Retreiving Xml child nodes using DataSet and XmlDataDocument

[Anas M. Nebuchadnezzar XXXVII]

I am loading an Xml file from a sql server. I have one field that
contains child nodes stored as text. The problem is, when I try to imort
the dataset into a XmlDataDocument , it loads the whole thing as a text
element. (with all of the special characters converted) The best
solution I've found so far is to do a string replace on all of the
special characters. Like so:

DataSet ds = new DataSet();
ds.EnforceConst raints = false;
SqlDataAdapter da = new SqlDataAdapter( query, connStr);
da.Fill(ds, "Keywords") ;
XmlDataDocument xmlDoc = new XmlDataDocument (ds);
string s = xmlDoc.SelectSi ngleNode("NewDa taSet/Keywords/Entry").InnerTe xt;
s = UnConvertEntiti es(s);
xmlDoc.SelectSi ngleNode("NewDa taSet/Keywords/Entry").InnerXm l = s;

// UnConvertEntiti es(string s)
s.Trim();
s.Replace("&amp ;", "&");
s.Replace("&lt; ", "<");
s.Replace("&gt; ", ">");
s.Replace("&apo s;", "'");
s.Replace("&quo t;", "\"");
return s;

Is there any better way to do this, or am I stuck with this method?
Also, I've found that this method adds a large amount of whitespace at
the end of the last element, but that may be a seperate problem.

[Oleg Tkachenko]

Anas M. Nebuchadnezzar XXXVII wrote:

I am loading an Xml file from a sql server. I have one field that
contains child nodes stored as text. The problem is, when I try to imort
the dataset into a XmlDataDocument , it loads the whole thing as a text
element.

I don't see any problem here. If it is stored as text, why it should be
not text node? Escaped markup is always a trouble waiting to happen,
always avoid it.
--
Oleg Tkachenko
XmlInsider
http://blog.tkachenko.com

[Anas M. Nebuchadnezzar XXXVII]

Oleg Tkachenko wrote:

Anas M. Nebuchadnezzar XXXVII wrote:

I am loading an Xml file from a sql server. I have one field that
contains child nodes stored as text. The problem is, when I try to
imort the dataset into a XmlDataDocument , it loads the whole thing as
a text element.

I don't see any problem here. If it is stored as text, why it should be
not text node? Escaped markup is always a trouble waiting to happen,
always avoid it.

My problem is, It's not stored in the database escaped. I'm storing it
in a char field for lack of knowing which type works better. I've stored
the literal characters (ie. '<') and keep getting the escaped characters
(ie. '&lt;') I was wondering if there was a better way to store this
data, and convert it from text, to part of the document.