org.geotools.data
Interface DataStoreFactorySpi

All Superinterfaces:

Factory

All Known Subinterfaces:

FileDataStoreFactorySpi

All Known Implementing Classes:

AbstractDataStoreFactory, ArcSDEDataStoreFactory, BBOXDataStoreFactory, DB2DataStoreFactory, DirectoryDataStoreFactory, GeoMediaDataStoreFactory, GMLDataStoreFactory, HsqlDataStoreFactory, IndexedShapefileDataStoreFactory, JDBCDataStoreFactory, LocationsXYDataStoreFactory, MIFDataStoreFactory, MySQLDataStoreFactory, OracleDataStoreFactory, OracleOCIDataStoreFactory, PostgisDataStoreFactory, PropertyDataStoreFactory, ShapefileDataStoreFactory, TigerDataStoreFactory, VPFDataStoreFactory, VPFFileFactory

public interface DataStoreFactorySpi

extends Factory

Constructs a live DataStore from a set of parameters.

An instance of this interface should exist for all data stores which want to take advantage of the dynamic plugin system. In addition to implementing this interface datastores should have a services file:

META-INF/services/org.geotools.data.DataStoreFactorySpi

The file should contain a single line which gives the full name of the implementing class.

example:
e.g. org.geotools.data.mytype.MyTypeDataSourceFacotry

The factories are never called directly by client code, instead the DataStoreFinder class is used.

The following example shows how a user might connect to a PostGIS database, and maintain the resulting datastore in a registry:

 HashMap params = new HashMap();
 params.put("namespace", "leeds"); 
 params.put("dbtype", "postgis");
 params.put("host","feathers.leeds.ac.uk");
 params.put("port", "5432");
 params.put("database","postgis_test");
 params.put("user","postgis_ro");
 params.put("passwd","postgis_ro");
 
 DefaultRegistry registry = new DefaultRegistry();
 registry.addDataStore("leeds", params);
 
 DataStore postgis = registry.getDataStore( "leeds" );
 FeatureSource = postgis.getFeatureSource( "table" );
 

• Jody - can we please get something better then Param to describe what is allowed?
  Jody - ISO19119 has something that looks okay, WSDL/SOAP could be used?
• Jody - can we seperate out Identification of a Service from configration of the service?
  Jody - this is mostly a problem when managing user supplied configurations in GeoServer and uDig.
  Jody - the "Catalog API" has now been ported and contains a URI as indentification, while still allowing configuration using a Map of parameters

Author:

Jody Garnett, Refractions Research

Nested Class Summary

static class

DataStoreFactorySpi.Param Data class used to capture Parameter requirements.

Method Summary

boolean	canProcess(java.util.Map params) Test to see if this factory is suitable for processing the data pointed to by the params map.
DataStore	createDataStore(java.util.Map params) Construct a live data source using the params specifed.
DataStore	createNewDataStore(java.util.Map params)
java.lang.String	getDescription() Describe the nature of the datasource constructed by this factory.
java.lang.String	getDisplayName() Name suitable for display to end user.
DataStoreFactorySpi.Param[]	getParametersInfo() MetaData about the required Parameters (for createDataStore).
boolean	isAvailable() Test to see if this datastore is available, if it has all the appropriate libraries to construct a datastore.

Methods inherited from interface org.geotools.factory.Factory

getImplementationHints

Method Detail

createDataStore

public DataStore createDataStore(java.util.Map params)
                          throws java.io.IOException

Construct a live data source using the params specifed.

You can think of this as setting up a connection to the back end data source.

Magic Params: the following params are magic and are honoured by convention by the GeoServer and uDig application.

• "user": is taken to be the user name
• "passwd": is taken to be the password
• "namespace": is taken to be the namespace prefix (and will be kept in sync with GeoServer namespace management.

When we eventually move over to the use of OpperationalParam we will have to find someway to codify this convention.

Parameters:

params - The full set of information needed to construct a live data store. Typical key values for the map include: url - location of a resource, used by file reading datasources. dbtype - the type of the database to connect to, e.g. postgis, mysql

Returns:

The created DataStore, this may be null if the required resource was not found or if insufficent parameters were given. Note that canProcess() should have returned false if the problem is to do with insuficent parameters.

Throws:

IOException - if there were any problems setting up (creating or connecting) the datasource.

createNewDataStore

public DataStore createNewDataStore(java.util.Map params)
                             throws java.io.IOException

Throws:

java.io.IOException

getDisplayName

public java.lang.String getDisplayName()

Name suitable for display to end user.

A non localized display name for this data store type.

Returns:

A short name suitable for display in a user interface.

getDescription

public java.lang.String getDescription()

Describe the nature of the datasource constructed by this factory.

A non localized description of this data store type.

Returns:

A human readable description that is suitable for inclusion in a list of available datasources.

getParametersInfo

public DataStoreFactorySpi.Param[] getParametersInfo()

MetaData about the required Parameters (for createDataStore).

Interpretation of FeatureDescriptor values:

• getDisplayName(): Gets the localized display name of this feature.
• getName(): Gets the programmatic name of this feature (used as the key in params)
• getShortDescription(): Gets the short description of this feature.

This should be the same as:

 Object params = factory.getParameters();
 BeanInfo info = getBeanInfo( params );
 
 return info.getPropertyDescriptors();
 

Returns:

Param array describing the Map for createDataStore

canProcess

public boolean canProcess(java.util.Map params)

Test to see if this factory is suitable for processing the data pointed to by the params map.

If this datasource requires a number of parameters then this mehtod should check that they are all present and that they are all valid. If the datasource is a file reading data source then the extentions or mime types of any files specified should be checked. For example, a Shapefile datasource should check that the url param ends with shp, such tests should be case insensative.

Parameters:

params - The full set of information needed to construct a live data source.

Returns:

booean true if and only if this factory can process the resource indicated by the param set and all the required params are pressent.

isAvailable

public boolean isAvailable()

Test to see if this datastore is available, if it has all the appropriate libraries to construct a datastore. Most datastores should return true, because geotools will distribute the appropriate libraries. Though it's not a bad idea for DataStoreFactories to check to make sure that the libraries are there. OracleDataStoreFactory is an example of one that may generally return false, since geotools can not distribute the oracle jars, they must be added by the client. One may ask how this is different than canProcess, and basically available is used by the DataStoreFinder getAvailableDataStore method, so that DataStores that can not even be used do not show up as options in gui applications.

Returns:

true if and only if this factory has all the appropriate jars on the classpath to create DataStores.