<?xml version="1.0"?>
<ex:x xmlns:ex="http://dummy-ns">
<ex:paragraph>
  <ex:START/>
  <ex:span>
    Some text C..
  </ex:span>
</ex:paragraph>
<ex:paragraph>
  Some text D..
  <ex:END/>
  <ex:span>
    Some text F..
  </ex:span>
</ex:paragraph>
</ex:x>

<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
               xmlns:ex="http://dummy-ns"
               version="1.0">

<xsl:template match="/">
 <ex:result>
   <xsl:apply-templates select="//ex:START" />
 </ex:result>
</xsl:template>

<xsl:template match="ex:START">
 <xsl:call-template name="OutputContent">
   <xsl:with-param name="node-set"
select="following-sibling::node()[not(self::text() and
normalize-space(.) = '')] |
../following-sibling::*/node()[not(self::text() and normalize-space(.)
= '')]" />
 </xsl:call-template>
</xsl:template>

<xsl:template name="OutputContent">
 <xsl:param name="node-set" />

 <xsl:if test="not($node-set[1]/self::ex:END)">
   <xsl:copy-of select="$node-set[1]" />
   <xsl:call-template name="OutputContent">
     <xsl:with-param name="node-set" select="$node-set[position() &gt; 1]" />
   </xsl:call-template>
 </xsl:if>
</xsl:template>

<ex:result xmlns:ex="http://dummy-ns">
  <ex:span>
     Some text C..
  </ex:span>
  Some text D..
</ex:result>

The stylesheet is not rigorously tested. But I feel, you could use
this as it is, or perhaps adapt it.

Imagine XML data, where continuous selection of data - randomly made by
the user - is being marked by using a <start/> and <end/> element.

For example:

<ex:paragraph>
   <ex:START/>
   <ex:span>
       Some text C..
   <ex:span>
</ex:paragraph>
<ex:paragraph>
   Some text D..
   <ex:END/>
   <ex:span>
       Some text F..
   <ex:span>
</ex:paragraph>

The desired nodes would be
   <ex:span>
       Some text C..
   <ex:span>
   Some text D..

Is there an easy way to process this kind of selection with XSLT?

Best regards,
Svante

--
Regards,
Mukul Gandhi