Query language can be used to access a portion of an XML document and process it. Queries can be used to retrieve row-wise information as shown. The star (*) symbol implies that all the columns of the records that satisfy the where condition are retrieved.

Select * from <table> where <column>=value"

Queries can be used to retrieve column-wise information as shown. Here only columns column1, column2, column3 of all rows are retrieved.

Select column1, column2, column3 from <table>

Summarized information of a computational column can be obtained as shown.

Select sum(computational_column) from <table>

Joins can be performed. That is, columns from more than one table can be retrieved as shown.

Select column1,column2,column3, column4 from <table1>,<table2>

One of the general forms of manipulating table contents using standard Structured Query Language is as shown.

Update <table> set <column>=<value1>, <column2>=<value2>

In 1998, a couple of groups started working on proposals to the W3C mechanism to query XML documents. Two earlier query languages submitted for proposals were XML-QL and XQL. But there were some short- comings in those querying languages.

The current state of querying technology uses XSLT (Extensible style sheet language transformation) and XPath to query XML documents. XSLT may be used to transform one XML document into another, addressing some of the querying needs.

XPath provides flexible mechanism for addressing individual portions of an XML document to any granularity, even down to individual characters in text elements. XPointer extends the functionality of XPath in two ways.

XPointer extends XPath to allow for some additional functionality. It introduces the concept of a point location, as distinct from a node.

XPointer also defines the concept of a range location, which is defined to be the XML structure and content between two points. The ability to specify ranges would allow a pointer to point to all occurrences of a particular word in a target document.

#xpointer (<locator> to <locator>)

The extensible style sheet language (XSL) is an XML language designed to transform an XML document into another XML document or to transform an XML document into rendition objects.
XSLT is written in XML. This means that an XSLT sheet is in itself a well-formed XML document.

XPath is used to access individual nodes of an XML document at the same time, preserving the hierarchy and structure of the original document.XSLT uses XPath to filter nodes out of document, while XSLT is used to present the data in any number of forms.

Here is the code "catalog.xml". We shall later see the various queries performed on this XML document.

<?XML version="1.0" encoding="utf-8" standalone="yes"?>
<!--===== The Timman press book catalog application =======-->

<Catalog>
<Book>
     <Title> C++ Programmers' reference </Title>
     <Authors>
        <Author> Stroup stru </Author>
     </Authors>
     <Publishers> Timman Press, Ltd.</Publisher>
     <PubDate>August 1999</PubDate>
     <Abstract> Object oriented programming using C++ </Abstract>
     <Pages> 480 </Pages>
<ISBN> 1-861001-57-6 </ISBN>
<RecSubjCategories>
     <Category> Object oriented </Category>
     <Category> Applications </Category>
     </RecSubjCategories>
<Price>$49.99</Price>
</Book>
…
<Book>
…
</Book>
<Catalog>

In XML, the equivalent of row-wise restriction of information would be the restriction of elements based on their content. XPath expression may be used to look for a specific value in a field. The following style sheet can be used to return all books in our catalog with an author called Bjarne Strustroup. The portion of the template that performs row-wise selection is the <xsl:for-each> block in the first template.

<xsl:stylesheet xmnls:xsl=http://www.w3.org/1999/XSL /Transform version="1.0">
        <xsl:template match="/Catalog">
                <xsl:copy>
                      <xsl:for-each select="//Book[Author='Bjarne Stroustroup']">
                          <xsl:copy>
                               <xsl:apply-templates name="childnodes"/>
                          </xsl:copy>
                      </xsl:for-each>
                </xsl:copy>
       </xsl:template>
       <xsl:template name="childnodes" match="*">
                <xsl:copy>
                      <xsl:apply-templates name="childnodes"/>
                </xsl:copy>
       </xsl:template>
</xsl:stylesheet>

The output of this script when run against catalog.xml document returns three books, whose author is Strustroup.

<?xml version="1.0" encoding ="utf-8"?>
<Catalog>
<Book>
     <Title>C++ Programmer ' s reference </Title>
     <Authors>
         <Author> Bjarne Stroustroup </Author>
     </Authors>
     <Publishers>Timman Press, Ltd.</Publisher>
     <PubDate>August 1999</PubDate>
     <Abstract> Object oriented programming using C++ </Abstract>
     <Pages> 480 </Pages>
<ISBN> 1-861001-57-6 </ISBN>
<RecSubjCategories>
     <Category> Object oriented </Category>
     <Category> Applications </Category>
</RecSubjCategories>
<Price>$49.99</Price>
</Book>
<Book>
     <Title Professional C++</Title>
     …
</Book>
<Book>
     <Title> C++ for beginners </Title>
     …
</Book>
<Catalog>

For XML, column-wise restriction of returned information is achieved by restricting the text-level elements or attributes returned by a style sheet. Here we display only the titles of the book written by Bjarne Stroustroup. We use the style sheet specified here.

<xsl:stylesheet xmnls:xsl=http://www.w3.org/1999/XSL /Transform version="1.0">
        <xsl:template match="/Catalog">
                <xsl:copy>
                       <xsl:for-each select="//Book[./Authors/Author = 'Bjarne Strustroup']">
                           <xsl:copy>
                                <xsl:apply-templates select ="Title" name="childnodes"/>
                           </xsl:copy>
                       </xsl:for-each>
                </xsl:copy>
        </xsl:template>
        <xsl:template name="childnodes" match="*">
                <xsl:copy>
                     <xsl:apply-templates name="childnodes"/>
                </xsl:copy>
        </xsl:template>
</xsl:stylesheet>

The output of the query associated with the xml document, catalog.xml is as shown.

<?xml version="1.0" encoding ="utf-8"?>
<Catalog>
<Book>
     <Title>C++' s reference </Title>
</Book>
<Book>
     <Title> Professional C++</Title>
</Book>
<Book>
     <Title> C++ for Beginners </Title>
</Book>
</Catalog>

Let us sort the titles of the books written by the author Bjarne Strustroup. The following style sheet is applied across the catalog.xml document. We have simply added <xsl:sort> element to the <xsl:for-each> element and hence would be sorted by Title.

<xsl:stylesheet xmnls:xsl=http://www.w3.org/1999/XSL /Transform version="1.0">
        <xsl:template match="/Catalog">
                 <xsl:copy>
                        <xsl:for-each select="//Book[./Authors/Author = 'Bjarne Strustroup']">
                                <xsl:sort select ="Title"/>
                                <xsl:copy>
                                    <xsl:apply-templates select ="Title" name="childnodes"/>
                        </xsl:copy>
                        </xsl:for-each>
                 </xsl:copy>
        </xsl:template>
        <xsl:template name="childnodes" match="*">
            <xsl:copy>
                  <xsl:apply-templates name="childnodes"/>
            </xsl:copy>
        </xsl:template>
</xsl:stylesheet>

Final output of the transformation is as shown. Thus we see that the titles of the books are ordered. Similarly we can perform all the standard operations of SQL on an XML document.

<?xml version="1.0" encoding ="utf-8"?>
<Catalog>
<Book>
      <Title> C++ for Beginners </Title>
</Book>
<Book>
      <Title> C++' s reference </Title>
</Book>
<Book>
      <Title> Professional C++ </Title>
</Book>
</Catalog>