• From: richard@i... (Richard Tobin)
• To: xml-dev@l...
• Date: Tue, 30 Sep 2008 12:11:19 +0100 (BST)

In article <43A9274214C240AB9F3000273156C996@Sealion> you write:

>The best argument is that people who adopt this approach usually fail to
>check for the presence of "]]>", which isn't allowed in CDATA sections. Once
>you start checking for that and dealing with it properly, it turns out to be
>easier to check for & and < and escape them as &_amp; and &_lt;
>respectively. (Underscores inserted to prevent misformatting).

You need to add > to your list of characters to escape, or else check
for ]]> even when not using CDATA, because that sequence is not legal
in text content.

-- Richard

• References:
  • RE: Tradeoffs of XML encoding by enclosing all content in CDATA blocks
    • From: "Michael Kay" <mike@s...>