Modified Files:
Utility.py -- Added a parameter to Collection class constructor,
'name' is the default attribute used for keys but one
can specify whatever key they want.
XMLSchema.py -- Used the above parameter to make Collection
instances use the appropriate 'attribute' as key.
----------------------------------------------------------------------
WSDLTools.py -- just a few minor changes so the new schema class
will be used instead of the schema tns placeholder.
Might want to add an argument to WSDL.load method so that
programmer can specify the placeholder or actual
implementation.
XMLSchema.py -- mostly new, unused original code is commented out at the bottom.
I fixed that fact that Notation and Entity were not be found. I changed them
to use the qualified name like the rest of the symbols from xml.dom.minidom.
I also discovered that a RecursionError was being thrown, but RecursionError
didn't exist. I created simple sub-class of Exception to rectify this.
XMLSchema.py
----------------------------------------------------------------------
fixed an obvious bug, added a SchemaError class so it can actually
be thrown.
----------------------------------------------------------------------
commit a little while ago. :) That is, wstools is now independant
of SOAPpy and ZSI. This commit does the following:
wstools/ServiceProxy.py is now ZSI/ServiceProxy.py, so some
imports and ZSI docs had to change.
ZSI needs some changing, in case I didn't patch up all the imports
right.
(which doesn't), so that SOAPpy can use wstools without needing ZSI
around... which is kinda the point of generic common-code. :)
class SOAPCallInfo:
class ParameterInfo:
class HeaderInfo(ParameterInfo):
def callInfoFromWSDL(port, name):
Next step is to move what's left of wstools/ServiceProxy.py into the ZSI
module (and fix up the imports), so that wstools has *no* soap-stack-specific
code in it.