mmx metadata framework
...the DNA of your data
MMX metadata framework is a lightweight implementation of OMG Metadata Object Facility built on relational database technology. MMX framework
is based on three general concepts:
Metamodel | MMX Metamodel provides a storage mechanism for various knowledge models. The data model underlying the metadata framework is more abstract in nature than metadata models in general. The model consists of only a few abstract entities... see more.
Access layer | Object oriented methods can be exploited using inheritance to derive the whole data access layer from a small set of primitives created in SQL. MMX Metadata Framework provides several diverse methods of data access to fulfill different requirements... see more.
Generic transformation | A large part of relationships between different objects in metadata model are too complex to be described through simple static relations. Instead, universal data transformation concept is put to use enabling definition of transformations, mappings and transitions of any complexity... see more.

The X Is For eXtensibility

September 11, 2011 19:24 by mmx

XDTL stands for eXtensible Data Transformation Language. Extensibility here means that new language elements can be easily added without having to make changes to XML schema defining the core XDTL language. These extensions can, for example, be coded in XDTL and stored as XDTL packages with task names identifying the extension elements. XDTL Runtime expects to find the extension element libraries in directories listed in extensions.path parameter in xdtlrt.xml configuration file: the pathlist is scanned sequentially until a task with a matching name is found. 

During package execution an extension is provided with a full copy of the current context. An extension gets access to every variable value that is present in the calling context, as well as its attribute values that get converted into variables with names of the attributes. From the extension's point of view all those values are 'read-write', but only those passed as variables retain their values after the extension element finishes. So, considering passing values to an extension, variables can be seen as 'globals' that return values and extension element attributes as 'locals' that get discarded.

XDTL syntax definition (XML schema) includes an any element that allows XDTL language to be extended with elements not specified directly in the schema. any element in XDTL is defined as

<xs:any namespace="##other" processContents="lax"/>

##other means that only elements from namespace other than the namespace of the parent element are allowed. In other words, when parser sees an unknown element it will not complain but assume that it could be defined in some other schema. This prevents ambiguity in XML Schema (Unique Particle Attribution). Setting processControl attribute of an any element to "lax" states that if that 'other' schema cannot be obtained, parser will not generate an error.


So how does this work? We assume that our main script is referencing an external XML schema, elements of which are qualified with prefix 'ext':


In this external schema a single element, "show" with attribute "text" is defined. Here are some examples of what works and what doesn't.

<ext:show text="sometext"/> works, as the external namespace with element "show" is referenced by the prefix 'ext'.

<show xmlns="" text="sometext"/> also works, as the namespace reference is 'embedded' in the "show" element.

<show text="sometext"/> does not validate, as the parser looks for element "show" in the current schema (error message Invalid content was found starting with element 'show' is produced).

<ext:show nottext="sometext"/> does not validate either (Attribute 'nottext' is not allowed to appear in element 'ext:show').

<ext:notshow text="sometext"/> validates but still does not work! As the processContents attribute of any element is "lax", although the element is not found the parser ignores this. However, the XDTL Runtime complains as it cannot find element definition in extension pathlist.


What if we would want to use extensions without XML schema accompanying it? For that we remove the reference to the external schema from the script header and run the examples once again.

<ext:show text="sometext"/> would not validate any more as the prefix 'ext' is not defined. The same applies to all the other examples with prefix in front of the extension element.

<show text="sometext"/> would not validate either as the parser looks for extension element in the current schema.

<show xmlns="" text="sometext"/>, however, validates and works! Although the parser cannot find the schema it does not complain due to "lax" processContents attribute. As long as XDTL Runtime is able to find the library package containing the extension in the pathlist everything is fine, otherwise it would give Command 'Extension' failed error.


So here's the summary. Extended elements (commands) can be well-defined (having their syntax definitions in form of an XML schema) or undefined (just the package, no XML schema), just as a transformation designer sees fit. In the former case, extended elements will be validated exactly as the core language elements would, in the latter case they will pass without validation. If an undefined and non-validated extension element is executed and does not match its invocation a run-time error would be generated.


Moving Data in XDTL: Wooden, Painted Red

August 4, 2011 00:37 by mmx

Transferring data from a place to another is the cornerstone of almost every data transformation job. XDTL has a very flexible and feature-ladden fetch command for that job. fetch element has the following attributes:


a query to a source database. source can contain either an implicit query or a variable containing the query. 


target table or output text file. Specifying destination connection attribute indicates that target is a table. In case target is a text file, all the common attributes of text files (delimiter, quote, null, encoding) also apply and can optionally be present.


overwrite="1" causes an automatic TRUNCATE command against the target table. 


specifies an XDTL variable (internally a Java array) to store the resultset of the source query. rowset and target are mutually exclusive and should not be present in the same fetch element. 


connection of the source query. 


connection of the target table. 


The most obvious use case for fetch would be transferring data from a database table to a text file. The following xdtl command would fetch the data specified by the query (source attribute) into a CSV (target attribute) file with ; as the delimiter and " as the quote symbol (defaults, they can be overriden):

<xdtl:fetch source="SELECT col, col, ... FROM table WHERE cond" target="textfile" connection="$conn"/> 

In case we need to process the results of the fetch operation further inside the xdtl package, fetch can store the result set in a variable (technically a Java array): 

<xdtl:fetch source="SELECT col, col, ... FROM table WHERE cond" rowset="myrowset" connection="$conn"/>
<xdtl:for item="myrow" rowset="$myrowset"> something here. columns have to be addressed via Java array syntax: ${myrow[0]}, ${myrow[1]}, etc...

We can go even further and let the read command - normally used to read data from text file(s) into database table(s) - to consider this rowset variable as its input and store data into a table with compliant structure, ie. matching the rowset by position  (type="ROWSET" tells read that input data is coming from a Java array instead of a text file):

<xdtl:fetch source="SELECT * FROM sourcetable WHERE cond" rowset="myrowset" connection="$conn"/>
<xdtl:read source="$myrowset" target="targettable" type="ROWSET" overwrite="1" connection="$dest"/>

Finally, in case we have no need to process the data in the xdtl package, use fetch command to achieve the behavior of a typical DataPump task in most of the graphical ETL tools in just one step (here, the second connection attribute, destination is used to specify target connection):

<xdtl:fetch source="SELECT * FROM sourcetable WHERE cond" target="targettable" connection="$conn" destination="$dest"/> 

Caution! fetch can transfer data between different servers that could come from different vendors. It can do automatic datatype conversions within reasonable limits (*), but all 'weird', non-standard data types (**) need to be taken care of in the source query statement. Source and target data structures are assumed to be identical, this requirement must be handled in source query by presenting data in suitable format and structure for the target table. And, mondo importante: data columns are matched based on position, not column names!

Taking all this into consideration, all the basic data transfer tasks can be handled with ease. For more complex transformations extensions or customized library tasks should be used. Naturally fetch would work best for reasonable data amounts; for large data other approaches (eg. bulkload tools) should be employed.

(*) Datatype conversion is done in JDBC via Java Primitive Data Types.


MMX Framework: Extending with Extended Properties

December 10, 2010 13:36 by mmx

Pretty often there is a need to add some extra attributes to a metamodel that has already been deployed. The need for those changes usually arises from changed requirements for an application. So how to enable those changes without extending the metamodel (M2 layer) itself?

There is a special property type, AnyProperty, to handle this scenario. Whan an object type (class) has a property of type AnyProperty, an object (instance of the class) can have an arbitrary number of different properties of type AnyProperty that all share the same property_type_cd. Then how to distinguish between those properties? 

All the instances of an AnyProperty descend from the same property type of a class in M2 layer, but an application who 'owns' them can give them new names. Alternatively, it is possible to use the same name, but a distinguishing prefix directly in the value field of the property. So an AnyProperty instance 'belongs' to an application and its completely up to the application how it chooses to use, denote and present this property.

Downsides (there obviously are some, right?). First, the inheritance mechanisms built into MMX Framework (MMXMD API) obviously cannot provide full support to those properties, as the names of them are not known to M2 layer. Second, AnyProperty is defined as String in M2 so an application cannot rely on the type information if it chooses to have an instance of AnyProperty that is implemented with some other datatype (eg. an XML document or a date formatted as a string).

Anyway, this mechanism (which bears some resemblance with anyAttribute element in XML Schema) enables making run-time changes in metamodel when design-time is not an option (run-time and design-time here refer to modeling, not application design), just remember to use it with care.