lempinen.neatseeker.core
Class Repository

java.lang.Object
  |
  +--lempinen.neatseeker.core.Repository

public class Repository

extends java.lang.Object

Class for handling a repository of Buckets.

The Repository class provides a layer of abstraction over the individual Buckets. A Bucket is a TreeMap containing keyword data for a particular fragment of the entire index (typically, words beginning with same initial letters).

The Repository class is responsible for handling incoming and outgoing keyword requests and directing them to the correct Bucket.

As a secondary task, the Repository also manages a TargetIndex, which complements the keyword data in the Buckets and provides information about the indexed documents (e.g. their URIs, titles etc.)

A Repository can be created read-only or read-write, depending on which constructor you choose. Read-write mode is required for indexing, whereas for searches, you can use the read-only mode. This allows you to use NeatSeeker e.g. in a servlet environment, where the user account that runs the server has only write access to the index.

Version:

$Id: Repository.java,v 1.5 2000/10/05 16:51:39 lempinen Exp $

Author:

Sami Lempinen

See Also:

Bucket, TargetIndex

Field Summary

private java.util.HashMap	buckets The HashMap that holds the individual buckets.
protected Configuration	conf The Configuration object.
protected TargetIndex	documents The TargetIndex object for storing details about the indexed targets.
protected boolean	readonly The read-only flag.
protected java.lang.String	root The root directory of the Bucket repository.
protected Stemmer	stemmer The Stemmer object.

Constructor Summary

Repository(Configuration c)
Creates a new read-write Repository with the given configuration.

Repository(Configuration c, boolean ro)
Creates a new Repository (read or read-write).

Method Summary

void	add(Entry entry) Adds a single key-value pair to this repository.
void	cleanup() Cleans up the repository.
private java.lang.String	getBucket(java.lang.String key) Returns the hashed lowercase bucket name for a particular key.
long	getNextID() Returns the ID of the next Target.
Target	getTarget(long id) Returns a Target from the TargetIndex by its ID.
private void	initRepository() Initialises the repository directory.
long	put(Target t) Adds an object into the TargetIndex and returns a unique ID.
void	save() Writes all Buckets and the TargetIndex onto disk.
PointerList	searchSimple(java.lang.String key) Searches for a particular key in the repository.

Methods inherited from class java.lang.Object

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

Field Detail

buckets

private java.util.HashMap buckets

The HashMap that holds the individual buckets. This HashMap provides a cache of Buckets, and takes away the overhead of loading Buckets from disk whenever they are accessed.

root

protected java.lang.String root

The root directory of the Bucket repository.

documents

protected TargetIndex documents

The TargetIndex object for storing details about the indexed targets.

readonly

protected boolean readonly

The read-only flag.

conf

protected Configuration conf

The Configuration object.

stemmer

protected Stemmer stemmer

The Stemmer object.

Constructor Detail

Repository

public Repository(Configuration c)
           throws java.io.IOException

Creates a new read-write Repository with the given configuration.

Repository

public Repository(Configuration c,
                  boolean ro)
           throws java.io.IOException

Creates a new Repository (read or read-write).

Method Detail

add

public void add(Entry entry)

Adds a single key-value pair to this repository.

searchSimple

public PointerList searchSimple(java.lang.String key)

Searches for a particular key in the repository.

save

public void save()

Writes all Buckets and the TargetIndex onto disk.

getTarget

public Target getTarget(long id)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException

Returns a Target from the TargetIndex by its ID.

getNextID

public long getNextID()
               throws java.io.IOException

Returns the ID of the next Target.

put

public long put(Target t)
         throws java.io.IOException

Adds an object into the TargetIndex and returns a unique ID.

getBucket

private java.lang.String getBucket(java.lang.String key)

Returns the hashed lowercase bucket name for a particular key. Tune this to control serialized Bucket size.

cleanup

public void cleanup()

Cleans up the repository.

initRepository

private void initRepository()

Initialises the repository directory.