using namespace std;
-/* "128 bits of thread-id ought to be enough for anybody" */
-#define NOTMUCH_THREAD_ID_BITS 128
-#define NOTMUCH_THREAD_ID_DIGITS (NOTMUCH_THREAD_ID_BITS / 4)
-typedef struct _thread_id {
- char str[NOTMUCH_THREAD_ID_DIGITS + 1];
-} thread_id_t;
+#define ARRAY_SIZE(arr) (sizeof (arr) / sizeof (arr[0]))
-static void
-thread_id_generate (thread_id_t *thread_id)
+typedef struct {
+ const char *name;
+ const char *prefix;
+} prefix_t;
+
+/* Here's the current schema for our database:
+ *
+ * We currently have two different types of documents: mail and timestamps.
+ *
+ * Mail document
+ * -------------
+ * A mail document is associated with a particular email message file
+ * on disk. It is indexed with the following prefixed terms:
+ *
+ * Single terms of given prefix:
+ *
+ * type: mail
+ *
+ * id: Unique ID of mail, (from Message-ID header or generated
+ * as "notmuch-sha1-<sha1_sum_of_entire_file>.
+ *
+ * Multiple terms of given prefix:
+ *
+ * ref: The message IDs from all In-Reply-To and References
+ * headers in the message.
+ *
+ * tag: Any tags associated with this message by the user.
+ *
+ * thread: The thread ID of all threads to which the mail belongs
+ *
+ * A mail document also has two values:
+ *
+ * TIMESTAMP: The time_t value corresponding to the message's
+ * Date header.
+ *
+ * MESSAGE_ID: The unique ID of the mail mess (see "id" above)
+ *
+ * Timestamp document
+ * ------------------
+ * A timestamp document is used by a client of the notmuch library to
+ * maintain data necessary to allow for efficient polling of mail
+ * directories. The notmuch library does no interpretation of
+ * timestamps, but merely allows the user to store and retrieve
+ * timestamps as name/value pairs.
+ *
+ * The timestamp document is indexed with a single prefixed term:
+ *
+ * timestamp: The user's key value (likely a directory name)
+ *
+ * and has a single value:
+ *
+ * TIMETAMPS: The time_t value from the user.
+ */
+
+/* With these prefix values we follow the conventions published here:
+ *
+ * http://xapian.org/docs/omega/termprefixes.html
+ *
+ * as much as makes sense. Note that I took some liberty in matching
+ * the reserved prefix values to notmuch concepts, (for example, 'G'
+ * is documented as "newsGroup (or similar entity - e.g. a web forum
+ * name)", for which I think the thread is the closest analogue in
+ * notmuch. This in spite of the fact that we will eventually be
+ * storing mailing-list messages where 'G' for "mailing list name"
+ * might be even a closer analogue. I'm treating the single-character
+ * prefixes preferentially for core notmuch concepts (which will be
+ * nearly universal to all mail messages).
+ */
+
+prefix_t BOOLEAN_PREFIX_INTERNAL[] = {
+ { "type", "T" },
+ { "thread", "G" },
+ { "ref", "XREFERENCE" },
+ { "timestamp", "XTIMESTAMP" },
+};
+
+prefix_t BOOLEAN_PREFIX_EXTERNAL[] = {
+ { "tag", "K" },
+ { "id", "Q" }
+};
+
+const char *
+_find_prefix (const char *name)
{
- static int seeded = 0;
- FILE *dev_random;
- uint32_t value;
- char *s;
- int i;
-
- if (! seeded) {
- dev_random = fopen ("/dev/random", "r");
- if (dev_random == NULL) {
- srand (time (NULL));
- } else {
- fread ((void *) &value, sizeof (value), 1, dev_random);
- srand (value);
- fclose (dev_random);
- }
- seeded = 1;
- }
+ unsigned int i;
+
+ for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_INTERNAL); i++)
+ if (strcmp (name, BOOLEAN_PREFIX_INTERNAL[i].name) == 0)
+ return BOOLEAN_PREFIX_INTERNAL[i].prefix;
+
+ for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++)
+ if (strcmp (name, BOOLEAN_PREFIX_EXTERNAL[i].name) == 0)
+ return BOOLEAN_PREFIX_EXTERNAL[i].prefix;
+
+ fprintf (stderr, "Internal error: No prefix exists for '%s'\n", name);
+ exit (1);
+
+ return "";
+}
- s = thread_id->str;
- for (i = 0; i < NOTMUCH_THREAD_ID_DIGITS; i += 8) {
- value = rand ();
- sprintf (s, "%08x", value);
- s += 8;
+const char *
+notmuch_status_to_string (notmuch_status_t status)
+{
+ switch (status) {
+ case NOTMUCH_STATUS_SUCCESS:
+ return "No error occurred";
+ case NOTMUCH_STATUS_XAPIAN_EXCEPTION:
+ return "A Xapian exception occurred";
+ case NOTMUCH_STATUS_FILE_ERROR:
+ return "Something went wrong trying to read or write a file";
+ case NOTMUCH_STATUS_FILE_NOT_EMAIL:
+ return "File is not an email";
+ case NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID:
+ return "Message ID is identical to a message in database";
+ case NOTMUCH_STATUS_NULL_POINTER:
+ return "Erroneous NULL pointer";
+ case NOTMUCH_STATUS_TAG_TOO_LONG:
+ return "Tag value is too long (exceeds NOTMUCH_TAG_MAX)";
+ default:
+ case NOTMUCH_STATUS_LAST_STATUS:
+ return "Unknown error status value";
}
}
+/* XXX: We should drop this function and convert all callers to call
+ * _notmuch_message_add_term instead. */
static void
add_term (Xapian::Document doc,
const char *prefix_name,
}
static void
-find_messages_by_term (Xapian::Database *db,
- const char *prefix_name,
- const char *value,
- Xapian::PostingIterator *begin,
- Xapian::PostingIterator *end)
+find_doc_ids (notmuch_database_t *notmuch,
+ const char *prefix_name,
+ const char *value,
+ Xapian::PostingIterator *begin,
+ Xapian::PostingIterator *end)
{
Xapian::PostingIterator i;
char *term;
term = g_strdup_printf ("%s%s", _find_prefix (prefix_name), value);
- *begin = db->postlist_begin (term);
+ *begin = notmuch->xapian_db->postlist_begin (term);
- if (end)
- *end = db->postlist_end (term);
+ *end = notmuch->xapian_db->postlist_end (term);
free (term);
}
-Xapian::Document
-find_message_by_docid (Xapian::Database *db, Xapian::docid docid)
+static notmuch_private_status_t
+find_unique_doc_id (notmuch_database_t *notmuch,
+ const char *prefix_name,
+ const char *value,
+ unsigned int *doc_id)
{
- return db->get_document (docid);
+ Xapian::PostingIterator i, end;
+
+ find_doc_ids (notmuch, prefix_name, value, &i, &end);
+
+ if (i == end) {
+ *doc_id = 0;
+ return NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND;
+ } else {
+ *doc_id = *i;
+ return NOTMUCH_PRIVATE_STATUS_SUCCESS;
+ }
}
-Xapian::Document
-find_message_by_message_id (Xapian::Database *db, const char *message_id)
+static Xapian::Document
+find_document_for_doc_id (notmuch_database_t *notmuch, unsigned doc_id)
{
- Xapian::PostingIterator i, end;
+ return notmuch->xapian_db->get_document (doc_id);
+}
- find_messages_by_term (db, "msgid", message_id, &i, &end);
+static notmuch_private_status_t
+find_unique_document (notmuch_database_t *notmuch,
+ const char *prefix_name,
+ const char *value,
+ Xapian::Document *document,
+ unsigned int *doc_id)
+{
+ notmuch_private_status_t status;
- if (i != end)
- return find_message_by_docid (db, *i);
- else
- return Xapian::Document ();
+ status = find_unique_doc_id (notmuch, prefix_name, value, doc_id);
+
+ if (status) {
+ *document = Xapian::Document ();
+ return status;
+ }
+
+ *document = find_document_for_doc_id (notmuch, *doc_id);
+ return NOTMUCH_PRIVATE_STATUS_SUCCESS;
}
+/* XXX: Should rewrite this to accept a notmuch_message_t* instead of
+ * a Xapian:Document and then we could just use
+ * notmuch_message_get_thread_ids instead of duplicating its logic
+ * here. */
static void
insert_thread_id (GHashTable *thread_ids, Xapian::Document doc)
{
string value_string;
- const char *value, *id, *comma;
-
- value_string = doc.get_value (NOTMUCH_VALUE_THREAD);
- value = value_string.c_str();
- if (strlen (value)) {
- id = value;
- while (*id) {
- comma = strchr (id, ',');
- if (comma == NULL)
- comma = id + strlen (id);
- g_hash_table_insert (thread_ids,
- strndup (id, comma - id), NULL);
- id = comma;
- if (*id)
- id++;
- }
+ Xapian::TermIterator i;
+ const char *prefix_str = _find_prefix ("thread");
+ char prefix;
+
+ assert (strlen (prefix_str) == 1);
+
+ prefix = *prefix_str;
+
+ i = doc.termlist_begin ();
+ i.skip_to (prefix_str);
+
+ while (1) {
+ if (i == doc.termlist_end ())
+ break;
+ value_string = *i;
+ if (value_string.empty () || value_string[0] != prefix)
+ break;
+ g_hash_table_insert (thread_ids,
+ strdup (value_string.c_str () + 1), NULL);
+ i++;
}
}
+notmuch_message_t *
+notmuch_database_find_message (notmuch_database_t *notmuch,
+ const char *message_id)
+{
+ notmuch_private_status_t status;
+ unsigned int doc_id;
+
+ status = find_unique_doc_id (notmuch, "id", message_id, &doc_id);
+
+ if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
+ return NULL;
+
+ return _notmuch_message_create (notmuch, notmuch, doc_id);
+}
+
/* Return one or more thread_ids, (as a GPtrArray of strings), for the
* given message based on looking into the database for any messages
* referenced in parents, and also for any messages in the database
* Caller should free all strings in the array and the array itself,
* (g_ptr_array_free) when done. */
static GPtrArray *
-find_thread_ids (Xapian::Database *db,
+find_thread_ids (notmuch_database_t *notmuch,
GPtrArray *parents,
const char *message_id)
{
thread_ids = g_hash_table_new_full (g_str_hash, g_str_equal,
free, NULL);
- find_messages_by_term (db, "ref", message_id, &child, &children_end);
+ find_doc_ids (notmuch, "ref", message_id, &child, &children_end);
for ( ; child != children_end; child++) {
- doc = find_message_by_docid (db, *child);
+ doc = find_document_for_doc_id (notmuch, *child);
insert_thread_id (thread_ids, doc);
}
for (i = 0; i < parents->len; i++) {
+ notmuch_message_t *parent;
+ notmuch_thread_ids_t *ids;
+
parent_message_id = (char *) g_ptr_array_index (parents, i);
- doc = find_message_by_message_id (db, parent_message_id);
- insert_thread_id (thread_ids, doc);
+ parent = notmuch_database_find_message (notmuch, parent_message_id);
+ if (parent == NULL)
+ continue;
+
+ for (ids = notmuch_message_get_thread_ids (parent);
+ notmuch_thread_ids_has_more (ids);
+ notmuch_thread_ids_advance (ids))
+ {
+ const char *id;
+
+ id = notmuch_thread_ids_get (ids);
+ g_hash_table_insert (thread_ids, strdup (id), NULL);
+ }
+
+ notmuch_message_destroy (parent);
}
result = g_ptr_array_new ();
struct stat st;
int err;
char *local_path = NULL;
+ unsigned int i;
if (path == NULL)
path = local_path = notmuch_database_default_path ();
notmuch->query_parser = new Xapian::QueryParser;
notmuch->query_parser->set_default_op (Xapian::Query::OP_AND);
notmuch->query_parser->set_database (*notmuch->xapian_db);
+
+ for (i = 0; i < ARRAY_SIZE (BOOLEAN_PREFIX_EXTERNAL); i++) {
+ prefix_t *prefix = &BOOLEAN_PREFIX_EXTERNAL[i];
+ notmuch->query_parser->add_boolean_prefix (prefix->name,
+ prefix->prefix);
+ }
} catch (const Xapian::Error &error) {
fprintf (stderr, "A Xapian exception occurred: %s\n",
error.get_msg().c_str());
return notmuch->path;
}
+notmuch_private_status_t
+find_timestamp_document (notmuch_database_t *notmuch, const char *db_key,
+ Xapian::Document *doc, unsigned int *doc_id)
+{
+ return find_unique_document (notmuch, "timestamp", db_key, doc, doc_id);
+}
+
+/* We allow the user to use arbitrarily long keys for timestamps,
+ * (they're for filesystem paths after all, which have no limit we
+ * know about). But we have a term-length limit. So if we exceed that,
+ * we'll use the SHA-1 of the user's key as the actual key for
+ * constructing a database term.
+ *
+ * Caution: This function returns a newly allocated string which the
+ * caller should free() when finished.
+ */
+static char *
+timestamp_db_key (const char *key)
+{
+ int term_len = strlen (_find_prefix ("timestamp")) + strlen (key);
+
+ if (term_len > NOTMUCH_TERM_MAX)
+ return notmuch_sha1_of_string (key);
+ else
+ return strdup (key);
+}
+
+notmuch_status_t
+notmuch_database_set_timestamp (notmuch_database_t *notmuch,
+ const char *key, time_t timestamp)
+{
+ Xapian::Document doc;
+ unsigned int doc_id;
+ notmuch_private_status_t status;
+ notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
+ char *db_key = NULL;
+
+ db_key = timestamp_db_key (key);
+
+ try {
+ status = find_timestamp_document (notmuch, db_key, &doc, &doc_id);
+
+ doc.add_value (NOTMUCH_VALUE_TIMESTAMP,
+ Xapian::sortable_serialise (timestamp));
+
+ if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND) {
+ char *term = talloc_asprintf (NULL, "%s%s",
+ _find_prefix ("timestamp"), db_key);
+ doc.add_term (term);
+ talloc_free (term);
+
+ notmuch->xapian_db->add_document (doc);
+ } else {
+ notmuch->xapian_db->replace_document (doc_id, doc);
+ }
+
+ } catch (Xapian::Error &error) {
+ fprintf (stderr, "A Xapian exception occurred: %s.\n",
+ error.get_msg().c_str());
+ ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
+ }
+
+ if (db_key)
+ free (db_key);
+
+ return ret;
+}
+
+time_t
+notmuch_database_get_timestamp (notmuch_database_t *notmuch, const char *key)
+{
+ Xapian::Document doc;
+ unsigned int doc_id;
+ notmuch_private_status_t status;
+ char *db_key = NULL;
+ time_t ret = 0;
+
+ db_key = timestamp_db_key (key);
+
+ try {
+ status = find_timestamp_document (notmuch, db_key, &doc, &doc_id);
+
+ if (status == NOTMUCH_PRIVATE_STATUS_NO_DOCUMENT_FOUND)
+ goto DONE;
+
+ ret = Xapian::sortable_unserialise (doc.get_value (NOTMUCH_VALUE_TIMESTAMP));
+ } catch (Xapian::Error &error) {
+ goto DONE;
+ }
+
+ DONE:
+ if (db_key)
+ free (db_key);
+
+ return ret;
+}
+
notmuch_status_t
notmuch_database_add_message (notmuch_database_t *notmuch,
const char *filename)
{
- Xapian::WritableDatabase *db = notmuch->xapian_db;
- Xapian::Document doc;
- notmuch_message_file_t *message;
+ notmuch_message_file_t *message_file;
+ notmuch_message_t *message;
+ notmuch_status_t ret = NOTMUCH_STATUS_SUCCESS;
GPtrArray *parents, *thread_ids;
const char *refs, *in_reply_to, *date, *header;
- const char *from, *to, *subject;
+ const char *from, *to, *subject, *old_filename;
char *message_id;
- time_t time_value;
unsigned int i;
- message = notmuch_message_file_open (filename);
+ message_file = notmuch_message_file_open (filename);
+ if (message_file == NULL) {
+ ret = NOTMUCH_STATUS_FILE_ERROR;
+ goto DONE;
+ }
- notmuch_message_file_restrict_headers (message,
+ notmuch_message_file_restrict_headers (message_file,
"date",
"from",
"in-reply-to",
"message-id",
"references",
"subject",
+ "to",
(char *) NULL);
try {
- doc.set_data (filename);
+ /* The first order of business is to find/create a message ID. */
- add_term (doc, "type", "mail");
+ header = notmuch_message_file_get_header (message_file, "message-id");
+ if (header) {
+ message_id = parse_message_id (header, NULL);
+ /* So the header value isn't RFC-compliant, but it's
+ * better than no message-id at all. */
+ if (message_id == NULL)
+ message_id = xstrdup (header);
+ } else {
+ /* No message-id at all, let's generate one by taking a
+ * hash over the file's contents. */
+ char *sha1 = notmuch_sha1_of_file (filename);
+
+ /* If that failed too, something is really wrong. Give up. */
+ if (sha1 == NULL) {
+ ret = NOTMUCH_STATUS_FILE_ERROR;
+ goto DONE;
+ }
+
+ message_id = g_strdup_printf ("notmuch-sha1-%s", sha1);
+ free (sha1);
+ }
+ /* Now that we have a message ID, we get a message object,
+ * (which may or may not reference an existing document in the
+ * database). */
+
+ /* Use NULL for owner since we want to free this locally. */
+
+ /* XXX: This call can fail by either out-of-memory or an
+ * "impossible" Xapian exception. We should rewrite it to
+ * allow us to propagate the error status. */
+ message = _notmuch_message_create_for_message_id (NULL, notmuch,
+ message_id);
+ if (message == NULL) {
+ fprintf (stderr, "Internal error. This shouldn't happen.\n\n");
+ fprintf (stderr, "I mean, it's possible you ran out of memory, but then this code path is still an internal error since it should have detected that and propagated the status value up the stack.\n");
+ exit (1);
+ }
+
+ /* Has a message previously been added with the same ID? */
+ old_filename = notmuch_message_get_filename (message);
+ if (old_filename && strlen (old_filename)) {
+ ret = NOTMUCH_STATUS_DUPLICATE_MESSAGE_ID;
+ goto DONE;
+ } else {
+ _notmuch_message_set_filename (message, filename);
+ _notmuch_message_add_term (message, "type", "mail");
+ }
+
+ /* Next, find the thread(s) to which this message belongs. */
parents = g_ptr_array_new ();
- refs = notmuch_message_file_get_header (message, "references");
+ refs = notmuch_message_file_get_header (message_file, "references");
parse_references (parents, refs);
- in_reply_to = notmuch_message_file_get_header (message, "in-reply-to");
+ in_reply_to = notmuch_message_file_get_header (message_file, "in-reply-to");
parse_references (parents, in_reply_to);
for (i = 0; i < parents->len; i++)
- add_term (doc, "ref", (char *) g_ptr_array_index (parents, i));
+ _notmuch_message_add_term (message, "ref",
+ (char *) g_ptr_array_index (parents, i));
- header = notmuch_message_file_get_header (message, "message-id");
- if (header) {
- message_id = parse_message_id (header, NULL);
- /* So the header value isn't RFC-compliant, but it's
- * better than no message-id at all. */
- if (message_id == NULL)
- message_id = xstrdup (header);
- } else {
- /* XXX: Should generate a message_id here, (such as a SHA1
- * sum of the message itself) */
- message_id = NULL;
- }
+ thread_ids = find_thread_ids (notmuch, parents, message_id);
- thread_ids = find_thread_ids (db, parents, message_id);
+ free (message_id);
for (i = 0; i < parents->len; i++)
g_free (g_ptr_array_index (parents, i));
g_ptr_array_free (parents, TRUE);
- if (message_id) {
- add_term (doc, "msgid", message_id);
- doc.add_value (NOTMUCH_VALUE_MESSAGE_ID, message_id);
- }
if (thread_ids->len) {
unsigned int i;
for (i = 0; i < thread_ids->len; i++) {
id = (char *) thread_ids->pdata[i];
- add_term (doc, "thread", id);
+ _notmuch_message_add_thread_id (message, id);
if (i == 0)
thread_id = g_string_new (id);
else
free (id);
}
- doc.add_value (NOTMUCH_VALUE_THREAD, thread_id->str);
g_string_free (thread_id, TRUE);
- } else if (message_id) {
- /* If not part of any existing thread, generate a new thread_id. */
- thread_id_t thread_id;
-
- thread_id_generate (&thread_id);
- add_term (doc, "thread", thread_id.str);
- doc.add_value (NOTMUCH_VALUE_THREAD, thread_id.str);
+ } else {
+ _notmuch_message_ensure_thread_id (message);
}
g_ptr_array_free (thread_ids, TRUE);
- free (message_id);
+ date = notmuch_message_file_get_header (message_file, "date");
+ _notmuch_message_set_date (message, date);
- date = notmuch_message_file_get_header (message, "date");
- time_value = notmuch_parse_date (date, NULL);
-
- doc.add_value (NOTMUCH_VALUE_DATE,
- Xapian::sortable_serialise (time_value));
-
- from = notmuch_message_file_get_header (message, "from");
- subject = notmuch_message_file_get_header (message, "subject");
- to = notmuch_message_file_get_header (message, "to");
+ from = notmuch_message_file_get_header (message_file, "from");
+ subject = notmuch_message_file_get_header (message_file, "subject");
+ to = notmuch_message_file_get_header (message_file, "to");
if (from == NULL &&
subject == NULL &&
to == NULL)
{
- notmuch_message_file_close (message);
- return NOTMUCH_STATUS_FILE_NOT_EMAIL;
+ ret = NOTMUCH_STATUS_FILE_NOT_EMAIL;
+ goto DONE;
} else {
- db->add_document (doc);
+ _notmuch_message_sync (message);
}
} catch (const Xapian::Error &error) {
fprintf (stderr, "A Xapian exception occurred: %s.\n",
error.get_msg().c_str());
- return NOTMUCH_STATUS_XAPIAN_EXCEPTION;
+ ret = NOTMUCH_STATUS_XAPIAN_EXCEPTION;
+ goto DONE;
}
- notmuch_message_file_close (message);
+ DONE:
+ if (message)
+ notmuch_message_destroy (message);
+ if (message_file)
+ notmuch_message_file_close (message_file);
- return NOTMUCH_STATUS_SUCCESS;
+ return ret;
}
projects / notmuch / blobdiff