* Ended the role of $wgLinkCache in link updates. Instead, links (and related entities) are registered in the ParserOutput object during a parse. The LinksUpdate constructor now takes a ParserOutput object as a parameter. $wgLinkCache is still used, but only as a cache of article IDs.
* Because the link list is now saved and restored in the parser cache, meta tag keywords now work on parser cache hits. Some refactoring took place in this area.
* Rendering of the HTML for category links has moved from Parser to OutputPage.
* Did some general pottering around in Article.php, such as allowing an Article object to be created with a specified revision ID, thereby optionally removing the dependence on $wgRequest. Not used at the current time.
* A few documentation tweaks.
dealing with moves and deletes of widely-linked pages.
The updaters should be fixed up to understand future versions without the
tables there without breaking upgrades.
* DB_WRITE now called DB_MASTER, DB_READ now called DB_SLAVE
* Converted to use SQL wrapper functions instead of direct SQL in various places
* Experimental method for preserving the chronological order of events when slave servers are used. Untested.
* Fixes to the new post-parse existence test feature
* Some.. other stuff
_page_ links to category pages like [[:category:Some cat]] will stay in
links/brokenlinks and not be listed in the category page list anymore.
A link can optionally specify a sort key like this: [[category:cat|Sort me]].
The page will then be sorted in category lists according to the given text
instead of the page name; thus categories may be sorted arbitrarily, by
last name or whatever.
There is also a timestamp field included on gwicke's request; this is not
used yet. Also it will currently be updated at every edit of the page,
this can be changed by implementing differential updating for the cat links.
TODO:
* Make sure that automatic sort keys are updated on page rename.
* Make sure cateory pages get cache-invalidated and purged
* Use proper text sorting instead of raw binary sort
* Allow specification of the display style on category pages: comma list
or bullet list or number list; by sortkey or by timestamp; etc
This saves trouble in a number of places where we can now do joins with the link tables to get other info (such as cur_is_redirect!) as well as the name, and fewer bits need to be juggled on page renaming, as outgoing links no longer have to be changed (cur_id remains the same when a page is renamed).
rebuildLinks.inc and some of the tools in the 'maintenance page' still need to be updated to work with the new setup. (Special:Maintenance needs a *lot* of cleanup in general. It's kind of a catch-all of vaguely defined features which suck performance like a hydroelectric dam.)
Also I've slipped in some extra debug code. And, I think 'indexes.sql' is a big waste of time and should all be moved into tables.sql. Building indexes separately doesn't help on InnoDB and won't do anything on MyISAM either if you're just going to replace the table after it's built with an imported one from a dump which creates it with indexes.
corrupt entries in the table. Also, include the column names in
the REPLACE statement to guard against potential ordering problems
in databases that are manually munged around.
