Package ai.preferred.cerebro.index.builder

Class PersonalizedDocFactory

java.lang.Object

ai.preferred.cerebro.index.builder.PersonalizedDocFactory

public class PersonalizedDocFactory
extends java.lang.Object

This class handles the creation of Document object to ensure that there is no conflict in field name and that all the hashcoding behaves as intended.

Method Summary

Modifier and Type	Method	Description
void	addField(org.apache.lucene.index.IndexableField... fields)
boolean	checkReservedFieldName(java.lang.String fieldname)
void	create(java.lang.Object ID, double[] features)
void	create(java.lang.Object ID, org.apache.lucene.index.IndexableField... fields)
org.apache.lucene.document.Document	getDoc()

Methods inherited from class java.lang.Object

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

Method Detail

create

public void create(java.lang.Object ID,
                   double[] features)
            throws java.lang.Exception

Parameters:

ID - unique ID of the document.

features - the latent feature vector to index.

Throws:

DocNotClearedException - this exception is triggered when a call to create(Object, double[]) is not paired with a call to getDoc(). Only use this function to construct a Document containing latent vector. To add additional fields to the Document, use addField(IndexableField...). Call getDoc() to pass the Document to IndexWriter or before creating a new Document.

java.lang.Exception

create

public void create(java.lang.Object ID,
                   org.apache.lucene.index.IndexableField... fields)
            throws SameNameException,
                   DocNotClearedException

Parameters:

ID - unique ID of the document

fields - the custom fields

Throws:

SameNameException - this is triggered when one of your custom field has name identical to Cerebro reserved word. See more detail at IndexConst.

DocNotClearedException - this exception is triggered when a call to create(Object, double[]) is not paired with a call to getDoc(). Call this function to construct a generic text-only Document. Should you need to add latent vector later call getDoc and start anew with the other create method.

addField

public void addField(org.apache.lucene.index.IndexableField... fields)
              throws SameNameException

Parameters:

fields -

Throws:

SameNameException - this is triggered when one of your custom field has name identical to Cerebro reserved word. See more detail at IndexConst. After calling create(Object, double[]) to create a document with latent vector if you still want add more custom fields to a Document then use this function.

getDoc

public org.apache.lucene.document.Document getDoc()

Returns:

the Document object being built since the last create(Object, double[]) or create(Object, IndexableField...) call. After calling this function the pointer doc become null again.

checkReservedFieldName

public boolean checkReservedFieldName(java.lang.String fieldname)

Parameters:

fieldname -

Returns:

true if the fieldname is the similar to one of the reserved words.