XmlSerializer: deserialize against xsd generated class

[=?Utf-8?B?ai5hLiBoYXJyaW1hbg==?=]

Hi,

I have a schema that has an optional element, fieldTag4000Fie ld. If the
element is omitted from the XML request, when it is deserialized, it will be
null when I check it - which is fine.

What happens when the element is supplied as <fieldTag4000Fi eld/(empty),
it does not equate to null. I want to be able handle this at the
deserialization level rahter than in my edits later.

Is there a way to alter the behavior so when I deserialize it, it is null?

I also want to add that I am calling the "CanDeserialize " method of the
XmlSerializer object, passing in the request as a XmlNodeReader.

Thanks.

The following C# class snippet was generated by xsd.exe V2:

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd", "2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageType {

private Request_TypeMes sageTypeFieldTa g4000 fieldTag4000Fie ld;

public Request_TypeMes sageTypeFieldTa g4000 FieldTag4000 {
get {
return this.fieldTag40 00Field;
}
set {
this.fieldTag40 00Field = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd", "2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageTypeFieldTa g4000 {

private IDType_Type intermediaryIDC odeField;

private bool intermediaryIDC odeFieldSpecifi ed;

private string intermediaryIde ntifierField;

public IDType_Type IntermediaryIDC ode {
get {
return this.intermedia ryIDCodeField;
}
set {
this.intermedia ryIDCodeField = value;
}
}
[System.Xml.Seri alization.XmlIg noreAttribute()]
public bool IntermediaryIDC odeSpecified {
get {
return this.intermedia ryIDCodeFieldSp ecified;
}
set {
this.intermedia ryIDCodeFieldSp ecified = value;
}
}
public string IntermediaryIde ntifier {
get {
return this.intermedia ryIdentifierFie ld;
}
set {
this.intermedia ryIdentifierFie ld = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd", "2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Xml.Seri alization.XmlTy peAttribute(Nam espace="MyNames pace")]
public enum IDType_Type {

B,

C,

D,

F,

U,
}

[Joe Fawcett]

Unless you want to implement IXmlSerializabl e then I don't think you can do
what you want. You can specify that a null element is shown as:
<myElement xsi:nil="true"/>
rather than being omitted entirely by adding the XmlElement(IsNu llable =
true) attribute to the field/property.

--

Joe Fawcett (MVP - XML)

http://joe.fawcett.name

"j.a. harriman" <je************ *******@nospam. nospamwrote in message
news:69******** *************** ***********@mic rosoft.com...

Hi,

I have a schema that has an optional element, fieldTag4000Fie ld. If the
element is omitted from the XML request, when it is deserialized, it will
be
null when I check it - which is fine.

What happens when the element is supplied as <fieldTag4000Fi eld/(empty),
it does not equate to null. I want to be able handle this at the
deserialization level rahter than in my edits later.

Is there a way to alter the behavior so when I deserialize it, it is null?

I also want to add that I am calling the "CanDeserialize " method of the
XmlSerializer object, passing in the request as a XmlNodeReader.

Thanks.

The following C# class snippet was generated by xsd.exe V2:

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd", "2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageType {

private Request_TypeMes sageTypeFieldTa g4000 fieldTag4000Fie ld;

public Request_TypeMes sageTypeFieldTa g4000 FieldTag4000 {
get {
return this.fieldTag40 00Field;
}
set {
this.fieldTag40 00Field = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd", "2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageTypeFieldTa g4000 {

private IDType_Type intermediaryIDC odeField;

private bool intermediaryIDC odeFieldSpecifi ed;

private string intermediaryIde ntifierField;

public IDType_Type IntermediaryIDC ode {
get {
return this.intermedia ryIDCodeField;
}
set {
this.intermedia ryIDCodeField = value;
}
}
[System.Xml.Seri alization.XmlIg noreAttribute()]
public bool IntermediaryIDC odeSpecified {
get {
return this.intermedia ryIDCodeFieldSp ecified;
}
set {
this.intermedia ryIDCodeFieldSp ecified = value;
}
}
public string IntermediaryIde ntifier {
get {
return this.intermedia ryIdentifierFie ld;
}
set {
this.intermedia ryIdentifierFie ld = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd", "2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Xml.Seri alization.XmlTy peAttribute(Nam espace="MyNames pace")]
public enum IDType_Type {

B,

C,

D,

F,

U,
}

[=?Utf-8?B?ai5hLiBoYXJyaW1hbg==?=]

Thanks for the answer Joe.

I also found some examples of using a style sheet and doing a transformation
(Using XslCompiledTran sform) of the original XmlDocument. I've run a test
and it seems to work in that the result is as if the original message hadn't
had them in there. Are there any reasons why not to implement this?

Also, I tried to locate good "beginner" examples of using IXmlSerializabl e
that might be similar to what I need to do, but came up short. I would like
to to look at this in further detail.

Do you have any links to examples or are you aware if any of the Microsoft
example downloads (such as SDK) have any?

Thanks. Jeff

"Joe Fawcett" wrote:

Unless you want to implement IXmlSerializabl e then I don't think you can do
what you want. You can specify that a null element is shown as:
<myElement xsi:nil="true"/>
rather than being omitted entirely by adding the XmlElement(IsNu llable =
true) attribute to the field/property.

--

Joe Fawcett (MVP - XML)

http://joe.fawcett.name

"j.a. harriman" <je************ *******@nospam. nospamwrote in message
news:69******** *************** ***********@mic rosoft.com...

Hi,

I have a schema that has an optional element, fieldTag4000Fie ld. If the
element is omitted from the XML request, when it is deserialized, it will
be
null when I check it - which is fine.

What happens when the element is supplied as <fieldTag4000Fi eld/(empty),
it does not equate to null. I want to be able handle this at the
deserialization level rahter than in my edits later.

Is there a way to alter the behavior so when I deserialize it, it is null?

I also want to add that I am calling the "CanDeserialize " method of the
XmlSerializer object, passing in the request as a XmlNodeReader.

Thanks.

The following C# class snippet was generated by xsd.exe V2:

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd", "2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageType {

private Request_TypeMes sageTypeFieldTa g4000 fieldTag4000Fie ld;

public Request_TypeMes sageTypeFieldTa g4000 FieldTag4000 {
get {
return this.fieldTag40 00Field;
}
set {
this.fieldTag40 00Field = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd", "2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageTypeFieldTa g4000 {

private IDType_Type intermediaryIDC odeField;

private bool intermediaryIDC odeFieldSpecifi ed;

private string intermediaryIde ntifierField;

public IDType_Type IntermediaryIDC ode {
get {
return this.intermedia ryIDCodeField;
}
set {
this.intermedia ryIDCodeField = value;
}
}
[System.Xml.Seri alization.XmlIg noreAttribute()]
public bool IntermediaryIDC odeSpecified {
get {
return this.intermedia ryIDCodeFieldSp ecified;
}
set {
this.intermedia ryIDCodeFieldSp ecified = value;
}
}
public string IntermediaryIde ntifier {
get {
return this.intermedia ryIdentifierFie ld;
}
set {
this.intermedia ryIdentifierFie ld = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd", "2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Xml.Seri alization.XmlTy peAttribute(Nam espace="MyNames pace")]
public enum IDType_Type {

B,

C,

D,

F,

U,
}

[Joe Fawcett]

Jeff

IXmlSerializabl e is reasonably straightforward .
The newer version needs three methods, ReadXml, WriteXml and one to find the
Schema which is pointed to by an attribute, I normally just include the
schema as an embedded resource.
In ReadXml you get an XmlReader containing the XML and use it to populate
the object's fields, either reading it directly or loading it into a
DomDocumnt/XPathDocument if that's easier. WriteXml takes the fields and
creates an XML document.
There is an example here: http://www.devx.com/dotnet/Article/29720

--

Joe Fawcett (MVP - XML)
http://joe.fawcett.name

"j.a. harriman" <je************ *******@nospam. nospamwrote in message
news:04******** *************** ***********@mic rosoft.com...

Thanks for the answer Joe.

I also found some examples of using a style sheet and doing a
transformation
(Using XslCompiledTran sform) of the original XmlDocument. I've run a test
and it seems to work in that the result is as if the original message
hadn't
had them in there. Are there any reasons why not to implement this?

Also, I tried to locate good "beginner" examples of using IXmlSerializabl e
that might be similar to what I need to do, but came up short. I would
like
to to look at this in further detail.

Do you have any links to examples or are you aware if any of the Microsoft
example downloads (such as SDK) have any?

Thanks. Jeff

"Joe Fawcett" wrote:

>Unless you want to implement IXmlSerializabl e then I don't think you can
do
what you want. You can specify that a null element is shown as:
<myElement xsi:nil="true"/>
rather than being omitted entirely by adding the XmlElement(IsNu llable =
true) attribute to the field/property.

--

Joe Fawcett (MVP - XML)

http://joe.fawcett.name

"j.a. harriman" <je************ *******@nospam. nospamwrote in message
news:69******* *************** ************@mi crosoft.com...

Hi,

I have a schema that has an optional element, fieldTag4000Fie ld. If
the
element is omitted from the XML request, when it is deserialized, it
will
be
null when I check it - which is fine.

What happens when the element is supplied as <fieldTag4000Fi eld/>
(empty),
it does not equate to null. I want to be able handle this at the
deserialization level rahter than in my edits later.

Is there a way to alter the behavior so when I deserialize it, it is
null?

I also want to add that I am calling the "CanDeserialize " method of the
XmlSerializer object, passing in the request as a XmlNodeReader.

Thanks.

The following C# class snippet was generated by xsd.exe V2:

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageType {

private Request_TypeMes sageTypeFieldTa g4000 fieldTag4000Fie ld;

public Request_TypeMes sageTypeFieldTa g4000 FieldTag4000 {
get {
return this.fieldTag40 00Field;
}
set {
this.fieldTag40 00Field = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageTypeFieldTa g4000 {

private IDType_Type intermediaryIDC odeField;

private bool intermediaryIDC odeFieldSpecifi ed;

private string intermediaryIde ntifierField;

public IDType_Type IntermediaryIDC ode {
get {
return this.intermedia ryIDCodeField;
}
set {
this.intermedia ryIDCodeField = value;
}
}
[System.Xml.Seri alization.XmlIg noreAttribute()]
public bool IntermediaryIDC odeSpecified {
get {
return this.intermedia ryIDCodeFieldSp ecified;
}
set {
this.intermedia ryIDCodeFieldSp ecified = value;
}
}
public string IntermediaryIde ntifier {
get {
return this.intermedia ryIdentifierFie ld;
}
set {
this.intermedia ryIdentifierFie ld = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Xml.Seri alization.XmlTy peAttribute(Nam espace="MyNames pace")]
public enum IDType_Type {

B,

C,

D,

F,

U,
}

[Sprotty]

Hi Jeff

The code produced by xsd.exe, copes well with simple schema's, but if
your working against anything a bit complicated then it can cause
issues.

If you continue to have problems then I suggest you take a look at
Liquid XML Data Binder

http://www.liquid-technologies.com/P...taBinding.aspx

This supports much more of the xsd standard, and the objects are more
strongly typed (it also comes with a free XSD editor).

Hope this helps

Regards Simon

On 23 Jul, 08:00, "Joe Fawcett" <joefawc...@new sgroup.nospamwr ote:

Jeff

IXmlSerializabl e is reasonably straightforward .
The newer version needs three methods, ReadXml, WriteXml and one to find the
Schema which is pointed to by an attribute, I normally just include the
schema as an embedded resource.
In ReadXml you get an XmlReader containing the XML and use it to populate
the object's fields, either reading it directly or loading it into a
DomDocumnt/XPathDocument if that's easier. WriteXml takes the fields and
creates an XML document.
There is an example here:http://www.devx.com/dotnet/Article/29720

--

Joe Fawcett (MVP - XML)http://joe.fawcett.name

"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message

news:04******** *************** ***********@mic rosoft.com...

Thanks for the answer Joe.

I also found some examples of using a style sheet and doing a
transformation
(Using XslCompiledTran sform) of the original XmlDocument. *I've run atest
and it seems to work in that the result is as if the original message
hadn't
had them in there. *Are there any reasons why not to implement this?

Also, I tried to locate good "beginner" examples of using IXmlSerializabl e
that might be similar to what I need to do, but came up short. *I would
like
to to look at this in further detail.

Do you have any links to examples or are you aware if any of the Microsoft
example downloads (such as SDK) have any?

Thanks. Jeff

"Joe Fawcett" wrote:

Unless you want to implement IXmlSerializabl e then I don't think you can
do
what you want. You can specify that a null element is shown as:
<myElement xsi:nil="true"/>
rather than being omitted entirely by adding the XmlElement(IsNu llable=
true) attribute to the field/property.

--

Joe Fawcett (MVP - XML)

>http://joe.fawcett.name

"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message
news:69******* *************** ************@mi crosoft.com...
Hi,

I have a schema that has an optional element, fieldTag4000Fie ld. *If
the
element is omitted from the XML request, when it is deserialized, it
will
be
null when I check it - which is fine.

What happens when the element is supplied as <fieldTag4000Fi eld/>
(empty),
it does not equate to null. *I want to be able handle this at the
deserialization level rahter than in my edits later.

Is there a way to alter the behavior so when I deserialize it, it is
null?

I also want to add that I am calling the "CanDeserialize " method of the
XmlSerializer object, passing in the request as a XmlNodeReader.

Thanks.

The following C# class snippet was generated by xsd.exe V2:

* *[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
* *[System.Serializ ableAttribute()]
* *[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
* *[System.Componen tModel.Designer CategoryAttribu te("code")]
* *[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
* *public partial class Request_TypeMes sageType {

* * * *private Request_TypeMes sageTypeFieldTa g4000 fieldTag4000Fie ld;

* * * *public Request_TypeMes sageTypeFieldTa g4000 FieldTag4000 {
* * * * * *get {
* * * * * * * *return this.fieldTag40 00Field;
* * * * * *}
* * * * * *set {
* * * * * * * *this.fieldTag4 000Field = value;
* * * * * *}
* * * *}
* *}

* *[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
* *[System.Serializ ableAttribute()]
* *[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
* *[System.Componen tModel.Designer CategoryAttribu te("code")]
* *[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
* *public partial class Request_TypeMes sageTypeFieldTa g4000 {

* * * *private IDType_Type intermediaryIDC odeField;

* * * *private bool intermediaryIDC odeFieldSpecifi ed;

* * * *private string intermediaryIde ntifierField;

* * * *public IDType_Type IntermediaryIDC ode {
* * * * * *get {
* * * * * * * *return this.intermedia ryIDCodeField;
* * * * * *}
* * * * * *set {
* * * * * * * *this.intermedi aryIDCodeField = value;
* * * * * *}
* * * *}

* * * *[System.Xml.Seri alization.XmlIg noreAttribute()]
* * * *public bool IntermediaryIDC odeSpecified {
* * * * * *get {
* * * * * * * *return this.intermedia ryIDCodeFieldSp ecified;
* * * * * *}
* * * * * *set {
* * * * * * * *this.intermedi aryIDCodeFieldS pecified= value;
* * * * * *}
* * * *}

* * * *public string IntermediaryIde ntifier {
* * * * * *get {
* * * * * * * *return this.intermedia ryIdentifierFie ld;
* * * * * *}
* * * * * *set {
* * * * * * * *this.intermedi aryIdentifierFi eld = value;
* * * * * *}
* * * *}
* *}

* *[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
* *[System.Serializ ableAttribute()]
* *[System.Xml.Seri alization.XmlTy peAttribute(Nam espace="MyNames pace")]
* *public enum IDType_Type {

* * * *B,

* * * *C,

* * * *D,

* * * *F,

* * * *U,
* *}- Hide quoted text -

- Show quoted text -

[=?Utf-8?B?ai5hLiBoYXJyaW1hbg==?=]

Simon,

Thanks for the link, I will take a look at it as it may be something we
could use.

In the meantime, I did find a code snippet that removes the "empty" nodes
but it removes ALL the empty nodes in the XML and I didn't state (earlier)
that I want to do this beginning with a particular child node only.

When I use the following XPath statement in XMLSpy's XPath Evaluator, it
locates the "empty" node mentioned in the previous posts:
/parentnode/secondchildnode//*[not(node())]

When I use the same statement in my C# code, it will not locate any empty
nodes "under" the "secondchildnod e" (count = 0):
XmlNodeList element =
doc.SelectNodes ("/parentnode/secondchildnode//*[not(node())]");

The plan is to loop through "element" and remove the nodes using the
"ParentNode.Rem oveChild" of an XmlNode - setting it to each array element,
etc.

Any ideas as to why this is not working in .NET VS2005 2.0 framework?

Thanks.

Jeff

"Sprotty" wrote:

Hi Jeff

The code produced by xsd.exe, copes well with simple schema's, but if
your working against anything a bit complicated then it can cause
issues.

If you continue to have problems then I suggest you take a look at
Liquid XML Data Binder

http://www.liquid-technologies.com/P...taBinding.aspx

This supports much more of the xsd standard, and the objects are more
strongly typed (it also comes with a free XSD editor).

Hope this helps

Regards Simon

On 23 Jul, 08:00, "Joe Fawcett" <joefawc...@new sgroup.nospamwr ote:

Jeff

IXmlSerializabl e is reasonably straightforward .
The newer version needs three methods, ReadXml, WriteXml and one to find the
Schema which is pointed to by an attribute, I normally just include the
schema as an embedded resource.
In ReadXml you get an XmlReader containing the XML and use it to populate
the object's fields, either reading it directly or loading it into a
DomDocumnt/XPathDocument if that's easier. WriteXml takes the fields and
creates an XML document.
There is an example here:http://www.devx.com/dotnet/Article/29720

--

Joe Fawcett (MVP - XML)http://joe.fawcett.name

"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message

news:04******** *************** ***********@mic rosoft.com...

Thanks for the answer Joe.

I also found some examples of using a style sheet and doing a
transformation
(Using XslCompiledTran sform) of the original XmlDocument. I've run a test
and it seems to work in that the result is as if the original message
hadn't
had them in there. Are there any reasons why not to implement this?

Also, I tried to locate good "beginner" examples of using IXmlSerializabl e
that might be similar to what I need to do, but came up short. I would
like
to to look at this in further detail.

Do you have any links to examples or are you aware if any of the Microsoft
example downloads (such as SDK) have any?

Thanks. Jeff

"Joe Fawcett" wrote:

>Unless you want to implement IXmlSerializabl e then I don't think you can
>do
>what you want. You can specify that a null element is shown as:
><myElement xsi:nil="true"/>
>rather than being omitted entirely by adding the XmlElement(IsNu llable =
>true) attribute to the field/property.

>--

>Joe Fawcett (MVP - XML)

>>http://joe.fawcett.name

>"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message
>>news:69****** *************** *************@m icrosoft.com...
Hi,

I have a schema that has an optional element, fieldTag4000Fie ld. If
the
element is omitted from the XML request, when it is deserialized, it
will
be
null when I check it - which is fine.

What happens when the element is supplied as <fieldTag4000Fi eld/>
(empty),
it does not equate to null. I want to be able handle this at the
deserialization level rahter than in my edits later.

Is there a way to alter the behavior so when I deserialize it, it is
null?

I also want to add that I am calling the "CanDeserialize " method of the
XmlSerializer object, passing in the request as a XmlNodeReader.

Thanks.

The following C# class snippet was generated by xsd.exe V2:

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageType {

private Request_TypeMes sageTypeFieldTa g4000 fieldTag4000Fie ld;

public Request_TypeMes sageTypeFieldTa g4000 FieldTag4000 {
get {
return this.fieldTag40 00Field;
}
set {
this.fieldTag40 00Field = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageTypeFieldTa g4000 {

private IDType_Type intermediaryIDC odeField;

private bool intermediaryIDC odeFieldSpecifi ed;

private string intermediaryIde ntifierField;

public IDType_Type IntermediaryIDC ode {
get {
return this.intermedia ryIDCodeField;
}
set {
this.intermedia ryIDCodeField = value;
}
}

[System.Xml.Seri alization.XmlIg noreAttribute()]
public bool IntermediaryIDC odeSpecified {
get {
return this.intermedia ryIDCodeFieldSp ecified;
}
set {
this.intermedia ryIDCodeFieldSp ecified = value;
}
}

public string IntermediaryIde ntifier {
get {
return this.intermedia ryIdentifierFie ld;
}
set {
this.intermedia ryIdentifierFie ld = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Xml.Seri alization.XmlTy peAttribute(Nam espace="MyNames pace")]
public enum IDType_Type {

B,

C,

D,

F,

U,
}- Hide quoted text -

- Show quoted text -

[Sprotty]

You could remove the empy elements in the DOM before loading it as you
suggest.
Im not seeing a problem with that

XmlDocument xmlDoc = new XmlDocument();
xmlDoc.LoadXml( "<parentnode><s econdchildnode> <b/></
secondchildnode ></parentnode>");
XmlNodeList xnl = xmlDoc.SelectNo des("//*[not(node())]");
Debug.Assert(xn l.Count == 1);

or you could remove it from the object model after its been loaded
(deserialized).

If you were using Liquid XML then you may be able to add code in that
would remove the empty elements as part of the de-serialization.

Cheers Simon
On 31 Jul, 21:22, j.a. harriman <jeffrey_no_spa m_al...@nospam. nospam>
wrote:

Simon,

Thanks for the link, I will take a look at it as it may be something we
could use.

In the meantime, I did find a code snippet that removes the "empty" nodes
but it removes ALL the empty nodes in the XML and I didn't state (earlier)
that I want to do this beginning with a particular child node only.

When I use the following XPath statement in XMLSpy's XPath Evaluator, it
locates the "empty" node mentioned in the previous posts:
/parentnode/secondchildnode//*[not(node())]

When I use the same statement in my C# code, it will not locate any empty
nodes "under" the "secondchildnod e" (count = 0):
XmlNodeList element =
doc.SelectNodes ("/parentnode/secondchildnode//*[not(node())]");

The plan is to loop through "element" and remove the nodes using the
"ParentNode.Rem oveChild" of an XmlNode - setting it to each array element,
etc.

Any ideas as to why this is not working in .NET VS2005 2.0 framework?

Thanks.

Jeff

"Sprotty" wrote:

Hi Jeff

The code produced by xsd.exe, copes well with simple schema's, but if
your working against anything a bit complicated then it can cause
issues.

If you continue to have problems then I suggest you take a look at
Liquid XML Data Binder

http://www.liquid-technologies.com/P...taBinding.aspx

This supports much more of the xsd standard, and the objects are more
strongly typed (it also comes with a free XSD editor).

Hope this helps

Regards Simon

On 23 Jul, 08:00, "Joe Fawcett" <joefawc...@new sgroup.nospamwr ote:

Jeff

IXmlSerializabl e is reasonably straightforward .
The newer version needs three methods, ReadXml, WriteXml and one to find the
Schema which is pointed to by an attribute, I normally just include the
schema as an embedded resource.
In ReadXml you get an XmlReader containing the XML and use it to populate
the object's fields, either reading it directly or loading it into a
DomDocumnt/XPathDocument if that's easier. WriteXml takes the fields and
creates an XML document.
There is an example here:http://www.devx.com/dotnet/Article/29720

--

Joe Fawcett (MVP - XML)http://joe.fawcett.name

"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message

>news:04******* *************** ************@mi crosoft.com...

Thanks for the answer Joe.

I also found some examples of using a style sheet and doing a
transformation
(Using XslCompiledTran sform) of the original XmlDocument. *I've run a test
and it seems to work in that the result is as if the original message
hadn't
had them in there. *Are there any reasons why not to implement this?

Also, I tried to locate good "beginner" examples of using IXmlSerializabl e
that might be similar to what I need to do, but came up short. *Iwould
like
to to look at this in further detail.

Do you have any links to examples or are you aware if any of the Microsoft
example downloads (such as SDK) have any?

Thanks. Jeff

"Joe Fawcett" wrote:

Unless you want to implement IXmlSerializabl e then I don't think you can
do
what you want. You can specify that a null element is shown as:
<myElement xsi:nil="true"/>
rather than being omitted entirely by adding the XmlElement(IsNu llable =
true) attribute to the field/property.

--

Joe Fawcett (MVP - XML)

>http://joe.fawcett.name

"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message
>news:69******* *************** ************@mi crosoft.com...
Hi,

I have a schema that has an optional element, fieldTag4000Fie ld.*If
the
element is omitted from the XML request, when it is deserialized, it
will
be
null when I check it - which is fine.

What happens when the element is supplied as <fieldTag4000Fi eld/>
(empty),
it does not equate to null. *I want to be able handle this at the
deserialization level rahter than in my edits later.

Is there a way to alter the behavior so when I deserialize it, it is
null?

I also want to add that I am calling the "CanDeserialize " methodof the
XmlSerializer object, passing in the request as a XmlNodeReader.

Thanks.

The following C# class snippet was generated by xsd.exe V2:

* *[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
* *[System.Serializ ableAttribute()]
* *[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
* *[System.Componen tModel.Designer CategoryAttribu te("code")]
* *[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
* *public partial class Request_TypeMes sageType {

* * * *private Request_TypeMes sageTypeFieldTa g4000 fieldTag4000Fie ld;

* * * *public Request_TypeMes sageTypeFieldTa g4000 FieldTag4000 {
* * * * * *get {
* * * * * * * *return this.fieldTag40 00Field;
* * * * * *}
* * * * * *set {
* * * * * * * *this.fieldTag4 000Field = value;
* * * * * *}
* * * *}
* *}

* *[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
* *[System.Serializ ableAttribute()]
* *[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
* *[System.Componen tModel.Designer CategoryAttribu te("code")]
* *[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
* *public partial class Request_TypeMes sageTypeFieldTa g4000 {

* * * *private IDType_Type intermediaryIDC odeField;

* * * *private bool intermediaryIDC odeFieldSpecifi ed;

* * * *private string intermediaryIde ntifierField;

* * * *public IDType_Type IntermediaryIDC ode {
* * * * * *get {
* * * * * * * *return this.intermedia ryIDCodeField;
* * * * * *}
* * * * * *set {
* * * * * * * *this.intermedi aryIDCodeField = value;
* * * * * *}
* * * *}

* * * *[System.Xml.Seri alization.XmlIg noreAttribute()]
* * * *public bool IntermediaryIDC odeSpecified {
* * * * * *get {
* * * * * * * *return this.intermedia ryIDCodeFieldSp ecified;
* * * * * *}
* * * * * *set {
* * * * * * * *this.intermedi aryIDCodeFieldS pecified = value;
* * * * * *}
* * * *}

* * * *public string IntermediaryIde ntifier {
* * * * * *get {
* * * * * * * *return this.intermedia ryIdentifierFie ld;
* * * * * *}
* * * * * *set {
* * * * * * * *this.intermedi aryIdentifierFi eld = value;
* * * * * *}
* * * *}
* *}

* *[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
* *[System.Serializ ableAttribute()]
* *[System.Xml.Seri alization.XmlTy peAttribute(Nam espace="MyNames pace")]
* *public enum IDType_Type {

* * * *B,

* * * *C,

* * * *D,

* * * *F,

* * * *U,
* *}- Hide quoted text -

- Show quoted text -

[=?Utf-8?B?ai5hLiBoYXJyaW1hbg==?=]

Building off your example, here's a respresentation of my actaul XML:
<parentnode>
<firstchildnode >
<a/>
</firstchildnode>
<secondchildnod e>
<b/>
<c>
<d/>
<e>12345</e>
</c>
</secondchildnode >
</parentnode>

The XPath expression, "//*[not(node())]", removes all "empty" elements. I
need it to *only* remove "b" & "d" above. "a" would be left alone.

Thanks.

Jeff

"Sprotty" wrote:

>
You could remove the empy elements in the DOM before loading it as you
suggest.
Im not seeing a problem with that

XmlDocument xmlDoc = new XmlDocument();
xmlDoc.LoadXml( "<parentnode><s econdchildnode> <b/></
secondchildnode ></parentnode>");
XmlNodeList xnl = xmlDoc.SelectNo des("//*[not(node())]");
Debug.Assert(xn l.Count == 1);

or you could remove it from the object model after its been loaded
(deserialized).

If you were using Liquid XML then you may be able to add code in that
would remove the empty elements as part of the de-serialization.

Cheers Simon
On 31 Jul, 21:22, j.a. harriman <jeffrey_no_spa m_al...@nospam. nospam>
wrote:

Simon,

Thanks for the link, I will take a look at it as it may be something we
could use.

In the meantime, I did find a code snippet that removes the "empty" nodes
but it removes ALL the empty nodes in the XML and I didn't state (earlier)
that I want to do this beginning with a particular child node only.

When I use the following XPath statement in XMLSpy's XPath Evaluator, it
locates the "empty" node mentioned in the previous posts:
/parentnode/secondchildnode//*[not(node())]

When I use the same statement in my C# code, it will not locate any empty
nodes "under" the "secondchildnod e" (count = 0):
XmlNodeList element =
doc.SelectNodes ("/parentnode/secondchildnode//*[not(node())]");

The plan is to loop through "element" and remove the nodes using the
"ParentNode.Rem oveChild" of an XmlNode - setting it to each array element,
etc.

Any ideas as to why this is not working in .NET VS2005 2.0 framework?

Thanks.

Jeff

"Sprotty" wrote:

Hi Jeff

The code produced by xsd.exe, copes well with simple schema's, but if
your working against anything a bit complicated then it can cause
issues.

If you continue to have problems then I suggest you take a look at
Liquid XML Data Binder

>http://www.liquid-technologies.com/P...taBinding.aspx

This supports much more of the xsd standard, and the objects are more
strongly typed (it also comes with a free XSD editor).

Hope this helps

Regards Simon

On 23 Jul, 08:00, "Joe Fawcett" <joefawc...@new sgroup.nospamwr ote:
Jeff

IXmlSerializabl e is reasonably straightforward .
The newer version needs three methods, ReadXml, WriteXml and one to find the
Schema which is pointed to by an attribute, I normally just include the
schema as an embedded resource.
In ReadXml you get an XmlReader containing the XML and use it to populate
the object's fields, either reading it directly or loading it into a
DomDocumnt/XPathDocument if that's easier. WriteXml takes the fields and
creates an XML document.
There is an example here:http://www.devx.com/dotnet/Article/29720

--

Joe Fawcett (MVP - XML)http://joe.fawcett.name

"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message

news:04******** *************** ***********@mic rosoft.com...

Thanks for the answer Joe.

I also found some examples of using a style sheet and doing a
transformation
(Using XslCompiledTran sform) of the original XmlDocument. I've run a test
and it seems to work in that the result is as if the original message
hadn't
had them in there. Are there any reasons why not to implement this?

Also, I tried to locate good "beginner" examples of using IXmlSerializabl e
that might be similar to what I need to do, but came up short. I would
like
to to look at this in further detail.

Do you have any links to examples or are you aware if any of the Microsoft
example downloads (such as SDK) have any?

Thanks. Jeff

"Joe Fawcett" wrote:

>Unless you want to implement IXmlSerializabl e then I don't think you can
>do
>what you want. You can specify that a null element is shown as:
><myElement xsi:nil="true"/>
>rather than being omitted entirely by adding the XmlElement(IsNu llable =
>true) attribute to the field/property.

>--

>Joe Fawcett (MVP - XML)

>>http://joe.fawcett.name

>"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message
>>news:69****** *************** *************@m icrosoft.com...
Hi,

I have a schema that has an optional element, fieldTag4000Fie ld. If
the
element is omitted from the XML request, when it is deserialized, it
will
be
null when I check it - which is fine.

What happens when the element is supplied as <fieldTag4000Fi eld/>
(empty),
it does not equate to null. I want to be able handle this at the
deserialization level rahter than in my edits later.

Is there a way to alter the behavior so when I deserialize it, it is
null?

I also want to add that I am calling the "CanDeserialize " method of the
XmlSerializer object, passing in the request as a XmlNodeReader.

Thanks.

The following C# class snippet was generated by xsd.exe V2:

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageType {

private Request_TypeMes sageTypeFieldTa g4000 fieldTag4000Fie ld;

public Request_TypeMes sageTypeFieldTa g4000 FieldTag4000 {
get {
return this.fieldTag40 00Field;
}
set {
this.fieldTag40 00Field = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageTypeFieldTa g4000 {

private IDType_Type intermediaryIDC odeField;

private bool intermediaryIDC odeFieldSpecifi ed;

private string intermediaryIde ntifierField;

public IDType_Type IntermediaryIDC ode {
get {
return this.intermedia ryIDCodeField;
}
set {
this.intermedia ryIDCodeField = value;
}
}

[System.Xml.Seri alization.XmlIg noreAttribute()]
public bool IntermediaryIDC odeSpecified {
get {
return this.intermedia ryIDCodeFieldSp ecified;
}
set {
this.intermedia ryIDCodeFieldSp ecified = value;
}
}

public string IntermediaryIde ntifier {
get {
return this.intermedia ryIdentifierFie ld;
}
set {
this.intermedia ryIdentifierFie ld = value;
}
}
}

[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Xml.Seri alization.XmlTy peAttribute(Nam espace="MyNames pace")]
public enum IDType_Type {

B,

C,

D,

F,

U,
}- Hide quoted text -

- Show quoted text -

[=?Utf-8?B?ai5hLiBoYXJyaW1hbg==?=]

I may have tried to oversimplify my example.

The "real" XML has a number of "xmlns:" entries following the "parentnode ".
When I remove them & test, this statement,
"/parentnode/secondchildnode//*[not(node())]", works as I was expecting.

Is there something I need to do with adding namespaces (e.g
XmlNamespaceMan ager) or something else that I am unaware of?

If there are 5 namespaces in the XML do I need to add a namespace for each
one?

How are the namespaces used in the XPath statement above?

Thanks.

Jeff

"j.a. harriman" wrote:

Building off your example, here's a respresentation of my actaul XML:
<parentnode>
<firstchildnode >
<a/>
</firstchildnode>
<secondchildnod e>
<b/>
<c>
<d/>
<e>12345</e>
</c>
</secondchildnode >
</parentnode>

The XPath expression, "//*[not(node())]", removes all "empty" elements. I
need it to *only* remove "b" & "d" above. "a" would be left alone.

Thanks.

Jeff

"Sprotty" wrote:

You could remove the empy elements in the DOM before loading it as you
suggest.
Im not seeing a problem with that

XmlDocument xmlDoc = new XmlDocument();
xmlDoc.LoadXml( "<parentnode><s econdchildnode> <b/></
secondchildnode ></parentnode>");
XmlNodeList xnl = xmlDoc.SelectNo des("//*[not(node())]");
Debug.Assert(xn l.Count == 1);

or you could remove it from the object model after its been loaded
(deserialized).

If you were using Liquid XML then you may be able to add code in that
would remove the empty elements as part of the de-serialization.

Cheers Simon
On 31 Jul, 21:22, j.a. harriman <jeffrey_no_spa m_al...@nospam. nospam>
wrote:

Simon,
>
Thanks for the link, I will take a look at it as it may be something we
could use.
>
In the meantime, I did find a code snippet that removes the "empty" nodes
but it removes ALL the empty nodes in the XML and I didn't state (earlier)
that I want to do this beginning with a particular child node only.
>
When I use the following XPath statement in XMLSpy's XPath Evaluator, it
locates the "empty" node mentioned in the previous posts:
/parentnode/secondchildnode//*[not(node())]
>
When I use the same statement in my C# code, it will not locate any empty
nodes "under" the "secondchildnod e" (count = 0):
XmlNodeList element =
doc.SelectNodes ("/parentnode/secondchildnode//*[not(node())]");
>
The plan is to loop through "element" and remove the nodes using the
"ParentNode.Rem oveChild" of an XmlNode - setting it to each array element,
etc.
>
Any ideas as to why this is not working in .NET VS2005 2.0 framework?
>
Thanks.
>
Jeff
>
"Sprotty" wrote:
Hi Jeff
>
The code produced by xsd.exe, copes well with simple schema's, but if
your working against anything a bit complicated then it can cause
issues.
>
If you continue to have problems then I suggest you take a look at
Liquid XML Data Binder
>
http://www.liquid-technologies.com/P...taBinding.aspx
>
This supports much more of the xsd standard, and the objects are more
strongly typed (it also comes with a free XSD editor).
>
Hope this helps
>
Regards Simon
>
On 23 Jul, 08:00, "Joe Fawcett" <joefawc...@new sgroup.nospamwr ote:
Jeff
>
IXmlSerializabl e is reasonably straightforward .
The newer version needs three methods, ReadXml, WriteXml and one to find the
Schema which is pointed to by an attribute, I normally just include the
schema as an embedded resource.
In ReadXml you get an XmlReader containing the XML and use it to populate
the object's fields, either reading it directly or loading it into a
DomDocumnt/XPathDocument if that's easier. WriteXml takes the fields and
creates an XML document.
There is an example here:http://www.devx.com/dotnet/Article/29720
>
--
>
Joe Fawcett (MVP - XML)http://joe.fawcett.name
>
"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message
>
>news:04******* *************** ************@mi crosoft.com...
>
Thanks for the answer Joe.
>
I also found some examples of using a style sheet and doing a
transformation
(Using XslCompiledTran sform) of the original XmlDocument. I've run a test
and it seems to work in that the result is as if the original message
hadn't
had them in there. Are there any reasons why not to implement this?
>
Also, I tried to locate good "beginner" examples of using IXmlSerializabl e
that might be similar to what I need to do, but came up short. I would
like
to to look at this in further detail.
>
Do you have any links to examples or are you aware if any of the Microsoft
example downloads (such as SDK) have any?
>
Thanks. Jeff
>
"Joe Fawcett" wrote:
>
Unless you want to implement IXmlSerializabl e then I don't think you can
do
what you want. You can specify that a null element is shown as:
<myElement xsi:nil="true"/>
rather than being omitted entirely by adding the XmlElement(IsNu llable =
true) attribute to the field/property.
>
--
>
Joe Fawcett (MVP - XML)
>
>http://joe.fawcett.name
>
"j.a. harriman" <jeffrey_no_spa m_al...@nospam. nospamwrote in message
>news:69******* *************** ************@mi crosoft.com...
Hi,
>
I have a schema that has an optional element, fieldTag4000Fie ld. If
the
element is omitted from the XML request, when it is deserialized, it
will
be
null when I check it - which is fine.
>
What happens when the element is supplied as <fieldTag4000Fi eld/>
(empty),
it does not equate to null. I want to be able handle this at the
deserialization level rahter than in my edits later.
>
Is there a way to alter the behavior so when I deserialize it, it is
null?
>
I also want to add that I am calling the "CanDeserialize " method of the
XmlSerializer object, passing in the request as a XmlNodeReader.
>
Thanks.
>
The following C# class snippet was generated by xsd.exe V2:
>
[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageType {
>
private Request_TypeMes sageTypeFieldTa g4000 fieldTag4000Fie ld;
>
public Request_TypeMes sageTypeFieldTa g4000 FieldTag4000 {
get {
return this.fieldTag40 00Field;
}
set {
this.fieldTag40 00Field = value;
}
}
}
>
[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Diagnost ics.DebuggerSte pThroughAttribu te()]
[System.Componen tModel.Designer CategoryAttribu te("code")]
[System.Xml.Seri alization.XmlTy peAttribute(Ano nymousType=true ,
Namespace="MyNa mespace")]
public partial class Request_TypeMes sageTypeFieldTa g4000 {
>
private IDType_Type intermediaryIDC odeField;
>
private bool intermediaryIDC odeFieldSpecifi ed;
>
private string intermediaryIde ntifierField;
>
public IDType_Type IntermediaryIDC ode {
get {
return this.intermedia ryIDCodeField;
}
set {
this.intermedia ryIDCodeField = value;
}
}
>
[System.Xml.Seri alization.XmlIg noreAttribute()]
public bool IntermediaryIDC odeSpecified {
get {
return this.intermedia ryIDCodeFieldSp ecified;
}
set {
this.intermedia ryIDCodeFieldSp ecified = value;
}
}
>
public string IntermediaryIde ntifier {
get {
return this.intermedia ryIdentifierFie ld;
}
set {
this.intermedia ryIdentifierFie ld = value;
}
}
}
>
[System.CodeDom. Compiler.Genera tedCodeAttribut e("xsd",
"2.0.50727. 42")]
[System.Serializ ableAttribute()]
[System.Xml.Seri alization.XmlTy peAttribute(Nam espace="MyNames pace")]
public enum IDType_Type {
>
B,
>
C,
>
D,
>
F,
>
U,
}- Hide quoted text -
>
- Show quoted text -

[Martin Honnen]

j.a. harriman wrote:

The "real" XML has a number of "xmlns:" entries following the "parentnode ".
When I remove them & test, this statement,
"/parentnode/secondchildnode//*[not(node())]", works as I was expecting.

Is there something I need to do with adding namespaces (e.g
XmlNamespaceMan ager) or something else that I am unaware of?

If there are XML namespace declarations and you have elements or
attributes in those namespaces you want to select with XPath then you
indeed should use an XmlNamespaceMan ager and bind prefixes to the
namespace URIs to use the prefixes in your XPath expressions to qualify
element or attribute names. The XmlNamespaceMan ager is then passed as
the second argument to the SelectNodes or SelectSingleNod e calls.
If you need more concrete help then show us the relevant XML.

--

Martin Honnen --- MVP XML
http://JavaScript.FAQTs.com/