> Each searches on the wiki use this function. Scanning the index directory > each time is time consuming with a constant series of disk access. > Switching a normal search to use file_exists 1 or more times, and not > readdir all the directory. > Switching a wildcard search to use a lengths.idx file containing all the > word lengths used in the wiki, file generated if a new configuration > parameter $conf[readdircache] is not false and fixed to a time in second. I like the changes in general, but would ask for a few tweaks before applying them: 1) the changes contain trailing whitespaces and tabs, please remove the former and convert the latter to spaces 2) please provide config metadata for the new config option 3) $conf['readdircache'] should default to 0 instead of false to avoid problems with the config manager 4) skip the /* Readdir cache improvment */ comment in dokuwiki.php 5) I would move the whole get or create the length index algorithm to it's own function, idx_indexLenghts() is quite complex already > I write also the simpletest to see the time improvments on my local install. > I think it's not really good but it tests also the function works like > expected. Excellent, I love it when patches come with unit tests :-) Andi -- splitbrain.org -- DokuWiki mailing list - more info at http://www.dokuwiki.org/mailinglist