Omit XML declaration?

Question

I have an XElement that I have to parse to remove the white space in the closing tag. My code looks like this:

var stringBuilder = new StringBuilder();
using (var stringWriter = new StringWriter(stringBuilder))
{
    xelement.Save(stringWriter);
}
stringBuilder.Replace(" />", "/>");
var xml = stringBuilder.ToString();

Basically, I'm making a stringbuilder and replacing the unneeded white space. The resulting string looks fine, except it has the XML declaration. I know that on an XmlWriter, I can omit the declaration with OmitXmlDeclaration but StringWriter doesn't have this.

Is there a way to do this, or do I need to manually parse out the declaration from the resulting string?

For clarity, here is the before and after XML:

// Before
<actionitem actiontaken="none" target="0" targetvariable="0">
  <windowname>Popup Window</windowname>
  <windowposx>-1</windowposx>
  <windowposy>-1</windowposy>
  <windowwidth>-1</windowwidth>
  <windowheight>-1</windowheight>
  <noscrollbars>false</noscrollbars>
  <nomenubar />
  <notoolbar />
  <noresize />
  <nostatus />
  <nolocation />
  <browserWnd />
</actionitem>

// After
<?xml version="1.0" encoding="utf-16"?>
<actionitem actiontaken="none" target="0" targetvariable="0">
  <windowname>Popup Window</windowname>
  <windowposx>-1</windowposx>
  <windowposy>-1</windowposy>
  <windowwidth>-1</windowwidth>
  <windowheight>-1</windowheight>
  <noscrollbars>false</noscrollbars>
  <nomenubar/>
  <notoolbar/>
  <noresize/>
  <nostatus/>
  <nolocation/>
  <browserWnd/>
</actionitem>

EDIT: For those that asked, this is for a Department of Defense project. Their specifications are locked in. That means, no white space in the closing tag, no matter how much I protest. Regardless of what's right or not, they don't want it, and they're signing the paycheck. I just try to accommodate them.

Answer 1

Use ToString() instead of Save(). That eliminates the need for the StringBuilder too.

 string xml = xelement.ToString(); // no declaration element added
 xml = xml.Replace(" />", "/>");   // if you really think you must

Answer 2

Disclaimer: this answer applies to Kevin J and Kevin J alone. Do not perform string manipulation on XML.

If you still want to use the StringBuilder/StringWriter you can wire the StringWriter (which is a TextWriter) through a XmlWriter:

var xe = new XElement("test",
    new XElement("child1"),
    new XElement("child2"));

var sb = new StringBuilder();
using (var writer = new StringWriter(sb))
using (var xr = XmlWriter.Create(writer, new XmlWriterSettings()
{
    OmitXmlDeclaration = true
}))
{
    xe.Save(xr);
}

sb.Replace(" />", "/>");

As for the Replace I wrote a function that should be a bit more resilient against corner cases (comments, CData). It should also use less cycles and consume less memory.

static void StripClosingWhitespace(StringBuilder sb)
{
    var inComment = false;
    var inCData = false;

    for (var i = 0; i < sb.Length; i++)
    {
        var c = sb[i];
        if (inComment)
        {
            if (c == '>' && sb[i - 1] == '-' && sb[i - 2] == '-')
                inComment = false;
        }
        else if (inCData)
        {
            if (c == '>' && sb[i - 1] == ']' && sb[i - 2] == ']')
                inCData = false;
        }
        else if (i > 2 && c == '-' && sb[i - 1] == '-' && sb[i - 2] == '!' && sb[i - 3] == '<')
        {
            inComment = true;
        }
        else if (i > 7 && 
            c == '[' && 
            sb[i - 1] == 'A' && sb[i - 2] == 'T' && sb[i - 3] == 'A' && sb[i - 4] == 'D' && sb[i - 5] == 'C' &&
            sb[i - 6] == '[' && sb[i - 7] == '!' && sb[i - 8] == '<')
        {
            inCData = true;
        }
        else if (i > 2 && c == '>' && sb[i - 1] == '/' && char.IsWhiteSpace(sb[i - 2]))
        {
            sb.Remove(i - 2, 1);
            i--;
        }
        else
        {
            // Do nothing
        }
    }
}