Feature request: searching through MiniWiki
How it's done now
This Wiki site is starting to gather quite a few pages; 287 in fact. Currently, the template uses a simple HTMl form that relies on GoogleIndex for results.
The problem
When Google decides to index is unpredictable. And, results may differ depending on what particular cache and/or cache index farm is being used at that moment. So, the Search is quite useless. Currently, the closest bet is to go to List all Pages and use your browser's "Find in page" functionality to search for words like 'linux' or 'fvlug' or 'ssh'. It's the simplist thing that works, but... not the greatest way.
Possible Solutions?
- Use grep or maybe one of the Perl searching/finding modules to do a search through the RCS files. Or, should searching be done only on the working copies?
- Use perl regexes, or simple "fuzzy matching" on both the page name and the titles. There hopefully is a better way than calling a shell to do searches or using an external program. Maybe keep pages and page titles in a hash in memory that can be searched against? Each time a page is added simply do "$allpages{$node} = $firstline" each time a page is saved, so new pages are added and title changes are kept, and you have something you can search through quickly at any time. Memory considerations though. Maybe a delayed loading? --AlanBailward
Back to MiniWiki or FeaturesWantedInMiniWiki