On 11/2/07, Paul Dietrich <[EMAIL PROTECTED]> wrote:

> We're planning to submit more packet test libraries which will give us
> some good confidence that the libraries all have consistent XML and
> binary representations.
>

Paul, I noticed in the wiki it mentions an "xml equivalence" test. Can
you elaborate? That seems like thorny issue.

The problem isn't just inconsequential formatting. It's also XML
namespaces and prefixes, where xmlns declarations appear (in the
header, or later on in the document), unpreserved order of
subparameters at alternation points and extension points, and
attributes which may be optional. We can try to recommend a normalized
form for LTK but some toolkits may be unable to comply for practical
reasons.

We could just compare the binary forms, although this does not deal
with the "order" at alternation and extension points issue. We could
deal with some of that by validating the output against schema, but it
still wouldn't tell us if there was data loss or an extra field that
is not expected. Perhaps this doesn't actually come up? It doesn't for
LTK-Perl since I never convert to an object tree (the "order" problem
would occur if one modeled an alternation or extension point as an
unordered data structure, like a hash table).

You could use another toolkit to keep you honest.

So Binary->XML of Ma with LTK-A, you would

run LTK-A's Bin2XML on Ba, getting an XML document Xa.
Then pass Xa through LTK-B's XML2Bin = Ba'
Then binary compare Ba to Ba'

We could close the loop with a LLRP binary comparator that is aware of
arbitrary order issues. We could pick one toolkit as the reference and
implement the comparator on that.

-- John.

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
llrp-toolkit-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/llrp-toolkit-devel

Reply via email to