First of all a big thanks to Upayavira for listening to my ramblings on optimization strategies and creating well done concrete implementation of them with excellent additional ideas.
Since he was not satisfied, he has also converted the CLI to commons-cli, added a CLI config file (cli.xconf), and created the Cocoon bean.
Hey, do we have a new Cocooner here? :-)
- o -
Now for the actual figures.
The 'original' run is done by taking Cocoon-2.1 CVS and doing build docs. The numbers are taken from the CLI run-time figure, not the Ant one.
To try the 'optimized' one, edit cli.xconf by changing confirm-extensions="true" to false. What does this entail? That all files without extensions are outputted as-is, without appending a proper extension. In sites that always use explicit extension, it makes no difference.
+---------------+---------------+----------+
| original | optimized | diff |
+---------------+---------------+----------+
| | | |
memory | 65 Mb | 40 Mb | *-30%* |
time | 51 sec | 103 sec | *-51%* |
| | | |
+---------------+---------------+----------+These figures are in line with what was estimated before venturing on this path, so it figures.
This amount though gets lower if PDFs are involved as in Forrest, because their links are the same of the respective pages, so they just add an overhead that is the same in both cases, ruining the % advantage.
51 seconds. A record.
And BTW it also makes links not have .html appended if mime-type is not recognized, as asked by many Forrest users.
Someone wants to try it on Forrest and see the speed increase? ;-)
We need though to make the Forrest sitemap much faster, because it's *terribly* (5x) slower. With these two enhancements, it can finally start to please more users with speed :-)
--
Nicola Ken Barozzi [EMAIL PROTECTED]
- verba volant, scripta manent -
(discussions get forgotten, just code remains)
---------------------------------------------------------------------