From: Austin Clements <amdragon@MIT.EDU>
To: notmuch@notmuchmail.org
Subject: [PATCH 4/5] Add the file name list to the unified message metadata
	pass.
Date: Thu,  9 Dec 2010 15:59:55 -0500
Message-Id: <1291928396-27937-5-git-send-email-amdragon@mit.edu>
In-Reply-To: <1291928396-27937-1-git-send-email-amdragon@mit.edu>
References: <1291928396-27937-1-git-send-email-amdragon@mit.edu>
Cc: Austin Clements <amdragon@mit.edu>
Precedence: list

Even if the caller never uses the file names, there is little cost to
simply fetching the file name terms.  However, retrieving the full
paths requires additional database work, so the expansion from terms
to full paths is performed lazily.

This also simplifies clearing the filename cache, since that's now
handled by the generic metadata cache code.

This further reduces my inbox search from 3.102 seconds before the
unified metadata pass to 2.206 seconds (1.4X faster).
---
 lib/message.cc |   52 +++++++++++++++++++++++++++++-----------------------
 1 files changed, 29 insertions(+), 23 deletions(-)

diff --git a/lib/message.cc b/lib/message.cc
index dbf683c..adb205f 100644
--- a/lib/message.cc
+++ b/lib/message.cc
@@ -32,6 +32,7 @@ struct _notmuch_message {
     char *message_id;
     char *thread_id;
     char *in_reply_to;
+    notmuch_string_list_t *filename_term_list;
     notmuch_string_list_t *filename_list;
     char *author;
     notmuch_message_file_t *message_file;
@@ -101,6 +102,7 @@ _notmuch_message_create_for_document (const void *talloc_owner,
     message->message_id = NULL;
     message->thread_id = NULL;
     message->in_reply_to = NULL;
+    message->filename_term_list = NULL;
     message->filename_list = NULL;
     message->message_file = NULL;
     message->author = NULL;
@@ -292,6 +294,7 @@ _notmuch_message_ensure_metadata (notmuch_message_t *message)
     Xapian::TermIterator i, end;
     const char *thread_prefix = _find_prefix ("thread"),
 	*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
@@ -314,8 +317,16 @@ _notmuch_message_ensure_metadata (notmuch_message_t *message)
 	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_get_terms_with_prefix (message, i, end, filename_prefix);
+
     /* Get reply to */
-    assert (strcmp (id_prefix, replyto_prefix) < 0);
+    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);
@@ -334,6 +345,12 @@ _notmuch_message_invalidate_metadata (notmuch_message_t *message,
 	message->thread_id = 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;
@@ -433,11 +450,6 @@ _notmuch_message_add_filename (notmuch_message_t *message,
     void *local = talloc_new (message);
     char *direntry;
 
-    if (message->filename_list) {
-	talloc_free (message->filename_list);
-	message->filename_list = NULL;
-    }
-
     if (filename == NULL)
 	INTERNAL_ERROR ("Message filename cannot be NULL.");
 
@@ -504,21 +516,18 @@ _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_string_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.
@@ -537,19 +546,13 @@ _notmuch_message_ensure_filename_list (notmuch_message_t *message)
 	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);
 
@@ -577,6 +580,9 @@ _notmuch_message_ensure_filename_list (notmuch_message_t *message)
 
 	talloc_free (local);
     }
+
+    talloc_free (message->filename_term_list);
+    message->filename_term_list = NULL;
 }
 
 const char *
-- 
1.7.2.3