Undoing restriction attributes for some complex xxxRefStructure, fixing problems in the spp,fixing a missing include in two files for VehicleRef, fixes problem with WSDL generation

[TuThoThai]

This fixes small bugs I have found after we merged #975, mostly due to the fact that it creates a double contraint on xxxRefStructure for items for which the restriction is defined just after in the XSD by using:
<xsd:restriction base="xxxRefStructure"/> that is linked to a complexType in which the attributes restrictions are alreay listed.

Namely, I am undoing the restriction for:

• AllCountriesRef
• AllTransportOrganisationsRefStructure
• AllPublicTransportOrganisationsRefStructure
• AllOperatorsRefStructure

It also removes some files in the project file that no longer exist.

It fixes some problems with the WSDL generation from the project

[skinkie]

Your interpretation is not correct. A restriction must name everything under it, hence if it is not named, it does not exist. There is no double constraint.

So while I can agree with the change, because it seems that those element in fact have a meaning, the reason why you are proposing it, is wrong.

[TuThoThai]

Your interpretation is not correct. A restriction must name everything under it, hence if it is not named, it does not exist. There is no double constraint.

So while I can agree with the change, because it seems that those element in fact have a meaning, the reason why you are proposing it, is wrong.

Chances are very high that I express my understanding wrongly. But, the way I read it is the above mentioned xxxRef or xxxRefStructure were part of Substition Groups for which the restriction was already existing.

[skinkie]

Chances are very high that I express my understanding wrongly. But, the way I read it is the above mentioned xxxRef or xxxRefStructure were part of Substition Groups for which the restriction was already existing.

Then you just understand it wrong. There are never double constraints.

[TuThoThai]

Then you just understand it wrong. There are never double constraints.

Apologies if the way I express what I see is using the wrong word. Sharing how I found the errors and the interpertation I had.
Steps:

1. Download the entire XSD from the next branch
2. Open it on XMLSpy
3. Cross-check some of my datasets
   I admit that I should have done that as we were reviewing your PRs @skinkie, so my bad.

At step 2, these errors were flagged:

And, when I went to the XSD schema, I read this:

So, the way I interpreted it is the 2 times restriction on TypeOfValueIdType was creating an issue. Hence me expressing it with "double contraint", which is not the most appropriate.
Regardless of how my brain was labelling it, in this PR I propose changes so that the only the 2nd restriction in the XSD is kept. So, the XSD can validate while the logic of our new checks preserved.

Is is clearer for review, @skinkie?

[skinkie]

Lets say first, these errors should be reported in xmllint too, they are not on this topic, and I am very sure they were on others. So that is interesting.

The first error mentions that an anonymous (thus a ComplexType as part of an element) is not valid. What I think that it wants to explain you is that the AllTransportOrganisationsRefStructure has a different restriction. Or that name="nameOfRefClass" type="NameOfClass" may not be the correct restriction.

This is not a fix, it is just reverting what it was. And since extension essentially copies the restriction 1:1 it resolves it.

My suggestion if you are in xmlspy maybe there is an option to convert the extension to a restriction, and see what the differences are.

@ue71603 anything you know that is possible?

[ue71603]

@TuThoThai @skinkie I see the error as well

[ue71603]

I just replicated the work. so I agree

[skinkie]

@ue71603 this is not something to agree with. The proposed is not a fix, it is just reverting what has been created. The question was: can xmlspy suggest what the restriction should look like to be correctly derived.

[ue71603]

a lot was missing the fixed="All" in ref.

[TuThoThai]

@skinkie, I did spent a bit of time trying all combination that you have in the drop-down list of XMLSpy to make a valid schema. And the only correct sequence is:

	<xsd:element name="AllCountriesRef">
		<xsd:annotation>
			<xsd:documentation>Reference to a country ISO 3166-1 Note that GB is used for UK . May be qualified with a 3166-2 subdivision e.g. gb +m ENG, GB + SCT, GB See www.iso.org/iso/country_codes/iso_3166_code_lists.htm.</xsd:documentation>
		</xsd:annotation>
		<xsd:complexType>
			<xsd:simpleContent>
				<xsd:restriction base="AllCountriesRefStructure" />
			</xsd:simpleContent>
		</xsd:complexType>	
	</xsd:element>
	<xsd:complexType name="AllCountriesRefStructure">
		<xsd:annotation>
			<xsd:documentation>Type for a reference All Countries.</xsd:documentation>
		</xsd:annotation>
		<xsd:simpleContent>
			<xsd:restriction base="TypeOfValueRefStructure">
				<xsd:attribute name="ref" type="TypeOfValueIdType" use="required" fixed="All">
					<xsd:annotation>
						<xsd:documentation>Iso 3166-1 Two Character country code.</xsd:documentation>
					</xsd:annotation>
				</xsd:attribute>
			</xsd:restriction>
		</xsd:simpleContent>
	</xsd:complexType>

Which equals to reverting your PR on 4 elements only. I think it would be best to accept this for now and rework that part for v2.1 or v3.0.

[skinkie]

				<xsd:restriction base="AllCountriesRefStructure" />

This is wrong. Because all attributes below are then ommited.

[TuThoThai]

This is wrong. Because all attributes below are then ommited.

The way I read it is that the XSD sends you to the restrictions listed below in the AllCountriesRefStructure which then is:

<xsd:restriction base="TypeOfValueRefStructure">
  <xsd:attribute name="ref" type="TypeOfValueIdType" use="required" fixed="All">
	  <xsd:annotation>
		  <xsd:documentation>Iso 3166-1 Two Character country code.</xsd:documentation>
	  </xsd:annotation>
  </xsd:attribute>
</xsd:restriction>

[TuThoThai]

Thanks for fixing the last items, @ue71603! Looks good to me 🙏

[ue71603]

@skinkie sorry. still not clear. the fixed="All" is gone at that location.
fixed="All" we only have now here:

[skinkie]

@skinkie sorry. still not clear. the fixed="All" is gone at that location. fixed="All" we only have now here:

Remove the entire ref attribute, if possible.

[skinkie]

I am unhappy with the complexContent to simpleContent. But if it works, I am ok.

[thbar]

LGTM