«

Hi !
i am trying to serialise object which contains some unicode objects
but looks like there is no way to do it.

File
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/site-packages/
_xmlplus/marshal/generic.py", line 92, in _marshal
return getattr(self, meth)(value, dict)
AttributeError: WDDXMarshaller instance has no attribute 'm_unicode'

WDDX is perfect for me for exchange between python and php application.
but maybe there is a better way to do it.


Any clue?

Thanks in advance.

"isthar" <> wrote in message
news: oups.com...
> Hi !
> i am trying to serialise object which contains some unicode objects
> but looks like there is no way to do it.
>

hi, I'm sure you'll get better answers for the unicode part of your problem
(I'd start with a look at the codecs module), but I wanted you to know that
I'm using the wddx connection between Python and PHP with no problem. Here's
the python code that writes a fairly complex data structure to the wddx
file:

def writeDocReport(self):
from xml.marshal import wddx
fileName = '%s/%s.doc.wddx' % (os.path.join(adsWeb,'jobs','wddx'),
self.name)
tmpFile = open(fileName,'wb')
tmpFile.write(wddx.dumps(self.getDocData()))
tmpFile.close()

and the PHP:

function getInputData($){
$file = "wddx/$prodName.wddx";

if (!is_file($file)) {
exit("<h1>Report is not available for $prodName</h1>");
}

$output = wddx_deserialize(file_get_contents($file));

isthar wrote:
> WDDX is perfect for me for exchange between python and php application.
> but maybe there is a better way to do it.

It appears that Unicode objects where forgotten in the WDDX
implementation. I suggest to define the following classes:

class UWDDXMarshaller(xml.marshal.wddx.WDDXMarshaller):
def m_unicode(self, data, dict):
return self.m_string(data.encode("utf-8"), dict)

class UWDDXUnmarshaller(xml.marshal.wddx.WDDXUnmarshalle r):
def um_end_string(self, name):
ds = self.data_stack
ds[-1] = u"".join(ds[-1])
self.accumulating_chars = 0

The m_unicode part should get integrated into PyXML;
the um_end_string should probably return ASCII strings
if possible, else Unicode strings.

Regards,
Martin

Ok. but how I suppose to use them. I am currently using marshaller
indirectly via wddx.dump().

Anyway, thanks

isthar wrote:
> Ok. but how I suppose to use them. I am currently using marshaller
> indirectly via wddx.dump().

Do UWDDXMarshaller().dump()

Regards,
Martin