Re: RE: Python 2.3 and aggregator

Georg Bauer wrote:
> Hmm - what error do you get now? Still the stuff on None not having 
> lower()? Or something different?

My apologies with regard to my code posting; I had thought that was the 
code that worked but it actually does not. That was on another machine. 
The open_http code in DownstreamTool that works for me (which I had to 
reconstruct) is actually:


	def open_http(self, url, data=None):
		numheaders = len(self.addheaders)
		self.isHTTP = 1
		self.lastURL = self.getTheUrl(url)
		try:
			theurl = self.getTheUrl(url)
			self.message = _('opening url: <a href="%s">%s</a>') % (theurl, theurl)
			if not(self.force):
				for h in self.cache._getUrlHeaders(theurl):
					apply(self.addheader, h)
					self.message += _('<br>adding Header "%s: %s"') % h
			url = theurl[5:]
			res = urllib.URLopener.open_http(self, url, data)
			self.message = self.message.replace('%', '%%')
			if self.verbose:
				self.cache.logVerbose(self.message)
			else:
				self.cache.log(self.message)
			return res
		finally:
			self.addheaders = self.addheaders[:numheaders]


Sorry about the wrapping, but note the "url = theurl[5:]" line; 
open_http for some reason wants to see something like 
'//slashdot.org/slashdot.rdf', not 'http://slashdot.org/slashdot.rdf'. 
Passing in the host as ('slashdot.org', 
'http://slashdot.org/slashdot.rdf') doesn't work, you get a 403.

This is working for me again now, and I'm back to my usual set of "time 
out" errors for editthispage.com sites ;-)