listMemberInfo - 'lists' field format problems

[DavidMcL, Perceptive Logic]

Hi Jesse,

If you remember back in July I had an issue with listMemberInfo and
the xml output from the lists field...

As a memory jogger...
-----------------------------------
David -
I see why this is happening, but will have to see about fixing it and
what repercussions it may have elsewhere in the API. Basically, for
whatever reason, the library that serializing to XML decides that if
the
id starts with a number, it's meant to be used as a key, otherwise it
must be the tag name. I'll try to be sure to let you know when I can
get
that corrected.

Jesse Peterson

---------------------------

... well it appears that it creates a problem with the xml-rpc output
as well, slightly different, but worse.

In the case of xml-rpc -- the listid just disappears, giving a null
element name -- boom!

Since this happens at the point that the xml-rpc parser is called,
there is no fix (other than writing an http handler which seems
excessive at this point) and there isn't any list id anyway, so all I
could do is stop it from crashing.

I can't use [XmlRpcMissingMapping( MappingAction.Ignore )] -- simply
using the Ignore won't work because without the element name (which is
null) the xml-rpc parser doesn't know what to ignore.

The only 'out' for xml-rpc is to remove the definition for the 'lists'
field altogether (and then 'Ignore' it)

what you see using fiddler is

<member>
<name>lists</name>
<value>
<struct>
<member>
<name/>
<value>
<string>subscribed</string>
</value>
</member>
<member>
<name>c41ded4d93</name>
<value>
<string>subscribed</string>
</value>
</member>
</struct>
</value>
</member>


Notice how the first 'name' element is empty...

Hope you can resolve this one.

Thanks,

David

[jesse]

David & others -

I've tracked down a workaround to fix this problem for the
listMemberInfo method using xml-rpc and the xml output format. If you
guys could verify that it works as expected now, that would be great.

While I've checked for other places this could happen and not
identified any, if anyone runs across this again I'll at least need
the method they are seeing it in as this is a very directed fix
specific to this method and those formats.

jesse

On Dec 16, 12:54 pm, "DavidMcL, Perceptive Logic"

[DavidMcL, Perceptive Logic]

Hi Jesse,

I tried 5 or 6 of the 'problem' records and they all work as expected,
in all output formats.
Will run through a couple long lists later to verify, but I suspect
that all is well now.

Great support as always...

Thanks again,

David

PS. I haven't seen this weirdness anywhere else either.