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 */
32 notmuch_status_to_string (notmuch_status_t status)
35 case NOTMUCH_STATUS_SUCCESS:
36 return "No error occurred";
37 case NOTMUCH_STATUS_XAPIAN_EXCEPTION:
38 return "A Xapian exception occurred";
39 case NOTMUCH_STATUS_FILE_NOT_EMAIL:
40 return "File is not an email";
41 case NOTMUCH_STATUS_NULL_POINTER:
42 return "Erroneous NULL pointer";
43 case NOTMUCH_STATUS_TAG_TOO_LONG:
44 return "Tag value is too long";
46 case NOTMUCH_STATUS_LAST_STATUS:
47 return "Unknown error status value";
51 /* "128 bits of thread-id ought to be enough for anybody" */
52 #define NOTMUCH_THREAD_ID_BITS 128
53 #define NOTMUCH_THREAD_ID_DIGITS (NOTMUCH_THREAD_ID_BITS / 4)
54 typedef struct _thread_id {
55 char str[NOTMUCH_THREAD_ID_DIGITS + 1];
59 thread_id_generate (thread_id_t *thread_id)
61 static int seeded = 0;
68 dev_random = fopen ("/dev/random", "r");
69 if (dev_random == NULL) {
72 fread ((void *) &value, sizeof (value), 1, dev_random);
80 for (i = 0; i < NOTMUCH_THREAD_ID_DIGITS; i += 8) {
82 sprintf (s, "%08x", value);
87 /* XXX: We should drop this function and convert all callers to call
88 * _notmuch_message_add_term instead. */
90 add_term (Xapian::Document doc,
91 const char *prefix_name,
100 prefix = _find_prefix (prefix_name);
102 term = g_strdup_printf ("%s%s", prefix, value);
104 if (strlen (term) <= NOTMUCH_TERM_MAX)
111 find_messages_by_term (Xapian::Database *db,
112 const char *prefix_name,
114 Xapian::PostingIterator *begin,
115 Xapian::PostingIterator *end)
117 Xapian::PostingIterator i;
120 term = g_strdup_printf ("%s%s", _find_prefix (prefix_name), value);
122 *begin = db->postlist_begin (term);
125 *end = db->postlist_end (term);
131 find_message_by_docid (Xapian::Database *db, Xapian::docid docid)
133 return db->get_document (docid);
137 notmuch_database_find_message (notmuch_database_t *notmuch,
138 const char *message_id)
140 Xapian::PostingIterator i, end;
142 find_messages_by_term (notmuch->xapian_db,
143 "msgid", message_id, &i, &end);
148 return _notmuch_message_create (notmuch, notmuch, *i);
151 /* Return one or more thread_ids, (as a GPtrArray of strings), for the
152 * given message based on looking into the database for any messages
153 * referenced in parents, and also for any messages in the database
154 * referencing message_id.
156 * Caller should free all strings in the array and the array itself,
157 * (g_ptr_array_free) when done. */
159 find_thread_ids (notmuch_database_t *notmuch,
161 const char *message_id)
163 Xapian::WritableDatabase *db = notmuch->xapian_db;
164 Xapian::PostingIterator child, children_end;
165 Xapian::Document doc;
166 GHashTable *thread_ids;
169 const char *parent_message_id;
172 thread_ids = g_hash_table_new_full (g_str_hash, g_str_equal,
175 find_messages_by_term (db, "ref", message_id, &child, &children_end);
176 for ( ; child != children_end; child++) {
177 const char *thread_id;
178 doc = find_message_by_docid (db, *child);
180 thread_id = doc.get_value (NOTMUCH_VALUE_THREAD).c_str ();
181 if (strlen (thread_id) == 0) {
182 fprintf (stderr, "Database error: Message with doc_id %u has empty thread-id value (value index %d)\n",
183 *child, NOTMUCH_VALUE_THREAD);
185 g_hash_table_insert (thread_ids, strdup (thread_id), NULL);
189 for (i = 0; i < parents->len; i++) {
190 notmuch_message_t *parent;
191 notmuch_thread_ids_t *ids;
193 parent_message_id = (char *) g_ptr_array_index (parents, i);
194 parent = notmuch_database_find_message (notmuch, parent_message_id);
198 for (ids = notmuch_message_get_thread_ids (parent);
199 notmuch_thread_ids_has_more (ids);
200 notmuch_thread_ids_advance (ids))
204 id = notmuch_thread_ids_get (ids);
205 g_hash_table_insert (thread_ids, strdup (id), NULL);
208 notmuch_message_destroy (parent);
211 result = g_ptr_array_new ();
213 keys = g_hash_table_get_keys (thread_ids);
214 for (l = keys; l; l = l->next) {
215 char *id = (char *) l->data;
216 g_ptr_array_add (result, id);
220 /* We're done with the hash table, but we've taken the pointers to
221 * the allocated strings and put them into our result array, so
222 * tell the hash not to free them on its way out. */
223 g_hash_table_steal_all (thread_ids);
224 g_hash_table_unref (thread_ids);
229 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
230 * a (potentially nested) parenthesized sequence with '\' used to
231 * escape any character (including parentheses).
233 * If the sequence to be skipped continues to the end of the string,
234 * then 'str' will be left pointing at the final terminating '\0'
238 skip_space_and_comments (const char **str)
243 while (*s && (isspace (*s) || *s == '(')) {
244 while (*s && isspace (*s))
249 while (*s && nesting) {
265 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
266 * comments, and the '<' and '>' delimeters.
268 * If not NULL, then *next will be made to point to the first character
269 * not parsed, (possibly pointing to the final '\0' terminator.
271 * Returns a newly allocated string which the caller should free()
274 * Returns NULL if there is any error parsing the message-id. */
276 parse_message_id (const char *message_id, const char **next)
281 if (message_id == NULL)
286 skip_space_and_comments (&s);
288 /* Skip any unstructured text as well. */
289 while (*s && *s != '<')
300 skip_space_and_comments (&s);
303 while (*end && *end != '>')
312 if (end > s && *end == '>')
317 result = strndup (s, end - s + 1);
319 /* Finally, collapse any whitespace that is within the message-id
325 for (r = result, len = strlen (r); *r; r++, len--)
326 if (*r == ' ' || *r == '\t')
327 memmove (r, r+1, len);
333 /* Parse a References header value, putting a copy of each referenced
334 * message-id into 'array'. */
336 parse_references (GPtrArray *array,
345 ref = parse_message_id (refs, &refs);
348 g_ptr_array_add (array, ref);
353 notmuch_database_default_path (void)
355 if (getenv ("NOTMUCH_BASE"))
356 return strdup (getenv ("NOTMUCH_BASE"));
358 return g_strdup_printf ("%s/mail", getenv ("HOME"));
362 notmuch_database_create (const char *path)
364 notmuch_database_t *notmuch = NULL;
365 char *notmuch_path = NULL;
368 char *local_path = NULL;
371 path = local_path = notmuch_database_default_path ();
373 err = stat (path, &st);
375 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
376 path, strerror (errno));
380 if (! S_ISDIR (st.st_mode)) {
381 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
386 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
388 err = mkdir (notmuch_path, 0755);
391 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
392 notmuch_path, strerror (errno));
396 notmuch = notmuch_database_open (path);
408 notmuch_database_open (const char *path)
410 notmuch_database_t *notmuch = NULL;
411 char *notmuch_path = NULL, *xapian_path = NULL;
414 char *local_path = NULL;
417 path = local_path = notmuch_database_default_path ();
419 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
421 err = stat (notmuch_path, &st);
423 fprintf (stderr, "Error opening database at %s: %s\n",
424 notmuch_path, strerror (errno));
428 xapian_path = g_strdup_printf ("%s/%s", notmuch_path, "xapian");
430 notmuch = talloc (NULL, notmuch_database_t);
431 notmuch->path = talloc_strdup (notmuch, path);
434 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
435 Xapian::DB_CREATE_OR_OPEN);
436 notmuch->query_parser = new Xapian::QueryParser;
437 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
438 notmuch->query_parser->set_database (*notmuch->xapian_db);
439 } catch (const Xapian::Error &error) {
440 fprintf (stderr, "A Xapian exception occurred: %s\n",
441 error.get_msg().c_str());
456 notmuch_database_close (notmuch_database_t *notmuch)
458 delete notmuch->query_parser;
459 delete notmuch->xapian_db;
460 talloc_free (notmuch);
464 notmuch_database_get_path (notmuch_database_t *notmuch)
466 return notmuch->path;
470 notmuch_database_add_message (notmuch_database_t *notmuch,
471 const char *filename)
473 Xapian::WritableDatabase *db = notmuch->xapian_db;
474 Xapian::Document doc;
475 notmuch_message_file_t *message;
477 GPtrArray *parents, *thread_ids;
479 const char *refs, *in_reply_to, *date, *header;
480 const char *from, *to, *subject;
486 message = notmuch_message_file_open (filename);
488 notmuch_message_file_restrict_headers (message,
498 doc.set_data (filename);
500 add_term (doc, "type", "mail");
502 parents = g_ptr_array_new ();
504 refs = notmuch_message_file_get_header (message, "references");
505 parse_references (parents, refs);
507 in_reply_to = notmuch_message_file_get_header (message, "in-reply-to");
508 parse_references (parents, in_reply_to);
510 for (i = 0; i < parents->len; i++)
511 add_term (doc, "ref", (char *) g_ptr_array_index (parents, i));
513 header = notmuch_message_file_get_header (message, "message-id");
515 message_id = parse_message_id (header, NULL);
516 /* So the header value isn't RFC-compliant, but it's
517 * better than no message-id at all. */
518 if (message_id == NULL)
519 message_id = xstrdup (header);
521 /* XXX: Should generate a message_id here, (such as a SHA1
522 * sum of the message itself) */
526 thread_ids = find_thread_ids (notmuch, parents, message_id);
528 for (i = 0; i < parents->len; i++)
529 g_free (g_ptr_array_index (parents, i));
530 g_ptr_array_free (parents, TRUE);
532 add_term (doc, "msgid", message_id);
533 doc.add_value (NOTMUCH_VALUE_MESSAGE_ID, message_id);
536 if (thread_ids->len) {
541 for (i = 0; i < thread_ids->len; i++) {
542 id = (char *) thread_ids->pdata[i];
543 add_term (doc, "thread", id);
545 thread_id = g_string_new (id);
547 g_string_append_printf (thread_id, ",%s", id);
551 doc.add_value (NOTMUCH_VALUE_THREAD, thread_id->str);
552 g_string_free (thread_id, TRUE);
553 } else if (message_id) {
554 /* If not part of any existing thread, generate a new thread_id. */
555 thread_id_t thread_id;
557 thread_id_generate (&thread_id);
558 add_term (doc, "thread", thread_id.str);
559 doc.add_value (NOTMUCH_VALUE_THREAD, thread_id.str);
562 g_ptr_array_free (thread_ids, TRUE);
566 date = notmuch_message_file_get_header (message, "date");
567 time_value = notmuch_parse_date (date, NULL);
569 doc.add_value (NOTMUCH_VALUE_DATE,
570 Xapian::sortable_serialise (time_value));
572 from = notmuch_message_file_get_header (message, "from");
573 subject = notmuch_message_file_get_header (message, "subject");
574 to = notmuch_message_file_get_header (message, "to");
580 notmuch_message_file_close (message);
581 return NOTMUCH_STATUS_FILE_NOT_EMAIL;
583 db->add_document (doc);
585 } catch (const Xapian::Error &error) {
586 fprintf (stderr, "A Xapian exception occurred: %s.\n",
587 error.get_msg().c_str());
588 return NOTMUCH_STATUS_XAPIAN_EXCEPTION;
591 notmuch_message_file_close (message);
593 return NOTMUCH_STATUS_SUCCESS;