X-Git-Url: https://git.notmuchmail.org/git?p=notmuch;a=blobdiff_plain;f=lib%2Fmessage.cc;h=20b77580227b8c94a1b12b1a9e3cac050a2236ce;hp=ad2208976729d653fc5162ec8b6d0823e6b5e1a5;hb=e4379c43e21ff69dcdb66064fe3de83ae0642fc0;hpb=36161181df961bb09da8090c13c234145270e5bf diff --git a/lib/message.cc b/lib/message.cc index ad220897..20b77580 100644 --- a/lib/message.cc +++ b/lib/message.cc @@ -25,20 +25,23 @@ #include -struct _notmuch_message { +struct visible _notmuch_message { notmuch_database_t *notmuch; Xapian::docid doc_id; int frozen; char *message_id; char *thread_id; char *in_reply_to; - notmuch_filename_list_t *filename_list; + notmuch_string_list_t *tag_list; + notmuch_string_list_t *filename_term_list; + notmuch_string_list_t *filename_list; char *author; notmuch_message_file_t *message_file; notmuch_message_list_t *replies; unsigned long flags; Xapian::Document doc; + Xapian::termcount termpos; }; #define ARRAY_SIZE(arr) (sizeof (arr) / sizeof (arr[0])) @@ -101,6 +104,8 @@ _notmuch_message_create_for_document (const void *talloc_owner, message->message_id = NULL; message->thread_id = NULL; message->in_reply_to = NULL; + message->tag_list = NULL; + message->filename_term_list = NULL; message->filename_list = NULL; message->message_file = NULL; message->author = NULL; @@ -123,6 +128,7 @@ _notmuch_message_create_for_document (const void *talloc_owner, talloc_set_destructor (message, _notmuch_message_destructor); message->doc = doc; + message->termpos = 0; return message; } @@ -207,7 +213,6 @@ _notmuch_message_create_for_message_id (notmuch_database_t *notmuch, { notmuch_message_t *message; Xapian::Document doc; - Xapian::WritableDatabase *db; unsigned int doc_id; char *term; @@ -227,7 +232,6 @@ _notmuch_message_create_for_message_id (notmuch_database_t *notmuch, if (notmuch->mode == NOTMUCH_DATABASE_MODE_READ_ONLY) INTERNAL_ERROR ("Failure to ensure database is writable."); - db = static_cast (notmuch->xapian_db); try { doc.add_term (term, 0); talloc_free (term); @@ -254,41 +258,137 @@ _notmuch_message_create_for_message_id (notmuch_database_t *notmuch, return message; } -unsigned int -_notmuch_message_get_doc_id (notmuch_message_t *message) +static char * +_notmuch_message_get_term (notmuch_message_t *message, + Xapian::TermIterator &i, Xapian::TermIterator &end, + const char *prefix) { - return message->doc_id; -} + int prefix_len = strlen (prefix); + const char *term = NULL; + char *value; -const char * -notmuch_message_get_message_id (notmuch_message_t *message) -{ - Xapian::TermIterator i; + i.skip_to (prefix); - if (message->message_id) - return message->message_id; + if (i != end) + term = (*i).c_str (); - i = message->doc.termlist_begin (); - i.skip_to (_find_prefix ("id")); - - if (i == message->doc.termlist_end ()) - INTERNAL_ERROR ("Message with document ID of %d has no message ID.\n", - message->doc_id); + if (!term || strncmp (term, prefix, prefix_len)) + return NULL; - message->message_id = talloc_strdup (message, (*i).c_str () + 1); + value = talloc_strdup (message, term + prefix_len); #if DEBUG_DATABASE_SANITY i++; - if (i != message->doc.termlist_end () && - strncmp ((*i).c_str (), _find_prefix ("id"), - strlen (_find_prefix ("id"))) == 0) - { - INTERNAL_ERROR ("Mail (doc_id: %d) has duplicate message IDs", - message->doc_id); + if (i != end && strncmp ((*i).c_str (), prefix, prefix_len) == 0) { + INTERNAL_ERROR ("Mail (doc_id: %d) has duplicate %s terms: %s and %s\n", + message->doc_id, prefix, value, + (*i).c_str () + prefix_len); } #endif + return value; +} + +void +_notmuch_message_ensure_metadata (notmuch_message_t *message) +{ + Xapian::TermIterator i, end; + const char *thread_prefix = _find_prefix ("thread"), + *tag_prefix = _find_prefix ("tag"), + *id_prefix = _find_prefix ("id"), + *filename_prefix = _find_prefix ("file-direntry"), + *replyto_prefix = _find_prefix ("replyto"); + + /* We do this all in a single pass because Xapian decompresses the + * term list every time you iterate over it. Thus, while this is + * slightly more costly than looking up individual fields if only + * one field of the message object is actually used, it's a huge + * win as more fields are used. */ + + i = message->doc.termlist_begin (); + end = message->doc.termlist_end (); + + /* Get thread */ + if (!message->thread_id) + message->thread_id = + _notmuch_message_get_term (message, i, end, thread_prefix); + + /* Get tags */ + assert (strcmp (thread_prefix, tag_prefix) < 0); + if (!message->tag_list) { + message->tag_list = + _notmuch_database_get_terms_with_prefix (message, i, end, + tag_prefix); + _notmuch_string_list_sort (message->tag_list); + } + + /* Get id */ + assert (strcmp (tag_prefix, id_prefix) < 0); + if (!message->message_id) + message->message_id = + _notmuch_message_get_term (message, i, end, id_prefix); + + /* Get filename list. Here we get only the terms. We lazily + * expand them to full file names when needed in + * _notmuch_message_ensure_filename_list. */ + assert (strcmp (id_prefix, filename_prefix) < 0); + if (!message->filename_term_list && !message->filename_list) + message->filename_term_list = + _notmuch_database_get_terms_with_prefix (message, i, end, + filename_prefix); + + /* Get reply to */ + assert (strcmp (filename_prefix, replyto_prefix) < 0); + if (!message->in_reply_to) + message->in_reply_to = + _notmuch_message_get_term (message, i, end, replyto_prefix); + /* It's perfectly valid for a message to have no In-Reply-To + * header. For these cases, we return an empty string. */ + if (!message->in_reply_to) + message->in_reply_to = talloc_strdup (message, ""); +} + +static void +_notmuch_message_invalidate_metadata (notmuch_message_t *message, + const char *prefix_name) +{ + if (strcmp ("thread", prefix_name) == 0) { + talloc_free (message->thread_id); + message->thread_id = NULL; + } + + if (strcmp ("tag", prefix_name) == 0) { + talloc_unlink (message, message->tag_list); + message->tag_list = NULL; + } + + if (strcmp ("file-direntry", prefix_name) == 0) { + talloc_free (message->filename_term_list); + talloc_free (message->filename_list); + message->filename_term_list = message->filename_list = NULL; + } + + if (strcmp ("replyto", prefix_name) == 0) { + talloc_free (message->in_reply_to); + message->in_reply_to = NULL; + } +} + +unsigned int +_notmuch_message_get_doc_id (notmuch_message_t *message) +{ + return message->doc_id; +} + +const char * +notmuch_message_get_message_id (notmuch_message_t *message) +{ + if (!message->message_id) + _notmuch_message_ensure_metadata (message); + if (!message->message_id) + INTERNAL_ERROR ("Message with document ID of %u has no message ID.\n", + message->doc_id); return message->message_id; } @@ -327,89 +427,19 @@ notmuch_message_get_header (notmuch_message_t *message, const char *header) const char * _notmuch_message_get_in_reply_to (notmuch_message_t *message) { - const char *prefix = _find_prefix ("replyto"); - int prefix_len = strlen (prefix); - Xapian::TermIterator i; - std::string in_reply_to; - - if (message->in_reply_to) - return message->in_reply_to; - - i = message->doc.termlist_begin (); - i.skip_to (prefix); - - if (i != message->doc.termlist_end ()) - in_reply_to = *i; - - /* It's perfectly valid for a message to have no In-Reply-To - * header. For these cases, we return an empty string. */ - if (i == message->doc.termlist_end () || - strncmp (in_reply_to.c_str (), prefix, prefix_len)) - { - message->in_reply_to = talloc_strdup (message, ""); - return message->in_reply_to; - } - - message->in_reply_to = talloc_strdup (message, - in_reply_to.c_str () + prefix_len); - -#if DEBUG_DATABASE_SANITY - i++; - - in_reply_to = *i; - - if (i != message->doc.termlist_end () && - strncmp ((*i).c_str (), prefix, prefix_len) == 0) - { - INTERNAL_ERROR ("Message %s has duplicate In-Reply-To IDs: %s and %s\n", - notmuch_message_get_message_id (message), - message->in_reply_to, - (*i).c_str () + prefix_len); - } -#endif - + if (!message->in_reply_to) + _notmuch_message_ensure_metadata (message); return message->in_reply_to; } const char * notmuch_message_get_thread_id (notmuch_message_t *message) { - const char *prefix = _find_prefix ("thread"); - Xapian::TermIterator i; - std::string id; - - /* This code is written with the assumption that "thread" has a - * single-character prefix. */ - assert (strlen (prefix) == 1); - - if (message->thread_id) - return message->thread_id; - - i = message->doc.termlist_begin (); - i.skip_to (prefix); - - if (i != message->doc.termlist_end ()) - id = *i; - - if (i == message->doc.termlist_end () || id[0] != *prefix) - INTERNAL_ERROR ("Message with document ID of %d has no thread ID.\n", + if (!message->thread_id) + _notmuch_message_ensure_metadata (message); + if (!message->thread_id) + INTERNAL_ERROR ("Message with document ID of %u has no thread ID.\n", message->doc_id); - - message->thread_id = talloc_strdup (message, id.c_str () + 1); - -#if DEBUG_DATABASE_SANITY - i++; - id = *i; - - if (i != message->doc.termlist_end () && id[0] == *prefix) - { - INTERNAL_ERROR ("Message %s has duplicate thread IDs: %s and %s\n", - notmuch_message_get_message_id (message), - message->thread_id, - id.c_str () + 1); - } -#endif - return message->thread_id; } @@ -434,60 +464,152 @@ notmuch_status_t _notmuch_message_add_filename (notmuch_message_t *message, const char *filename) { + const char *relative, *directory; notmuch_status_t status; void *local = talloc_new (message); char *direntry; - if (message->filename_list) { - _notmuch_filename_list_destroy (message->filename_list); - message->filename_list = NULL; - } - if (filename == NULL) INTERNAL_ERROR ("Message filename cannot be NULL."); + relative = _notmuch_database_relative_path (message->notmuch, filename); + + status = _notmuch_database_split_path (local, relative, &directory, NULL); + if (status) + return status; + status = _notmuch_database_filename_to_direntry (local, message->notmuch, filename, &direntry); if (status) return status; + /* New file-direntry allows navigating to this message with + * notmuch_directory_get_child_files() . */ _notmuch_message_add_term (message, "file-direntry", direntry); + /* New terms allow user to search with folder: specification. */ + _notmuch_message_gen_terms (message, "folder", directory); + talloc_free (local); return NOTMUCH_STATUS_SUCCESS; } -/* Change a particular filename for 'message' from 'old_filename' to - * 'new_filename' +/* Remove a particular 'filename' from 'message'. * * This change will not be reflected in the database until the next * call to _notmuch_message_sync. - */ -static notmuch_status_t -_notmuch_message_rename (notmuch_message_t *message, - const char *old_filename, - const char *new_filename) + * + * If this message still has other filenames, returns + * NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID. + * + * Note: This function does not remove a document from the database, + * even if the specified filename is the only filename for this + * message. For that functionality, see + * _notmuch_database_remove_message. */ +notmuch_status_t +_notmuch_message_remove_filename (notmuch_message_t *message, + const char *filename) { + const char *direntry_prefix = _find_prefix ("file-direntry"); + int direntry_prefix_len = strlen (direntry_prefix); + const char *folder_prefix = _find_prefix ("folder"); + int folder_prefix_len = strlen (folder_prefix); void *local = talloc_new (message); + char *zfolder_prefix = talloc_asprintf(local, "Z%s", folder_prefix); + int zfolder_prefix_len = strlen (zfolder_prefix); char *direntry; notmuch_private_status_t private_status; notmuch_status_t status; - - status = _notmuch_message_add_filename (message, new_filename); - if (status) - return status; + Xapian::TermIterator i, last; status = _notmuch_database_filename_to_direntry (local, message->notmuch, - old_filename, &direntry); + filename, &direntry); if (status) return status; + /* Unlink this file from its parent directory. */ private_status = _notmuch_message_remove_term (message, "file-direntry", direntry); status = COERCE_STATUS (private_status, "Unexpected error from _notmuch_message_remove_term"); + if (status) + return status; + + /* Re-synchronize "folder:" terms for this message. This requires: + * 1. removing all "folder:" terms + * 2. removing all "folder:" stemmed terms + * 3. adding back terms for all remaining filenames of the message. */ + + /* 1. removing all "folder:" terms */ + while (1) { + i = message->doc.termlist_begin (); + i.skip_to (folder_prefix); + + /* Terminate loop when no terms remain with desired prefix. */ + if (i == message->doc.termlist_end () || + strncmp ((*i).c_str (), folder_prefix, folder_prefix_len)) + { + break; + } + + try { + message->doc.remove_term ((*i)); + } catch (const Xapian::InvalidArgumentError) { + /* Ignore failure to remove non-existent term. */ + } + } + + /* 2. removing all "folder:" stemmed terms */ + while (1) { + i = message->doc.termlist_begin (); + i.skip_to (zfolder_prefix); + + /* Terminate loop when no terms remain with desired prefix. */ + if (i == message->doc.termlist_end () || + strncmp ((*i).c_str (), zfolder_prefix, zfolder_prefix_len)) + { + break; + } + + try { + message->doc.remove_term ((*i)); + } catch (const Xapian::InvalidArgumentError) { + /* Ignore failure to remove non-existent term. */ + } + } + + /* 3. adding back terms for all remaining filenames of the message. */ + i = message->doc.termlist_begin (); + i.skip_to (direntry_prefix); + + for (; i != message->doc.termlist_end (); i++) { + unsigned int directory_id; + const char *direntry, *directory; + char *colon; + + /* Terminate loop at first term without desired prefix. */ + if (strncmp ((*i).c_str (), direntry_prefix, direntry_prefix_len)) + break; + + /* Indicate that there are filenames remaining. */ + status = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID; + + direntry = (*i).c_str (); + direntry += direntry_prefix_len; + + directory_id = strtol (direntry, &colon, 10); + + if (colon == NULL || *colon != ':') + INTERNAL_ERROR ("malformed direntry"); + + directory = _notmuch_database_get_directory_path (local, + message->notmuch, + directory_id); + if (strlen (directory)) + _notmuch_message_gen_terms (message, "folder", directory); + } talloc_free (local); @@ -509,21 +631,18 @@ _notmuch_message_clear_data (notmuch_message_t *message) static void _notmuch_message_ensure_filename_list (notmuch_message_t *message) { - const char *prefix = _find_prefix ("file-direntry"); - int prefix_len = strlen (prefix); - Xapian::TermIterator i; + notmuch_string_node_t *node; if (message->filename_list) return; - message->filename_list = _notmuch_filename_list_create (message); + if (!message->filename_term_list) + _notmuch_message_ensure_metadata (message); - i = message->doc.termlist_begin (); - i.skip_to (prefix); + message->filename_list = _notmuch_string_list_create (message); + node = message->filename_term_list->head; - if (i == message->doc.termlist_end () || - strncmp ((*i).c_str (), prefix, prefix_len)) - { + if (!node) { /* A message document created by an old version of notmuch * (prior to rename support) will have the filename in the * data of the document rather than as a file-direntry term. @@ -537,24 +656,18 @@ _notmuch_message_ensure_filename_list (notmuch_message_t *message) if (data == NULL) INTERNAL_ERROR ("message with no filename"); - _notmuch_filename_list_add_filename (message->filename_list, data); + _notmuch_string_list_append (message->filename_list, data); return; } - for (; i != message->doc.termlist_end (); i++) { + for (; node; node = node->next) { void *local = talloc_new (message); const char *db_path, *directory, *basename, *filename; char *colon, *direntry = NULL; unsigned int directory_id; - /* Terminate loop at first term without desired prefix. */ - if (strncmp ((*i).c_str (), prefix, prefix_len)) - break; - - direntry = talloc_strdup (local, (*i).c_str ()); - - direntry += prefix_len; + direntry = node->string; directory_id = strtol (direntry, &colon, 10); @@ -578,11 +691,13 @@ _notmuch_message_ensure_filename_list (notmuch_message_t *message) filename = talloc_asprintf (message, "%s/%s", db_path, basename); - _notmuch_filename_list_add_filename (message->filename_list, - filename); + _notmuch_string_list_append (message->filename_list, filename); talloc_free (local); } + + talloc_free (message->filename_term_list); + message->filename_term_list = NULL; } const char * @@ -594,12 +709,12 @@ notmuch_message_get_filename (notmuch_message_t *message) return NULL; if (message->filename_list->head == NULL || - message->filename_list->head->filename == NULL) + message->filename_list->head->string == NULL) { INTERNAL_ERROR ("message with no filename"); } - return message->filename_list->head->filename; + return message->filename_list->head->string; } notmuch_filenames_t * @@ -645,10 +760,20 @@ notmuch_message_get_date (notmuch_message_t *message) notmuch_tags_t * notmuch_message_get_tags (notmuch_message_t *message) { - Xapian::TermIterator i, end; - i = message->doc.termlist_begin(); - end = message->doc.termlist_end(); - return _notmuch_convert_tags(message, i, end); + notmuch_tags_t *tags; + + if (!message->tag_list) + _notmuch_message_ensure_metadata (message); + + tags = _notmuch_tags_create (message, message->tag_list); + /* _notmuch_tags_create steals the reference to the tag_list, but + * in this case it's still used by the message, so we add an + * *additional* talloc reference to the list. As a result, it's + * possible to modify the message tags (which talloc_unlink's the + * current list from the message) while still iterating because + * the iterator will keep the current list alive. */ + talloc_reference (message, message->tag_list); + return tags; } const char * @@ -738,12 +863,14 @@ _notmuch_message_add_term (notmuch_message_t *message, talloc_free (term); + _notmuch_message_invalidate_metadata (message, prefix_name); + return NOTMUCH_PRIVATE_STATUS_SUCCESS; } /* Parse 'text' and add a term to 'message' for each parsed word. Each * term will be added both prefixed (if prefix_name is not NULL) and - * also unprefixed). */ + * also non-prefixed). */ notmuch_private_status_t _notmuch_message_gen_terms (notmuch_message_t *message, const char *prefix_name, @@ -755,11 +882,13 @@ _notmuch_message_gen_terms (notmuch_message_t *message, return NOTMUCH_PRIVATE_STATUS_NULL_POINTER; term_gen->set_document (message->doc); + term_gen->set_termpos (message->termpos); if (prefix_name) { const char *prefix = _find_prefix (prefix_name); term_gen->index_text (text, 1, prefix); + message->termpos = term_gen->get_termpos (); } term_gen->index_text (text); @@ -801,6 +930,8 @@ _notmuch_message_remove_term (notmuch_message_t *message, talloc_free (term); + _notmuch_message_invalidate_metadata (message, prefix_name); + return NOTMUCH_PRIVATE_STATUS_SUCCESS; } @@ -1154,8 +1285,17 @@ notmuch_message_tags_to_maildir_flags (notmuch_message_t *message) if (err) continue; - new_status = _notmuch_message_rename (message, - filename, filename_new); + new_status = _notmuch_message_remove_filename (message, + filename); + /* Hold on to only the first error. */ + if (! status && new_status + && new_status != NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID) { + status = new_status; + continue; + } + + new_status = _notmuch_message_add_filename (message, + filename_new); /* Hold on to only the first error. */ if (! status && new_status) { status = new_status; @@ -1202,6 +1342,7 @@ notmuch_message_remove_all_tags (notmuch_message_t *message) if (! message->frozen) _notmuch_message_sync (message); + talloc_free (tags); return NOTMUCH_STATUS_SUCCESS; }