#! /usr/bin/python
import cgi
import MySQLdb
import wikipedia

import cgitb; cgitb.enable()
db=None

def replag():
	""" Lazy man's Replag, determines the date of the newest entry
	"""
	c = db.cursor()
	c.execute("SELECT time_to_sec(timediff(now(),rev_timestamp)) as replag FROM revision ORDER BY rev_timestamp DESC LIMIT 1")
	return c.fetchone()[0]

def printLinks(offset, limit, queryString):
	basehref="/~dispenser/cgi-bin/geosearch.py?"+queryString
	if offset > 0:
		print '<p>View (<a href="%s&amp;offset=%d&amp;limit=%d">previous %d</a>) '% (basehref, offset-limit >=0 and offset-limit or 0, limit, limit)
	else:
		print '<p>View (previous %d) '% (limit, )
	if True:
		print '(<a href="%s&amp;offset=%d&amp;limit=%d">next %d</a>)' % (basehref, offset+limit, limit, limit)
	else:
		print '(next %d)' % (limit,)
	print '(<a href="%(basehref)s&amp;offset=%(offset)d&amp;limit=20">20</a> | <a href="%(basehref)s&amp;offset=%(offset)d&amp;limit=50">50</a> | <a href="%(basehref)s&amp;offset=%(offset)d&amp;limit=100">100</a> | <a href="%(basehref)s&amp;offset=%(offset)d&amp;limit=250">250</a> | <a href="%(basehref)s&amp;offset=%(offset)d&amp;limit=500">500</a>)</p>'% locals()

def main():
	form = cgi.FieldStorage()
	#SHOULDBE: /query -> ?regexp=query
	namespace = int(form.getfirst('namespace', 0))
	regexp	= form.getfirst('regexp', form.getfirst('page', ""))
	site	= wikipedia.Site(form.getfirst('site', 'en'))
	isRegex	= form.getfirst('type', 'like')=='regex'
	invert  = bool(form.getfirst('invert', False))
	limit	= int(form.getfirst('limit', 100))
	offset	= int(form.getfirst('offset', 0))

	print r'''
<script type="text/javascript" language="JavaScript">/*<![CDATA[*/
Autocheck = true
function checkregex(n){
	n.form.type.checked=(Autocheck && n.value.match(/[^\\][$^()|\[\].?*+{}]/));
	document.getElementById('slashmatch').style.backgroundColor=((n.form.type.checked && n.value.match(/[^\\](\\\\)*\\[bBdDsSwW]/))?"lightyellow":"");
}
/*]]>*/</script>
<form action="/~dispenser/cgi-bin/geosearch.py" style="text-align:left;">
<fieldset><legend>GeoHack External links</legend>
<p style="font-size:x-small;" id="slashmatch"><a href="http://en.wikipedia.org/wiki/Wikipedia:WikiProject_Geographical_coordinates/coordinates_search_tool" class="extiw">Examples</a>,
Cheat sheet: \w &#8594; [[:alnum:]], \d &#8594; [[:digit:]], \s &#8594; [[:space:]], \b &#8594; [[:&lt;:]] or [[:&gt;:]], search of parameter values: params=[^&amp;=]+ </p>
<p><label for="regexp">Search expression</label>:  <input type="text" name="regexp" id="regexp" size="50" value="%s" onkeyup="checkregex(this)" />
<input type="submit" value="Search" /></p>''' % (wikipedia.escape(regexp), )
	print '<input type="checkbox" name="type" id="type" %s value="regex" onclick="Autocheck=this.checked" /><label for="type">Regular expression</label>' % (isRegex and 'checked="checked" ' or '')
	print '<input type="checkbox" name="invert" id="invert" %s /><label for="invert">Invert logic</label>' % (invert and 'checked="checked" ' or '')
	if site.language() != 'en':
		print '<input type="hidden" name="site" value="%s" />' % site.language()
	print '</fieldset></form>'
	
#	if not regexp: return
	
	queryString = 'regexp='+wikipedia.urllib.quote(regexp)+(site.language()!='en' and '&site='+site.language() or '')
	
	global db
	db = MySQLdb.connect(db=site.dbName()+'_p', host=site.dbName().replace('_', '-') + '-p.db.toolserver.org', read_default_file="/home/dispenser/.my.cnf")

	c = db.cursor()
	c.execute("SET SESSION TRANSACTION ISOLATION LEVEL READ UNCOMMITTED")
#	subReg = regexp.replace('%', '\%').replace('.*', '%').replace('.', '_')
#	regex_charset = '$^*()+{}|[].?'
#	for char in regex_charset:
#		if char in subReg:
#			regexp=regexp
#			break
#	else:
#		# all Chars are non regex
#		isRegex = False
	if isRegex: # REGEXP 
		c.execute("""
SELECT el_to 
FROM externallinks 
/* JOIN page on el_from = page_id */
WHERE /* page_namespace=%s
AND */ el_to LIKE %s 
AND el_to """+(invert and 'NOT' or '')+""" REGEXP %s LIMIT %s,%s;""",
			(namespace, "http://stable.toolserver.org/geohack/geohack.php"+"?%", regexp, offset, limit))
	else: # LIKE
		c.execute("""
SELECT el_to 
FROM externallinks 
/* JOIN page on el_from = page_id */
WHERE /* page_namespace=%s
AND */ el_to LIKE %s 
AND el_to """+(invert and 'NOT' or '')+""" LIKE %s LIMIT %s,%s;""",
			(namespace, "http://stable.toolserver.org/geohack/geohack.php"+"?%", "%"+regexp+"%", offset, limit))

	printLinks(offset, limit, queryString)
		
	print '<table class="wikitable sortable">'
	print '<tr>'
	print '<th></th><th>Title</th><th>Coordinate</th><th> Parameters</th>'
	print '</tr>'
	idx = offset
	for (el_to,) in c.fetchmany(limit):
		print '<tr>'
		qs = el_to[len("http://stable.toolserver.org/geohack/geohack.php?"):].split('&')
		pagename = ''
		title    = ''
		params = ''
		for s in qs:
			if s.startswith('pagename='):
				pagename = s[9:]
				title = wikipedia.urllib.unquote(s[9:])
			elif s.startswith('params='):
				params = s[7:]
			elif s.startswith('title=') and pagename == '':
				pagename = s[6:]
				title = wikipedia.urllib.unquote(s[6:])

		iProps = min(params.find('_W_')+1 or len(params)-1, params.find('_E_')+1 or len(params)-1)+1
		idx+=1
		print ('<td>%d</td><td><a href="http://%s/wiki/%s" class="extiw">%s</a></td><td><a href="%s" class="external">%s</a></td><td><code>%s</code></td>' % (
			idx,
			site.hostname(),
			pagename,
			title.replace('_', ' '),
			el_to, 
			params[:iProps], 
			params[iProps:].replace('_', ' '),
			) ).replace('&', '&amp;')
		print '</tr>'
	print '</table>'

	printLinks(offset, limit, queryString)
		
if __name__ == "__main__" and wikipedia.handleUrlAndHeader():
	import time
	startTime = time.time()
	try:
		wikipedia.startContent(form=False)
		main()
	finally:
		wikipedia.endContent(replag = db and replag() or 0)
		wikipedia.stopme()