- this plugin is based on sqlite and in description of project I found this sentence from author: "However there might be some bugs making this (refresh of data from txt page) not always 100% reliable." Not very nice to read that!
Problem is, that external edits not via the browser are not recognized by the plugin. a "normal" setup should do the job ;-)
- there is no automatic way to delete and rebuild whole database from pages
Thats true. To rebuild the data manual you have to delete the database and visit all pages with data in the browser. Workaround is something like to crawl the hole page with wget or lynx. May dokuwikis own cli indexer in the bin folder do the job.
beside that - to rebuild the hole database is not the normal use case with that plugin.
- only one table can be used for one page
i'm not sure about it, but maybe we fixed that. you can simply try that and look at the result. maybe its just working ;-)
Basically, I like this plugin, but I'm afraid to use it. And it seems to me that basic idea to have all data in simple txt files is the most powerfull advantage of dokuwiki, and this plugin do not follow that idea.
That really akward on the data plugin. it basically duplicates the data of the wiki syntax. The data will be ever in the page source. The SQLite database is only used as a "cache" to render the aggregated views. But at least the sqlite database is stored in one file.
When using only the wiki pages you will have big performance issues.
In one of the latest releases DokuWiki has introduced to search the metadata of pages with the fulltext search engine. Its a quite new feature and quite unknown to developers. sadly the data plugin was developed before that feature. that would be a quite nice alternative.