- From: Simon Pieters <simonp@opera.com>
- Date: Tue, 18 Mar 2008 21:21:23 +0100
- To: public-html <public-html@w3.org>
The serializing of HTML fragments algorithm currently looks up the
ancestor chain for Text and CDATASection nodes for CDATA elements (or
<plaintext>) in order to decide when to escape text.
Firefox takes a different approach, it seems: when finding an (R)CDATA
element (or <plaintext>), emit the data of the child Text and CDATASection
nodes (escaped or not as appropriate) instead of recursing the algorithm
on its subtree.
I think Firefox's approach is a bit more performant (it would probably be
less checking in the general case), and it roundtrips a bit better; as the
spec notes:
For instance, if a textarea element to which a Comment node has been
appended is serialised and the output is then reparsed, the comment
will end up being displayed in the text field.
--
Simon Pieters
Opera Software
Received on Tuesday, 18 March 2008 20:22:02 UTC