Whitespace again (ODF all version)

["Alex Brown" <alexb@griffinbrown.co.uk>]

Dear all,

1.5 has:

---------begin extract
"In conformance with the W3C XML specification [XML1.0], optional white-space characters that are contained in elements that have element content are ignored. This applies to the following white-space and end-of-line (EOL) [UNICODE] characters:

HORIZONTAL TABULATION (0x0009)
LINE FEED (0x000A)
CARRIAGE RETURN (0x000D)
SPACE (0x0020)

For any other element, white-spaces are preserved by default. Unless otherwise stated, there is no special processing for any of the four white-space characters."
---------end extract

* How is the behaviour in the first sentence in conformance with XML 1.0? "optional whitespace" is only mentioned there in relation to (DTD) validity. I think this phrase should be removed.

* Consider this fragment:

<text:p><text:span>Hello</text:span> <text:span>world</text:span></text:p>

Since the <text:p> element has element content, according to extract above the space between "Hello" and "world" has to be ignored. Surely not.

I think what the author intended here was "whitespace characters in elements DECLARED TO HAVE ONLY element content are ignored".

* Note the repertoire of whitespace characters differs between eds 4 and 5 of XML 1.0. It might be better to remove this list of whitespace characters as they are already defined by XML.

* "For any other elements" -> "For elements declared to have mixed content" (for the same reason as above).

* The last sentence in the extract is unhelpful and could usefully be removed.

Also, throughout ODF, whitespace is referred to as "whitespace", "white space" or "white-space". If ODF could standardise on just one form it would go one better than the XML 1.0 Recommendation :-)

- Alex.