From: SourceForge.net <no...@so...> - 2008-07-06 05:53:49
|
Plugin Patches item #2011261, was opened at 2008-07-05 15:34 Message generated for change (Comment added) made by shlomy You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=997937&aid=2011261&group_id=588 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Priority: 5 Private: No Submitted By: Kazutoshi Satoda (k_satoda) Assigned to: Nobody/Anonymous (nobody) Summary: CtagsSideKick: Avoid quadratic parsing time with large tags Initial Comment: I had encountered quadratic time consumption with large tags. This patch will avoid it. Such tags comes from a machine generated header file which has 40000+ macro definitions. The real problem is large memory consumption with such tags rather than time consumption. It climbs 50M+ and 100M+ for repeated parse. This patch make it easy to investigate the problem. ---------------------------------------------------------------------- >Comment By: Shlomy Reinstein (shlomy) Date: 2008-07-06 08:53 Message: Logged In: YES user_id=1477607 Originator: NO First, thanks for the patch! This patch replaces the current, linear child-lookup algorithm with a hash lookup; it trades memory for performance. Personally, I don't like maintaining the same data in two different places / structures. I currently don't have time for this, but maybe we can think of a better way to store the nodes in the first place, so we don't need two separate data structures? I also need a patch such as this, because I also work with several very large files. I just prefer to avoid redundancy when not really needed. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=997937&aid=2011261&group_id=588 |