On Wed, Oct 19, 2011 at 11:53 AM, Marcin Tustin <marcin.tus...@gmail.com>wrote:
> OK, how should I provide the schema to pyxb so that it can decode such
> documents? The XSD which I need my documents to conform to is the one from
> which I'm generating the code.
As far as I can tell, the schema isn't the problem. (You will have to also
locate the schema for and generate bindings for
http://www.govtalk.gov.uk/CM/core, since the envelope's schema references
types defined in that namespace.)
The problem is the example document is lying about what namespace its
content conforms to. Change:
<GovTalkMessage xmlns="
http://www.govtalk.gov.uk/schemas/govtalk/govtalkheader";
at the top of the document you're parsing to:
<GovTalkMessage xmlns="http://www.govtalk.gov.uk/CM/envelope";
Peter
On Wed, Oct 19, 2011 at 17:46, Peter Bigot <big...@acm.org> wrote:
>
>> On Wed, Oct 19, 2011 at 11:22 AM, Marcin Tustin
>> <marcin.tus...@gmail.com>wrote:
>>
>>> Hi Peter,
>>>
>>> Thanks for this. Unless I'm mistaken, namespace URIs are not required to
>>> map to any resource at all, let alone an XSD. However, the xsd from which
>>> I'm generating the code is at
>>> http://xmlgw.companieshouse.gov.uk/v2-1/schema/Egov_ch-v2-0.xsd.
>>>
>>
>> That schema is for namespace http://www.govtalk.gov.uk/CM/envelope.
>>
>> What is PyXB expecting in its input that could be leading to this? Does it
>>> rely on namespaces naming XSD files?
>>>
>>
>> No, PyXB treats namespaces as URIs, not URLs, and has no expectation that
>> they can be dereferenced. (It may, when generating bindings, make some
>> attempt to locate the schema from an xsi:schemaLocation attribute, but that
>> shouldn't happen when parsing.)
>>
>> The example documents (from the link you provided) do not reference the
>> namespace http://www.govtalk.gov.uk/CM/
>> envelope. They both specify that the default namespace is
>> http://www.govtalk.gov.uk/schemas/govtalk/govtalkheader. Unless you have
>> a schema for that namespace, PyXB won't be able to decode the content of the
>> document.
>>
>> If the document doesn't specify a default namespace I believe it would
>> fall back to the one you used as the CreateFromDocument call, but when the
>> document does specify a namespace it's expected to be correct.
>>
>> Try editing the document and correcting the xmlns= attribute at the root
>> so it has the same namespace as the module you're using to parse it.
>>
>> Peter
>>
>>
>>>
>>> On Wed, Oct 19, 2011 at 16:49, Peter Bigot <big...@acm.org> wrote:
>>>
>>>> On Wed, Oct 19, 2011 at 10:32 AM, Marcin Tustin <
>>>> marcin.tus...@gmail.com> wrote:
>>>>
>>>>> Hi all,
>>>>>
>>>>> I'm making my first contact with pyxb, and so far it seems to only
>>>>> exist to generate odd errors.
>>>>>
>>>>> I've tried to have it parse a document with:
>>>>>
>>>>> > import govtalkenvelope as g
>>>>> > g.CreateFromDocument("""[xml taken from
>>>>> http://xmlgw.companieshouse.gov.uk/example_http.html]""";)
>>>>> Traceback (most recent call last):
>>>>> File "<stdin>", line 35, in <module>
>>>>> File "govtalkenvelope.py", line 32, in CreateFromDocument
>>>>> saxer.parse(StringIO.StringIO(xml_text))
>>>>> File "C:\Python27\Lib\xml\sax\expatreader.py", line 107, in parse
>>>>> xmlreader.IncrementalParser.parse(self, source)
>>>>> File "C:\Python27\Lib\xml\sax\xmlreader.py", line 123, in parse
>>>>> self.feed(buffer)
>>>>> File "C:\Python27\Lib\xml\sax\expatreader.py", line 207, in feed
>>>>> self._parser.Parse(data, isFinal)
>>>>> File "C:\Python27\Lib\xml\sax\expatreader.py", line 338, in
>>>>> start_element_ns
>>>>> AttributesNSImpl(newattrs, qnames))
>>>>> File
>>>>> "C:\Users\Marcin\Documents\oneclickcos\lib\site-packages\pyxb\binding\sax
>>>>> er.py", line 310, in startElementNS
>>>>> element_binding = name_en.elementBinding()
>>>>> File
>>>>> "C:\Users\Marcin\Documents\oneclickcos\lib\site-packages\pyxb\namespace\_
>>>>> _init__.py", line 96, in __getattr__
>>>>> category_value =
>>>>> self.namespace().categoryMap(name).get(self.localName())
>>>>> File
>>>>> "C:\Users\Marcin\Documents\oneclickcos\lib\site-packages\pyxb\namespace\_
>>>>> _init__.py", line 293, in categoryMap
>>>>> raise pyxb.NamespaceError(self, '%s has no category %s' % (self,
>>>>> category))
>>>>> pyxb.exceptions_.NamespaceError:
>>>>> http://www.govtalk.gov.uk/schemas/govtalk/govta
>>>>> lkheader has no category elementBinding
>>>>>
>>>>
>>>> It's probably saying it doesn't know how to deal with the
>>>> http://www.govtalk.gov.uk/schemas/govtalk/govtalkheader namespace.
>>>> Certainly it isn't clear to me where to find a schema for that namespace;
>>>> the xsiLocation URI in the sample resolves to a web page, not a schema.
>>>>
>>>> Did you generate bindings for that namespace? How?
>>>>
>>>>
>>>>>
>>>>> Needless to say, it's less than clear what this is supposed to mean.
>>>>> Any suggestions as to what might be causing this, or indeed how in general
>>>>> to deal with pyxb's exception messages?
>>>>>
>>>>
>>>> Experience is the only guide. Sorry, PyXB development is unfunded, so
>>>> making it user friendly hasn't gotten much attention.
>>>>
>>>> Peter
>>>>
>>>>
>>>>>
>>>>> Marcin
>>>>>
>>>>> --
>>>>> Marcin Tustin
>>>>> Tel: 07773 787 105
>>>>>
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> All the data continuously generated in your IT infrastructure contains
>>>>> a
>>>>> definitive record of customers, application performance, security
>>>>> threats, fraudulent activity and more. Splunk takes this data and makes
>>>>> sense of it. Business sense. IT sense. Common sense.
>>>>> http://p.sf.net/sfu/splunk-d2d-oct
>>>>> _______________________________________________
>>>>> pyxb-users mailing list
>>>>> pyxb-users@lists.sourceforge.net
>>>>> https://lists.sourceforge.net/lists/listinfo/pyxb-users
>>>>>
>>>>>
>>>>
>>>
>>>
>>> --
>>> Marcin Tustin
>>> Tel: 07773 787 105
>>>
>>>
>>
>
>
> --
> Marcin Tustin
> Tel: 07773 787 105
>
>
------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2d-oct
_______________________________________________
pyxb-users mailing list
pyxb-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/pyxb-users
