1. Speaking as a former HTML Working Group member: quirks mode is a
concept that, for srcdoc documents, does not exist. That is: HTML5.x
says that the author may include or exclude the DOCTYPE, it does not
matter, the document must anyhow rendered by the parser in what HTML5.x
refers to as **no-quirks mode**. This is unique for srcdoc documents.
And browsers do implement it (verified by checking Firefox and Safari).
However, the parser of XMLmind XML editor does not conform to this: I
have noticed that when DOCTYPE is excluded, then the document is
rendered - within XXE - in quirks-mode. (I used <div style='width:50'>
to check for quirks-mode: in no-quirks mode, that code will not have any
effect, but in quirks mode it does - in XXe - have effect. In standard
browsers, there is no such effect.)
(Tidbit: The reason @srcdoc does not have quirks mode is because it was
a fresh start - there did not exist legacy code to care about.)
2. Some might think (ahem - I am looking at myself ...) that a srcdoc
document would be a XHTML document when inside a XHTML document, and a
HTML document when inside a HTML document. But, actually (and of course:
my bad), a srcdoc document is always a 'text/html' HTML document. This
means, says the HTML5.x spec, that you are allowed to skip the start and
end tag of a lot of elements, including the root element.
However, XMLmind XML editor does not implement this: Unless the <html>
start tag and end tag is included in the srcdoc attriute, then XXe
erroneously reports that the content is not HTML. By contrast, the
HTML5.x specification says that the @srcdoc does not need to contain any
code - it could simply simply be this:
srcdoc="XMLmind must update its srcdoc HTML mind"
The HTML parser will then, itself, insert all the lacking elements when
the srcdoc document is rendered.
3. Another thing: XXe escapes the code of srcdoc cod more than
necessary. E.g
<html><head>
becomes
<html><head>
whereas it would have been enough to let it become
<html><head>
May be it is just me, but I think that code becomes easier to read if
it is not escaped more than necesary. So would prefer that behavior ...
It also safes a few code bytes ...
--
leif halvard silli
--
XMLmind XML Editor Support List
xmleditor-support@xmlmind.com
http://www.xmlmind.com/mailman/listinfo/xmleditor-support
