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_strdup_printf, g_free, GPtrArray, GHashTable */
31 /* "128 bits of thread-id ought to be enough for anybody" */
32 #define NOTMUCH_THREAD_ID_BITS 128
33 #define NOTMUCH_THREAD_ID_DIGITS (NOTMUCH_THREAD_ID_BITS / 4)
34 typedef struct _thread_id {
35 char str[NOTMUCH_THREAD_ID_DIGITS + 1];
39 thread_id_generate (thread_id_t *thread_id)
41 static int seeded = 0;
48 dev_random = fopen ("/dev/random", "r");
49 if (dev_random == NULL) {
52 fread ((void *) &value, sizeof (value), 1, dev_random);
60 for (i = 0; i < NOTMUCH_THREAD_ID_DIGITS; i += 8) {
62 sprintf (s, "%08x", value);
68 add_term (Xapian::Document doc,
69 const char *prefix_name,
78 prefix = _find_prefix (prefix_name);
80 term = g_strdup_printf ("%s%s", prefix, value);
82 if (strlen (term) <= NOTMUCH_TERM_MAX)
89 find_messages_by_term (Xapian::Database *db,
90 const char *prefix_name,
92 Xapian::PostingIterator *begin,
93 Xapian::PostingIterator *end)
95 Xapian::PostingIterator i;
98 term = g_strdup_printf ("%s%s", _find_prefix (prefix_name), value);
100 *begin = db->postlist_begin (term);
103 *end = db->postlist_end (term);
109 find_message_by_docid (Xapian::Database *db, Xapian::docid docid)
111 return db->get_document (docid);
115 notmuch_database_find_message (notmuch_database_t *notmuch,
116 const char *message_id)
118 Xapian::PostingIterator i, end;
120 find_messages_by_term (notmuch->xapian_db,
121 "msgid", message_id, &i, &end);
126 return _notmuch_message_create (notmuch, notmuch, *i);
129 /* Return one or more thread_ids, (as a GPtrArray of strings), for the
130 * given message based on looking into the database for any messages
131 * referenced in parents, and also for any messages in the database
132 * referencing message_id.
134 * Caller should free all strings in the array and the array itself,
135 * (g_ptr_array_free) when done. */
137 find_thread_ids (notmuch_database_t *notmuch,
139 const char *message_id)
141 Xapian::WritableDatabase *db = notmuch->xapian_db;
142 Xapian::PostingIterator child, children_end;
143 Xapian::Document doc;
144 GHashTable *thread_ids;
147 const char *parent_message_id;
150 thread_ids = g_hash_table_new_full (g_str_hash, g_str_equal,
153 find_messages_by_term (db, "ref", message_id, &child, &children_end);
154 for ( ; child != children_end; child++) {
155 const char *thread_id;
156 doc = find_message_by_docid (db, *child);
158 thread_id = doc.get_value (NOTMUCH_VALUE_THREAD).c_str ();
159 if (strlen (thread_id) == 0) {
160 fprintf (stderr, "Database error: Message with doc_id %u has empty thread-id value (value index %d)\n",
161 *child, NOTMUCH_VALUE_THREAD);
163 g_hash_table_insert (thread_ids, strdup (thread_id), NULL);
167 for (i = 0; i < parents->len; i++) {
168 notmuch_message_t *parent;
169 notmuch_thread_ids_t *ids;
171 parent_message_id = (char *) g_ptr_array_index (parents, i);
172 parent = notmuch_database_find_message (notmuch, parent_message_id);
176 for (ids = notmuch_message_get_thread_ids (parent);
177 notmuch_thread_ids_has_more (ids);
178 notmuch_thread_ids_advance (ids))
182 id = notmuch_thread_ids_get (ids);
183 g_hash_table_insert (thread_ids, strdup (id), NULL);
186 notmuch_message_destroy (parent);
189 result = g_ptr_array_new ();
191 keys = g_hash_table_get_keys (thread_ids);
192 for (l = keys; l; l = l->next) {
193 char *id = (char *) l->data;
194 g_ptr_array_add (result, id);
198 /* We're done with the hash table, but we've taken the pointers to
199 * the allocated strings and put them into our result array, so
200 * tell the hash not to free them on its way out. */
201 g_hash_table_steal_all (thread_ids);
202 g_hash_table_unref (thread_ids);
207 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
208 * a (potentially nested) parenthesized sequence with '\' used to
209 * escape any character (including parentheses).
211 * If the sequence to be skipped continues to the end of the string,
212 * then 'str' will be left pointing at the final terminating '\0'
216 skip_space_and_comments (const char **str)
221 while (*s && (isspace (*s) || *s == '(')) {
222 while (*s && isspace (*s))
227 while (*s && nesting) {
243 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
244 * comments, and the '<' and '>' delimeters.
246 * If not NULL, then *next will be made to point to the first character
247 * not parsed, (possibly pointing to the final '\0' terminator.
249 * Returns a newly allocated string which the caller should free()
252 * Returns NULL if there is any error parsing the message-id. */
254 parse_message_id (const char *message_id, const char **next)
259 if (message_id == NULL)
264 skip_space_and_comments (&s);
266 /* Skip any unstructured text as well. */
267 while (*s && *s != '<')
278 skip_space_and_comments (&s);
281 while (*end && *end != '>')
290 if (end > s && *end == '>')
295 result = strndup (s, end - s + 1);
297 /* Finally, collapse any whitespace that is within the message-id
303 for (r = result, len = strlen (r); *r; r++, len--)
304 if (*r == ' ' || *r == '\t')
305 memmove (r, r+1, len);
311 /* Parse a References header value, putting a copy of each referenced
312 * message-id into 'array'. */
314 parse_references (GPtrArray *array,
323 ref = parse_message_id (refs, &refs);
326 g_ptr_array_add (array, ref);
331 notmuch_database_default_path (void)
333 if (getenv ("NOTMUCH_BASE"))
334 return strdup (getenv ("NOTMUCH_BASE"));
336 return g_strdup_printf ("%s/mail", getenv ("HOME"));
340 notmuch_database_create (const char *path)
342 notmuch_database_t *notmuch = NULL;
343 char *notmuch_path = NULL;
346 char *local_path = NULL;
349 path = local_path = notmuch_database_default_path ();
351 err = stat (path, &st);
353 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
354 path, strerror (errno));
358 if (! S_ISDIR (st.st_mode)) {
359 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
364 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
366 err = mkdir (notmuch_path, 0755);
369 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
370 notmuch_path, strerror (errno));
374 notmuch = notmuch_database_open (path);
386 notmuch_database_open (const char *path)
388 notmuch_database_t *notmuch = NULL;
389 char *notmuch_path = NULL, *xapian_path = NULL;
392 char *local_path = NULL;
395 path = local_path = notmuch_database_default_path ();
397 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
399 err = stat (notmuch_path, &st);
401 fprintf (stderr, "Error opening database at %s: %s\n",
402 notmuch_path, strerror (errno));
406 xapian_path = g_strdup_printf ("%s/%s", notmuch_path, "xapian");
408 notmuch = talloc (NULL, notmuch_database_t);
409 notmuch->path = talloc_strdup (notmuch, path);
412 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
413 Xapian::DB_CREATE_OR_OPEN);
414 notmuch->query_parser = new Xapian::QueryParser;
415 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
416 notmuch->query_parser->set_database (*notmuch->xapian_db);
417 } catch (const Xapian::Error &error) {
418 fprintf (stderr, "A Xapian exception occurred: %s\n",
419 error.get_msg().c_str());
434 notmuch_database_close (notmuch_database_t *notmuch)
436 delete notmuch->query_parser;
437 delete notmuch->xapian_db;
438 talloc_free (notmuch);
442 notmuch_database_get_path (notmuch_database_t *notmuch)
444 return notmuch->path;
448 notmuch_database_add_message (notmuch_database_t *notmuch,
449 const char *filename)
451 Xapian::WritableDatabase *db = notmuch->xapian_db;
452 Xapian::Document doc;
453 notmuch_message_file_t *message;
455 GPtrArray *parents, *thread_ids;
457 const char *refs, *in_reply_to, *date, *header;
458 const char *from, *to, *subject;
464 message = notmuch_message_file_open (filename);
466 notmuch_message_file_restrict_headers (message,
476 doc.set_data (filename);
478 add_term (doc, "type", "mail");
480 parents = g_ptr_array_new ();
482 refs = notmuch_message_file_get_header (message, "references");
483 parse_references (parents, refs);
485 in_reply_to = notmuch_message_file_get_header (message, "in-reply-to");
486 parse_references (parents, in_reply_to);
488 for (i = 0; i < parents->len; i++)
489 add_term (doc, "ref", (char *) g_ptr_array_index (parents, i));
491 header = notmuch_message_file_get_header (message, "message-id");
493 message_id = parse_message_id (header, NULL);
494 /* So the header value isn't RFC-compliant, but it's
495 * better than no message-id at all. */
496 if (message_id == NULL)
497 message_id = xstrdup (header);
499 /* XXX: Should generate a message_id here, (such as a SHA1
500 * sum of the message itself) */
504 thread_ids = find_thread_ids (notmuch, parents, message_id);
506 for (i = 0; i < parents->len; i++)
507 g_free (g_ptr_array_index (parents, i));
508 g_ptr_array_free (parents, TRUE);
510 add_term (doc, "msgid", message_id);
511 doc.add_value (NOTMUCH_VALUE_MESSAGE_ID, message_id);
514 if (thread_ids->len) {
519 for (i = 0; i < thread_ids->len; i++) {
520 id = (char *) thread_ids->pdata[i];
521 add_term (doc, "thread", id);
523 thread_id = g_string_new (id);
525 g_string_append_printf (thread_id, ",%s", id);
529 doc.add_value (NOTMUCH_VALUE_THREAD, thread_id->str);
530 g_string_free (thread_id, TRUE);
531 } else if (message_id) {
532 /* If not part of any existing thread, generate a new thread_id. */
533 thread_id_t thread_id;
535 thread_id_generate (&thread_id);
536 add_term (doc, "thread", thread_id.str);
537 doc.add_value (NOTMUCH_VALUE_THREAD, thread_id.str);
540 g_ptr_array_free (thread_ids, TRUE);
544 date = notmuch_message_file_get_header (message, "date");
545 time_value = notmuch_parse_date (date, NULL);
547 doc.add_value (NOTMUCH_VALUE_DATE,
548 Xapian::sortable_serialise (time_value));
550 from = notmuch_message_file_get_header (message, "from");
551 subject = notmuch_message_file_get_header (message, "subject");
552 to = notmuch_message_file_get_header (message, "to");
558 notmuch_message_file_close (message);
559 return NOTMUCH_STATUS_FILE_NOT_EMAIL;
561 db->add_document (doc);
563 } catch (const Xapian::Error &error) {
564 fprintf (stderr, "A Xapian exception occurred: %s.\n",
565 error.get_msg().c_str());
566 return NOTMUCH_STATUS_XAPIAN_EXCEPTION;
569 notmuch_message_file_close (message);
571 return NOTMUCH_STATUS_SUCCESS;