Class AbstractSimpleBeanDefinitionParser

All Implemented Interfaces:
BeanDefinitionParser
Direct Known Subclasses:
GroovyMarkupConfigurerBeanDefinitionParser, ScriptTemplateConfigurerBeanDefinitionParser

public abstract class AbstractSimpleBeanDefinitionParser extends AbstractSingleBeanDefinitionParser
Convenient base class for when there exists a one-to-one mapping between attribute names on the element that is to be parsed and the property names on the Class being configured.

Extend this parser class when you want to create a single bean definition from a relatively simple custom XML element. The resulting BeanDefinition will be automatically registered with the relevant BeanDefinitionRegistry.

An example will hopefully make the use of this particular parser class immediately clear. Consider the following class definition:

public class SimpleCache implements Cache {

     public void setName(String name) {...}
     public void setTimeout(int timeout) {...}
     public void setEvictionPolicy(EvictionPolicy policy) {...}

     // remaining class definition elided for clarity...
 }

Then let us assume the following XML tag has been defined to permit the easy configuration of instances of the above class;

<caching:cache name="..." timeout="..." eviction-policy="..."/>

All that is required of the Java developer tasked with writing the parser to parse the above XML tag into an actual SimpleCache bean definition is the following:

public class SimpleCacheBeanDefinitionParser extends AbstractSimpleBeanDefinitionParser {

     protected Class getBeanClass(Element element) {
         return SimpleCache.class;
     }
 }

Please note that the AbstractSimpleBeanDefinitionParser is limited to populating the created bean definition with property values. if you want to parse constructor arguments and nested elements from the supplied XML element, then you will have to implement the postProcess(org.springframework.beans.factory.support.BeanDefinitionBuilder, org.w3c.dom.Element) method and do such parsing yourself, or (more likely) subclass the AbstractSingleBeanDefinitionParser or AbstractBeanDefinitionParser classes directly.

The process of actually registering the SimpleCacheBeanDefinitionParser with the Spring XML parsing infrastructure is described in the Spring Framework reference documentation (in one of the appendices).

For an example of this parser in action (so to speak), do look at the source code for the UtilNamespaceHandler.PropertiesBeanDefinitionParser; the observant (and even not so observant) reader will immediately notice that there is next to no code in the implementation. The PropertiesBeanDefinitionParser populates a PropertiesFactoryBean from an XML element that looks like this:

<util:properties location="jdbc.properties"/>

The observant reader will notice that the sole attribute on the <util:properties/> element matches the PropertiesLoaderSupport.setLocation(org.springframework.core.io.Resource) method name on the PropertiesFactoryBean (the general usage thus illustrated holds true for any number of attributes). All that the PropertiesBeanDefinitionParser needs actually do is supply an implementation of the AbstractSingleBeanDefinitionParser.getBeanClass(org.w3c.dom.Element) method to return the PropertiesFactoryBean type.

Since:
2.0
Author:
Rob Harrop, Rick Evans, Juergen Hoeller
See Also:
  • Constructor Details

    • AbstractSimpleBeanDefinitionParser

      public AbstractSimpleBeanDefinitionParser()
  • Method Details

    • doParse

      protected void doParse(Element element, ParserContext parserContext, BeanDefinitionBuilder builder)
      Parse the supplied Element and populate the supplied BeanDefinitionBuilder as required.

      This implementation maps any attributes present on the supplied element to PropertyValue instances, and adds them to the builder.

      The extractPropertyName(String) method is used to reconcile the name of an attribute with the name of a JavaBean property.

      Overrides:
      doParse in class AbstractSingleBeanDefinitionParser
      Parameters:
      element - the XML element being parsed
      builder - used to define the BeanDefinition
      parserContext - the object encapsulating the current state of the parsing process
      See Also:
    • isEligibleAttribute

      protected boolean isEligibleAttribute(Attr attribute, ParserContext parserContext)
      Determine whether the given attribute is eligible for being turned into a corresponding bean property value.

      The default implementation considers any attribute as eligible, except for the "id" attribute and namespace declaration attributes.

      Parameters:
      attribute - the XML attribute to check
      parserContext - the ParserContext
      See Also:
    • isEligibleAttribute

      protected boolean isEligibleAttribute(String attributeName)
      Determine whether the given attribute is eligible for being turned into a corresponding bean property value.

      The default implementation considers any attribute as eligible, except for the "id" attribute.

      Parameters:
      attributeName - the attribute name taken straight from the XML element being parsed (never null)
    • extractPropertyName

      protected String extractPropertyName(String attributeName)
      Extract a JavaBean property name from the supplied attribute name.

      The default implementation uses the Conventions.attributeNameToPropertyName(String) method to perform the extraction.

      The name returned must obey the standard JavaBean property name conventions. For example for a class with a setter method 'setBingoHallFavourite(String)', the name returned had better be 'bingoHallFavourite' (with that exact casing).

      Parameters:
      attributeName - the attribute name taken straight from the XML element being parsed (never null)
      Returns:
      the extracted JavaBean property name (must never be null)
    • postProcess

      protected void postProcess(BeanDefinitionBuilder beanDefinition, Element element)
      Hook method that derived classes can implement to inspect/change a bean definition after parsing is complete.

      The default implementation does nothing.

      Parameters:
      beanDefinition - the parsed (and probably totally defined) bean definition being built
      element - the XML element that was the source of the bean definition's metadata