child_thread_id)) {
1944 _notmuch_message_remove_term (child_message, "reference",
1946 _notmuch_message_sync (child_message);
1947 ret = _merge_threads (notmuch, *thread_id, child_thread_id);
1952 notmuch_message_destroy (child_message);
1953 child_message = NULL;
1958 notmuch_message_destroy (child_message);
1963 /* Fetch and clear the stored thread_id for message, or NULL if none. */
1965 _consume_metadata_thread_id (void *ctx, notmuch_database_t *notmuch,
1966 notmuch_message_t *message)
1968 const char *message_id;
1972 message_id = notmuch_message_get_message_id (message);
1973 metadata_key = _get_metadata_thread_id_key (ctx, message_id);
1975 /* Check if we have already seen related messages to this one.
1976 * If we have then use the thread_id that we stored at that time.
1978 stored_id = notmuch->xapian_db->get_metadata (metadata_key);
1979 if (stored_id.empty ()) {
1982 Xapian::WritableDatabase *db;
1984 db = static_cast <Xapian::WritableDatabase *> (notmuch->xapian_db);
1986 /* Clear the metadata for this message ID. We don't need it
1988 db->set_metadata (metadata_key, "");
1990 return talloc_strdup (ctx, stored_id.c_str ());
1994 /* Given a (mostly empty) 'message' and its corresponding
1995 * 'message_file' link it to existing threads in the database.
1997 * The first check is in the metadata of the database to see if we
1998 * have pre-allocated a thread_id in advance for this message, (which
1999 * would have happened if a message was previously added that
2000 * referenced this one).
2002 * Second, we look at 'message_file' and its link-relevant headers
2003 * (References and In-Reply-To) for message IDs.
2005 * Finally, we look in the database for existing message that
2006 * reference 'message'.
2008 * In all cases, we assign to the current message the first thread_id
2009 * found (through either parent or child). We will also merge any
2010 * existing, distinct threads where this message belongs to both,
2011 * (which is not uncommon when messages are processed out of order).
2013 * Finally, if no thread ID has been found through parent or child, we
2014 * call _notmuch_message_generate_thread_id to generate a new thread
2015 * ID. This should only happen for new, top-level messages, (no
2016 * References or In-Reply-To header in this message, and no previously
2017 * added message refers to this message).
2019 static notmuch_status_t
2020 _notmuch_database_link_message (notmuch_database_t *notmuch,
2021 notmuch_message_t *message,
2022 notmuch_message_file_t *message_file)
2024 void *local = talloc_new (NULL);
2025 notmuch_status_t status;
2026 const char *thread_id;
2028 /* Check if the message already had a thread ID */
2029 thread_id = _consume_metadata_thread_id (local, notmuch, message);
2031 _notmuch_message_add_term (message, "thread", thread_id);
2033 status = _notmuch_database_link_message_to_parents (notmuch, message,
2039 status = _notmuch_database_link_message_to_children (notmuch, message,
2044 /* If not part of any existing thread, generate a new thread ID. */
2045 if (thread_id == NULL) {
2046 thread_id = _notmuch_database_generate_thread_id (notmuch);
2048 _notmuch_message_add_term (message, "thread", thread_id);
2052 talloc_free (local);
2058 notmuch_database_add_message (notmuch_database_t *notmuch,
2059 const char *filename,
2060 notmuch_message_t **message_ret)
2062 notmuch_message_file_t *message_file;
2063 notmuch_message_t *message = NULL;
2064 notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS, ret2;
2065 notmuch_private_status_t private_status;
2067 const char *date, *header;
2068 const char *from, *to, *subject;
2069 char *message_id = NULL;
2072 *message_ret = NULL;
2074 ret = _notmuch_database_ensure_writable (notmuch);
2078 message_file = _notmuch_message_file_open (filename);
2079 if (message_file == NULL)
2080 return NOTMUCH_STATUS_FILE_ERROR;
2082 /* Adding a message may change many documents. Do this all
2084 ret = notmuch_database_begin_atomic (notmuch);
2088 /* Parse message up front to get better error status. */
2089 ret = _notmuch_message_file_parse (message_file);
2094 /* Before we do any real work, (especially before doing a
2095 * potential SHA-1 computation on the entire file's contents),
2096 * let's make sure that what we're looking at looks like an
2097 * actual email message.
2099 from = _notmuch_message_file_get_header (message_file, "from");
2100 subject = _notmuch_message_file_get_header (message_file, "subject");
2101 to = _notmuch_message_file_get_header (message_file, "to");
2103 if ((from == NULL || *from == '\0') &&
2104 (subject == NULL || *subject == '\0') &&
2105 (to == NULL || *to == '\0'))
2107 ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
2111 /* Now that we're sure it's mail, the first order of business
2112 * is to find a message ID (or else create one ourselves). */
2114 header = _notmuch_message_file_get_header (message_file, "message-id");
2115 if (header && *header != '\0') {
2116 message_id = _parse_message_id (message_file, header, NULL);
2118 /* So the header value isn't RFC-compliant, but it's
2119 * better than no message-id at all. */
2120 if (message_id == NULL)
2121 message_id = talloc_strdup (message_file, header);
2124 if (message_id == NULL ) {
2125 /* No message-id at all, let's generate one by taking a
2126 * hash over the file's contents. */
2127 char *sha1 = _notmuch_sha1_of_file (filename);
2129 /* If that failed too, something is really wrong. Give up. */
2131 ret = NOTMUCH_STATUS_FILE_ERROR;
2135 message_id = talloc_asprintf (message_file,
2136 "notmuch-sha1-%s", sha1);
2140 /* Now that we have a message ID, we get a message object,
2141 * (which may or may not reference an existing document in the
2144 message = _notmuch_message_create_for_message_id (notmuch,
2148 talloc_free (message_id);
2150 if (message == NULL) {
2151 ret = COERCE_STATUS (private_status,
2152 "Unexpected status value from _notmuch_message_create_for_message_id");
2156 _notmuch_message_add_filename (message, filename);
2158 /* Is this a newly created message object? */
2159 if (private_status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
2160 _notmuch_message_add_term (message, "type", "mail");
2162 ret = _notmuch_database_link_message (notmuch, message,
2167 date = _notmuch_message_file_get_header (message_file, "date");
2168 _notmuch_message_set_header_values (message, date, from, subject);
2170 ret = _notmuch_message_index_file (message, message_file);
2174 ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
2177 _notmuch_message_sync (message);
2178 } catch (const Xapian::Error &error) {
2179 fprintf (stderr, "A Xapian exception occurred adding message: %s.\n",
2180 error.get_msg().c_str());
2181 notmuch->exception_reported = TRUE;
2182 ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
2188 if ((ret == NOTMUCH_STATUS_SUCCESS ||
2189 ret == NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID) && message_ret)
2190 *message_ret = message;
2192 notmuch_message_destroy (message);
2196 _notmuch_message_file_close (message_file);
2198 ret2 = notmuch_database_end_atomic (notmuch);
2199 if ((ret == NOTMUCH_STATUS_SUCCESS ||
2200 ret == NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID) &&
2201 ret2 != NOTMUCH_STATUS_SUCCESS)
2208 notmuch_database_remove_message (notmuch_database_t *notmuch,
2209 const char *filename)
2211 notmuch_status_t status;
2212 notmuch_message_t *message;
2214 status = notmuch_database_find_message_by_filename (notmuch, filename,
2217 if (status == NOTMUCH_STATUS_SUCCESS && message) {
2218 status = _notmuch_message_remove_filename (message, filename);
2219 if (status == NOTMUCH_STATUS_SUCCESS)
2220 _notmuch_message_delete (message);
2221 else if (status == NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID)
2222 _notmuch_message_sync (message);
2224 notmuch_message_destroy (message);
2231 notmuch_database_find_message_by_filename (notmuch_database_t *notmuch,
2232 const char *filename,
2233 notmuch_message_t **message_ret)
2236 const char *prefix = _find_prefix ("file-direntry");
2237 char *direntry, *term;
2238 Xapian::PostingIterator i, end;
2239 notmuch_status_t status;
2241 if (message_ret == NULL)
2242 return NOTMUCH_STATUS_NULL_POINTER;
2244 if (! (notmuch->features & NOTMUCH_FEATURE_FILE_TERMS))
2245 return NOTMUCH_STATUS_UPGRADE_REQUIRED;
2247 /* return NULL on any failure */
2248 *message_ret = NULL;
2250 local = talloc_new (notmuch);
2253 status = _notmuch_database_filename_to_direntry (
2254 local, notmuch, filename, NOTMUCH_FIND_LOOKUP, &direntry);
2255 if (status || !direntry)
2258 term = talloc_asprintf (local, "%s%s", prefix, direntry);
2260 find_doc_ids_for_term (notmuch, term, &i, &end);
2263 notmuch_private_status_t private_status;
2265 *message_ret = _notmuch_message_create (notmuch, notmuch, *i,
2267 if (*message_ret == NULL)
2268 status = NOTMUCH_STATUS_OUT_OF_MEMORY;
2270 } catch (const Xapian::Error &error) {
2271 fprintf (stderr, "Error: A Xapian exception occurred finding message by filename: %s\n",
2272 error.get_msg().c_str());
2273 notmuch->exception_reported = TRUE;
2274 status = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
2278 talloc_free (local);
2280 if (status && *message_ret) {
2281 notmuch_message_destroy (*message_ret);
2282 *message_ret = NULL;
2287 notmuch_string_list_t *
2288 _notmuch_database_get_terms_with_prefix (void *ctx, Xapian::TermIterator &i,
2289 Xapian::TermIterator &end,
2292 int prefix_len = strlen (prefix);
2293 notmuch_string_list_t *list;
2295 list = _notmuch_string_list_create (ctx);
2296 if (unlikely (list == NULL))
2299 for (i.skip_to (prefix); i != end; i++) {
2300 /* Terminate loop at first term without desired prefix. */
2301 if (strncmp ((*i).c_str (), prefix, prefix_len))
2304 _notmuch_string_list_append (list, (*i).c_str () + prefix_len);
2311 notmuch_database_get_all_tags (notmuch_database_t *db)
2313 Xapian::TermIterator i, end;
2314 notmuch_string_list_t *tags;
2317 i = db->xapian_db->allterms_begin();
2318 end = db->xapian_db->allterms_end();
2319 tags = _notmuch_database_get_terms_with_prefix (db, i, end,
2320 _find_prefix ("tag"));
2321 _notmuch_string_list_sort (tags);
2322 return _notmuch_tags_create (db, tags);
2323 } catch (const Xapian::Error &error) {
2324 fprintf (stderr, "A Xapian exception occurred getting tags: %s.\n",
2325 error.get_msg().c_str());
2326 db->exception_reported = TRUE;