1 /* thread.cc - Results of thread-based searches from a notmuch database
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 "notmuch-private.h"
22 #include "database-private.h"
26 #include <gmime/gmime.h>
27 #include <glib.h> /* GHashTable */
29 struct _notmuch_thread {
30 notmuch_database_t *notmuch;
33 GHashTable *authors_hash;
35 char *nonmatched_authors;
38 notmuch_message_list_t *message_list;
39 GHashTable *message_hash;
47 _notmuch_thread_destructor (notmuch_thread_t *thread)
49 g_hash_table_unref (thread->authors_hash);
50 g_hash_table_unref (thread->tags);
51 g_hash_table_unref (thread->message_hash);
57 _thread_add_author (notmuch_thread_t *thread,
63 if (g_hash_table_lookup_extended (thread->authors_hash,
67 g_hash_table_insert (thread->authors_hash, xstrdup (author), NULL);
70 thread->authors = talloc_asprintf (thread, "%s, %s",
74 thread->authors = talloc_strdup (thread, author);
78 * move authors of matched messages in the thread to
79 * the front of the authors list, but keep them in
80 * existing order within their group
83 _thread_move_matched_author (notmuch_thread_t *thread,
88 char *last_pipe,*next_pipe;
89 int idx,nm_start,author_len,authors_len;
91 if (thread->authors == NULL || author == NULL)
93 if (thread->nonmatched_authors == NULL)
94 thread->nonmatched_authors = thread->authors;
95 author_len = strlen(author);
96 authors_len = strlen(thread->authors);
97 if (author_len == authors_len) {
99 thread->nonmatched_authors += author_len;
102 current_author = strcasestr(thread->authors, author);
103 if (current_author == NULL)
105 /* how far inside the nonmatched authors is our author? */
106 idx = current_author - thread->nonmatched_authors;
108 /* already among matched authors */
111 /* are there any authors in the list after our author? */
112 if (thread->nonmatched_authors + author_len < thread->authors + authors_len) {
113 /* we have to make changes, so let's get a temp copy */
114 authors_copy = talloc_strdup(thread,thread->authors);
115 /* nm_start is the offset into where the non-matched authors start */
116 nm_start = thread->nonmatched_authors - thread->authors;
117 /* copy this author and add the "| " - the if clause above tells us there's more */
118 strncpy(thread->nonmatched_authors,author,author_len);
119 strncpy(thread->nonmatched_authors+author_len,"| ",2);
120 thread->nonmatched_authors += author_len+2;
122 /* we are actually moving authors around, not just changing the separator
123 * first copy the authors that came BEFORE our author */
124 strncpy(thread->nonmatched_authors, authors_copy+nm_start, idx-2);
125 /* finally, if there are authors AFTER our author, copy those */
126 if(author_len+nm_start+idx < authors_len) {
127 strncpy(thread->nonmatched_authors + idx - 2,", ",2);
128 strncpy(thread->nonmatched_authors + idx, authors_copy+nm_start + idx + author_len + 2,
129 authors_len - (nm_start + idx + author_len + 2));
132 /* finally let's make sure there's just one '|' in the authors string */
133 last_pipe = strchr(thread->authors,'|');
135 next_pipe = strchr(last_pipe+1,'|');
138 last_pipe = next_pipe;
141 thread->nonmatched_authors += author_len;
142 /* so now all authors are matched - let's remove the '|' */
143 last_pipe = strchr(thread->authors,'|');
150 /* clean up the uggly "Lastname, Firstname" format that some mail systems
151 * (most notably, Exchange) are creating to be "Firstname Lastname"
152 * To make sure that we don't change other potential situations where a
153 * comma is in the name, we check that we match one of these patterns
154 * "Last, First" <first.last@company.com>
155 * "Last, First MI" <first.mi.last@company.com>
158 _thread_cleanup_author (notmuch_thread_t *thread,
159 const char *author, const char *from)
161 char *clean_author,*test_author;
166 clean_author = talloc_strdup(thread, author);
167 if (clean_author == NULL)
169 comma = strchr(author,',');
171 /* let's assemble what we think is the correct name */
172 lname = comma - author;
173 fname = strlen(author) - lname - 2;
174 strncpy(clean_author, comma + 2, fname);
175 *(clean_author+fname) = ' ';
176 strncpy(clean_author + fname + 1, author, lname);
177 *(clean_author+fname+1+lname) = '\0';
178 /* make a temporary copy and see if it matches the email */
179 test_author = talloc_strdup(thread,clean_author);
181 blank=strchr(test_author,' ');
182 while (blank != NULL) {
184 blank=strchr(test_author,' ');
186 if (strcasestr(from, test_author) == NULL)
187 /* we didn't identify this as part of the email address
188 * so let's punt and return the original author */
189 strcpy (clean_author, author);
195 /* Add 'message' as a message that belongs to 'thread'.
197 * The 'thread' will talloc_steal the 'message' and hold onto a
201 _thread_add_message (notmuch_thread_t *thread,
202 notmuch_message_t *message)
204 notmuch_tags_t *tags;
206 InternetAddressList *list = NULL;
207 InternetAddress *address;
208 const char *from, *author;
211 _notmuch_message_list_add_message (thread->message_list,
212 talloc_steal (thread, message));
213 thread->total_messages++;
215 g_hash_table_insert (thread->message_hash,
216 xstrdup (notmuch_message_get_message_id (message)),
219 from = notmuch_message_get_header (message, "from");
221 list = internet_address_list_parse_string (from);
224 address = internet_address_list_get_address (list, 0);
226 author = internet_address_get_name (address);
227 if (author == NULL) {
228 InternetAddressMailbox *mailbox;
229 mailbox = INTERNET_ADDRESS_MAILBOX (address);
230 author = internet_address_mailbox_get_addr (mailbox);
232 clean_author = _thread_cleanup_author (thread, author, from);
233 _thread_add_author (thread, clean_author);
234 notmuch_message_set_author (message, clean_author);
236 g_object_unref (G_OBJECT (list));
239 if (! thread->subject) {
241 subject = notmuch_message_get_header (message, "subject");
242 thread->subject = talloc_strdup (thread, subject ? subject : "");
245 for (tags = notmuch_message_get_tags (message);
246 notmuch_tags_valid (tags);
247 notmuch_tags_move_to_next (tags))
249 tag = notmuch_tags_get (tags);
250 g_hash_table_insert (thread->tags, xstrdup (tag), NULL);
255 _thread_set_subject_from_message (notmuch_thread_t *thread,
256 notmuch_message_t *message)
259 const char *cleaned_subject;
261 subject = notmuch_message_get_header (message, "subject");
265 if ((strncasecmp (subject, "Re: ", 4) == 0) ||
266 (strncasecmp (subject, "Aw: ", 4) == 0) ||
267 (strncasecmp (subject, "Vs: ", 4) == 0) ||
268 (strncasecmp (subject, "Sv: ", 4) == 0)) {
270 cleaned_subject = talloc_strndup (thread,
272 strlen(subject) - 4);
274 cleaned_subject = talloc_strdup (thread, subject);
278 talloc_free (thread->subject);
280 thread->subject = talloc_strdup (thread, cleaned_subject);
284 _thread_add_matched_message (notmuch_thread_t *thread,
285 notmuch_message_t *message,
289 notmuch_message_t *hashed_message;
291 date = notmuch_message_get_date (message);
293 if (date < thread->oldest || ! thread->matched_messages)
294 thread->oldest = date;
296 if (date > thread->newest || ! thread->matched_messages)
297 thread->newest = date;
299 thread->matched_messages++;
301 if (g_hash_table_lookup_extended (thread->message_hash,
302 notmuch_message_get_message_id (message), NULL,
303 (void **) &hashed_message)) {
304 notmuch_message_set_flag (hashed_message,
305 NOTMUCH_MESSAGE_FLAG_MATCH, 1);
307 _thread_move_matched_author (thread,notmuch_message_get_author(hashed_message));
309 if ((sort == NOTMUCH_SORT_OLDEST_FIRST && date <= thread->newest) ||
310 (sort != NOTMUCH_SORT_OLDEST_FIRST && date == thread->newest))
312 _thread_set_subject_from_message (thread, message);
317 _resolve_thread_relationships (unused (notmuch_thread_t *thread))
319 notmuch_message_node_t **prev, *node;
320 notmuch_message_t *message, *parent;
321 const char *in_reply_to;
323 prev = &thread->message_list->head;
324 while ((node = *prev)) {
325 message = node->message;
326 in_reply_to = _notmuch_message_get_in_reply_to (message);
327 if (in_reply_to && strlen (in_reply_to) &&
328 g_hash_table_lookup_extended (thread->message_hash,
333 if (thread->message_list->tail == &node->next)
334 thread->message_list->tail = prev;
336 _notmuch_message_add_reply (parent, node);
338 prev = &((*prev)->next);
342 /* XXX: After scanning through the entire list looking for parents
343 * via "In-Reply-To", we should do a second pass that looks at the
344 * list of messages IDs in the "References" header instead. (And
345 * for this the parent would be the "deepest" message of all the
346 * messages found in the "References" list.)
348 * Doing this will allow messages and sub-threads to be positioned
349 * correctly in the thread even when an intermediate message is
350 * missing from the thread.
354 /* Create a new notmuch_thread_t object for the given thread ID,
355 * treating any messages matching 'query_string' as "matched".
357 * Creating the thread will trigger two database searches. The first
358 * is for all messages belonging to the thread, (to get the first
359 * subject line, the total count of messages, and all authors). The
360 * second search is for all messages that are in the thread and that
361 * also match the given query_string. This is to allow for a separate
362 * count of matched messages, and to allow a viewer to display these
363 * messages differently.
365 * Here, 'ctx' is talloc context for the resulting thread object.
367 * This function returns NULL in the case of any error.
370 _notmuch_thread_create (void *ctx,
371 notmuch_database_t *notmuch,
372 const char *thread_id,
373 const char *query_string,
376 notmuch_thread_t *thread;
377 const char *thread_id_query_string;
378 notmuch_query_t *thread_id_query;
380 notmuch_messages_t *messages;
381 notmuch_message_t *message;
382 notmuch_bool_t matched_is_subset_of_thread;
384 thread_id_query_string = talloc_asprintf (ctx, "thread:%s", thread_id);
385 if (unlikely (query_string == NULL))
388 /* Under normal circumstances we need to do two database
389 * queries. One is for the thread itself (thread_id_query_string)
390 * and the second is to determine which messages in that thread
391 * match the original query (matched_query_string).
393 * But under two circumstances, we use only the
394 * thread_id_query_string:
396 * 1. If the original query_string *is* just the thread
399 * 2. If the original query_string matches all messages ("" or
402 * In either of these cases, we can be more efficient by running
403 * just the thread_id query (since we know all messages in the
404 * thread will match the query_string).
406 * Beyond the performance advantage, in the second case, it's
407 * important to not try to create a concatenated query because our
408 * parser handles "" and "*" as special cases and will not do the
409 * right thing with a query string of "* and thread:<foo>".
411 matched_is_subset_of_thread = 1;
412 if (strcmp (query_string, thread_id_query_string) == 0 ||
413 strcmp (query_string, "") == 0 ||
414 strcmp (query_string, "*") == 0)
416 matched_is_subset_of_thread = 0;
419 thread_id_query = notmuch_query_create (notmuch, thread_id_query_string);
420 if (unlikely (thread_id_query == NULL))
423 thread = talloc (ctx, notmuch_thread_t);
424 if (unlikely (thread == NULL))
427 talloc_set_destructor (thread, _notmuch_thread_destructor);
429 thread->notmuch = notmuch;
430 thread->thread_id = talloc_strdup (thread, thread_id);
431 thread->subject = NULL;
432 thread->authors_hash = g_hash_table_new_full (g_str_hash, g_str_equal,
434 thread->authors = NULL;
435 thread->nonmatched_authors = NULL;
436 thread->tags = g_hash_table_new_full (g_str_hash, g_str_equal,
439 thread->message_list = _notmuch_message_list_create (thread);
440 if (unlikely (thread->message_list == NULL))
443 thread->message_hash = g_hash_table_new_full (g_str_hash, g_str_equal,
446 thread->total_messages = 0;
447 thread->matched_messages = 0;
451 notmuch_query_set_sort (thread_id_query, NOTMUCH_SORT_OLDEST_FIRST);
453 for (messages = notmuch_query_search_messages (thread_id_query);
454 notmuch_messages_valid (messages);
455 notmuch_messages_move_to_next (messages))
457 message = notmuch_messages_get (messages);
459 _thread_add_message (thread, message);
461 if (! matched_is_subset_of_thread)
462 _thread_add_matched_message (thread, message, sort);
464 _notmuch_message_close (message);
467 notmuch_query_destroy (thread_id_query);
469 if (matched_is_subset_of_thread)
471 const char *matched_query_string;
472 notmuch_query_t *matched_query;
474 matched_query_string = talloc_asprintf (ctx, "%s AND (%s)",
475 thread_id_query_string,
477 if (unlikely (matched_query_string == NULL))
480 matched_query = notmuch_query_create (notmuch, matched_query_string);
481 if (unlikely (matched_query == NULL))
484 for (messages = notmuch_query_search_messages (matched_query);
485 notmuch_messages_valid (messages);
486 notmuch_messages_move_to_next (messages))
488 message = notmuch_messages_get (messages);
489 _thread_add_matched_message (thread, message, sort);
490 _notmuch_message_close (message);
493 notmuch_query_destroy (matched_query);
496 _resolve_thread_relationships (thread);
502 notmuch_thread_get_toplevel_messages (notmuch_thread_t *thread)
504 return _notmuch_messages_create (thread->message_list);
508 notmuch_thread_get_thread_id (notmuch_thread_t *thread)
510 return thread->thread_id;
514 notmuch_thread_get_total_messages (notmuch_thread_t *thread)
516 return thread->total_messages;
520 notmuch_thread_get_matched_messages (notmuch_thread_t *thread)
522 return thread->matched_messages;
526 notmuch_thread_get_authors (notmuch_thread_t *thread)
528 return thread->authors;
532 notmuch_thread_get_subject (notmuch_thread_t *thread)
534 return thread->subject;
538 notmuch_thread_get_oldest_date (notmuch_thread_t *thread)
540 return thread->oldest;
544 notmuch_thread_get_newest_date (notmuch_thread_t *thread)
546 return thread->newest;
550 notmuch_thread_get_tags (notmuch_thread_t *thread)
552 notmuch_tags_t *tags;
555 tags = _notmuch_tags_create (thread);
556 if (unlikely (tags == NULL))
559 keys = g_hash_table_get_keys (thread->tags);
561 for (l = keys; l; l = l->next)
562 _notmuch_tags_add_tag (tags, (char *) l->data);
566 _notmuch_tags_prepare_iterator (tags);
572 notmuch_thread_destroy (notmuch_thread_t *thread)
574 talloc_free (thread);