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_MAX_TERM)
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 find_message_by_message_id (Xapian::Database *db, const char *message_id)
117 Xapian::PostingIterator i, end;
119 find_messages_by_term (db, "msgid", message_id, &i, &end);
122 return find_message_by_docid (db, *i);
124 return Xapian::Document ();
128 insert_thread_id (GHashTable *thread_ids, Xapian::Document doc)
131 const char *value, *id, *comma;
133 value_string = doc.get_value (NOTMUCH_VALUE_THREAD);
134 value = value_string.c_str();
135 if (strlen (value)) {
138 comma = strchr (id, ',');
140 comma = id + strlen (id);
141 g_hash_table_insert (thread_ids,
142 strndup (id, comma - id), NULL);
150 /* Return one or more thread_ids, (as a GPtrArray of strings), for the
151 * given message based on looking into the database for any messages
152 * referenced in parents, and also for any messages in the database
153 * referencing message_id.
155 * Caller should free all strings in the array and the array itself,
156 * (g_ptr_array_free) when done. */
158 find_thread_ids (Xapian::Database *db,
160 const char *message_id)
162 Xapian::PostingIterator child, children_end;
163 Xapian::Document doc;
164 GHashTable *thread_ids;
167 const char *parent_message_id;
170 thread_ids = g_hash_table_new_full (g_str_hash, g_str_equal,
173 find_messages_by_term (db, "ref", message_id, &child, &children_end);
174 for ( ; child != children_end; child++) {
175 doc = find_message_by_docid (db, *child);
176 insert_thread_id (thread_ids, doc);
179 for (i = 0; i < parents->len; i++) {
180 parent_message_id = (char *) g_ptr_array_index (parents, i);
181 doc = find_message_by_message_id (db, parent_message_id);
182 insert_thread_id (thread_ids, doc);
185 result = g_ptr_array_new ();
187 keys = g_hash_table_get_keys (thread_ids);
188 for (l = keys; l; l = l->next) {
189 char *id = (char *) l->data;
190 g_ptr_array_add (result, id);
194 /* We're done with the hash table, but we've taken the pointers to
195 * the allocated strings and put them into our result array, so
196 * tell the hash not to free them on its way out. */
197 g_hash_table_steal_all (thread_ids);
198 g_hash_table_unref (thread_ids);
203 /* Advance 'str' past any whitespace or RFC 822 comments. A comment is
204 * a (potentially nested) parenthesized sequence with '\' used to
205 * escape any character (including parentheses).
207 * If the sequence to be skipped continues to the end of the string,
208 * then 'str' will be left pointing at the final terminating '\0'
212 skip_space_and_comments (const char **str)
217 while (*s && (isspace (*s) || *s == '(')) {
218 while (*s && isspace (*s))
223 while (*s && nesting) {
239 /* Parse an RFC 822 message-id, discarding whitespace, any RFC 822
240 * comments, and the '<' and '>' delimeters.
242 * If not NULL, then *next will be made to point to the first character
243 * not parsed, (possibly pointing to the final '\0' terminator.
245 * Returns a newly allocated string which the caller should free()
248 * Returns NULL if there is any error parsing the message-id. */
250 parse_message_id (const char *message_id, const char **next)
255 if (message_id == NULL)
260 skip_space_and_comments (&s);
262 /* Skip any unstructured text as well. */
263 while (*s && *s != '<')
274 skip_space_and_comments (&s);
277 while (*end && *end != '>')
286 if (end > s && *end == '>')
291 result = strndup (s, end - s + 1);
293 /* Finally, collapse any whitespace that is within the message-id
299 for (r = result, len = strlen (r); *r; r++, len--)
300 if (*r == ' ' || *r == '\t')
301 memmove (r, r+1, len);
307 /* Parse a References header value, putting a copy of each referenced
308 * message-id into 'array'. */
310 parse_references (GPtrArray *array,
319 ref = parse_message_id (refs, &refs);
322 g_ptr_array_add (array, ref);
327 notmuch_database_default_path (void)
329 if (getenv ("NOTMUCH_BASE"))
330 return strdup (getenv ("NOTMUCH_BASE"));
332 return g_strdup_printf ("%s/mail", getenv ("HOME"));
336 notmuch_database_create (const char *path)
338 notmuch_database_t *notmuch = NULL;
339 char *notmuch_path = NULL;
342 char *local_path = NULL;
345 path = local_path = notmuch_database_default_path ();
347 err = stat (path, &st);
349 fprintf (stderr, "Error: Cannot create database at %s: %s.\n",
350 path, strerror (errno));
354 if (! S_ISDIR (st.st_mode)) {
355 fprintf (stderr, "Error: Cannot create database at %s: Not a directory.\n",
360 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
362 err = mkdir (notmuch_path, 0755);
365 fprintf (stderr, "Error: Cannot create directory %s: %s.\n",
366 notmuch_path, strerror (errno));
370 notmuch = notmuch_database_open (path);
382 notmuch_database_open (const char *path)
384 notmuch_database_t *notmuch = NULL;
385 char *notmuch_path = NULL, *xapian_path = NULL;
388 char *local_path = NULL;
391 path = local_path = notmuch_database_default_path ();
393 notmuch_path = g_strdup_printf ("%s/%s", path, ".notmuch");
395 err = stat (notmuch_path, &st);
397 fprintf (stderr, "Error opening database at %s: %s\n",
398 notmuch_path, strerror (errno));
402 xapian_path = g_strdup_printf ("%s/%s", notmuch_path, "xapian");
404 notmuch = talloc (NULL, notmuch_database_t);
405 notmuch->path = talloc_strdup (notmuch, path);
408 notmuch->xapian_db = new Xapian::WritableDatabase (xapian_path,
409 Xapian::DB_CREATE_OR_OPEN);
410 notmuch->query_parser = new Xapian::QueryParser;
411 notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
412 notmuch->query_parser->set_database (*notmuch->xapian_db);
413 } catch (const Xapian::Error &error) {
414 fprintf (stderr, "A Xapian exception occurred: %s\n",
415 error.get_msg().c_str());
430 notmuch_database_close (notmuch_database_t *notmuch)
432 delete notmuch->query_parser;
433 delete notmuch->xapian_db;
434 talloc_free (notmuch);
438 notmuch_database_get_path (notmuch_database_t *notmuch)
440 return notmuch->path;
444 notmuch_database_add_message (notmuch_database_t *notmuch,
445 const char *filename)
447 Xapian::WritableDatabase *db = notmuch->xapian_db;
448 Xapian::Document doc;
449 notmuch_message_file_t *message;
451 GPtrArray *parents, *thread_ids;
453 const char *refs, *in_reply_to, *date, *header;
454 const char *from, *to, *subject;
460 message = notmuch_message_file_open (filename);
462 notmuch_message_file_restrict_headers (message,
472 doc.set_data (filename);
474 add_term (doc, "type", "mail");
476 parents = g_ptr_array_new ();
478 refs = notmuch_message_file_get_header (message, "references");
479 parse_references (parents, refs);
481 in_reply_to = notmuch_message_file_get_header (message, "in-reply-to");
482 parse_references (parents, in_reply_to);
484 for (i = 0; i < parents->len; i++)
485 add_term (doc, "ref", (char *) g_ptr_array_index (parents, i));
487 header = notmuch_message_file_get_header (message, "message-id");
489 message_id = parse_message_id (header, NULL);
490 /* So the header value isn't RFC-compliant, but it's
491 * better than no message-id at all. */
492 if (message_id == NULL)
493 message_id = xstrdup (header);
495 /* XXX: Should generate a message_id here, (such as a SHA1
496 * sum of the message itself) */
500 thread_ids = find_thread_ids (db, parents, message_id);
502 for (i = 0; i < parents->len; i++)
503 g_free (g_ptr_array_index (parents, i));
504 g_ptr_array_free (parents, TRUE);
506 add_term (doc, "msgid", message_id);
507 doc.add_value (NOTMUCH_VALUE_MESSAGE_ID, message_id);
510 if (thread_ids->len) {
515 for (i = 0; i < thread_ids->len; i++) {
516 id = (char *) thread_ids->pdata[i];
517 add_term (doc, "thread", id);
519 thread_id = g_string_new (id);
521 g_string_append_printf (thread_id, ",%s", id);
525 doc.add_value (NOTMUCH_VALUE_THREAD, thread_id->str);
526 g_string_free (thread_id, TRUE);
527 } else if (message_id) {
528 /* If not part of any existing thread, generate a new thread_id. */
529 thread_id_t thread_id;
531 thread_id_generate (&thread_id);
532 add_term (doc, "thread", thread_id.str);
533 doc.add_value (NOTMUCH_VALUE_THREAD, thread_id.str);
536 g_ptr_array_free (thread_ids, TRUE);
540 date = notmuch_message_file_get_header (message, "date");
541 time_value = notmuch_parse_date (date, NULL);
543 doc.add_value (NOTMUCH_VALUE_DATE,
544 Xapian::sortable_serialise (time_value));
546 from = notmuch_message_file_get_header (message, "from");
547 subject = notmuch_message_file_get_header (message, "subject");
548 to = notmuch_message_file_get_header (message, "to");
554 notmuch_message_file_close (message);
555 return NOTMUCH_STATUS_FILE_NOT_EMAIL;
557 db->add_document (doc);
559 } catch (const Xapian::Error &error) {
560 fprintf (stderr, "A Xapian exception occurred: %s.\n",
561 error.get_msg().c_str());
562 return NOTMUCH_STATUS_XAPIAN_EXCEPTION;
565 notmuch_message_file_close (message);
567 return NOTMUCH_STATUS_SUCCESS;