Bernd,
Thank you for your notes on this issue.
I'm still trying to think this one through. I'm still a bit
confused about it.
However, here is one item that may help. The `namespace_` parameter
is mis-named I suppose; its value should be a namespace prefix and a
colon. That value should be something like this: "abc:".
If you were to pass something like "xyz:" as the value of the
`namespace_` parameter, would that fix one of your issues?
By the way: I believe that the documentation included near the top
of a generated module, is not correct. It incorrectly reads:
# A sample table is:
#
# # File: generatedsnamespaces.py
#
# GenerateDSNamespaceDefs = {
# "ElementtypeA": "http://www.xxx.com/namespaceA";,
# "ElementtypeB": "http://www.xxx.com/namespaceB";,
# }
whereas, it should be:
# A sample table is:
#
# # File: generatedsnamespaces.py
#
# GenerateDSNamespaceDefs = {
# "ElementtypeA": 'xmlns:nsa="http://www.xxx.com/namespaceA";',
# "ElementtypeB": 'xmlns:nsb="http://www.xxx.com/namespaceB";',
# }
With respect to your issue 2:
> 2.)
> OUTPUT with testxml.export(output, 0,
> 'xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData";')
> results in:
> <xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData"RetainedDataMessage
> xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData";>
> </xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData"RetainedDataMessage>
>
> This is not correct.
I believe that the namespace definition that you are passing in as
the 2nd parameter is getting attached to the namespace_ parameter,
whereas you actually want to set the namespacedef_ parameter. You
might consider using a keyword argument to make this explicit.
Example:
testxml.export(
output,
0,
namespacedef_='xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData";'
)
While trying to straighten my head out on this, I wrote up the
following notes on the parameters to the `export` methods.
Arguments to the generated export method:
- outfile -- A file like object open for writing.
- level -- the indentation level. If the pretty_print argument is
True, the (generated) function showIndent is used to prefix each
exported line with 4 spaces for each level of indent.
- namespace_ -- An empty string or an XML namespace prefix plus a
colon, example "abc:". This value is printed immediately in front
of the tag name.
- name_ -- The element tag name. Note that the tag name can be
overridden by the original_tagname_, which can be set by the class
constructor.
- namespacedef_ -- Zero or more namespace prefix definitions.
Actually, its value can be any attribute-value pairs.
Examples:
''
'xmlns:abc="http://www.abc.com"http://www.def.com";
'xmlns:abc="http://www.abc.com xmlns:def="http://www.def.com";
or, because it is printed where the attributes occur, even:
'size="25" color="blue"'
For more on namespacedef_, see:
http://www.davekuhlman.org/generateDS.html#namespaces-inserting-namespace-definition-in-exported-documents
- pretty_print -- If True, exported output is printed with
indentation and newlines. If False, indentation and newlines are
omitted, which produces a more compact representation.
I've added the above notes to the generateDS document. See:
http://www.davekuhlman.org/generateDS.html#method-export
Sigh. XML namespaces make my head hurt.
I'll study this and your notes a bit more. There must be *some* way
we can make sense of this and get it to work correctly.
Hope your weekend is (or was) a good one.
Dave
On Thu, Aug 09, 2018 at 10:14:57AM +0000, Zimmermann, Bernd wrote:
> Hello Dave,
>
> thank you very much for the patch for the validate patterns, it
> works 😉 so it could go productive on your next release.
>
> Just a few additional remarks on the export issue:
>
> Code: def export(self, outfile, level, namespace_='',
> name_='RetainedDataMessage', namespacedef_='', pretty_print=True):
> There are three possible parameters: namespace_, name_, namespacedef_
>
> The line: outfile.write('<%s%s%s' % (namespace_, name_, namespacedef_ and ' '
> + namespacedef_ or '', ))
> will print:
> <namespace_ , name_, namespacedef_> if all three values are set.
> Note: namespacedef_ may be set via GenerateDSNamespaceDefs option.
>
> Here is just a first problem, because the namespace must be a second value in
> the xml tag
> <OBJECT_NAME NAMESPACE>
> </OBJECT_NAME>
> in a valid xml, not the first.
>
> TESTing:
> 1.)
> using: testxml.export(output, 0, '', 'retainedDataMessage')
> results in:
> <retainedDataMessage xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData";>
> ...
> </retainedDataMessage>
>
> this ist Ok.
> Note that GenerateDSNamespaceDefs is set in my test case.
> Parameter namespace_ is empty.
> Parameter name_ is set
> Parameter namespacedef_ is not used but overwritten by:
> imported_ns_def_ = GenerateDSNamespaceDefs_.get('RetainedDataMessage')
> if imported_ns_def_ is not None:
> namespacedef_ = imported_ns_def_
>
> 2.)
> OUTPUT with testxml.export(output, 0,
> 'xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData";')
> results in:
> <xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData"RetainedDataMessage
> xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData";>
> </xmlns="http://uri.etsi.org/02657/v1.18.1#/RetainedData"RetainedDataMessage>
>
> This is not correct.
> There is the problem as stated above that the namespace_ is set
> und printed out at first in the opening tag before the name_
>
> The logical problem is: The parameter namespace_ is set (the first
> one) and there is an GenerateDSNamespaceDefs setting, so there is
> a logic missing, to decide which parameter to use in the export if
> there are 2 possibilities. My suggestion would be to always use
> namespacedef_ it set and use the value via GenerateDSNamespaceDefs
> setting only if namespacedef_ parameter is not used or set.
>
> So for me it looks like to adjust the lines with outfile.write('<
> and outfile.write('> to match an valid XML form like:
> <OBJECT_NAME NAMESPACE>
> </OBJECT_NAME>
>
> Hope you guess what I mean.
>
> Best wishes,
> Bernd
>
> -----Ursprüngliche Nachricht-----
> Von: Dave Kuhlman [mailto:dkuhl...@davekuhlman.org]
> Gesendet: Dienstag, 24. Juli 2018 23:03
> An: Zimmermann, Bernd <bernd.zimmerm...@qsc.de>
> Cc: generateds-users <generateds-users@lists.sourceforge.net>
> Betreff: Re: [Generateds-users] validate_XXX_patterns generated wrong from
> XSD and Problems with export and namspaces
>
> Bernd,
>
> With respect to the validation of time and the
> validate_GeneralizedTime_patterns_, I've attached a patch. With this patch,
> I believe that generateDS.py produces the pattern that works for you. These
> regular expressions are complicated enough, and I'm in the dark about what
> they should and should not be checking. So, if this patch works for you,
> let's go with it. We can make adjustments when someone else reports a
> problem with it.
>
> Here is a little explanation, as far as I can understand it: This patch
> removes the code that was replacing "|" with "$|^". That replacement was
> intended to enable us to handle alternatives in regular expressions by
> breaking them up into several concatenated expressions. *But*, maybe the
> ability to handle those alternatives is (or should be) written into the
> regular expression itself, *if* the author of the XML schema writes them
> correctly. If so, the replacement that we are removing is not needed, and
> this patch is the right thing to do. At least, that's what I'm hoping.
>
> About your other issue, I'll take a look tomorrow. But, at first glance, it
> looks like the values you are passing in to the export method are being
> attached to the wrong arguments. You are passing positional arguments. You
> could try using the keyword arguments:
>
> namespace_
> name_
> namespacedef_
>
> You will have to look at one of the export methods to see how each of those
> parameters is used.
>
> By the way, your suggestion in your latest email that a module generated by
> generateDS.py be able to do an automatic validation of the exported XML
> instance document seems like a good idea. Although adding something like
> this line:
>
> xmllint --schema test.xsd test01.xml
>
> to a batch file does seem like an easy enough way to handle that.
> So, I'll put that feature on a "do" list, but it might not be at the very top
> of the list.
>
> More later. Thanks again for your help with these issues.
>
> Dave
>
> On Thu, Jul 19, 2018 at 12:19:52PM +0000, Zimmermann, Bernd wrote:
> > Hello Dave,
> >
> > very great stuff you've done with generateDS, works like a charm but only
> > after a resolving a few little things.
> >
> > First on, this might be a bug:
> >
> > XSD Pattern:
> >
> > <xsd:simpleType name="GeneralizedTime">
> > <xsd:restriction base="xsd:string">
> > <xsd:pattern
> > value="\d{4}(0[1-9]|1[012])(0[1-9]|[12][0-9]|3[01])([01][0-9]|2[0-3])([0-5][0-9])([0-5][0-9])(.[0-9][0-9]*)?(Z|[\+|\-][0-9]{4})"/>
> > </xsd:restriction>
> > </xsd:simpleType>
> >
> > results in:
> >
> >
> > validate_GeneralizedTime_patterns_ =
> > [['^\\d{4}(0[1-9]$|^1[012])(0[1-9]$|^[12][0-9]$|^3[01])([01][0-9]$|^2[
> > 0-3])([0-5][0-9])([0-5][0-9])(.[0-9][0-9]*)?(Z$|^[\\+$|^\\-][0-9]{4})$
> > ']]
> >
> > which give the warning:
> >
> >
> > /var/etsi/work/V1.18.1/Natparas2V18.py:7644: UserWarning: Value
> > "b'20180719123801+0200'" does not match xsd pattern
> > restrictions:
> > [['^\\d{4}(0[1-9]$|^1[012])(0[1-9]$|^[12][0-9]$|^3[01])([01][0-9]$|^2[
> > 0-3])([0-5][0-9])([0-5][0-9])(.[0-9][0-9]*)?(Z$|^[\\+$|^\\-][0-9]{4})$
> > ']] warnings_.warn('Value "%s" does not match xsd pattern
> > restrictions: %s' % (value.encode('utf-8'),
> > self.validate_GeneralizedTime_patterns_, ))
> >
> > Adjusting to:
> >
> >
> > validate_GeneralizedTime_patterns_ =
> > [['^\\d{4}(0[1-9]|1[012])(0[1-9]|[12][0-9]|3[01])([01][0-9]|2[0-3])([0
> > -5][0-9])([0-5][0-9])(.[0-9][0-9]*)?(Z|[\\+|\\-][0-9]{4})$']]
> >
> > gives no warning.
> >
> > So it might be a problem with the separator | which results in a wrong ^|$
> > pattern?
> >
> >
> > Second one, discovering a problem with:
> >
> > testxml.export(sys.stdout, 0, '', 'retainedDataMessage')
> > at:
> > outfile.write('<%s%s%s' % (namespace_, name_, namespacedef_ and ' ' +
> > namespacedef_ or '', ))
> >
> > This works correct, printing out <retainedDataMessage> or with a
> > generatednamespaces.py with GenerateDSNamespaceDefs the
> > <retainedDataMessage NSGIVEN>
> >
> > But:
> >
> > testxml.export(sys.stdout, 0, 'NSDATA', 'retainedDataMessage')
> >
> > gives: <NSDATAretainedDataMessage> which is not correct, it should be
> > <retainedDataMessage NSDATA>, so it's printed in the wrong order and a
> > space is missing between the two arguments.
> >
> > Perhaps a:
> > outfile.write('<%s%s%s' % (name_, namespace_ and ' ' + namespace_ or
> > '', namespacedef_ and ' ' + namespacedef_ or '', ))
> >
> > could fix this. Note: the closing tag must be adjusted too.
> >
> > Perhaps the export could also be adjusted to overwrite the namespacedef_ if
> > a namespace_ is given or vice versa ...
> >
> > I am using the latest version 2.29.17 installed via pip under Python
> > 3.7
> >
> > Hope you could fix it.
> >
> > Again, great work!
> >
> > Best regards,
> > Bernd
> >
>
> > ----------------------------------------------------------------------
> > -------- Check out the vibrant tech community on one of the world's
> > most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>
> > _______________________________________________
> > generateds-users mailing list
> > generateds-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/generateds-users
>
>
>
> --
>
> Dave Kuhlman
> http://www.davekuhlman.org
--
Dave Kuhlman
http://www.davekuhlman.org
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
generateds-users mailing list
generateds-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/generateds-users
