org.apache.solr.search

Class SolrDocumentFetcher

java.lang.Object

org.apache.solr.search.SolrDocumentFetcher

public class SolrDocumentFetcher
extends Object

A helper class of SolrIndexSearcher for stored Document related matters including DocValue substitutions.

Method Summary

Modifier and Type	Method and Description
void	decorateDocValueFields(SolrDocumentBase doc, int docid, Set<String> fields) This will fetch and add the docValues fields to a given SolrDocument/SolrInputDocument
Document	doc(int docId)
Document	doc(int i, Set<String> fields) Retrieve the Document instance corresponding to the document id.
void	doc(int docId, StoredFieldVisitor visitor)
Set<String>	getAllStored()
SolrCache<Integer,Document>	getDocumentCache()
Set<String>	getDvsCanSubstituteStored()
Set<String>	getNonStoredDVs(boolean onlyUseDocValuesAsStored) Returns an unmodifiable set of non-stored docValues field names.
Set<String>	getNonStoredDVsWithoutCopyTargets() Returns an unmodifiable set of names of non-stored docValues fields, except those that are targets of a copy field.
Collection<String>	getStoredHighlightFieldNames() Returns a collection of the names of all stored fields which can be highlighted the index reader knows about.
boolean	isLazyFieldLoadingEnabled()
SolrDocument	solrDoc(int luceneDocId, SolrReturnFields solrReturnFields) This is an optimized version for populating a SolrDocument that: 1.

Methods inherited from class java.lang.Object

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

Method Detail

isLazyFieldLoadingEnabled

public boolean isLazyFieldLoadingEnabled()

getDocumentCache

public SolrCache<Integer,Document> getDocumentCache()

getStoredHighlightFieldNames

public Collection<String> getStoredHighlightFieldNames()

Returns a collection of the names of all stored fields which can be highlighted the index reader knows about.

doc

public Document doc(int docId)
             throws IOException

Throws:

IOException

See Also:

SolrIndexSearcher.doc(int)

doc

public Document doc(int i,
                    Set<String> fields)
             throws IOException

Retrieve the Document instance corresponding to the document id.

NOTE: the document will have all fields accessible, but if a field filter is provided, only the provided fields will be loaded (the remainder will be available lazily).

Throws:

IOException

See Also:

SolrIndexSearcher.doc(int, Set)

solrDoc

public SolrDocument solrDoc(int luceneDocId,
                            SolrReturnFields solrReturnFields)

This is an optimized version for populating a SolrDocument that: 1. fetches all fields from docValues if possible. If no decompression of the stored data is necessary, we can avoid a disk seek and decompression cycle. This step is only used if all requested fields are {code docValues=true stored=false multiValued=false}. This last restriction because multiValued docValues fields do not faithfully reflect the input order in all cases. the values are returned and no decompression is necessary. 2. if 1 is impossible, try to fetch all requested fields from the stored values. If the stored data has to be decompressed anyway, it's more efficient to just get all field values from the stored values. If we got all the requested fields, return. 3. add fields where docValues=true stored=false thus could not be fetched in step 2

Parameters:

luceneDocId - The Lucene doc ID

solrReturnFields - the structure holding the fields to be returned. The first time this method is called for a particular document list, it will be modified by adding a RetrieveFieldsOptimizer for use in future calls.

Returns:

The SolrDocument with values requested.

This method is designed to be as simple as possible to use, just call it. e.g. {code SolrDocument sdoc = docFetcher.solrDoc(id, solrReturnFields);} then process the resulting SolrDocument as usual. Subsequent calls with the same solrReturnFields will re-use the optimizer created the first time. NOTE: DO NOT re-use the same SolrReturnFields object if the fields requested change.

doc

public void doc(int docId,
                StoredFieldVisitor visitor)
         throws IOException

Throws:

IOException

See Also:

SolrIndexSearcher.doc(int, StoredFieldVisitor)

decorateDocValueFields

public void decorateDocValueFields(SolrDocumentBase doc,
                                   int docid,
                                   Set<String> fields)
                            throws IOException

This will fetch and add the docValues fields to a given SolrDocument/SolrInputDocument

Parameters:

doc - A SolrDocument or SolrInputDocument instance where docValues will be added

docid - The lucene docid of the document to be populated

fields - The fields with docValues to populate the document with. DocValues fields which do not exist or not decodable will be ignored.

Throws:

IOException

getDvsCanSubstituteStored

public Set<String> getDvsCanSubstituteStored()

getAllStored

public Set<String> getAllStored()

getNonStoredDVs

public Set<String> getNonStoredDVs(boolean onlyUseDocValuesAsStored)

Returns an unmodifiable set of non-stored docValues field names.

Parameters:

onlyUseDocValuesAsStored - If false, returns all non-stored docValues. If true, returns only those non-stored docValues which have the SchemaField.useDocValuesAsStored() flag true.

getNonStoredDVsWithoutCopyTargets

public Set<String> getNonStoredDVsWithoutCopyTargets()

Returns an unmodifiable set of names of non-stored docValues fields, except those that are targets of a copy field.