[I originally wrote this for StackOverflow, hence all the backticks; I've left them in to deliminate strings.]

Not sure if this is a bug or me failing to understand `CDATA`.

I have some Python 2 `unicode` objects I want to include in the text of my XML document; I've been asked to put them into `CDATA`, so I used `element.text = lxml.etree.CDATA(mytext)`.  However, `ş` (repr: `u'\xfe'`) and `þ` (repr: `u'þ'`) seem to produce the same XML output, despite being very different strings:

`<root><![CDATA[&#254;]]></root>`

which lxml parses as `&#254;`.

Can someone confirm it's a bug, or correct my wrongheadedness about using CDATA?

BTW: `element.text = mytext` (i.e. no CDATA involved) works as I'd expect.

_test code:_

    import lxml.etree

    import sys

    literal=sys.argv[1].decode('utf-8')

    print repr(literal)

    literalroot=lxml.etree.fromstring("<root></root>")

    literalroot.text = lxml.etree.CDATA(literal)

    xml = lxml.etree.tostring(literalroot)

    print xml

    print literal, lxml.etree.fromstring(xml).text

    $ python test.py "ş"

    u'\xfe'

    <root><![CDATA[&#254;]]></root>

    ş &#254;

    $ python test.py "&#254;"

    u'&#254;'

    <root><![CDATA[&#254;]]></root>

    &#254; &#254;