Can't understand schema validation error

[Brad Wood]

I'm using a 2.0 XmlReaderSettings object with setting.ValidationType set
to ValdationType.Schema to validate a document against a schema.

Following is a schema fragment (names altered):
<xsd:element name="Stuff">
<xsd:complexType>
<xsd:sequence>
<xsd:element ref="RequiredSimpleThing1"/>
<xsd:element ref="RequiredSimpleThing2"/>
<xsd:element name="OptionalSimpleThing1" type="OptionalType"
minOccurs="0"/>
<xsd:element name="OptionalSimpleThing2" type="OptionalType"
minOccurs="0"/>
<xsd:element name="ComplexThing" minOccurs="0">
<xsd:complexType>
<xsd:sequence maxOccurs="unbounded">
<xsd:element name="ChildThing" type="ChildType"/>
</xsd:sequence>
</xsd:complexType>
</xsd:element>
</xsd:sequence>
</xsd:complexType>
</xsd:element>

Following is a document fragment:
<Stuff>
<RequiredSimpleThing1>86876</RequiredSimpleThing1>
<RequiredSimpleThing2>MC</RequiredSimpleThing2>
<ComplexThing>
<!-- several ChildThing elements -->
</ComplexThing>
</Stuff>

The error I recieve from the validator is "The element 'Stuff' has
invalid child element 'ComplexThing'. List of possible elements
expected: 'OptionalSimpleThing1, OptionalSimpleThing2'."

Why is the validator requiring these optional (minOccurs = 0) elements?

[Nick Hounsome]

"Brad Wood" <bradley|.wood|@ndsu|.edu> wrote in message
news:uz**************@TK2MSFTNGP09.phx.gbl...

I'm using a 2.0 XmlReaderSettings object with setting.ValidationType set
to ValdationType.Schema to validate a document against a schema.

Following is a schema fragment (names altered):
<xsd:element name="Stuff">
<xsd:complexType>
<xsd:sequence>
<xsd:element ref="RequiredSimpleThing1"/>
<xsd:element ref="RequiredSimpleThing2"/>
<xsd:element name="OptionalSimpleThing1" type="OptionalType"
minOccurs="0"/>
<xsd:element name="OptionalSimpleThing2" type="OptionalType"
minOccurs="0"/>
<xsd:element name="ComplexThing" minOccurs="0">
<xsd:complexType>
<xsd:sequence maxOccurs="unbounded">
<xsd:element name="ChildThing" type="ChildType"/>
</xsd:sequence>
</xsd:complexType>
</xsd:element>
</xsd:sequence>
</xsd:complexType>
</xsd:element>

Following is a document fragment:
<Stuff>
<RequiredSimpleThing1>86876</RequiredSimpleThing1>
<RequiredSimpleThing2>MC</RequiredSimpleThing2>
<ComplexThing>
<!-- several ChildThing elements -->
</ComplexThing>
</Stuff>

The error I recieve from the validator is "The element 'Stuff' has invalid
child element 'ComplexThing'. List of possible elements expected:
'OptionalSimpleThing1, OptionalSimpleThing2'."

Why is the validator requiring these optional (minOccurs = 0) elements?

Looks OK to me. Perhaps you have oversimplified in your post.
Post some real files (cut down and names changed if necessary) that produce
the error.

[Peter Morris [Droopy eyes software]]

I am new to XSD, but here I go anyway, speaking up and showing myself to be
a fool :-)

01) I use xs:all rather than xs:sequence if the order of the elements is not
important. (Not really relevant)
02) You have "xsd:" everywhere, a newsgroup typo?
03) "xs:sequence maxOccurs", shouldn't the maxOccures not be in xs;sequence
but in the xs:element beneath it?

Pete

[Nick Hounsome]

"Peter Morris [Droopy eyes software]" <pete@droopyeyes[dot]comREMOVETHIS>
wrote in message news:%2****************@TK2MSFTNGP11.phx.gbl...

I am new to XSD, but here I go anyway, speaking up and showing myself to be
a fool :-)

01) I use xs:all rather than xs:sequence if the order of the elements is
not important. (Not really relevant)

There are good reasons not to use "all".

Yes - in abstract modelling terms it is correct but in practice it is never
a good idea.

[Peter Morris [Droopy eyes software]]

Hi Nick

There are good reasons not to use "all".

Yes - in abstract modelling terms it is correct but in practice it is
never a good idea.

I'd be interested in hearing them, as I am using xs:all quite a lot at the
moment.
Pete

[Brad Wood]

Peter Morris [Droopy eyes software] wrote:

01) I use xs:all rather than xs:sequence if the order of the elements is not
important. (Not really relevant)
02) You have "xsd:" everywhere, a newsgroup typo?
03) "xs:sequence maxOccurs", shouldn't the maxOccures not be in xs;sequence
but in the xs:element beneath it?

The xsd is generated by Altova's XMLSpy.

[Nick Hounsome]

"Peter Morris [Droopy eyes software]" <pete@droopyeyes[dot]comREMOVETHIS>
wrote in message news:u$*************@tk2msftngp13.phx.gbl...

Hi Nick

There are good reasons not to use "all".

Yes - in abstract modelling terms it is correct but in practice it is
never a good idea.

I'd be interested in hearing them, as I am using xs:all quite a lot at the
moment.

Get "XML Schema" by Eric Van Der Vlist published by O'Reilly.

I quote briefly:

"unordered content models not only increase the risk of non-deterministic
content models but are also an important complexity factor for schema
processors. For the sake of implementation simplicity the recommendation has
imposed huge limitations on the xs:all element which makes it hardly usable
in practice."

In particular xs:all

1) cannot be used as a particle
2) cannot have a number of occurences > 1
3) the particles included within xs:all must be xs:element and must not
specify maxOccurences > 1

In simple schemas you will be OK but once you get into groups, choices,
extension and restriction you will wish that you hadn't used them.

The book explains the limitations in detail but it is easier to just
remember not to use it.

[Priscilla Walmsley]

Hi Brad,

LicensingStatus and InsuranceStatus are required (you don't have
minOccurs="0" on them), and they are missing.

Hope that helps,
Priscilla
------------------------------------------------------------------
Priscilla Walmsley
Author, Definitive XML Schema / XML in Office 2003
http://www.datypic.com
------------------------------------------------------------------

"Brad Wood" <bradley|.wood|@ndsu|.edu> wrote in message
news:uJ*************@tk2msftngp13.phx.gbl...

Nick Hounsome wrote:

Looks OK to me. Perhaps you have oversimplified in your post.
Post some real files (cut down and names changed if necessary) that
produce
the error.

The schema will be public in a day or 2, so I attached it. The basic
validation code I'm using can be found a couple posts prior in this
group with subject "validating XML with XmlValidatingReader" (my
response to question).

Here is the doc I'm validating:
<?xml version="1.0" encoding="utf-16"?>
<LicensingInsurance>
<CompanyUSDOTNum>123</CompanyUSDOTNum>
<CompanyDocketPrefixCode>MC</CompanyDocketPrefixCode>
<CompanyDocketNum>123</CompanyDocketNum>
<CompanyLegalName>Company A</CompanyLegalName>
<CompanyDBAName>Company A</CompanyDBAName>
<InsurancePolicies>
<InsurancePolicy>
<InsuranceTypeCode>1</InsuranceTypeCode>
<InsurerName>INSURANCE COMPANY</InsurerName>
<PolicyNumber>123</PolicyNumber>
<PolicyEffectiveDate>2004-10-03</PolicyEffectiveDate>
<PolicyCoverageLowAmount>0</PolicyCoverageLowAmount>
<PolicyCoverageHighAmount>5000000</PolicyCoverageHighAmount>
</InsurancePolicy>
</InsurancePolicies>
</LicensingInsurance>

[Priscilla Walmsley]

Or buy my book, Definitive XML Schema. It's on page 289. :-)

I too avoid xs:all, for the reasons described below.

------------------------------------------------------------------
Priscilla Walmsley
Author, Definitive XML Schema / XML in Office 2003
http://www.datypic.com
------------------------------------------------------------------

"Nick Hounsome" <nh***@nickhounsome.me.uk> wrote in message
news:Tb*****************@fe2.news.blueyonder.co.uk ...

"Peter Morris [Droopy eyes software]" <pete@droopyeyes[dot]comREMOVETHIS>
wrote in message news:u$*************@tk2msftngp13.phx.gbl...

Hi Nick

There are good reasons not to use "all".

Yes - in abstract modelling terms it is correct but in practice it is
never a good idea.

I'd be interested in hearing them, as I am using xs:all quite a lot at
the moment.

Get "XML Schema" by Eric Van Der Vlist published by O'Reilly.

I quote briefly:

"unordered content models not only increase the risk of non-deterministic
content models but are also an important complexity factor for schema
processors. For the sake of implementation simplicity the recommendation
has imposed huge limitations on the xs:all element which makes it hardly
usable in practice."

In particular xs:all

1) cannot be used as a particle
2) cannot have a number of occurences > 1
3) the particles included within xs:all must be xs:element and must not
specify maxOccurences > 1

In simple schemas you will be OK but once you get into groups, choices,
extension and restriction you will wish that you hadn't used them.

The book explains the limitations in detail but it is easier to just
remember not to use it.

[Priscilla Walmsley]

Hi Peter,

There's nothing wrong with using the xsd: prefix instead of xs:, as long as
you declare it. Prefixes are technically irrelevant. You could use "bob:"
if you wanted.

About maxOccurs, it can occur on the xs:sequence. In this case, since there
is one element child, it doesn't make a difference whether you put it on the
xs:sequence or on the xs:element. If there were multiple element
declarations in the sequence group, you could use maxOccurs on the sequence
to allow the entire sequence of elements to appear multiple times.

Hope that helps,
Priscilla

------------------------------------------------------------------
Priscilla Walmsley
Author, Definitive XML Schema / XML in Office 2003
http://www.datypic.com
------------------------------------------------------------------
"Peter Morris [Droopy eyes software]" <pete@droopyeyes[dot]comREMOVETHIS>
wrote in message news:%2****************@TK2MSFTNGP11.phx.gbl...

I am new to XSD, but here I go anyway, speaking up and showing myself to be
a fool :-)

01) I use xs:all rather than xs:sequence if the order of the elements is
not important. (Not really relevant)
02) You have "xsd:" everywhere, a newsgroup typo?
03) "xs:sequence maxOccurs", shouldn't the maxOccures not be in
xs;sequence but in the xs:element beneath it?

Pete

[Brad Wood]

Priscilla Walmsley wrote:

Hi Brad,

LicensingStatus and InsuranceStatus are required (you don't have
minOccurs="0" on them), and they are missing.

OK, this is what threw me for a loop: the error I get is, "The element
'LicensingInsurance' has invalid child element 'InsurancePolicies'. List
of possible elements expected: 'CarrierClassPropertyFlag,
CarrierClassPassengerFlag, CarrierClassHouseholdFlag,
CarrierBOC3FormFlag, CompanyAddress, CompanyContact, LicensingStatus'."

So the list of expected elements includes all of the optional elements
and only one out of two of the required elements. I was being dumb and
didn't even notice the required element in the list - only all the
optional ones. It is kind of strange that one of the required elements
is missing from the list in the error message.

[Brad Wood]

Now when I try to validate, I get the error, "The 'PolicyEffectiveDate'
element is invalid - The value '2003-10-4' is invalid according to its
datatype 'http://www.w3.org/2001/XMLSchema:date' - The string
'2003-10-4' is not a valid XsdDateTime value."

Judging from the spec, '2003-10-4' is a valid date. The data type for
the element is date (not DateTime), but the end of the error message
refers to XsdDateTime??

[Priscilla Walmsley]

Hi Brad,

I think the reason it didn't list the required InsuranceStatus element is
that it couldn't possibly appear there (where InsurancePolicies erroneously
appears) - LicensingStatus would have to appear first. So, it was telling
you what the valid elements after CompanyDBAName.

Hope that helps,
Priscilla

------------------------------------------------------------------
Priscilla Walmsley
Author, Definitive XML Schema / XML in Office 2003
http://www.datypic.com
------------------------------------------------------------------

"Brad Wood" <bradley|.wood|@ndsu|.edu> wrote in message
news:Om**************@TK2MSFTNGP09.phx.gbl...

Priscilla Walmsley wrote:

Hi Brad,

LicensingStatus and InsuranceStatus are required (you don't have
minOccurs="0" on them), and they are missing.

OK, this is what threw me for a loop: the error I get is, "The element
'LicensingInsurance' has invalid child element 'InsurancePolicies'. List
of possible elements expected: 'CarrierClassPropertyFlag,
CarrierClassPassengerFlag, CarrierClassHouseholdFlag, CarrierBOC3FormFlag,
CompanyAddress, CompanyContact, LicensingStatus'."

So the list of expected elements includes all of the optional elements and
only one out of two of the required elements. I was being dumb and didn't
even notice the required element in the list - only all the optional ones.
It is kind of strange that one of the required elements is missing from
the list in the error message.

[Priscilla Walmsley]

Hi Again,

The day has to be two digits long, so it should be 2003-10-04. I'm not
sure why it says XsdDateTime.

Hope that helps,
Priscilla

------------------------------------------------------------------
Priscilla Walmsley
Author, Definitive XML Schema / XML in Office 2003
http://www.datypic.com
------------------------------------------------------------------

"Brad Wood" <bradley|.wood|@ndsu|.edu> wrote in message
news:%2****************@tk2msftngp13.phx.gbl...

Now when I try to validate, I get the error, "The 'PolicyEffectiveDate'
element is invalid - The value '2003-10-4' is invalid according to its
datatype 'http://www.w3.org/2001/XMLSchema:date' - The string '2003-10-4'
is not a valid XsdDateTime value."

Judging from the spec, '2003-10-4' is a valid date. The data type for the
element is date (not DateTime), but the end of the error message refers to
XsdDateTime??

[Nick Hounsome]

If you are going to plug your book then do it properly :-)

See:
http://www.amazon.com/gp/product/013...books&n=283155

Looks OK.

As far as I can tell most books are fairly similar.

It was the practical advice and guidlines that I was most looking for when I
was looking for a book (sorry - didn't see yours at the time) such as
avoiding xs:all and the advantages of groups over extensions.

What is probably considered out of scope but which would be very useful is a
description of the mapping of schemas to classes by xsd.exe. This is
woefully underdocumented anywhere that I know of yet is of great interest to
the .NET community.

"Priscilla Walmsley" <no****@datypic.com> wrote in message
news:eb*************@tk2msftngp13.phx.gbl...

Or buy my book, Definitive XML Schema. It's on page 289. :-)

I too avoid xs:all, for the reasons described below.

------------------------------------------------------------------
Priscilla Walmsley
Author, Definitive XML Schema / XML in Office 2003
http://www.datypic.com
------------------------------------------------------------------

"Nick Hounsome" <nh***@nickhounsome.me.uk> wrote in message
news:Tb*****************@fe2.news.blueyonder.co.uk ...

"Peter Morris [Droopy eyes software]" <pete@droopyeyes[dot]comREMOVETHIS>
wrote in message news:u$*************@tk2msftngp13.phx.gbl...

Hi Nick

There are good reasons not to use "all".

Yes - in abstract modelling terms it is correct but in practice it is
never a good idea.

I'd be interested in hearing them, as I am using xs:all quite a lot at
the moment.

Get "XML Schema" by Eric Van Der Vlist published by O'Reilly.

I quote briefly:

"unordered content models not only increase the risk of non-deterministic
content models but are also an important complexity factor for schema
processors. For the sake of implementation simplicity the recommendation
has imposed huge limitations on the xs:all element which makes it hardly
usable in practice."

In particular xs:all

1) cannot be used as a particle
2) cannot have a number of occurences > 1
3) the particles included within xs:all must be xs:element and must not
specify maxOccurences > 1

In simple schemas you will be OK but once you get into groups, choices,
extension and restriction you will wish that you hadn't used them.

The book explains the limitations in detail but it is easier to just
remember not to use it.

[Stan Kitsis [MSFT]]

> As far as I can tell most books are fairly similar.
Some books are more equal than others

I think (personal opinion) that Priscilla's book is by far the best book on
XML Schema. It is written in a language that is easy to understand, has
good examples and best practice advise, and is great for reference. I have
gone through a number of schema books and don't think anybody else comes
close to this one. If you want to learn XML Schema and for some reason
don't want to read the spec (which I don't blame you if you don't), read
this book.

--
Stan Kitsis
Program Manager, XML Technologies
Microsoft Corporation

This posting is provided "AS IS" with no warranties, and confers no rights.

"Nick Hounsome" <nh***@nickhounsome.me.uk> wrote in message
news:HU*************@fe1.news.blueyonder.co.uk...

If you are going to plug your book then do it properly :-)

See:
http://www.amazon.com/gp/product/013...books&n=283155

Looks OK.

As far as I can tell most books are fairly similar.

It was the practical advice and guidlines that I was most looking for when
I was looking for a book (sorry - didn't see yours at the time) such as
avoiding xs:all and the advantages of groups over extensions.

What is probably considered out of scope but which would be very useful is
a description of the mapping of schemas to classes by xsd.exe. This is
woefully underdocumented anywhere that I know of yet is of great interest
to the .NET community.

"Priscilla Walmsley" <no****@datypic.com> wrote in message
news:eb*************@tk2msftngp13.phx.gbl...

Or buy my book, Definitive XML Schema. It's on page 289. :-)

I too avoid xs:all, for the reasons described below.

------------------------------------------------------------------
Priscilla Walmsley
Author, Definitive XML Schema / XML in Office 2003
http://www.datypic.com
------------------------------------------------------------------

"Nick Hounsome" <nh***@nickhounsome.me.uk> wrote in message
news:Tb*****************@fe2.news.blueyonder.co.uk ...

"Peter Morris [Droopy eyes software]"
<pete@droopyeyes[dot]comREMOVETHIS> wrote in message
news:u$*************@tk2msftngp13.phx.gbl...
Hi Nick

> There are good reasons not to use "all".
>
> Yes - in abstract modelling terms it is correct but in practice it is
> never a good idea.

I'd be interested in hearing them, as I am using xs:all quite a lot at
the moment.
Get "XML Schema" by Eric Van Der Vlist published by O'Reilly.

I quote briefly:

"unordered content models not only increase the risk of
non-deterministic content models but are also an important complexity
factor for schema processors. For the sake of implementation simplicity
the recommendation has imposed huge limitations on the xs:all element
which makes it hardly usable in practice."

In particular xs:all

1) cannot be used as a particle
2) cannot have a number of occurences > 1
3) the particles included within xs:all must be xs:element and must not
specify maxOccurences > 1

In simple schemas you will be OK but once you get into groups, choices,
extension and restriction you will wish that you hadn't used them.

The book explains the limitations in detail but it is easier to just
remember not to use it.

[John Harcourt]

Pardon me for jumping in so late...

I understand the reasons for not using xs:all, but I am using it. For one
reason, my XML is generated automatically from my C# classes by using the
XMLSerializer. I couldn't figure out how to control the order of the elements
when serializing the classes, which caused validation errors.
"Priscilla Walmsley" wrote:

Or buy my book, Definitive XML Schema. It's on page 289. :-)

I too avoid xs:all, for the reasons described below.

------------------------------------------------------------------
Priscilla Walmsley
Author, Definitive XML Schema / XML in Office 2003
http://www.datypic.com
------------------------------------------------------------------

"Nick Hounsome" <nh***@nickhounsome.me.uk> wrote in message
news:Tb*****************@fe2.news.blueyonder.co.uk ...

"Peter Morris [Droopy eyes software]" <pete@droopyeyes[dot]comREMOVETHIS>
wrote in message news:u$*************@tk2msftngp13.phx.gbl...

Hi Nick

There are good reasons not to use "all".

Yes - in abstract modelling terms it is correct but in practice it is
never a good idea.

I'd be interested in hearing them, as I am using xs:all quite a lot at
the moment.

Get "XML Schema" by Eric Van Der Vlist published by O'Reilly.

I quote briefly:

"unordered content models not only increase the risk of non-deterministic
content models but are also an important complexity factor for schema
processors. For the sake of implementation simplicity the recommendation
has imposed huge limitations on the xs:all element which makes it hardly
usable in practice."

In particular xs:all

1) cannot be used as a particle
2) cannot have a number of occurences > 1
3) the particles included within xs:all must be xs:element and must not
specify maxOccurences > 1

In simple schemas you will be OK but once you get into groups, choices,
extension and restriction you will wish that you hadn't used them.

The book explains the limitations in detail but it is easier to just
remember not to use it.