[shadow] Escape non-ascii chars in urls

_This is a _shadow issue* for Issue 31 on Google Code (from which this project was moved).
Added 2009-10-08T09:52:12.000Z by atregoub...@gmail.com.
Labels: Type-Defect, Priority-High.
Please make updates to the bug there.*

Original description

<b>What steps will reproduce the problem?</b>
1. import markdown2
2. markdown2.markdown(u'[link](http://example.com/тест)')

<b>What is the expected output? What do you see instead?</b>
now this produces following output: 
>>> u'<p><a href=&quot;http://example.com/\xd1\x82\xd0\xb5\xd1\x81\xd1\x82&quot;>link</a></p>\n'
it would be nice to see more standards compliant
>>> u'<p><a href=&quot;http://example.com/%D1%82%D0%B5%D1%81%D1%82&quot;>link</a></p>\n'

<b>What version of the product are you using? On what operating system?</b>
1.0.1.15 on ubuntu

<b>Please provide any additional information below.</b>
this can be done quite easily with urllib2.quote 

Reproduce the issue by importing markdown2 and rendering a link whose URL contains non-ASCII characters, such as http://example.com/тест. Compare the generated href with the expected percent-encoded URL; the work is done when the output uses standards-compliant URL escaping. The issue directs updates to the linked Google Code issue.