Hi.
I was wondering if it was the right behavior to have for suds, and
if
so, how I could mitigate that ? I felt on that article
(
http://stackoverflow.com/questions/7500492/how-do-i-prevent-suds-from-fet...)
and used the chosen answer, which works and seems to be the good one.
Yup, that should make suds use a local XSD data instead of attempting
to download it from the net.
This can be made a bit more localized by using the DocumentStore
object as patched in suds-jurko to support URLs with protocol other than
suds but the end effect should be the same in your case.
But then I felt on this :
http://www.w3.org/blog/systeam/2008/02/08/w3c_s_excessive_dtd_traffic/,
and I wondered if suds didn't fall in the categories of "badly written
xml libraries". I'm not blaming suds in any way, it does an amazing
jobs, I'm just trying to help.
Jurko, what do you think ? Do you think Suds has the bad behavior the
article describe ? If so, do you think a fix is possible ?
I think it downloads the XSD schema correctly - when referenced from
a WSDL given to it. It could perhaps be hardcoded internally in suds so
it knows about it and never downloads it at all, but I do not think this
should have much of an impact. As it stands now, suds should download
this document just once and then cache it. So unless you're explicitly
messing with its caching or unless you've actually run into a bug :-) -
it should already be skipping this download after getting it the first
time (which, by default, I think is once per day).
Hope this helps.
Best regards,
Jurko Gospodnetić