#!/usr/bin/env python
# -*- coding: utf-8  -*-
import wikipedia
import oursql
import cgitb; cgitb.enable(logdir='tracebacks')

connections = {}
def getConn(wiki):
	if not wiki.endswith('_p'): wiki+='_p'
	if wiki not in connections:
		connections[wiki] = oursql.connect(db=wiki, host=wiki.replace('_', '-') + '.userdb.toolserver.org', read_default_file='/home/dispenser/.my.cnf', charset=None, use_unicode=False)
	return connections[wiki]
def reconnect(wiki, host=""):
	if not wiki.endswith('_p'): wiki+='_p'
	del connection[host,wiki]
	return getConn(wiki, host)

def main():
	SysArgs = wikipedia.SysArgs
	limit  = SysArgs.get('limit', '250')
	offset = SysArgs.get('offset', '0')
	page   = wikipedia.Page(
			wikipedia.getSite(),
			SysArgs.get('category', SysArgs.get('page', 'Main_Page')), 
			defaultNamespace=14
	)
	site   = page.site()
	category = page.titleWithoutNamespace(underscore=True)
	categories = category.split('|')
	

	def createLink(title, label=None):
		return '<a href="//%s/wiki/%s" title="%s">%s</a>'%tuple(wikipedia.escape(s) for s in (site.hostname(), wikipedia.urllib.quote(title.replace(' ', '_'), safe=";@$!*(),/:-_."), title.replace('_', ' '), label or title.replace('_', ' '), ))
	
	conn = getConn(site.dbName())
	cursor = conn.cursor()
	try:
		cursor.execute("""/* LIMIT:20 */
SELECT cat_pages
FROM category
WHERE cat_title=?
""", (category,))
		((cat_pages,),) = cursor.fetchall()
		# FIXME cat_pages might be < limit
		print "<p>The following %s articles are in %s, out of %d total.  This list may not reflect recent changes (%s).</p>" % (
			wikipedia.escape(limit),
			createLink("Category:%s"%category),
			cat_pages,
			createLink("Wikipedia:FAQ/Categories#Why_might_a_category_list_not_be_up_to_date?", "learn more")
		)
	except Exception, e:
		wikipedia.output(repr(e))
	try:
		cursor.execute("""/* categorder LIMIT:300 NM */
SELECT 
  page_title AS "Article title",
  IFNULL(blc_count, "") AS "Incoming links",
  ROUND(page_len / 1024.0, 1) AS "KB"
FROM categorylinks
JOIN page ON page_id=cl_from
LEFT JOIN u_dispenser_p.backlinkcount_enwiki_p ON blc_from=cl_from
WHERE cl_to IN (%s)
AND cl_type="page"
AND page_namespace=0
ORDER BY blc_count DESC, cl_sortkey ASC
LIMIT ?, ?
""" % ','.join(('?',)*len(categories)), tuple(categories)+(offset, limit))
	except oursql.Error as (errno, strerror, extra):
		if errno == 1317:
			print '<p class="errormsg"> Query timed out. </p>'
			return
		else:
			(categories, offset, limit)
			raise 
	else:
		print '<table class="wikitable sortable" style="margin:1em auto;">'
		print '<tr>'
		for tup in cursor.description:
			print '<th>%s</th>'%tup[0]
		print '</tr>'
		for page_title, blc_count, size in cursor:
			print '''<tr>
<td>%s</td>
<td>%s</td>
<td>%6s</td>
</tr>''' % (
				createLink(page_title),
				blc_count,
				size,
			)
		print '</table>'
	
	cursor.execute("""/* categorder LIMIT:20 */
SELECT TIMESTAMPDIFF(DAY, CREATE_TIME, NOW())
FROM information_schema.tables
WHERE TABLE_SCHEMA='u_dispenser_p' AND TABLE_NAME="backlinkcount_enwiki_p"
""")
	print 'Counts of incoming links were updated %s days ago.'%cursor.fetchall().pop()


if __name__ == "__main__" and wikipedia.handleUrlAndHeader():
	try:
		wikipedia.startContent(form=True)
		main()
	except oursql.Error as (errno, strerror, extra):
		if errno in (1040, 1226, 2013): # Too many connection / Max connections / Lost connection
			print '<p>Database operational error, retrying a few minutes.</p><blockquote>%s</blockquote>'%(wikipedia.escape(strerror),)
			print '<script type="text/javascript">setTimeout("window.location.reload()", (Math.random()*4+1)*60*1000);</script>'
		else:
			raise
	finally:
		wikipedia.endContent()
		wikipedia.stopme()