Package org.apache.solr.response.transform

Class DocTransformer

java.lang.Object

org.apache.solr.response.transform.DocTransformer

Direct Known Subclasses:

BaseEditorialTransformer, DocTransformer.NoopFieldTransformer, DocTransformers, RenameFieldTransformer, ScoreAugmenter, ValueSourceAugmenter

public abstract class DocTransformer
extends Object

A DocTransformer can add, remove or alter a Document before it is written out to the Response. For instance, there are implementations that can put explanations inline with a document, add constant values and mark items as being artificially boosted (see QueryElevationComponent)

New instance for each request

See Also:

TransformerFactory

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	DocTransformer.NoopFieldTransformer	Trivial Impl that ensure that the specified field is requested as an "extra" field, but then does nothing during the transformation phase.

Field Summary

Fields

Modifier and Type	Field	Description
protected ResultContext	context

Constructor Summary

Constructors

Constructor	Description
DocTransformer()

Method Summary

Modifier and Type	Method	Description
String[]	getExtraRequestFields()	When a transformer needs access to fields that are not automatically derived from the input fields names, this option lets us explicitly say the field names that we hope will be in the SolrDocument.
abstract String	getName()
Collection<String>	getRawFields()	If this transformer wants to bypass escaping in the TextResponseWriter and write content directly to output for certain field(s), the names of any such field(s) should be returned
boolean	needsSolrIndexSearcher()	Indicates if this transformer requires access to the underlying index to perform it's functions.
void	setContext(ResultContext context)	This is called before transform(org.apache.solr.common.SolrDocument, int, float) to provide context for any subsequent transformations.
String	toString()
abstract void	transform(org.apache.solr.common.SolrDocument doc, int docid)	This is where implementations do the actual work.
void	transform(org.apache.solr.common.SolrDocument doc, int docid, float score)	This is where implementations do the actual work.

Methods inherited from class java.lang.Object

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

Field Detail

context

protected ResultContext context

Constructor Detail

DocTransformer

public DocTransformer()

Method Detail

getName

public abstract String getName()

Returns:

The name of the transformer

setContext

public void setContext(ResultContext context)

This is called before transform(org.apache.solr.common.SolrDocument, int, float) to provide context for any subsequent transformations.

Parameters:

context - The ResultContext stores information about how the documents were produced.

See Also:

needsSolrIndexSearcher()

getRawFields

public Collection<String> getRawFields()

If this transformer wants to bypass escaping in the TextResponseWriter and write content directly to output for certain field(s), the names of any such field(s) should be returned

NOTE: normally this will be conditional on the `wt` param in the request, as supplied to the DocTransformer's parent TransformerFactory at the time of transformer creation.

Returns:

Collection containing field names to be written raw; if no field names should be written raw, an empty collection should be returned. Any collection returned collection need not be externally modifiable -- i.e., Collections.singleton(Object) is acceptable.

needsSolrIndexSearcher

public boolean needsSolrIndexSearcher()

Indicates if this transformer requires access to the underlying index to perform it's functions.

In some situations (notably RealTimeGet) this method may be called before setContext(org.apache.solr.response.ResultContext) to determine if the transformer must be given a "full" ResultContext and accurate docIds that can be looked up using ResultContext.getSearcher(), or if optimizations can be taken advantage of such that ResultContext.getSearcher() may return null, and docIds passed to transform(org.apache.solr.common.SolrDocument, int, float) may be negative.

The default implementation always returns false.

See Also:

ResultContext.getSearcher(), transform(org.apache.solr.common.SolrDocument, int, float)

transform

public void transform(org.apache.solr.common.SolrDocument doc,
                      int docid,
                      float score)
               throws IOException

This is where implementations do the actual work. If implementations require a valid docId and index access, the needsSolrIndexSearcher() method must return true

Default implementation calls transform(SolrDocument, int).

Parameters:

doc - The document to alter

docid - The Lucene internal doc id, or -1 in cases where the doc did not come from the index

score - the score for this document

Throws:

IOException - If there is a low-level I/O error.

See Also:

needsSolrIndexSearcher()

transform

public abstract void transform(org.apache.solr.common.SolrDocument doc,
                               int docid)
                        throws IOException

This is where implementations do the actual work. If implementations require a valid docId and index access, the needsSolrIndexSearcher() method must return true

Parameters:

doc - The document to alter

docid - The Lucene internal doc id, or -1 in cases where the doc did not come from the index

Throws:

IOException - If there is a low-level I/O error.

See Also:

needsSolrIndexSearcher()

getExtraRequestFields

public String[] getExtraRequestFields()

When a transformer needs access to fields that are not automatically derived from the input fields names, this option lets us explicitly say the field names that we hope will be in the SolrDocument. These fields will be requested from the SolrIndexSearcher but may or may not be returned in the final QueryResponseWriter

Returns:

a list of extra lucene fields

toString

public String toString()

Overrides:

toString in class Object