CloverETL Release Notes

• Schema with local imports can now be properly used
• Mapping generation from XSD no longer prefills element types as values
• XSD root selection combo default selection changed
• Drop of a field onto element and binding tweaked to be more user-friendly
• Root element can contain a binding to a port as long as only one record per file is set. A warning is reported otherwise.

• see http://server-demo-ec2.cloveretl.com/clover/docs/osgi.html
• Engine plugins library isolation, fixes clashing custom libraries with Server default libraries

• for reading structured and unstructured flat files
• similar to MultiLevelReader, yet more powerful and user-friendly

• brand new mapping (incompatible)
• old XMLWriter still available as a deprecated component
• supports disk caching for large amounts of data

• brand new XMLWriter interface with visual and source editors

• automatic data denormalization
• metadata creation out of XMLExtract dialog
• ability to add ports and assign metadata
• annotations
• improved error reporting

• handy description bar
• component autosize for component name
• improved edge handling

• drag'n'drop features for dragging files to graph, outline, etc.

• tasks now can be executed manually instead just by the scheduler

Now you can use list of records and maps of records in your code.
Example:
	// lists
	MyDataRecord[] myList;
	MyDataRecord a;
	MyDataRecord b;
	myList[2] = a;
	b = myList[2];
	// maps
	map[string, MyDataRecord] myMap;
	myMap["hello"] = a;
(For maps, only values can be records)

When reading XML files with XML Extract, you can encouter the need to know whether an element or attribute is present in the source
or just has an empty value.
XMLExtract now sets either "null" value whenever the element/attribute is omitted or empty value ("") 
when the element/attribute is present but without a value

Now you can easily see whether the aggration function produce the integer overflow and thus you're not getting the right results.

DataIntersection now stops immediately when it sees the input is not sorted properly ensuring data integrity.

Quoting support in readers/writers has been around for long time but now you can even specify the quoting character. In previous CloverETL
versions there has been built-in quote defined as either `'` or `"` (quote or double quote).
Now we present more fine-grained setting where the quote character can be defined.   

• Introducing next generation of CTL (CloverETL Transformation Languge)
• Typed languge, close to Java, can be transparently converted to Java
• up to 30% speed increase

• Dictionary can be declared in the Designer

A graph can now run on multiple nodes in a cluster

• CloverETL Designer