---------------------------------
* Improvements:
* Support XPath namespace prefixes (``namespace:element``) and CSS namespace
prefixes (``namespace|element``) in DefaultSelectorHandler
* new built-in extension function ``parslepy:strip()`` mapped
to Python's strip() for strings
* new built-in extension function ``parslepy:attrname()``
that takes an attribute selector and returns the attribute's name
* support for user-defined extension functions, which take an additional
context parameter when called (context is passed either at selector
handler instantiation or when calling Parselet.parse())
* use ``smart_strings=False`` for XPath compiled expressions,
except for user-defined extensions and some built-in extensions
(see http://lxml.de/xpathxslt.htmlxpath-return-values)
* Bug fixes:
* 2: XPath namespace prefixes raise cssselect.xpath.ExpressionError
with DefaultSelectorHandler
* 3: Docs suggest using \*.js files when they are JSON documents
* 4: The example usage should not have both url_css and url_xpath
* 5: In example usage, skip lines between "configuration" and "execution"
* 6: add underscore to _version__
* 7: Empty result set on boolean or numerical selectors