All:
Has anyone implemented any usage reporting systems? Daily/weekly/ monthly periodic analysis?
- Something to track daily page creations / changes counts. - List new users - Maybe also fetch and deliver excerpts from housekeeping special:pages like: Special:Recentchanges Special:Newpages Special:Lonelypages - Pages with no link to them, orpaned Special:Deadendpages - Pages that have no external links Special:Uncategorizedpages - Pages lacking a category Special:Wantedpages Special:BrokenRedirects Special:Wantedcategories Dead End Pages ...
A lot of this could be gleaned from the xml.php API system -- a small PHP script using xml_*() functions could parse it out into human readable pages.
Thoughts?
Thanks,
[Intentional Top Post]
Here is example output from the code I've hacked together.
For some stats, I was able to use the XML API.php + SimpleXML:
$General_Stats_Suffix="api.php?format=xml&action=query&meta=siteinfo&siprop=statistics"; $today_pagecount=$doc->query->statistics['pages'];
For other things, I had to form custom queries and count() output array sizes:
$Count_Categories_Suffix="api.php?format=xml&action=query&list=allpages&apnamespace=14"; $Count_Template_Suffix="api.php?format=xml&action=query&list=allpages&apnamespace=10"; $today_categorycount=count($doc->query->allpages->p);
Some housekeeping pages provide nice simple RSS/ATOM XML feeds you can parse (New Pages, Recent Changes):
$doc = simplexml_load_file("$WIKIPrefixURI" . "index.php/Special:Newpages?feed=rss&limit=20");
For other housekeeping pages that don't provide this, you have to scrape HTML data structures using stupidity like:
$doc = simplexml_load_file($WIKI . 'index.php?title=Special:Wantedpages&limit=500&offset=0'); foreach ($doc->body->div->div->div->div->div[2]->ol->li as $listItem) { print " o " . $listItem->a . "\n"; }
If anyone wants to jointly develop a more-mature version, let me know.
Thanks, ~BAS
------------- Example Output -----------------------
Starting: : wiki_periodic_report.php 278 2008-02-07 19:32:42Z seklecki $
- Content Summary o Total Page Count: 344 (344) o Unique Articles: 122 (122) o Images/Media Objects: 140 (140) o Templates: 9 (9) o Category: 10 (10) - Activity Summary o Total Edits: 4874 (4874) o Total Views: 11149 (11149) - User Summary o Total Admins: 3 (3) o Total Users: 19 (19)
-- New Pages (20) --
o first page o second page o ....
-- Most Recent Changes (Not Yet Implemented) --
-- Begin Housekeeping Items --
-- Orphaned/Lonely Pages -- o Sex o Drugs o Unix...
-- Dead - End Pages -- o Windows users o Windows admins o ....
-- Uncategorized Pages -- o Prison Sex o PHP + SimpleXML + XHTML
-- Wanted Pages -- o SOAP-XML o SOAP-XML o SOAP-XML o SOAP-XML o SOAP-XML o SOAP-XML o SOAP-XML o ....
-- Wanted Categories -- o Systems Operations o C0ff33 o ....
On Tue, 2008-02-05 at 17:05 -0500, Brian A. Seklecki wrote:
All:
Has anyone implemented any usage reporting systems? Daily/weekly/ monthly periodic analysis?
- Something to track daily page creations / changes counts.
- List new users
- Maybe also fetch and deliver excerpts from housekeeping special:pages like: Special:Recentchanges Special:Newpages Special:Lonelypages - Pages with no link to them, orpaned Special:Deadendpages - Pages that have no external links Special:Uncategorizedpages - Pages lacking a category Special:Wantedpages Special:BrokenRedirects Special:Wantedcategories Dead End Pages ...
A lot of this could be gleaned from the xml.php API system -- a small PHP script using xml_*() functions could parse it out into human readable pages.
Thoughts?
Thanks,
mediawiki-enterprise@lists.wikimedia.org