Re: Y4 Exclusive C145n interop; was Re: c14n/exc-c14n interop samples

Hi Christian,

r/geuer-pollmann@nue.et-inf.uni-siegen.de/2002.05.31/17:34:39
>Hi Merlin,
>
>merlin-c14n-three.tar.gz shows the same failing results:
>
> 0,  4,  5,  9, 11, 12, 13, 14, 15, 16, 17, 18, 20, 21, 22, 23 and 25 have 
>been validated successfully.
>
>These fail because of the "superflous namespaces" issue which
>Aleksey already mentioned:
>
> -  2
> - 10
> - 19

While they may appear superfluous, according to the letter
of the spec they should be emitted. And, given that this is
an unusual case of little use, I don't think that any
aesthetic concerns are of particular import.

>These fail because I don't output an namespace as
>a TEXT node if it's owner element is not document
>sub set.
>
> -  3
> -  6
> -  7
> - 24

They are not output as text nodes, per se; they are output
as any attribute or namespace node. If you attempt to parse
this directly, then you will get a text node. If you wrap
this in an element declaration, then you will get an
attribute node.

>unknown reason: I don't know yet why we have differences.
>
> -  1
> -  8
> - 26

I would suspect because the node set omits all namespace nodes
from the foo:Nothing element (the first instance of the foo:
prefix).

These examples are among those situations where no one
really cares because no one should be producing node sets
of this form; however, interop of the spec, as written,
is relatively important.

And you did ask for them ;}

Merlin

>Regards,
>Christian
>
>--On Freitag, 31. Mai 2002 13:18 +0100 merlin <merlin@baltimore.ie> wrote:
>
>> Hi Christian,
>>
>> I've tweaked the input document slightly to show a few more
>> edge cases if you're interested; see attached:
>>
>> Merlin
>
>
>

Received on Friday, 31 May 2002 12:52:22 UTC