Shortcuts
Documentation
Community
Resources

Using XQuery on command line with DRB APIŽ

Purpose This tutorial aims to help beginners to work with DRB APIŽ using XQuery especially on the command line.
Note: The presented samples are ordered by increasing complexity.
Target audience Beginner
Required knowledge Before you continue you should have a basic understanding of the following:
  • XML
  • XQuery
  • XPath
If you want to study these subjects first, find the corresponding tutorials on the dedicated list .
What you will learn in this tutorial In this tutorial you will learn how to use the DRB APIŽ XQuery engine on command line.

Prerequisites

You may find usefull to config your environnment by using an alias, so that the command of this tutorial appears simpler.
The following definition is for Bash shell: 

alias drb='java -jar [DRB_INSTALLATION_HOME]/lib/java/drb-[M]-[m]-[t].jar'

Test your installation

Getting the DRB APIŽ version

drb --version

May returns something like:

Data Request Broker - DRB API - 2-3-rc-2 (GAEL Consultant)

Getting the DRB APIŽ help

drb --help

Handling computation

Simple computation

drb --query '2000 + 9'

May return:

2009

Using 'math:' functions

The math: namespace is used for mathematical function available in the XQuery language.

For example computing a cosinus value:

drb --query 'math:cos(45)'

May return:

0.5253219888177297

Another valuable function is math:random() which return a random value from 0 to 1.

drb --query 'math:random()'

May return:

0.30315802310561635

Getting all the available functions

drb --help xql

Working with simple XML file

Open the XML file

Having the hotel.xml :

<hotel>
   <name>Hotel du Parc</name>
   <address>
      <street>7 place de la Concorde</street>
      <zipcode>75008</zipcode>
      <city>Paris</city>
      <country>France</country>
   </address>
   <phone>+33 1 4280 5000</phone>
   <fax>+33 1 4280 5001</fax>
   <rooms>
      <room>
         <name>Royal Suite</name>
         <price>99</price>
      </room>
      <room>
         <name>Standard Suite</name>
         <price>45</price>
      </room>
   </rooms>
</hotel>

You open the file with the command:

drb --query 'hotel.xml'

May return:

<hotel.xml directory="false" size="474" modified="16-MAR-2009 15:40:16.000000" readable="true" writable="true" hidden="false">
   <hotel>
      <name>Hotel du Parc</name>
      <address>
         <street>7 place de la Concorde</street>
         <zipcode>75008</zipcode>
         <city>Paris</city>
         <country>France</country>
      </address>
      <phone>+33 1 4280 5000</phone>
      <fax>+33 1 4280 5001</fax>
      <rooms>
         <room>
            <name>Royal Suite</name>
            <price>99</price>
         </room>
         <room>
            <name>Standard Suite</name>
            <price>45</price>
         </room>
      </rooms>
   </hotel>
</hotel.xml>

Retrieve a node

The command for retrieving the hotel name node :

drb --query 'hotel.xml/hotel/name'

May return:

<name>Hotel du Parc</name>

Retrieve a node value

The command for retrieving the hotel name value :

drb --query 'fn:data(hotel.xml/hotel/name)'

May return:

Hotel du Parc

Retrieve a node using predicate

The predicate (introduce with the symbols []) is used for selecting a specific node. For example in case of several node with the same name we can select the second one like in the following command:

drb --query 'hotel.xml/hotel/rooms/room[2]'

May return:

<room>
   <name>Standard Suite</name>
   <price>45</price>
</room>

We can also retrieve node regarding some test on values. In the following example we select all room nodes that have a price node value strictly superior to 50.0

drb --query 'hotel.xml/hotel/rooms/room[xs:float(price) > xs:float(50.0)]'

May return:

<room>
   <name>Royal Suite</name>
   <price>99</price>
</room>

Retrieve a node value and make some computation

Take care that the retrieve node value has to be converted to the appropriate type (here xs:integer) for the computation to be possible:

drb --query 'xs:integer(fn:data(hotel.xml/hotel/rooms/room[2]/price)) - 10'

May return:

35

Create a structured output (XML)

Simple structured output

The query has to be placed between { } :

drb --query '<root><a>{ math:random() }</a></root>'

May return:

<root>
   <a>0.047975992783876764</a>
</root>

Structured output using a value extracted from an XML file

drb --query '<root><a>{ fn:data(hotel.xml/hotel/name) }</a></root>'

May return:

<root>
   <a>Hotel du Parc</a>
</root>

Structured output using a node extracted from an XML file

drb --query '<root><a>{ hotel.xml/hotel/name }</a></root>'

May return:

<root>
   <a>
      <name>Hotel du Parc</name>
   </a>
</root>

Conditionnal expressions and loop

If

drb --query 'if (5 < 10) then "Five is less than ten." else "Five is not less than ten!!! Arrrggggg..."'

May return:

Five is less than ten.

For

drb --query 'for $i in 1 to 5 return fn:concat("iteration nb#", $i)'

May return:

iteration nb#1, iteration nb#2, iteration nb#3, iteration nb#4, iteration nb#5

Redirect the output (Saving query result)

The purpose of the redirection is for example to save the result of your query in a file.

Caution: This action depends on your environmment.

In Bash shell you can use the following command (note the > symbol):

drb --query '<root><a>{ math:random() }</a></root>' > output.xml

If you list the content of the file output.xml:

<root>
   <a>0.21997404015814814</a>
</root>

Create and execute a query script

A query script is a file containing XQuery instructions.

Having the following script file (script_random.xql ):

xquery version "1.0";

fn:concat("Hello, here is a random number: ", math:random())

You can execute it using the --query-file option:

drb --query-file script_random.xql

May return:

Hello, here is a random number: 0.22337426002117922