com.googlecode.fascinator.harvester.filesystem

Class DerbyCache

java.lang.Object

com.googlecode.fascinator.harvester.filesystem.DerbyCache

public class DerbyCache
extends Object

This class is designed to encapsulate all logic required to run a cache backed into a derby database in support of the file system harvester. There are two caches available.

• Basic: The file is considered 'cached' if the last modified date matches the database entry. On some operating systems (like linux) this can provide a minimum of around 2 seconds of granularity. For most purposes this is sufficient, and this cache is the most efficient.
• Hashed: The entire contents of the file are SHA hashed and the hash is stored in the database. The file is considered cached if the old hash matches the new hash. This approach will only trigger a harvest if the contents of the file really change, but it is quite slow across large data sets and large files.

Some form of caching must be enabled to support deletion detection for this plugin.

Author:

Greg Pendlebury

Constructor Summary

Constructors

Constructor and Description
DerbyCache(JsonSimpleConfig config)

Method Summary

Methods

Modifier and Type	Method and Description
Set<String>	getUnsetFlags() Used to support deletion detection.
boolean	hasChanged(String oid, File file) Check whether or not the file has changed according to the configured cache.
void	purgeUnsetFlags() Used to support deletion detection.
void	resetFlags() Used to support deletion detection.
void	shutdown() Shutdown the database connections and cleanup.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

DerbyCache

public DerbyCache(JsonSimpleConfig config)
           throws Exception

Throws:

Exception

Method Detail

hasChanged

public boolean hasChanged(String oid,
                 File file)

Check whether or not the file has changed according to the configured cache. The response from this method should be used as an indicator on whether or not to proceed with harvesting the file. In that context:

• If there are any errors or exceptions accessing or assessing the file, the return value will be false.
• If no caches are configured at all, the return value will be true.
• If any cache is turned on and the file appears to have changed, the return value will be true.
• If any cache is turned on and the file does not appear to have changed, the return value will be false.

Parameters:

file - : The file to test.

Returns:

boolean : True if the harvest should proceed, otherwise false.

resetFlags

public void resetFlags()

Used to support deletion detection. Should be called at the beginning of a harvest to reset the flags in the database.

getUnsetFlags

public Set<String> getUnsetFlags()

Used to support deletion detection. This method is called after the harvest, and will return a list of all file paths

Returns:

Set: A list of all object IDs in the cache which have not been 'touched'

purgeUnsetFlags

public void purgeUnsetFlags()

Used to support deletion detection. This method is just for cleanup after the process completes.

shutdown

public void shutdown()
              throws Exception

Shutdown the database connections and cleanup.

Throws:

Exception - if there are errors