apache > cocoon
 

Portal: Configuring Coplets

Overview

This document gives an overview of how to configure coplets. The sample portal that comes with the Cocoon distribution contains several samples.

The configuration of a coplet is done in several steps that are outlined in the next chapters. We use the provided sample as a base. This sample stores all profiles in XML documents that are all stored in the samples/portal directory. So all file and directory names are relative to this directory. It's not required that you store the configuration in XML files, you can also store them in a database, an LDAP system etc. This is also configurable and customizable.

Configuring Coplets

Configuring coplets is like defining a class and creating their instances. So in fact, you define the available coplets (= classes) and each portal view gets some instances of these coplets.

In addition to the classes and the instances, we have coplet types which can be seen as a classification of the coplet classes. There are different coplet types, uri based coplets, JSR-168 coplets, application coplets etc. The coplet type defines, how the coplet works or how the coplet gets it's content. For example, the uri based coplet triggers a uri to get the coplet - this can be an http request or an internal cocoon pipeline request.

Another example is the JSR-168 coplet type. This type allows to use JSR-168 compliant portlets for fetching the content. So let's start with the coplet types:

Available Coplet Types

Before you can define your available coplets, you have to define the available coplet types, or the so called coplet base data. The current example contains an XML document for this (the file is in the profiles/copletbasedata directory and is called portal.xml). This is an excerpt from the file:

...
<coplets>
   <coplet-base-data id="URICoplet">
      <coplet-adapter>uri</coplet-adapter>
   </coplet-base-data>
</coplets>
...

In the example above, we define one coplet type, the URICoplet, that uses the uri coplet adapter. By this we define a type, that uses URIs to get the content of a coplet. A uri can either be targetted at a different server (using http, ftp etc.) or it can be an internal Cocoon pipeline (using cocoon:).

You can add different coplet types with additional configuration here, however you rarely have to touch this file as most is preconfigured already.

Available Coplets

Based on the coplet types, you can define the available coplets in your portal application (= classes). In the example portal a configuration file contains these so called coplet datas (it's in the profiles/copletdata directory and has the filename portal.xml). Here is an excerpt:

...
<coplets>
   <coplet-data id="CZ Weblog" name="standard">
      <title>CZ's Weblog</title>
      <coplet-base-data>URICoplet</coplet-base-data>
      <attribute>
         <name>uri</name>
      	 <value xsi:type="java:java.lang.String" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
      	    cocoon:/news/liverss?feed=http://radio.weblogs.com/0107211/rss.xml
	 </value>
      </attribute>
      <attribute>
         <name>buffer</name>
      	 <value xsi:type="java:java.lang.Boolean" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">true</value>
      </attribute>
      <attribute>
         <name>error-uri</name>
      	 <value xsi:type="java:java.lang.String" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">cocoon:/news/CZ_weblog.rss</value>
      </attribute>
   </coplet-data>
</coplets>
...
 

Each coplet data contains a unique ID and additional configuration. Configuration of the underlying coplet base data is required. In the example above, the URICoplet is used that we have configured in the previous section.

The above configured coplet data requires some configuration. This configuration depends on the coplet type. In this case we use the URICoplet and this needs of course the URI to fetch the content from for this coplet. This configuration is passed in the different attributes you see above. Each attribute has a name and value.

The set of coplet datas defines the set of available coplets a user can choose from. If a user chooses to view a coplet, an instance of this coplet data is created. If, e.g. the user chooses the same coplet twice, two instances are created. This is useful for configurable coplets where the user can choose the same coplet with different configurations.

Selected Coplets

The selected coplets are described by the set of coplet instance datas. This again is another XML document (that is stored in profiles/copletinstancedata and has the name portal.xml). Here is an excerpt:

...
<coplets>
   <coplet-instance-data id="CZ Weblog-1" name="standard">
      <coplet-data>CZ Weblog</coplet-data>
   </coplet-instance-data>
</coplets>
...
   

The coplet instance data refers to its coplet data (the class) by specifying the unique ID. The instance itself has a unique ID as well that is referenced from the portal view. This id of the instance has only to be unique within the scope of a user. Different users can have different instances with the same id.

In addition, a coplet instance data could have own configuration information. This configuration depends on the coplet type and on the coplet (class).

Common Coplet Configuration

As outlined above, a coplet can be configured at various places. In general, configuration values are most often configured at the coplet type (coplet base data) and may be overwritten by the coplet class (coplet data).

The coplet base data takes the information as config entries whereas the coplet data uses attributes. Apart from that the keys and the data types are the same. Let's have a look at an example:

...
<coplet-base-data id="URICoplet">
   <coplet-adapter>uri</coplet-adapter>
   <configuration>
      <name>buffer</name>
      <value xsi:type="java:java.lang.Boolean" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">true</value>
   </configuration>
</coplet-base-data>
...
<coplet-data id="CZ Weblog" name="standard">
   <title>CZ's Weblog</title>
   <coplet-base-data>URICoplet</coplet-base-data>
   <attribute>
      <name>buffer</name>
      <value xsi:type="java:java.lang.Boolean" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">false</value>
   </attribute>
</coplet-data>
...
   

As you can see in the example above, there is no real difference in configuring a coplet type (base data) and a coplet class (coplet data) except that the first one uses the element name configuration and the last one uses attribute.

Output Buffering

Each coplet can be configured with the boolean value buffer (default is false) that defines if the xml data stream from the coplet is buffered before it is streamed in the main portal pipeline.

If the stream is not buffered and an exception occurs during the streaming of the one coplet, then the whole portal might be rendered invalid. Therefore you should turn on buffering whenever you can't guarantee that the output of a coplet is always valid XML.

Delivery Timeout

Usually the portal waits forever(!) for a coplet to deliver it's content. If you want to fine-tune this behaviour, you can configure a timeout value. This integer value (default is endless) defines the maximum time (in seconds) the coplet has to deliver it's content.

If the timeout is reached the content is assumed as not retrievable. If you set a timeout, the content is automatically buffered and the buffer configuration is ignored.

Coplet Rendering

Each coplet can be configured for supporting different features, like minimizing or removing. These are aspect configurations that are configured similar to attributes.

Sizing

By default, the window size of a coplet can be changed by the user. The user can choose between maximized and minimized.

This is configurable by the boolean aspect sizable on the coplet data (class). The default is true:

...
   <coplet-data id="CZ Weblog" name="standard">
      <title>CZ's Weblog</title>
      <coplet-base-data>URICoplet</coplet-base-data>
      <aspect>
      	<name>sizable</name>
      	<value xsi:type="java:java.lang.Boolean" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
      	  false
      	</value>
      </aspect>
   </coplet-data>
...
   

In addition the size of a coplet (minimized/maximized) can be preconfigured as well by the integer aspect value size. The value 0 means minimized and 1 is maximized (with 1 as the default).

...
   <coplet-data id="CZ Weblog" name="standard">
      <title>CZ's Weblog</title>
      <coplet-base-data>URICoplet</coplet-base-data>
      <aspect>
      	<name>size</name>
      	<value xsi:type="java:java.lang.Integer" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
      	  0
      	</value>
      </aspect>
   </coplet-data>
...
   

Mandatory Coplets

By default, the user can delete any coplet from his portal view. However, important coplets can be configured as mandatory. In this case, the remove icon is not displayed, and the coplet can never be removed by the user.

This is configurable by the boolean aspect mandatory on the coplet data (class). The default is false:

...
   <coplet-data id="CZ Weblog" name="standard">
      <title>CZ's Weblog</title>
      <coplet-base-data>URICoplet</coplet-base-data>
      <aspect>
      	<name>mandatory</name>
      	<value xsi:type="java:java.lang.Boolean" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
      	  true
      	</value>
      </aspect>
   </coplet-data>
...