-> G'day Titus, -> -> The web server was consistently failing to send a Content-Type if twill -> was used to query it during evening hours. The debug http confirmed it. -> -> Apparently some browsers assume it is text/html and proceed regardless. -> -> The attached patch adds an assume_html() method to mechanize. I know -> you'd rather not take patches for something that is upstream of you, but -> I thought I'd mention how it was solved in the end. I could not find a -> way to solve it without this change.
James, I came across my own problems with mechanize, and needed a way to selectively override MechanizeBrowser. In the process of arranging that -- see the new 'browser.py', class 'PatchedMechanizeBrowser' -- I realized that I could just override 'viewing_html' to do the right thing for your situation. So, PatchedMechanizeBrowser now assumes that text/xml is HTML, and assumes that an absence of content-type is indicative of text/html. If other people run into issues with these changes, then I'll implement your boolean flag idea too. Let me know if you get a chance to try out the latest twill & find that it solves your problems. (I'll post to this list when I finish the hacking and post the changes.) thanks, --titus _______________________________________________ twill mailing list [EMAIL PROTECTED] http://lists.idyll.org/listinfo/twill
