Adding a source connector

LSC defines synchronization tasks as transformations from a source connector to a destination connector.

As of version 1.2, the only supported destination connector is a LDAP directory. Output to CSV and LDIF files is also possible via the log system.

The source connectors in LSC are completely modular. Several are provided:

• SimpleJndiSrcService: Commonly used connector to read from a LDAP directory
• SimpleJdbcSrcService: Commonly used connector to read from a JDBC-compatible database, via iBatis sqlMaps.

These source connectors are specified in a task definition in the configuration file (lsc.properties), by their full class name in Java (for example : org.lsc.jndi.SimpleJndiSrcService).

The rest of this page describes how to implement a new connector and use it for a synchronization.

A source connector is a Java class that implements the ISrcService interface. This interface defines two methods:

• Map<String,LscAttributes> getListPivots()
• IBean getBean(IBean taskBean, Entry<String,LscAttributes> obj)

The source connector may also implement a constructor with one parameter: a Properties object. This constructor will be called once, at the beginning of a synchronization task, and the properties will be the subset from lsc.properties under lsc.tasks.TaskName.srcService.

This method must return a list of all objects to be considered for synchronization in this task.

The return object is a Map (usually a HashMap, but whatever is fine) containing:

• The key is a name for the object. It is used for logging, so a clear name is helpful. LDAP source connectors use the DN here. Warning: this key must be guaranteed to be unique!
• The value is a LscAttributes object. This is basically another Map whose entries are attribute names and their values. Each attribute can only have one value. The names usually correspond to an LDAP attribute.

A simple representation of the returned object might be this (for a LDAP source connector):

"uid=jdoe,ou=people,o=example", {["sn", "Doe"], ["givenName", "John"]}
"uid=jane.doe,ou=people,o=example", {["sn", "Doe"], ["givenName", "Jane"]}
"uid=elilly,ou=people,o=example", {["sn", "Lilly"], ["givenName", "Evangeline"]}

Implementation of this method is completely up to you!

It could be used for anything - reading from another kind of data store, or returning only recently modified entries, etc.

This method must return the Bean passed as a parameter one (taskBean), once it has been filled with data from the source corresponding to one object, identified by parameter two (obj).

The taskBean is a newly-instantiated object of type IBean. At a minimum, you must set some attributes in it, with their names and values.

Continuing the example above for the first object (“John Doe”), you would do the equivalent of this:

taskBean.setAttribute("sn", new HashSet().add("Doe"));
taskBean.setAttribute("givenName", new HashSet().add("John"));
taskBean.setAttribute("homePhone", new HashSet().add(.....));

Of course, the actual values must be read from the source, by using the obj parameter to retrieve the object, then parse the attributes and add them to the Bean.

Let's say you just created a new source connector, called org.lsc.service.MySourceConnector. You can use it by defining these properties on a task called TaskName:

lsc.tasks.TaskName.srcService = org.lsc.service.MySourceConnector

To specify some properties to be used in the source connector, that will be passed in a Properties object to the constructor, just add them to the properties file:

lsc.tasks.TaskName.srcService.myproperty = something
lsc.tasks.TaskName.srcService.serviceURL = http://a.server.com/service

We hope that this document will help you get started writing a source connector. Of course, you'll need to play around a bit to get things working.

If you need some advice, feel free to ask on the mailing lists or on IRC. We'll be happy to help!